UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The theory of Long-term international capital flows and Canadian Corporate debt issues in the United… Stroetmann, Karl Antonius 1974

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1974_A1 S87_6.pdf [ 15.74MB ]
Metadata
JSON: 831-1.0100013.json
JSON-LD: 831-1.0100013-ld.json
RDF/XML (Pretty): 831-1.0100013-rdf.xml
RDF/JSON: 831-1.0100013-rdf.json
Turtle: 831-1.0100013-turtle.txt
N-Triples: 831-1.0100013-rdf-ntriples.txt
Original Record: 831-1.0100013-source.json
Full Text
831-1.0100013-fulltext.txt
Citation
831-1.0100013.ris

Full Text

THE THEORY OF LONG-TERM INTERNATIONAL CAPITAL FLOWS AND CANADIAN CORPORATE DEBT ISSUES IN THE UNITED STATES by KARL ANTONIUS STROETMANN Diplom-Kaufmann, Free University of B e r l i n , 1968 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY i n the Faculty of Commerce and Business Administration We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA May, 197^ I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an advanced degree a t the U n i v e r s i t y o f B r i t i s h C olumbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p urposes may be g r a n t e d by the Head o f my Department o r by h i s r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department o f Commerce and Business Administration The U n i v e r s i t y o f B r i t i s h Columbia Vancouver 8, Canada Date May 13, 1974 ABSTRACT This study presents a new attempt to gain a better under-standing of those forces that lead to the movement of funds from one country to another. Attention i s r e s t r i c t e d to the international market f o r long-term debt c a p i t a l . The empirical analysis focuses on c a p i t a l flows between Canada and the United States, p a r t i c u l a r l y on Canadian corporate borrowing i n the United States during the period from i 9 6 0 through May 1973. A model of the international term structure of interest rates i s developed. Differences i n time preferences between nations, exchange rate expectations and exchange r i s k , and transaction costs are shown to determine inter e s t rate d i f f e r -e n t i a l s and to influence international c a p i t a l flows. The inflow of long-term debt c a p i t a l into Canada i s almost exclusively due to the sale of new bond issues abroad by borrow-ers other than the federal government. A c t i v i t i e s of interna-t i o n a l investors i n secondary markets are of only minor import-ance. Therefore we have to rel y on an i n d i r e c t test of the basic features of our theory. We concentrate on an analysis of decisions by Canadian corporations to f l o a t U.S.-pay bonds. An examination of macro-economic data indicates that Cana-dians have a markedly higher demand f o r funds than Americans. An analysis of bond markets i n the two countries suggests that Canadian lenders prefer comparatively marketable s e c u r i t i e s . To further test f o r such differences i n time preferences, i t i s hypothesized that the a v a i l a b i l i t y of a well-functioning private placement market, of long-term forward commitments, and of longer maturities should be factors a t t r a c t i n g Canadian corporations to the U.S. bond market. Both discriminant analysis r e s u l t s and interviews with managers, underwriters, and l i f e insurance o f f i c e r s provide strong support f o r our assertions, except that longer terms to maturity a v a i l a b l e i n the U.S. are of lesser i n t e r e s t to Canadian firms. An analysis of exchange r i s k suggests that long terms to maturity and evenly distributed sinking fund payments should be preferred. Firms active i n export markets should regard foreign borrowing as a means to s e l l income denominated i n a foreign currency forward. Only weak s t a t i s t i c a l support f o r the asserted corporate behaviour i s found. Interviews revealed that exchange r i s k influences foreign borrowing, but i t s manage-ment i s not well understood. S l i g h t l y lower nominal interest costs seem to be a l l the protection against exchange r i s k firms require. Factors other than lower borrowing costs have become increasingly important f o r the choice between Canadian-pay and U .S.-pay bonds. In our model i t i s assumed that information and transaction costs are higher when two investors from d i f f e r e n t nations deal with each other than f o r purely domestic transactions. Causes of such differences and t h e i r impact on c a p i t a l flows are - i v -analyzed. Whereas the t y p i c a l American private placement i s small i n size and issued by a smaller, less f i n a n c i a l l y secure firm, Canadian U.S.-pay bonds are large i n size and sold by larger corporations or those with international connections. Continuing relationships with American lenders have also proven very b e n e f i c i a l . F i n a l l y , the in d i v i d u a l results are drawn together. It i s shown that the hypotheses derived from our model lead to the i d e n t i f i c a t i o n of variables that allow an almost perfect discrim-ination between Canadian-pay and U.S.-pay bonds issued by Cana-dian corporations. PREFACE International c a p i t a l flows and t h e i r impact on the well-being of nations have gained wide attention i n recent years not only among economic scholars but also among p o l i t i c i a n s and non-academic people. This study attempts to provide new insights into those factors that influence c a p i t a l movements from one country to another. We r e s t r i c t our attention to one p a r t i c u l a r component of international f i n a n c i a l transactions, the flow of long-term debt c a p i t a l . Corporations are a major group of trans-actors i n inte r n a t i o n a l c a p i t a l markets. The empirical analy-s i s concentrates on t h e i r a c t i v i t i e s , i n p a r t i c u l a r on Canadian corporate borrowing i n the United States. My intere s t i n t h i s subject arose because of the apparent d i f f i c u l t i e s encountered by economists i n t h e i r attempts to ex-p l a i n observed long-term c a p i t a l flows when using standard econ-ometric techniques. Such studies have usually employed macro-economic data. I hope that t h i s analysis of micro-economic behaviour w i l l lead to a better understanding of the forces behind international c a p i t a l movements and w i l l prove useful f o r future research i n t h i s area. My greatest indebtness i s to Professor Whatarangi Winiata. Many a sunny afternoon was spent i n hi s b e a u t i f u l garden d i s --v-- v i -cussing i n i t i a l drafts of Chapters 2 and 3 of t h i s d i s s e r t a -t i o n . His assistance i n setting up interviews and during the data c o l l e c t i o n process proved invaluable when we were faced with a seemingly endless number of obstacles. When he l e f t f o r a year of research abroad, Professor Bernard Schwab suc-ceeded him as Chairman of my thesis committee. His detailed review of my work l e d to numerous improvements i n the l o g i c a l rigour and the organization of t h i s study. The other members of my committee, Professors John G. Cragg, Maurice D. Levi, and James C. T. Mao also provided many help f u l comments, and through t h e i r penetrating c r i t i c i s m forced me to c l a r i f y and improve upon several points. Discussions i n the Finance Work-shop and suggestions received from Professors Michael J. Brennan and Alan Kraus were equally valuable. I am p a r t i c u l a r l y thankful to a l l those corporate f i n a n c i a l managers, investment bankers and l i f e insurance o f f i c e r s who provided data f o r t h i s study and granted us interviews to dis-cuss questions related to Canadian corporate borrowing i n the United States. Their interest i n t h i s work and t h e i r coopera-t i o n f a r exceeded our expectations. A considerable part of the information gathered f o r t h i s study i s considered c o n f i -dential by those who provided i t . Consequently t h e i r names cannot be revealed. This we promised to them to protect t h e i r i n t e r e s t s . - v i i -As part of i t s c u l t u r a l exchange programme with Germany, the Canadian Government, through the Canada Council, supported my studies and provided most of the financing f o r t h i s research. F i n a n c i a l assistance was also received from the L e s l i e G. J. Wong Memorial Foundation, The Vancouver C i t y Savings Credit Union,, and the Faculty of Commerce and Business Administration of the University of B r i t i s h Columbia. Mrs. Lynne Durward expertly typed t h i s report. F i n a l l y , I am most gr a t e f u l to my wife Barbara, not only f o r her a s s i s -tance i n the preparation of the numerous tables and graphs i n -cluded i n t h i s study but much more so f o r her moral support during my studies. I hope the future w i l l make up f o r the many hours my family had to spend without me. Vancouver, B.C. A p r i l , 197^ Karl A. Stroetmann C O N T E N T S Page ABSTRACT i i PREFACE v LIST OF TABLES x i LIST OF FIGURES x i i i LIST OF EXHIBITS xiv CHAPTER 1. INTRODUCTION 1 1.1 Short Review of the Development of the Theory of International Capital Flows 1 1.2 Purpose and Scope of Thesis 5 2. EXCHANGE RATE RISK, TRANSACTION COSTS AND EQUILI-BRIUM INTEREST RATES IN AN INTERNATIONAL LONG-TERM CAPITAL MARKET 14 2.1 Introduction 14 2.2 The Term Structure of Interest Rates i n a Closed Economy 1? 2.21 Ro l l ' s Approach 20 2 .3 The Term Structure of Interest Rates i n a Perfect International C a p i t a l Market with Fixed Exchange Rates 27 2.4 The Term Structure of Interest Rates i n a Perfect International Capital Market with Changing Exchange Rates 29 2.41 Risk-Neutral Investors 31 2.42 Risk-Averse Investors 39 - v i i i -- i x -2.4-21 The Measurement of Exchange Risk 2.422 An Extension of Roll's Model to Open Economies with Exchange Rate Risk 2.423 A Graphical Analysis 2.5 Information Costs, Transaction Costs, and Equilibrium Interest Rates i n an Internation-a l C a pital Market 2.6 Summary CORPORATE BORROWING IN INTERNATIONAL CAPITAL MAR-KETS 1 THE CANADIAN - UNITED STATES CASE 3.1 Introduction 3.2 The Canadian-United States Interest Rate Dif-f e r e n t i a l , Exchange Rate Expectations and Dif ferences i n the Aggregate Demand for Funds Be tween Canada and the United States 3.21 Canadian - United States Interest Rate D i f f e r e n t i a l s and Exchange Rate Be-haviour 3.22 Differences i n the Relative Demand for Funds Between Canada and the United States 3.23 Differences i n Time Preferences and Corporate Bond Issues i n the United States 3.24 The Bank of Canada's Influence on In-terest Rate D i f f e r e n t i a l s 3 . 3 Exchange Rate Changes and the Cost of Foreign Borrowing 3.31 Hedging Export Earnings through Inter-national Borrowing 3.32 Exchange Rate Risk and the Cost Advan-tage of Lower Interest Rates Abroad 3.4 Information and Transaction Costs i n an In-ternational Capital Market 3«4l Transaction Costs to Borrowers 3.42 Transaction Costs and Lending Beha-viour 3 . 5 Summary - X -Page 4. DATA COLLECTION AND SAMPLE DESCRIPTION 132 4.1 Data C o l l e c t i o n 132 4.2 Interviews l 4 l 5. A COMPARATIVE ANALYSIS OF CANADIAN CORPORATE DEBT ISSUES IN THE UNITED STATES AND CANADIAN BOND MARKETS 143 5.1 Introduction 143 5.2 S t a t i s t i c a l Methods 145 5«3 Interest Rate D i f f e r e n t i a l s , Differences i n Time Preferences and Canadian Corporate Borrowing i n the United States 149 5.4 Exchange Rate Risk and International Long-Term Financing by Canadian Corporations l6d 5.5 Transaction Costs Differences and Inter-national Debt Issues 179 5.6 Summary Analysis of Factors Influencing Canadian Corporate Borrowing Abroad 194 5.7 Non-Economic Influences on Corporate Bor-rowing Behaviour 201 6. SUMMARY AND CONCLUSIONS 205 BIBLIOGRAPHY 212 DATA SOURCES 218 Appendix 1» Equilibrium Interest Rates i n an Inter-national Capital Markets A P o r t f o l i o Approach 220 Appendix 2: Sample Selection, Data Sources and Ques-tionnaires 234 L I S T O F T A B L E S Table Page 3-1 P o r t f o l i o Transactions i n Canadian and United States Bonds between Canada and the United States, I960 - 1972 72 3-2 United States Do l l a r Exchange Rate i n Canada, 1927 - 1972 79 3-3 Exchange Rate Changes and Mean Annual Y i e l d Spreads between Canada and the United States, I960 - 1972 8 0 3 - 4 Total Funds Raised by Non-Financial Sectors as a Percentage of Gross National Product i n Canada and the United States, 1962 - 1972 85 3-5 Net Amount of Canadian Bonds Acquired by U.S. In-vestors as a Percentage of Total Funds Raised by Canadian Non-Financial Sectors, 1962 - 1972 86 3-6 Devaluation Percentages Required to Offset a F i f t y Basis Points Interest Advantage of Foreign Borrow-ing 109 3-7 Interest Rate D i f f e r e n t i a l s i n Favour of the For-eign C a p i t a l Market Required to Eliminate the Dis-advantage of Higher F l o t a t i o n Costs Abroad on a Twenty-year Bond 120 3 - 8 Average F l o t a t i o n Costs of Public Corporate Bond Issues as a Percentage of Gross Proceeds i n Canada and the United States and Cost D i f f e r e n t i a l s Be-tween the Two Countries 123 4 - 1 D e f i n i t i o n of Bond Samples 134 4-2 Description of Samples Covering Period 1968 to Mid-1973! Comparison of Number of Issues on which Information has been Obtained with Total Number of Issuss i n Each Sample According to Selected Issue Charact e r i s t i c s 136 4 - 3 Description of Samples Covering Period i 9 6 0 to 1967s Comparison of Number of Issues on which In-formation has been Obtained with Total Number of Issues i n Each Sample According to Selected Char-a c t e r i s t i c s 139 - x i -- x i i -Table Page 4- 4 Comparison of Corporate Interviews Sought with Interviews Granted by Industry C l a s s i f i c a t i o n . 142 5- 1 Mean Difference between Yields on Newly Issued Canadian Corporate Bonds and Yields on Long-Term Canadian Government Bonds at Date of Contract 151 5 - 2 a Discriminant Analysis Results, i 9 6 0 - 1967* Time Preference Measures 157 5-2b Discriminant Analysis Results, January 1968 - May 1970s Time Preference Measures 15 8 5 - 2 c Discriminant Analysis Results, June 1970 - May 1973s Time Preference Measures 159 5-3 Mean Terms to Maturity of Bonds Having Terms to Maturity of More than Twelve Years and which have been Sold by Corporations that have Borrowed i n the United States 164 5 - 4 a Discriminant Analysis Results, i 9 6 0 - 1967? Ex-change Risk Measures 173 5-4b Discriminant Analysis Results, January 1968 - May 1970s Exchange Risk Measures 174 5 - 5 a F l o t a t i o n Costs of P r i v a t e l y Placed Corporate Bonds 1968 - 1973 as a Percentage of Gross Proceeds 18 0 5 - 5 l 3 F l o t a t i o n Costs of P u b l i c l y Offered Corporate Bonds 1968 - 1973 as a Percentage of Gross Proceeds 181 5 - 6 a Discriminant Analysis Results, i 9 6 0 - 1967s Meas-ures Related to Transactions Costs 18 4 5-6b Discriminant Analysis Results, January 1968 - May 1970s Measures Related to Transactions Costs I85 5-6c Discriminant Analysis Results, June 1970 - May 1973s Measures Related to Transactions Costs 186 5-7 Mean Issue Size of Bonds Sold by Corporations which have Borrowed i n the United States 18 8 5 - 8 a Discriminant Analysis Results, i 9 6 0 - 1967s Com-bination of Several Measures 195 5-8b Discriminant Analysis Results, January 1968 - May 1970s Combination of Several Measures 197 5 - 8 c Discriminant Analysis Results, June 1970 - May 1973$ Combination of Several Measures 199 L I S T O F F I G U R E S Figure Page 25 2-1 Excess Supply Functions f o r Spot and Forward Loans 2-2a - International C a p i t a l Market Equilibrium f o r Forward Loanst Market i n Country D 58 2-2b International C a p i t a l Market Equilibrium f o r Forward Loanss Market i n Country F 60 2-3 A Domestic Investor's Excess Supply Functions of Spot and Forward Funds i n a Foreign Capital Market with Variable Transaction Costs Di f f e r e n -t i a l s 64 2- 4 International Capital Market Equilibrium for Forward Funds with Variable Transactions Costs D i f f e r e n t i a l s 66 3- 1 Devaluation Percentage Required to Offset the Interest Advantage of Foreign Borrowing as a Function of the Interest Rate D i f f e r e n t i a l 102 3-2 Devaluation Percentage Required to Offset a F i f t y Basis Points Interest Advantage of Foreign Borrowing as a Function of the Interest Rate Level 103 3-3 Devaluation Percentage Required to Offset a F i f t y Basis Points Interest Advantage of Foreign Borrowing as a Function of the Term to Maturity 106 3 - 4 Devaluation Percentage Required to Offset a F i f t y Basis Points Interest Advantage of Foreign Borrow-ing as a Function of the Year i n which the Exchange Rate Change Occurs 107 - x i i i -L I S T O F E X H I B I T S Exhibit Page A2-1 Data Questionnaire 24-5 A2-2 Interview Questionnaire, Canadian Corporations 251 A2-3 Summary of Answers Obtained from Canadian Corporations 267 A2-4 Interview Questionnaire, Underwriters 273 A2-5 Interview Questionnaire, Canadian L i f e Insurance Corporations 286 -xiv-CHAPTER 1 INTRODUCTION 1.1 SHORT REVIEW OF THE DEVELOPMENT OF THE THEORY OF INTER-NATIONAL CAPITAL FLOWS International c a p i t a l flows have always been an important component of economic interchange between d i f f e r e n t countries. For more than one hundred years, up to 1914, London served as a marketplace f o r international funds. A f t e r World War I, New York emerged as the world's f i n a n c i a l center. But i n the a f t e r -math of the Great Depression f i n a n c i a l transactions among r e s i -dents of d i f f e r e n t countries became t i g h t l y controlled and re-mained severely r e s t r i c t e d t i l l t h e late ' f i f t i e s , w h e n major i n d u s t r i a l i z e d nations allowed t h e i r currencies to become con-v e r t i b l e again. Though balance of payments and foreign exchange theories sometimes attempted to take the influence of c a p i t a l 2 movements into account, during those years problems of i n t e r -national trade and i t s influence on the domestic economy were of primary importance to scholars interested i n international 3 economics. "*"This does not necessarily apply to foreign investment i n United States equities or Canadian borrowing i n the United States. Cf. Charles P. Kindleberger, International Economics ( 4 t h ed.: Homewoodt Irwin, 1968), pp. 3 6 4 - 3 8 8 . 2 See, f o r example, the relevant papers i n Howard S. E l l i s and L. A. Metzler (eds.), Readings i n the Theory of Internation-a l Trade (Homewoodt Irwin, 1 9 5 0 ) . 3 In the l a s t sentence of his paper "Towards a General - 1 -During the early ' s i x t i e s , c a p i t a l flows again "became a major element of international economic transactions. New York experienced a r e v i v a l as the leading international c a p i t a l mar-ket. Mainly i n response to various U.S. c a p i t a l r e s t r a i n t pro-grammes and other s t r i c t controls l i k e SEC regulations, the Eurodollar and the Eurobond markets developed. Today there i s a strong tendency towards one i n t e r n a t i o n a l l y integrated money and c a p i t a l market, and the severe r e s t r i c t i o n s on internation-a l f i n a n c i a l transactions recently imposed by many countries w i l l temporarily hinder but probably not reverse t h i s development. The appearance of free flows of funds posed new problems f o r the conduct of economic p o l i c i e s . The question of r e l a t i v e ef-f i c a c y of f i s c a l and monetary p o l i c y i n an open economy with c a p i t a l mobility became a major target of t h e o r e t i c a l probing. Theory of the Balance of Payments" Harry G. Johnson notes that "when c a p i t a l account transactions are introduced into the analy-s i s , the choice between p o l i c y a l t e r n a t i v e s requires reference to growth considerations not r e a d i l y susceptible to economic analysis." See H. G. Johnson, International Trade and Economic  Growths: Studies i n Pure Theory (Cambridge, Mass.s Harvard Uni-v e r s i t y Press, 1961), pp. 153-168. Reprinted i n Richard E. Caves and Harry G. Johnson (eds.), Readings i n International Eco-nomics (Homewoodj Irwin, 1968), pp. 3 7 4 - 3 8 8 . Morris Mendelson, "The Eurobond and c a p i t a l Market Integra-t i o n , " Journal of Finance, XXVII (March, 1972), pp. 110-126. The implications of the American r e s t r a i n t programmes f o r Canada are discussed by Robert M. Dunn J r . i n Canada's Experience with Fixed  and F l e x i b l e Exchange Rates i n a North American Ca p i t a l Market (Washington and Montreal: Canadian-American Committee, 1971). pp. 28-43. -'For an up-to-date account .of such r e s t r i c t i o n s see the most recent issue of the International Monetary Fund's Annual  Report on Exchange Restrictions (Washington, D.C.). Caves and Johnson note i n Readings i n International Eco-nomics , p. X J "Dollar shortage, d o l l a r glut, the r e v i v a l of i n -ternational c a p i t a l markets, and the problem of international l i q u i d i t y have a l l l e d to s i g n i f i c a n t new t h e o r e t i c a l work aimed at i d e n t i f y i n g the ess e n t i a l elements of the p o l i c y problem i n -volved or t e s t i n g hypothetical solutions." Models i n t h i s t r a d i t i o n usually assume that international capi-t a l flows are determined s o l e l y "by i n t e r e s t rate d i f f e r e n t i a l s . The widely c i t e d a r t i c l e by Krueger on "The Impact of Alterna-t i v e Government P o l i c i e s Under Varying Exchange Systems" pro-vides an i n t e r e s t i n g analysis and c r i t i q u e of these theories.'' However, the assumption made i n these models that interest rates i n each country are determined independently of interna-t i o n a l c a p i t a l flows and that these flows only react to interest rate d i f f e r e n t i a l s (instead of postulating a simultaneous r e l a -tionship) i s not very s a t i s f a c t o r y except perhaps f o r a large country l i k e the United States. Also, these theories cannot ex-p l a i n the simultaneous flow of c a p i t a l into and out of the same country which we sometimes observe. Partly i n response to such c r i t i c i s m , a p o r t f o l i o model of international c a p i t a l movements has been developed. The perhaps Q best-known t h e o r e t i c a l contributions are those by Grubel and g Floyd . Many researchers have used p o r t f o l i o theory as a basis f o r deriving empirically testable models of international capi-fAnne 0 . Krueger, "The Impact of Alternative Government P o l i c i e s Under Varying Exchange Systems," Quarterly Journal of Economics, LXXIX (May, 1965), pp. 1 9 5 - 2 0 8 . See also her a r t i c l e on "Balance of Payments Theory," Journal of Economic Literature, VII (March, 1 9 6 9 ) , pp. 1-26. o Herbert G. Grubel, "Internationally D i v e r s i f i e d P o r t f o l i o s ! Welfare' Gains and Capital Flows," American Economic Review, LVIII (December, 1 9 6 8 ) , pp. 1301-14. o John E. Floyd, "International C a p i t a l Movements and Mone-tary Equilibrium," American Economic Review, LIX (September, 1 9 6 9 ) , pp. 4 7 2 - 9 2 . t a l flows.^" 0 By suggesting that investors w i l l tend to reduce the r i s k i n e s s of t h e i r p o r t f o l i o s by d i v e r s i f y i n g across nation-a l c a p i t a l markets these models have made a s i g n i f i c a n t c o n t r i -bution to the theory of international c a p i t a l movements. How-ever, the recent i n c l i n a t i o n of econometricians to use the port-f o l i o model whenever estimating the determinants of short-term"'""1" 12 and long-term c a p i t a l flows has come under considerable attack. Expecting that t h i s model can explain a l l international c a p i t a l flows quite probably overrates i t s general a p p l i c a b i l i t y . Fur-ther development of t h i s theory along the l i n e s undertaken by Solnik seems much more promising. He concentrates on interna-t i o n a l transactions i n equities and derives "an equilibrium mod-e l of the world c a p i t a l market" i n the C a p i t a l Asset P r i c i n g very useful summary and c r i t i q u e of such studies can be found i n Richard D. Haas, "A P o r t f o l i o Model of International Ca p i t a l Flows" (unpublished Ph.D. thesis, Department of Econo-mics, Duke University, 1971). See also William H. Branson and Raymond D. H i l l , J r . , "Capital Movements Among Major OECD Coun-t r i e s : Some Preliminary Results," Journal of Finance, XXVI (May, 1971), pp. 2 6 9 - 8 6 . 1 1 F o r a recent study of short-term c a p i t a l flows see Norman C. M i l l e r and Marina V. N. Whitman, "Alternative Theories and Tests of U.S. Short-Term Foreign Investment," Journal of Finance, XXVIII (December, 1973), pp. 1131-1150. 12 Robert M. Stern, when discussing Branson and H i l l ' s paper, c r i t i c i z e s t h e i r r e s u l t s and offers the following suggestions: "How then can the r e s u l t s be improved upon? . . . The most cru-c i a l problems are conceptual rather than f a c t u a l . What I would urge the authors to do therefore i s to seek to i d e n t i f y the ma-jor groups of transactors according to t h e i r d i f f e r e n t behavior-a l c h a r a c t e r i s t i c s and to formulate empirical s p e c i f i c a t i o n s that w i l l capture the variations i n these c h a r a c t e r i s t i c s . . . . My suggestion then i s to close the computer momentarily down and think more about what i t i s that we are t r y i n g to explain." See his "Discussion," Journal of Finance, XXVI (May, 1971), pp.308-309. Cf. also E r i c h S p i t a l l e r , "A Survey of Recent Quan-t i t a t i v e Studies of Long-Term Ca p i t a l Movements," IMF S t a f f  Papers, XVIII (March, 1971), pp. 189-220. - 5 -Model (CAPM) s p i r i t . 1 - ^ Theories of the determinants of dire c t foreign investment, on the other hand, have increasingly focused on factors other than differences i n rates of return on c a p i t a l 14 or those related to r i s k d i v e r s i f i c a t i o n . 1.2 PURPOSE AND SCOPE OF THESIS This study concentrates on an analysis of the international market f o r long-term debt c a p i t a l . The purpose of our research i s (1) to extend the theory of the term structure of interest rates to open economies with free c a p i t a l movements and changing exchange rates; (2) to allow f o r differences i n transaction and information costs between trading i n the domestic and trading i n foreign c a p i t a l markets; (3) to elucidate the re s u l t s obtained by deriving t h e i r implications f o r Canadian corporate borrowing i n the United States bond market; and (4J to test the theory by comparing the financing behaviour of Canadian corporations ap-proaching the United States bond market with t h e i r financing a c t i v i t i e s i n the Canadian market and with financing a c t i v i t i e s of those Canadian firms that issue bonds only i n Canada. Corporate borrowers are one of the major groups of trans-actors i n international c a p i t a l markets. By concentrating only on t h e i r a c t i v i t i e s and looking at t h e i r i n t e r n a t i o n a l .financing e f f o r t s from a managerial point of view we expect to gain a 1 3 Bruno H. Solnik, European C a p i t a l Markets (Lexington, Mass.j Lexington Books, 1973)» p. XI. 14 Giorgio Ragazzi, "Theories of the Determinants of Direct Foreign Investment," IMF Staff Papers, XX (July, 1973), pp. 471-498. - 6 -better understanding of those c a p i t a l flows that are mainly i n i -t i a t e d by firms which have ready access to bond markets abroad. In addition, by extending the theory of the term structure of interest rates to open economies we hope to shed new l i g h t on the determinants of international long-term c a p i t a l flows i n general and on the rela t i o n s h i p between the Canadian and the Uni-ted States c a p i t a l markets i n p a r t i c u l a r . Research on these questions seems espe c i a l l y relevant and timely i n view of the continuing dependence of the Canadian economy on free interna-t i o n a l trade and free international c a p i t a l movements and i n view of the unsettled international monetary scene. Several empirical studies have been undertaken to explain aggregate c a p i t a l flows between Canada and the United States. Usually they have concentrated on long-term c a p i t a l movements. 16 In 1962 H e l l e i n e r was the f i r s t researcher to use data dis-aggregated by type of borrower - federal government, p r o v i n c i a l governments, municipal governments and corporations - i n an attempt to analyze factors influencing the issuance of foreign-17 pay bonds. He found only a marginal response of the percent-age of t o t a l corporate issues floated i n the United States to 15 -'For a review of these studies see the l i t e r a t u r e c i t e d i n footnotes 10 and 12 above and Dunn, pp. 2 2 - 2 7 . 1 A Gerald K. Helleiner, "Connections Between United States* and Canadian Capital Markets, 1 9 5 2 - 1 9 6 0 , " Yale Economic Essays, II ( F a l l , 1962), pp. 350-400. 17 T r a d i t i o n a l l y the vast majority of Canadian bonds sold to foreign investors have been denominated i n U.S. do l l a r s and re-cently i n other foreign currencies. For a f u l l e r discussion of t h i s point, see Chapter 3 below. changes i n the y i e l d d i f f e r e n t i a l on federal government "bonds. Ripley, i n 1969, regressed the quarterly amount of corporate f l o t a t i o n s of foreign currency bonds, a flow variable, on Cana-dian and U.S. corporate bond y i e l d indexes, term structure v a r i -1 R ables and t o t a l corporate bond issues. This l a s t variable i s extremely s i g n i f i c a n t and explains most of the v a r i a t i o n i n the dependent vari a b l e . Unfortunately i t contains the dependent variable and thereby introduces a severe bias into the estimates. 19 One year l a t e r Freedman uses an almost i d e n t i c a l approach but 20 discards i t i n favour of a stock-adjustment model. Using t h i s l a t t e r approach h i s r e s u l t s f o r the " f i f t i e s when the Canadian d o l l a r floated are very encouraging. For the fixed exchange rate period t i l l 1966 he notes problems with detecting a s i g n i -f i c a n t influence of the interest rate d i f f e r e n t i a l and a lenth-21 ening i n the response lag to changes i n i t . The Bank of Canada research s t a f f estimated a si m i l a r model f o r the period from 1955 through I 9 6 8 . Besides quarterly dummies and binary variables f o r the f l e x i b l e exchange rate period and f o r the introduction -I o Eleanor D. Ripley, "United States Investment i n Canadian Securities 1958-1965" (unpublished Ph.D. thesis, Department of Economics, Harvard University, 1 9 6 9 ) . 19 Charles Freedman, "Long-Term C a p i t a l Flows Between the United States and Canada" (unpublished Ph.D. thesis, Massa-chusetts Ins t i t u t e of Technology, 1 9 7 0 ) . 20 For a concise discussion of stock versus flow models and a r e j e c t i o n of the l a t t e r because i t does not take into account adjustments of exis t i n g p o r t f o l i o s see John F. H e l l i w e l l et a l . , The Structure of RDX2 (Bank of Canada Research Studies No. 7; Ottawa: Bank of Canada, 1971), Part 1, pp. 196-99-21 Freedman, p. 156. As an index f o r the U.S. inter e s t rate he uses, the y i e l d on two public U.S.-pay bonds issued by Cana-dian corporations. He also t r i e s to incorporate the anticipated costs of expected exchange rate changes into the y i e l d d i f f e r -e n t i a l . - 8 -of the interest equalization tax i n the United States i t con-tains a weighted moving average over f i v e quarters of a Canadian 22 bank l i q u i d i t y variable and a measure of U.S. lender response. Experiments with interest rate d i f f e r e n t i a l s were not successful. Judging from these extensive e f f o r t s to detect those v a r i -ables that influence Canadian corporate borrowing i n the United States, i t has become increasingly d i f f i c u l t i n recent years to explain observed variations i n corporate f l o t a t i o n s of foreign-pay bonds. P a r t i c u l a r l y the obviously l i m i t e d response of corp-orate financing to changes i n the i n t e r e s t rate d i f f e r e n t i a l which has been measured by these researchers i n almost every conceivable way i s surprising. We think f i v e factors are prob-ably responsible f o r the li m i t e d success of these research ef-f o r t s : 1. O f f i c i a l s t a t i s t i c s record c a p i t a l movements only on the date the funds are a c t u a l l y transferred into Canada, As Canadian corporations have increasingly entered i n -to long-term forward contracts with U.S. investors, the time lag between the o f f e r i n g of a bond issue and the take-down of some or a l l of the funds i s up to four 2 3 years, p a r t i c u l a r l y on large issues. J Incomplete i n -formation available to us indicates, f o r example, that at least h a l f of the corporate foreign-pay bonds de-22 See H e l l i w e l l et a l . , Part 1, pp. 2 1 0 - 2 1 2 and Part 2, p. 1 2 0 . 2 3 Both Helleiner, p. 387, and Richard E. Caves and Grant L. Reuber, Ca p i t a l Transfers and Economic Policy: Canada, 1951- 1 9 6 2 (Cambridge, Mass.1 Harvard University Press, 1971 J , pp. 2 4 9 - 5 0 , noted t h i s problem already during the ' f i f t i e s but did not regard i t as serious enough to d i s t o r t t h e i r r e s u l t s . - 9 -l i v e r e d i n 1970 were offered six to twenty-eight months e a r l i e r . The contract f o r a sizeable part of the bond issues recorded f o r 1973 was signed i n 1969. Between the s t a r t of negotiations with U.S. lenders and the signing of a f i n a l contract sometimes several months elapse as well. Consequently i t i s almost impossible to detect by means of regression analysis whether capi-t a l flows observed during a p a r t i c u l a r quarter are a function of c a p i t a l market and other conditions pre-v a i l i n g during the same or the preceding quarter or those p r e v a i l i n g many quarters e a r l i e r . 2. The p o r t f o l i o model on which the more successful studies by Freedman and the Bank of Canada are based i s norma-t i v e i n nature. I f borrowers and lenders do not behave i n a manner prescribed by t h i s model than i t cannot f u l l y explain observed f i n a n c i a l transactions. For example, f i n a n c i a l non-call clauses i n bond indentures and long-term forward contracts make a "realignment of ex i s t i n g ["debt] p o r t f o l i o s i n response to a change i n 24 the international rate-of-return d i f f e r e n t i a l " v i r -t u a l l y impossible. 3. Measuring the interest rate d i f f e r e n t i a l as the d i f -ference between Canadian and United States bond y i e l d indices may not be appropriate. J Because of the 24 H e l l i w e l l et a l . , p. 196. 25 ^But note Freedman*s e f f o r t s to construct a y i e l d index fo r Canadian U.S.-pay bonds, see footnote 21 above. -10-" p o l i t i c a l r i s k " involved i n lending to foreigners, American investors can be expected to demand a higher y i e l d on Canadian U,S.-pay bonds than on otherwise com-parable domestic s e c u r i t i e s . A l i b e r has recently shown the bias that can be introduced into the computation of inte r e s t d i f f e r e n t i a l s i f such differences i n p o l i t i c a l 26 r i s k are not taken into account. 4. Capital flows do not only react to i n t e r e s t rate d i f f e r -entials? they also influence the money; supply and there-by i n t e r e s t rates. The neglect of such simultaneous re-lationships may p a r t l y explain the l i m i t e d significance of variables measuring the i n t e r e s t rate d i f f e r e n t i a l i n studies applying single equation ordinary l e a s t squares methods to c a p i t a l flow data. 5. Other factors l i k e information and transaction costs may have a considerable influence on corporate borrowing behaviour. Using macro-data regression analysis may not be able to detect the influence of such variables. Because of the problems noted above, p a r t i c u l a r l y the f i r s t point, and because our i n t e r e s t i s i n gaining a better understanding of micro-behaviour that leads to international c a p i t a l flows, t h i s study w i l l employ discriminant analysis and interviews as i t s main empirical techniques. Robert Z. A l i b e r , "The Interest Parity Theorem: A Rein-terpretation," Journal of P o l i t i c a l Economy, LXXXI (November/ December, 1973), pp. 1^51-59. -11-In Chapter 2, Ro l l ' s theory of equilibrium int e r e s t rates 2 i n an e f f i c i e n t bond market w i l l be extended to open economies. Bierwag and Grove have shown that a p o r t f o l i o model of the term 28 structure of inter e s t rates can be developed. However, because of i t s less r e s t r i c t i v e assumptions Roll's theory i s more gener-a l and better suited f o r an adaptation to international c a p i t a l 29 markets. We demonstrate that i n a world dominated by r i s k -averse investors and with changing exchange rates, international differences i n inter e s t rates do not only r e f l e c t exchange rate expectations but also r i s k premia necessary to reimburse i n -ternational lenders and borrowers f o r accepting exchange rate r i s k . The persistence of such interest rate d i f f e r e n t i a l s i s compatible with the assumption that the international market f o r long-term c a p i t a l i s perfect. Countries whose residents have r e l a t i v e l y high time preferences w i l l be net borrowers i n international c a p i t a l markets. I f information and transaction costs are higher when two traders deal with each other who are from d i f f e r e n t nations than when residents from the same country-transact business with each other, then in t e r n a t i o n a l interest d i f f e r e n t i a l s w i l l , i n addition, be a function of such d i f f e r -ences i n transaction costs. 27 'Richard R o l l , The Behavior of Interest Rates (New York! Basic Books, 1970), Chapter I I I . 28 G.O. Bierwag and M.A. Grove, "A Model of the Term Struct-ure of Interest Rates," Review of Economics and S t a t i s t i c s , IL (February, 1967), pp. 50-62. 29 In Appendix 1 we show how under rather r e s t r i c t i v e assump-tions a p o r t f o l i o model of the i n t e r n a t i o n a l term structure of interest rates can be developed. -12-In Chapter 3 the three conditions that are c r u c i a l to our conclusion that the persistence of i n t e r e s t rate d i f f e r e n t i a l s between national long-term c a p i t a l markets i s consistent with equilibrium i n the international c a p i t a l market are discussed i n more d e t a i l . These conditions are that ( l ) because of d i f -ferences i n time preferences between nations and r e s u l t i n g d i f -ferences i n the demand f o r funds inte r e s t rate l e v e l s tend to d i f f e r among countries, that (2) expectations of exchange rate changes and exchange r i s k cause inte r e s t rate d i f f e r e n t i a l s to p e r s i s t , even i f the international c a p i t a l market i s perfect, and that (3) differences i n transaction and information costs between operating i n the domestic and a foreign market w i l l fur-ther reduce the i n t e r e s t - e q u i l i b r a t i n g influence of internation-a l long-term c a p i t a l flows. Data taken from Canadian-United States experience w i l l be presented as preliminary evidence i n support of these assumptions. Their implications f o r corporate borrowing i n foreign c a p i t a l markets, i n p a r t i c u l a r f o r Canadian corporate debt issues i n the United States, w i l l be discussed. Hypotheses w i l l be derived that allow us to test the theory developed. Guided by our t h e o r e t i c a l conclusions c e r t a i n data on Cana-dian-pay and U.S.pay bonds and data on the issuing Canadian corporations have been co l l e c t e d . Chapter 4 discusses the data c o l l e c t i o n process and provides comparative s t a t i s t i c s about the bond issues which have been included i n the empirical analy-s i s . Information on the Canadian corporate bond market and the f i n a n c i a l behaviour of Canadian corporations was also obtained -13-through personal interviews with f i n a n c i a l managers, Canadian underwriters and investment o f f i c e r s of Canadian l i f e insurance companies. Chapter 5 presents s t a t i s t i c a l tests of the hyptheses de-veloped i n Chapter 3 and summarizes the information obtained through interviews. Discriminant analysis i s used to i s o l a t e those factors that lead to foreign currency borrowing f o r domes-t i c purposes.-^ 0 Supporting evidence f o r most of our hypotheses has been found. In addition, the interviews revealed that f a c t -ors not incorporated into our theory l i k e moral suasion by the Bank of Canada and the Federal Government and growing national-ism i n Canada have exerted a s u r p r i s i n g l y strong influence on Canadian corporate borrowing behaviour. F i n a l l y , Chapter 6 summarizes our res u l t s and points out areas i n need of further research. Ripley has used t h i s technique to d i f f e r e n t i a t e between Canadian-pay and U.S.-pay bonds issued by Canadian provinces and m u n i c i p a l i t i e s . His three discriminating variables, size of issue, maturity, and a s e r i a l bond dummy were selected be-cause t h i s information was readily available rather than be-cause of a t h e o r e t i c a l model postulating the relevance of these measures. See Duncan M. Ripley, "Some Determinants of Canadian Municipal and P r o v i n c i a l Bond Flotations i n the United States," Review of Economics and S t a t i s t i c s , LII (November, 19?0), pp. 417-26. CHAPTER 2 EXCHANGE RATE RISK, TRANSACTION COSTS AND EQUILIBRIUM INTEREST RATES IN AN INTERNATIONAL LONG-TERM CAPITAL MARKET 2.1 INTRODUCTION T r a d i t i o n a l l y , models employed i n i n t e r n a t i o n a l economic theory postulate that international c a p i t a l flows are mainly or s o l e l y a function of i n t e r e s t rate d i f f e r e n t i a l s observed be-tween a p a r t i c u l a r country and the rest of the world. Often theories d i f f e r , however, with regard to assumptions made about the interest e l a s t i c i t y of such c a p i t a l flows. Mundell, f o r example, when discussing the e f f e c t s of f i s c a l and monetary p o l i c i e s under fi x e d and f l e x i b l e exchange rates, assumes perfect c a p i t a l m o b i l i t y . 1 This i n f i n i t e l y interest elas-t i c mobility of international c a p i t a l implies that i n t e r e s t rate 2 d i f f e r e n t i a l s cannot exist; only one world inte r e s t rate p r e v a i l s . Other writers have attempted to draw t h e i r models closer to the r e a l world by asserting that i n t e r n a t i o n a l funds are only "^"Robert A. Mundell, International Economics (New York: MacMillan, 1968), pp. 250-251. But see also p. 234 where i t i s only assumed that " c a p i t a l flows are responsive to interest-rate d i f f e r e n t i a l s . " 2 Of course, t h i s holds only i f s e c u r i t i e s are perfect sub-s t i t u t e s f o r each other. Not only differences i n default r i s k but also differences i n p o l i t i c a l r i s k between bonds o r i g i n a t i n g i n d i f f e r e n t countries w i l l lead to d i f f e r e n t prices even i f the promised income stream i s i d e n t i c a l . See A l i b e r , "The Interest Parity Theorem: A Reinterpretation," on t h i s . Here we w i l l a l -ways assume that bonds are free of default r i s k . -14--15-imperfectly mobile so that int e r e s t rate d i f f e r e n t i a l s are nar-3 4 rowed but not eliminated among open economies. However, the precise conditions are never f u l l y spelled out under which only a p a r t i a l but not t o t a l equalization of interest rates across national f i n a n c i a l markets w i l l come about, even i f these mark-ets are perfect, and the factors determining a country's r e l a -t i v e l e v e l of interest rates v i s - a - v i s the rest of the v/orld. Clearly, interest rates observed at a given point i n time on comparable c a p i t a l market instruments l i k e government bonds do vary i n t e r n a t i o n a l l y , and though a tendency of interest rates i n national markets to follow i n t e r n a t i o n a l y i e l d trends has been noted such comovements are rather imperfect.^ It would appear, therefore, that the assumption of perfect c a p i t a l mobili-ty with only one world interest rate p r e v a i l i n g does not ap-proximate r e a l world phenomena to any reasonable extent. •^Cf., e.g. Lloyd A. Metzler, "The Process of International Adjustment under Conditions of F u l l Employments A Keynesian View", i n Caves and Johnson, pp. 4 7 4 - 7 6 . 4 A concise discussion of the assumption of perfect versus imperfect c a p i t a l mobility and extensive references to the l i t -erature can be found i n Akira Takayama, International Trade (New Yorks Holt, Rinehart and Winston, 1972), pp. 330-31. ^For evidence, see Sidney Homer and Richard I. Johannesen, The Price of Money, An A n a l y t i c a l Study of U.S. and Foreign In-terest Rates (New Brunswick, N.J.s Rutgers University Press, 1969), or Douglas R, Bohi, "The International Interdependence of Interest Rates", Kyklos, XXV(November, 1972), pp. 597-600. Extensive data on the relationships between Canadian and U.S. intere s t rates can be found i n Helleiner, pp. 362-76 or Ripley, "United States Investment i n Canadian Securities I 9 5 8 - I 9 6 5 " , pp. 41-44. ^For c e r t a i n s c i e n t i f i c purposes the realism of a p a r t i c u l a r assumption may be only of minor importance. Our goal here i s to develop a model that can explain observed international long-term interest rate d i f f e r e n t i a l s and the r e s u l t i n g flows of debt capi-t a l . For a recent review of the ongoing epistemological discus-sion amongst economists concerning "Theory and Realism" see Stan-ley Wong, "The 'F-Twist' and the Methodology of Paul Samuelson", American Economic Review, LXIII (June, 1973)1 PP« 312-325* -16-In t h i s chapter, our e f f o r t s w i l l be directed towards the development of a model of the i n t e r n a t i o n a l market f o r long-term debt c a p i t a l that takes into account lender and borrower beha-viour. The term structure of i n t e r e s t rates i n countries whose residents engage i n international f i n a n c i a l transactions w i l l be shown to be a function of domestic and foreign traders* ex-pectations of future domestic and foreign spot i n t e r e s t rates and of future rates of change i n exchange rates. Furthermore, t h e i r degree of r i s k aversion, differences in.time preferences, and differences i n costs incurred when transacting business with a foreign investor rather than with an investor of i d e n t i c a l nation-a l i t y w i l l be demonstrated to a f f e c t i n t e r e s t rates and c a p i t a l flows. 8 We w i l l s t a r t our discussion by b r i e f l y o u t l i n i n g R o l l ' s theory of the term structure of i n t e r e s t rates i n a closed econo-my. Then a perfect international c a p i t a l market with f i x e d and unchangeable exchange rates w i l l be considered. Given such a market, i n t e r e s t rates w i l l indeed be equalized across national sub-markets. When allowing exchange rates to change, a perfect international c a p i t a l market w i l l no longer lead to i n t e r e s t equalization. Rather, assuming investors to be r i s k - n e u t r a l , ^Note that our theory w i l l not address the question of what determines the national or international l e v e l of i n t e r e s t rates but rather how national term structures of i n t e r e s t rates are i n t e r r e l a t e d through c a p i t a l movements from one currency area to another. Our model i s a s t a t i c p a r t i a l equilibrium model that does not consider the influence of r e a l and monetary forces on the l e v e l of i n t e r e s t rates. 8 Throughout, the terms trader and investor are used i n t e r -changeably and r e f e r to any p a r t i c i p a n t i n the c a p i t a l market whether he i s a borrower or a lender. International trading i n commodities or services w i l l not be considered i n t h i s study. -17-i n t e r e s t d i f f e r e n t i a l s r e f l e c t expected exchange rate changes. I f traders are risk-averse this conclusion has to be q u a l i f i e d . Risk-averse investors w i l l move t h e i r funds abroad only i f they can expect a gain larger than i s necessary to cover expected exchange rate losses. They w i l l not accept the higher r i s k i n -herent i n for e i g n investment due to exchange rate r i s k unless they expect to earn a r i s k premium abroad. By extending R o l l ' s approach to open economies i t w i l l be shown how from excess sup-ply functions f o r funds i n domestic i.and foreign forward markets an approximate s o l u t i o n f o r the equilibrium i n t e r n a t i o n a l term structure of i n t e r e s t rates can be derived. In the f i n a l section our theory w i l l be extended by introducing imperfections into the model. I t w i l l be assumed that information and transaction costs are higher when two traders deal with each other who are from d i f f e r e n t nations than when residents from the same country trans-act business with each other. This leads to the conclusion that observed i n t e r e s t rate d i f f e r e n t i a l s do not only r e f l e c t exchange rate expectations and exchange r i s k premia but also differences i n transaction costs between trading i n a domestic and a foreign c a p i t a l market. 2.2 THE TERM STRUCTURE OF INTEREST RATES IN A CLOSED ECONOMY The theory of the term structure of i n t e r e s t rates i s con-cerned with analyzing those factors that determine the y i e l d to maturity on default-free bonds of d i f f e r e n t maturities. Y i e l d to maturity i s defined as that rate of i n t e r e s t at which the f u -ture income stream of a bond has to be discounted to equal the present market p r i c e . However, a n a l y t i c a l l y i t i s often more -18-convenient to concentrate on a study of forward or "futures" inte r e s t rates implied by the term structure of i n t e r e s t rates observed i n the market. As Hicks has shown, " i f we decide upon some minimum period of time, loans f o r le s s than which time we s h a l l be prepared to disregard, every loan of every duration can be reduced to a standard pattern - a loan f o r the mi imum period, combined with a given number of renewals f o r subsequent periods of the same length, contracted forward."^ Let R n be the y i e l d to maturity on an n-period bond expressed as (per cent per period)/lOO. Then, assuming in t e r e s t to be compounded at the end of each period, (1 + R n ) n = (1 + r j ) ( l + r 2 ) ... (1 + r n ) where the ^ • s are forward rates f o r a one-period loan out-standing during the nth period. Market forward rates f o r any future period n can be derived from observed y i e l d s as followsi 1 + r n = (1 + R n ) n / ( 1 + R n - l ) n " \ I t follows that an analysis of variables determining forward rates i s equivalent to a study of the term structure of inter e s t rates. The mathematical formulation can be s i m p l i f i e d by assuming intere s t to be compounded continually. I f bonds carry no coup-ons and are issued at a discount, then the current p r i c e of an n-period bond, P , equals P n = (face value)exp(-nR n) J . R. Hicks, Value and C a p i t a l , (2nd ed.; Londoni Oxford University Press, 19^6), pp. 144-45. -19-and nR n = + r 2 + ... + r n ) . (2-1) The nth forward rate i s given by r n =nR n - ( n - l ) R n - 1 . Identical r e s u l t s can be derived by assuming that a bond i s quoted (or issued) at par and pays a continuous coupon which can be reinvested at the contracted rate. Two major theories of the determinants of the term struc-ture have been developed. According to the pure expectations hypothesis forward rates represent unbiased estimates of future one-period spot fates. Adherents of the l i q u i d i t y preference t h e o r y 1 0 claim that the r i s k aversion of lenders makes them value the s t a b i l i t y of p r i n c i p a l more than the s t a b i l i t y of income and that therefore forward rates overestimate future spot rates. But t h i s l a t t e r conclusion does not necessarily hold i f univer-s a l r i s k aversion of both lenders and borrowers i s assumed. Telser has pointed out that such a premise imples that a l l trad-ers i n the market " w i l l attempt to hedge against the r i s k of changes i n int e r e s t rates by f i n a n c i a l transactions intended to approximately match the timing of payments and re c e i p t s . However, i t does not follow that such hedging r e s u l t s i n a bias that can 11 be deduced a p r i o r i . " For example, i f investors plan to lend during some future periods, they may decide to already commit *°As R o l l , Chapter IV, notes, the l i q u i d i t y preference theory can be regarded as a special variant of a more general market segmentation hypothesis. 11 See pp. 546-47 of the paper by L. G. Telser, "A Criti q u e of Some Recent Empirical Research on the Explanation of the Term Structure of Interest Rates", Journal of P o l i t i c a l Economy, LXXV (Supplement} August, 196?), pp. 546-561. -20-t h e i r funds now at p r e v a i l i n g forward rates even i f they expect to earn more money by investing l a t e r at the then r u l i n g spot rates. Whereas future spot rates are uncertain, forward rates are known with certainty, and the reduction i n r i s k may compen-sate lenders f o r the lower return. S i m i l a r l y , i f a trader plans to borrow during some future period, he may be prepared to bor-row already now i n the forward market rather than wait to ob-12 t a i n those funds l a t e r at a lower expected spot rate. For-ward rates may therefore be biased upwards or downwards depend-ing on the demand and supply f o r funds i n each forward market. 2.21 R0LL*S APPROACH R o l l has presented a comprehensive model of the determin-ants of the term structure of i n t e r e s t rates which makes no a p r i o r i assumption about possible biases i n forward r a t e s . 1 3 In t h i s section a short exposition of his theory w i l l be provided because l a t e r h i s model w i l l be used to extend the term st r u c t -ure theory to open economies. Most of our l a t e r t h e o r e t i c a l discussions are based on the premise that c a p i t a l markets are perfect. Therefore l e t us f i r s t state the assumptions made with regard to such markets and to the investors operating i n themi 1. Relevant information i s a free good, that i s , i t i s costless and available at the same time to every trader i n the market. Note that, e.g., issuing an n-period bond and buying an n-1 period bond of equal par value i s equivalent to borrowing the same amount forward f o r the nth period. For a mathemati-c a l proof that demand functions f o r forward loans are equival-ent to demand functions f o r bonds see R o l l , p. 26. 1 3 R o l l , Chapter I I I . -21-2. Transaction costs and taxes are zero,, 3. Buyers and s e l l e r s of s e c u r i t i e s a) take the prices of s e c u r i t i e s as givent b) act r a t i o n a l l y (that i s , prefer more wealth to l e s s and use a l l relevant information available) and be-l i e v e that other traders do likewise} 1**' c) possess subjective p r o b a b i l i t y d i s t r i b u t i o n s on f u -ture one-period spot i n t e r e s t rates expected to pre-v a i l at the s t a r t of each period up to and including period N} d) have an investment horizon of N periods or l e s s , that i s , what happens beyond period N i s of no concern to anybody. These assumptions are s i m i l a r to those usually made when defin-ing a perfect c a p i t a l market i n a closed economy.^ As future i n t e r e s t rates are not known with certainty, traders have to form subjective b e l i e f s about these rates i n order to a r r i v e at r a t i o n a l investment decisions. Such b e l i e f s are conveniently expressed by subjective p r o b a b i l i t y d i s t r i b u -t i o n s . 1 ^ The assumption that investors' time horizons do not extend beyond N periods i s an expedient used i n term structure of i n t e r e s t rates theory. However, t h i s does not impose any r e a l l i m i t a t i o n s to the theory because the length of a period and N are not defined i n s p e c i f i c terms. Market rates are a function of i n d i v i d u a l investor behavi-our. Therefore R o l l s t a r t s the development of his model by A , H'This concept of "symmetric market r a t i o n a l i t y " i s due to Merton H. M i l l e r and Franco Modigliani, "Dividend Policy, Growth and the Valuation of Shares", Journal of Business, XXXIV (Octo-ber, 1961), pp. 427-28. 1^Cf., e.g., Eugene F. Fama and Merton H. M i l l e r , The Theory  of Finance (New Yorki Holt, Rinehart and Winston, 1972), p. 21. See also R o l l , pp. 9, 19. 1^See Kenneth J . Arrow, Aspects of the Theory of Risk-Bear-ing (Helsinki 1 Yrjo Jahnssonin s a a t i f l , 1965), Lecture 1. -22-assuming that each trader i has N excess supply functions f o r 17 forward loans f o r every one of the N future periods. To simplify the exposition, l e t us use a two-period framework, that i s , N = 2. Then these supply functions are given by = i f j ( r i " i ri» r2 " i r2> f o r ^ = !• 2» (2-2) Here .q. i s the i t h investor's supply of one-period forward 18 loans to be outstanding during period j . I f .q. i s p o s i t i v e , the investor lends i n the forward market f o r period j , and he borrows i f i t i s negative. r n i s the market forward rate on loans during period n, and the personal forward rates of trad-er i , ^ r ^ , are chosen such that I f market forward rates equal the trader's respective personal rates he w i l l not enter forward markets.^ Each trader's personal forward rates are assumed to depend on four quantities: 17 R o l l ' s discussion i s i n terms of excess demand functions. However, as p o s i t i v e excess "demand" implies lending or a supply of funds during a c e r t a i n period ( c f . R o l l , p. 20), we prefer to use the term excess supply functions. 18 The a d d i t i o n a l subscript t customarily used i n term s t r u c t -ure l i t e r a t u r e as a reference to the date on which the i n t e r e s t rate i s f i x e d i s being dropped here whenever t would r e f e r to the s t a r t of the "current" period. * ^ R o l l adds an a d d i t i o n a l subscript j to personal forward rates. As both j and n can vary from 1 to N t h i s seems to imply that investors may have not one but N d i f f e r e n t forward rates f o r each d i s t i n c t i v e future period. As was pointed out to me by Professors M. Brennan and A. Kraus, i n t h i s case R o l l ' s sys-tem does not have a unique solution. Here we assume that, at a given points i n time, every investor has only one personal forward rate f o r each future period. 1. His current expectation about the one-period spot rate at the s t a r t of period n. 2. The degree of confidence he has i n his expectation. 3. His degree of r i s k aversion. 4. The time preference he has with respect to period n, that i s , h i s current assessment of the amount of spot lending or borrowing he would undertake at the sta r t of period n i f his expectation of the future spot rate R* „ were r e a l i z e d , l ,n R o l l assumes that these four quantities can be expressed i n an additive form by the t y p i c a l equation •R = E.(R« ) + ,L (2-3) i n i v l , n ' i n v J ' where n = 1, 2 i n our case. Here E^(R^ n ) i s the one-period spot rate expected by the i t h trader to rule at the s t a r t of period n and ^ L n i s his r i s k and time premium with respect to period n. In order to elucidate the c h a r a c t e r i s t i c s of the i n d i v i d u a l excess supply functions, l e t us assume that r^ = ^ r ^ and consid-er the case where n = 2. The r i s k and time premium, i L 2 = i r 2 " E i ( ^ l , 2 ) w i l l be p o s i t i v e f o r a risk-averse investor who, i f the expected future spot rate were to be r e a l i z e d , that i s , i f 2 = E^(R^ 2 ) would be a spot borrower during period 2. This follows because as a ri s k - a v e r t e r he i s w i l l i n g to pay a s l i g h t l y higher rate now to be c e r t a i n about the rate at which funds can be obtained during period 2 rather than to run the r i s k of paying perhaps even more l a t e r i n the spot market. Conversely, the premium w i l l be negative i f the trader, given that h i s expectation were re a l i z e d , would be a spot lender during period 2. -24-This idea that the forward supply of funds i s derived from the supply function f o r corresponding future spot loans, which 20 i s basic to Roll's theory, may be i l l u s t r a t e d by Figure 2 - 1 . Two supply curves, one (jj3 g ^S g) f o r spot loans and one (^S^ ^S^) f o r forward loans are drawn. They cross at E^CR^ 2 ) , ^ e one-period spot rate expected by the i t h investor to p r e v a i l at the s t a r t of period 2 , because (1 ) at a higher forward rate, say r x , he w i l l supply forward funds of at l e a s t q x j since he would make spot loans of t h i s amount during period 2 i f the spot rate, R^  2» were indeed equal to r„. I t follows that his supply curve f o r forward funds cannot be above his supply curve f o r future spot funds whenever r 2 > E^(R^ 2 ) . Indeed, the supply curve f o r f o r -ward funds w i l l be below the supply curve f o r spot funds at a rate l i k e r x because the investor w i l l also loan q x 2 - q x ^ f o r -ward f o r speculative reasons. This speculative forward loan w i l l be l i q u i d a t e d i n the spot market at the s t a r t of period 2 , and the expected rate of return on t h i s speculative engagement i s r x - E i ( R 1 ^ 2 ) . (2 ) At a forward rate lower than E i ( R 1 2 ) , say r y , an investor w i l l lend only q y 2 since he expects a higher return on future spot loans f o r the same period, that i s , at t h i s rate his supply curve f o r forward funds w i l l be above his spot supply curve. At a forward rate below ^ r 2 he w i l l even begin borrowing forward. Whenever the forward rate equals the trader's expected future spot rate he w i l l lend or borrow f o r -ward just what he plans to supply at the given rate. By d e f i n i -t i o n , he w i l l not enter the forward market when the market rate See R o l l ' s Figure 3-l» p. 2 2 . R o l l also assumes that the trader " i s c e r t a i n about a l l future income earned outside the c a p i t a l markets." -25-FIGURE 2-1 EXCESS SUPPLY FUNCTIONS FOR SPOT AND FORWARD LOANS -26-equals his personal forward rate. The angle 6 between the spot and forward supply functions w i l l be influenced by two factors, (1) the investor's confidence about his forecast of the future spot rate and (2) his degree of r i s k aversion. The more confident he i s about his forecast the wider w i l l the angle 9 be because the more he i s prepared to speculate, and the more risk-averse he i s the less he w i l l spec-ulate and the smaller the angle w i l l be. In addition, as R o l l has pointed out, two factors exogenous to t h i s theory w i l l also influence ^ L n and ^ r n i n d i r e c t l y by 21 t h e i r effect on the spot supply function: (1) the trader's wealth and (2) his savings-consumption plan f o r period n. The greater the investor's wealth, c e t e r i s paribus, the f l a t t e r w i l l be his demand function. That i s , the higher w i l l be his demand for or supply of loans at any given rate compared to a trader with fewer resources. And his time-preferences w i l l determine whether he borrows or lends at a given rate, that i s , they w i l l influence the l e v e l of the demand function f o r spot funds. For example, a trader who values consumption during a certa i n period very highly may intend to borrow spot at a given interest rate l e v e l whereas another investor i d e n t i c a l i n every respect ex-cept that he values consumption during the same period less may plan to be a spot lender at the same intere s t rate. In order to derive an approximate solution f o r the e q u i l i -brium term structure of interest rates, R o l l expands i n d i v i d u a l excess supply functions i n a Maclaurin series and adds them up R o l l , p. 24. f o r each period. Market equilibrium i s achieved i f the set of forward rates i s such that t o t a l excess supply i s zero. Drop-ping higher order terms, a solution f o r equilibrium forward rates can be obtained. Details w i l l not be reported here as our derivation of a s o l u t i o n f o r the equilibrium i n t e r n a t i o n a l term structure w i l l proceed along s i m i l a r l i n e s . 2 . 3 THE TERM STRUCTURE OP INTEREST RATES IN A PERFECT INTER-NATIONAL CAPITAL MARKET WITH FIXED EXCHANGE RATES 22 As mentioned e a r l i e r , c e r t a i n models used i n international economic theory postulate that i n t e r e s t rates are equalized i n -t e r n a t i o n a l l y and cannot change because c a p i t a l moves " f r i c t i o n -l e s s " from one country to another. Besides the assumptions made at the beginning of the preceding section with regard to perfect c a p i t a l markets, the following assumption i s necessary to a r r i v e at such a conclusion t 4 . Exchange rates between a l l currencies have been fixed and remain unchanged, and governments do not i n t e r f e r e i n the free market mechanism by introducing exchange controls or discriminatory taxes against foreigners, or by any other means. In other words, currencies are f r e e l y convertible into each other at f i x e d rates. This implies that the world i s one single cur-rency area with, i n e f f e c t , only a single money. The currency i n which a security i s denominated i s i r r e l e v a n t f o r i t s value and only of i n t e r e s t f o r accounting purposes. Given these i d e a l i z e d conditions, i n t e r n a t i o n a l i n t e r e s t arbitrage w i l l assure that the y i e l d on assets i d e n t i c a l with respect to r i s k and other c h a r a c t e r i s t i c s i s equalized i n t e r -See p . l 4 above -28-n a t i o n a l l y . Should ever a y i e l d d i f f e r e n t i a l occur, a r b i t r a -geurs could make a sure p r o f i t by s e l l i n g bonds short i n one country and buying i d e n t i c a l bonds i n another country. In other words, borrowing or lending i n a foreign c a p i t a l market cannot, i n equilibrium, benefit any market p a r t i c i p a n t and he w i l l be i n d i f f e r e n t between trading i n the domestic or a f o r -eign c a p i t a l market. From a purely f i n a n c i a l point of view, the world can be regarded as a closed economy, and convention-a l term structure of i n t e r e s t rate theory may be used to explain observed relationships between spot and forward i n t e r e s t rates. It would appear that t h i s and s i m i l a r models of a perfect international c a p i t a l market do not approximate real-world phenomena to any reasonable extent. P a r t i c u l a r l y assumption 4., that exchange rates are f i x e d and unchangeable, i s open to c r i t i -cism. Under the c l a s s i c a l gold standard exchange rates were indeed extremely stable and fluctuated only within a very narrow band determined by the so-called gold export and gold import points. However, the A r t i c l e s of Agreement of the International Monetary Fund as l a i d down at the Bretton Woods conference i n 1944 e x p l i c i t l y provide f o r changes i n exchange rates. And ex-perience with the international monetary system that has de-veloped since 1945 has shown that such changes may be quite d r a s t i c . The purpose of the next section i s to show that, i f we allow exchange rates to change, in t e r n a t i o n a l i n t e r e s t rate d i f f e r e n t i a l s and differences i n the term structures of i n t e r e s t rates between d i f f e r e n t countries are compatible with the assump-t i o n that there exists only one perfect i n t e r n a t i o n a l c a p i t a l market. -29-2.4 THE TERM STRUCTURE OF INTEREST RATES IN A PERFECT INTER-NATIONAL CAPITAL MARKET WITH CHANGING EXCHANGE RATES In t h i s section, assumption 4. which stated that exchange rates are f i x e d and do not change w i l l be dropped. Instead, i t w i l l be assumed henceforth that 4a. Exchange rates between a l l currencies may change at any point i n time. The implications of t h i s assumption f o r i n t e r n a t i o n a l c a p i t a l flows may be i l l u s t r a t e d by the following example. Let us as-sume that a Canadian investor has a choice between investing i n a one-year Canadian bond y i e l d i n g 7 per cent and a one-year American bond y i e l d i n g 9 per cent. Being a r a t i o n a l investor, he chooses to buy a U.S. bond. However, when l i q u i d a t i n g his investment one year l a t e r and t r a n s f e r i n g h i s wealth back into Canada f o r consumption purposes, he discovers that the United States d o l l a r has decreased i n value by 5 per cent. I f he payed o r i g i n a l l y one Canadian d o l l a r f o r one U.S. d o l l a r , he now pays U.S. $1 . 05 f o r Can. $1.00, or U.S. $1.00 = Can. $1.00/ 1 . 0 5 = Can. $ . 9 5 2 4 . In other words, f o r every one hundred Canadian d o l l a r s invested i n the United States our investor receives U.S. $109.00 = Can. $(109.00 x .9524) = Can. $103.81. His return on t h i s foreign investment, when expressed i n Cana-dian d o l l a r s , i s a mere 3*81 per cent, not the 9 per cent as he had expected. When choosing between domestic and foreign investment op-p o r t u n i t i e s , an investor must therefore form subjective b e l i e f s about future exchange rate changes i n order to be able to act r a t i o n a l l y . Such b e l i e f s are assumed to be expressable by sub--30-jective p r o b a b i l i t y d i s t r i b u t i o n s , and assumption 4a. makes i t necessary to add the following assumption to our character-i s t i c s of a perfect international c a p i t a l marketi 3. e) Buyers and s e l l e r s of s e c u r i t i e s possess subjective p r o b a b i l i t y d i s t r i b u t i o n s on rates of change of ex-change rates during every period from the present period up to and including period N. I t should be noted that assumption 4a. does not specify under what kind of exchange rate regime a country or the world i s operating. I t i s immaterial f o r our considerations whether the exchange rate i s pegged but adjustable as i t was s t i p u l a t e d i n the IMF A r t i c l e s of Agreement, whether the exchange rate i s allowed to be determined s o l e l y by market forces as under "a clean f l o a t " , or whether a crawling peg system p r e v a i l s . Im-portant i s only that the rate of return r e a l i z a b l e on foreign investments i s p o t e n t i a l l y influenced by exchange rate changes. Exchange rate r i s k seems also not to be a function of the p a r t i -c ular exchange rate systemj at l e a s t there i s no s c i e n t i f i c basis f o r such a claim. As Katz points out, "within the pro-fession as a whole, there i s no evidence of a consensus that aggregate exchange-risk would be less under either f i x e d or under f l e x i b l e rates." J In a less than perfect world, a d d i t i o n a l factors not en-countered when investing i n domestic s e c u r i t i e s w i l l influence the return r e a l i z a b l e on foreign investments. Discriminatory taxes on i n t e r e s t accruing to foreigners, exchange controls or moratoria w i l l a l l tend to decrease the y i e l d on foreign i n -23 •'Samuel I. Katz, "Exchange-Risk Under Fixed and F l e x i b l e Exchange Rates", The B u l l e t i n of the I n s t i t u t e of Finance, Graduate School of Business Administration, New York Univer-s i t y , Nos. 83-84 (June, 1 9 7 2 ) , p. 1 1 . -31-vestments. On the other hand, i t may be easier to evade domes-t i c l e v i e s when wealth i s held abroad. A l l such additional f a c t -ors influencing the return and r i s k of foreign investment w i l l be excluded from our analysis. To simplify the following analysis, i t w i l l be assumed that the world consists of two countries only which w i l l be c a l l e d 24 domestic country D and foreign country F. Bonds are assumed to be free of default r i s k and to have f i x e d maturities. Do-mestic s e c u r i t i e s are usually denominated i n l o c a l money, the numeraire "commodity" i n the domestic country. In an i n t e r -national c a p i t a l market s e c u r i t i e s w i l l also be denominated i n other numeraires which are generally foreign currencies. In our case we w i l l be concerned with only two currencies and one exchange rate determining the value rel a t i o n s h i p between country D's and country F's numeraire. 2.41 RISK-NEUTRAL INVESTORS If investors have u t i l i t y functions which are l i n e a r i n wealth, they are i n d i f f e r e n t between holding domestic bonds and bonds denominated i n the foreign currency as long as the expect-ed e f f e c t i v e rate of return, that i s , the rate of return a f t e r taking expected exchange rate changes into account, i s the same on both investment opportunities. In a perfect international c a p i t a l market inter e s t arbitrage w i l l ensure that indeed the expected e f f e c t i v e y i e l d on both types of bonds w i l l be i d e n t i -24 The assumption of two countries only i s solely made f o r a n a l y t i c a l convenience. The l a t e r analysis can e a s i l y be ex-tended to n countries but t h i s would not lead to conclusions fundamentally d i f f e r e n t from those presented here. -32-2 *5 c a l . J This e f f e c t i v e rate of return on s e c u r i t i e s denominated i n d i f f e r e n t currencies may be measured i n a "world numeraire", e.g. the U.S. d o l l a r , any other f r e e l y convertible currency or a commodity l i k e gold i f t h i s commodity i s traded i n a perfect international goods market. Independently of the p a r t i c u l a r common denominator chosen, the expected y i e l d on i d e n t i c a l as-sets w i l l be equal when expressed i n t h i s common numeraire. I t follows that expected nominal i n t e r e s t rates on s e c u r i t i e s de-nominated i n terms of d i f f e r e n t numeraires w i l l d i f f e r as long as exchange rates are expected to change. Interest equali-zation between d i f f e r e n t countries implies that e i t h e r the inter n a t i o n a l c a p i t a l market i s imperfect or that exchange rates are not expected to change. Let be the spot one-period i n t e r e s t rate i n country D, f the one-period i n t e r e s t rate i n country P, &.Q the spot price of country F's currency i n terms of D's currency, and E(a^) the expected spot exchange rate at the s t a r t of the next period. Ex-pressed i n mathematical terms, a trader w i l l be i n d i f f e r e n t be-tween investing X units of domestic money now i n the domestic market and receiving (1 + R^)X units a f t e r one period, and ex-changing domestic currency f o r ( l / a Q ) X units of foreign currency and receiving ( l / a Q ) X ( l + R^)E(a^) units of domestic currency a f t e r one period i f (1 + R^)X = ( l / a 0 ) X ( l + R^)E(a^). 2 ^ I f investors are r i s k - n e u t r a l , i n the event of divergent expectations market equilibrium w i l l be determinate only i f i t i s assumed that the command over resources by any i n d i v i d u a l trader i s l i m i t e d by, e.g., a ceiling on the amount he can borrow. -33-Recognizing that E ( a 1 ) / a Q = 1 + [E(a 1) - /SLQ and defining the r e l a t i v e rate of change of the exchange rate expected to occur during period one as E(C^) = [Efa^) - a 0 ] / a Q , we obtain (1 + R*) = (1 + R*) [ 1 + E ( C 1 ) ] . (2-4) Expression (2-4) i s very s i m i l a r to formal statements of the i n t e r e s t p a r i t y theorem f a m i l i a r from the theory of forward ex-change except that the expected future spot rate i s substituted 2 6 f o r the forward exchange rate. Before generalizing formula (2-4) i t seems worthwhile to pause f o r a moment to take a closer look at t h i s expression. F i r s t of a l l , i t i s important to remember that the foreign ex-change rate expected to p r e v a i l at the s t a r t of period n, E ( a n ) , i s defined as the price of one unit of the foreign country's currency i n terms of the domestic numeraire. I f one wants to express the value of one unit of domestic currency i n terms of 27 foreign money, one simply divides 1 by E ( a n ) . ' For example, 2 6 See, e.g., Robert M. Stern, The Balance of Payments Chicagoi Aldine, 1973)» pp. 42-44, on the i n t e r e s t p a r i t y theor-em. For an i n t e r e s t i n g discussion of the empirical v a l i d i t y of t h i s theorem see Robert Z. 'l i b e r , "The Interest P a r i t y Theoremi A Reinterpretation". r 2 ate that i n general E ( l / a n ) ^ 1/E(a n). This problem i s discussed i n Jeremy J . Spiegel's paper on "Risk, Interest Rates and the Forward Exchange", Quarterly Journal of Economics, LXXXVI (May, 1 9 7 2 ) , pp. 3 0 3 - 0 9 . Spiegel encounters t h i s mathematical problem when attempting to derive an expression s i m i l a r to our expression (2-4) f o r bonds where the coupon i s reinvested con-t i n u a l l y at the contracted rate. Though defining the exchange rate as "the spot price of the foreign currency i n terms of the domestic currency", his expression ( 1 ) , i n our notation, states that exp (R£) = exp(R^)E(a Q/a 1) which i s c l e a r l y wrong. When making correct use of the exchange rate d e f i n i t i o n the l a s t term i n Spiegel's expression i s E(a^/a Q) i n which case the mathemati-- 3 4 -i f the currency unit i n both country D and country F i s the d o l l a r and i f the exchange rate i s F$l = D$1.10, then an invest-or owning D$100 could exchange these f o r (1/1.10)100 = 90.91 F d o l l a r s . Investing these F d o l l a r s i n one-period bonds y i e l d -ing 10 per cent, he w i l l own F$100.00 one period hence. I f during t h i s period an appreciation i n the value of the D d o l l a r by 5 per cent i s anticipated, the expected exchange rate one period hence i s F$l = D$1.045, and our investor can expect to f i n a l l y r e a l i z e 100(1.045) = 104 .50 D d o l l a r s from t h i s f o r -eign investment opportunity. Consequently he w i l l be i n d i f f -erent between investing his funds i n domestic bonds y i e l d i n g a rate of return of 4.5 per cent and investing i n foreign bonds y i e l d i n g 10 per cent i f he expects the domestic currency to ap-preciate i n value by 5 per cent. Using (2-4) to express t h i s i n formal terms we obtain (1 + .045) = (1 + ,100)[l + ( - . 0 5)J . Thus i f the expected change i n the exchange rate i s negative the foreign currency i s assumed to lose i n value v i s - a - v i s the domestic currency. Note that a 5 per cent appreciation of the D d o l l a r i s equivalent to a 1/1.045 - l / l . l 0 0 ) / ( l / l . l 0 ) ] l 0 0 = 5 . 2 6 3 per cent devaluation of the F d o l l a r . This difference between the appreciation and depreciation values becomes more pronounced as the r e l a t i v e magnitude of the exchange rate change increases. Thus a 15 per cent appreciation of one cur-c a l problem disappears. Spiegel provides also an example which indicates that f o r a l l p r a c t i c a l purposes the difference between E ( l / a n ) and l / E ( a n ) i s n e g l i g i b l e . This problem can be avoided completely by conducting the analysis i n terms of expected r e l a -t i v e rates of changes i n exchange rates rather than i n terms of expected exchange rates. -35-rency i s equivalent to a 1 7 » 6 5 P e r cent devaluation of the other currency. Another i n t e r e s t i n g r e s u l t can be derived by solving ex-pression ( 2 - 4 ) f o r R?jJ, R J = R £ + M C ± ) + R^E(C 1) ( 2 - 5 ) and, assuming E C c ^ ) to be constant, by d i f f e r e n t i a t i n g equation ( 2 - 5 ) with respect to Rj^, dR^/dR* = 1 + E < S 1 ) . This shows that, f o r a given expected rate of change of the ex-change rate, the in t e r e s t rate d i f f e r e n t i a l w i l l increase as the general i n t e r e s t rate l e v e l increases. For example, i f R^ = . 0 5 and R^ = . 0 2 , the implied expected exchange rate change i s . 0 2 9 1 . I f R 1 increases to . 0 ? and exchange rate expectations remain unchanged, R^ w i l l increase by ( 1 . 0 2 9 1 ) ( . 0 5 ) = . 0 5 1 5 to . 1 0 1 5 and the i n t e r e s t rate d i f f e r e n t i a l w i l l increase from . 0 3 0 0 to . 0 3 1 5 . For low inter e s t rate l e v e l s and small expected exchange rate changes, equation ( 2 - 5 ) can reasonably be approxi-mated by R j = R* + E t S ^ ) ( 2 - 5 a ) because R 1E(C 1) w i l l be a n e g l i g i b l e amount. However, as our e a r l i e r example has shown, t h i s does no longer hold f o r today's 28 intere s t rate l e v e l s . Later i n Chapter 3 i t w i l l be demon-strated that, f o r a given expected rate of change of the ex-change rate, the in t e r e s t rate d i f f e r e n t i a l w i l l increase very considerably as the in t e r e s t rate l e v e l increases once long-term See pp. 33-34 above. -36-bonds rather than one-period bonds are considered. The preceding analysis can e a s i l y be generalized to n p e r i -ods. To simplify the algebra, l e t R n be the continually com-pounded y i e l d to maturity on an n-period domestic bond and R n the y i e l d on a foreign bond. A r i s k - n e u t r a l investor w i l l be i n d i f f e r e n t between holding a domestic or a foreign bond i f exp(nR^) = [E(a n)exp(nR^)]/a 0 where E(a n) i s the exchange rate expected to p r e v a i l at the end of the nth period. Taking logarithms on both sides and d i v i d -ing by n we obtain R n = R n + l n [ E ( a n ) / a Q ] / n . The l a s t term i n t h i s equation i s the expected instantaneous r e l a t i v e rate of change i n the exchange rate measured i n units of (percentage per p e r i o d ) / l 0 0 . 2 ^ Let us denote t h i s rate by E ( C n ) . I t follows immediately that, given r i s k - n e u t r a l invest-ors and perfect international c a p i t a l markets, the i n t e r e s t rate d i f f e r e n t i a l observed between two countries i s an unbiased estimate of the average r e l a t i v e rate of change i n the exchange rate expected by the market. Denote t h i s d i f f e r e n t i a l by C n. From our e a r l i e r discussion we know that nR* = r? + 4 + + *t n 1 £. n Let nC\ = c, + c9 + ... e n x c. n 29 I f a t = a o e x p ( t C t ) , then ( d a t / d t ) / a t = c t = l n ( a t / a Q ) / t , -37-where c n stands f o r the rate of change i n the exchange rate during period n. We showed that nR = n(R + C ). It follows n n n that r l + r2 + * * * + r n = r l + r2 + ••• + r n + c l * c2 + ••• + cn* Dividing both sides of t h i s equation by ( n * - l ) R n - l = ^ n " , 1 ^ R n - l + ,) we obtain r = r"; + c or n-x ,n n n r d - r f = c = E(C, ). (2-6) n n n l , n ' v ; This fundamental r e l a t i o n s h i p demonstrates that i n a perfect international c a p i t a l market with r i s k - n e u t r a l investors the i n t e r e s t rate d i f f e r e n t i a l between domestic and foreign forward in t e r e s t rates f o r any period as derivable from the term struc-ture of i n t e r e s t rates i n the respective country i s equal to the market's expectation regarding the rate of change i n the ex-change rate during the same future period. This generalizes equations (2-4) and (2-5a) derived e a r l i e r f o r one-period spot r a t e s . 3 0 This analysis can e a s i l y be extended to one domestic and two or more foreign c a p i t a l markets. Let us introduce a t h i r d f d ffd ef country, G, and l e t c n , c° and c° designate the rate of change i n the exchange rate between countries D and F, D and G, 30 J I f coupon payments are made on a discontinuous basis, (2-6) becomes (1 + r ) - (1 + r * ) ( l + c n ) . A derivation of t h i s re-s u l t can be found i n Michael G. Porter, "A Theoretical and Em-p i r i c a l Framework f o r Analysing the Term Structure-of Exchange Rate Expectations", IMF S t a f f Papers, XVIII (November, 1971), pp.613-642. -38-and G and F, respectively, during period n as implied by 31 the term structure of interest rates i n these countries. Then, from (2-6), i t w i l l hold that < - r n " ° " • < 2- 6 a> r « - «« - o f , (?-6b) r£ - «* - o f . (2-«o) j _ j x» Subtracting (2-6a) from (2-6b) we obtain r - r j - (r - r ) = r n ~ r n = °n d" c n d * Substituting from (2-6c), i t i s seen that ° f = ° f " c " . (2-7) n n n Thus the difference i n the expected rate of change i n the ex-change rate between countries D and G and countries D and F i s equal to the expected rate of change i n the exchange rate be-tween countries F and G and also equal to the i n t e r e s t rate 32 d i f f e r e n t i a l between these two countries. In a perfect i n -ternational c a p i t a l market in t e r e s t arbitrage w i l l always as-sure that s i m i l a r relationships hold between any number of countries. -^Note that whether the exchange rate i s defined i n terms of the domestic ( c ^ * * 0 ^ ) o r a ^ore^S^ currency ( c d f , c d g ) i s only relevant f o r the sign of the rate of change ( p o s i t i v e i de-valuation? negativei revaluation) but no\longer f o r the absol-ute value of the rate because i t i s defined as an instantan-eous rate of change. For example, a 10?o devaluation of D's currency i s equivalent to a 10% revaluation of F*s currency. 32 For the discontinuous case, equation (2-7) becomes c g f = cgd . f c f d ( 1 + r f ) ] / ( i + r g ) . n n I n v n ' J ' v n' -39-2.42 RISK-AVERSE INVESTORS In our analysis of the i n t e r n a t i o n a l c a p i t a l market i t has been assumed up to now that e i t h e r exchange rates are f i x e d (Section 2.3 above) or that investors are r i s k - n e u t r a l (Section 2.41 above) so that exchange r i s k considerations could be ne-glected i n our analysis. However, i n order to derive a capi-t a l market model more c l o s e l y r e f l e c t i n g actual relationships between national sub-markets both the p o s s i b i l i t y of changes i n exchange rates and of risk-averse investor behaviour should be allowed f o r . In our two-country international c a p i t a l market, any i n -vestor i s faced by uncertainties concerning three d i f f e r e n t variables, future domestic inte r e s t rates, future foreign i n -te r e s t rates and future rates of change i n the exchange rates. I f there existed f u l l y developed forward exchange markets f o r a l l future time periods up to period N, the r i s k due to uncer-t a i n t i e s about future exchange rates could be completely elim-inated. Whether t h i s would be at a cost comparable to an i n -surance premium or whether the forward exchange market should be compared to a zero-sum game i n that one p a r t i c i p a n t ' s loss 33 i s another parti c i p a n t ' s gain i s an unsolved issue. J How-ever, t h i s need not concern us because the best-developed forward markets are the 30-day and 90-day markets, and f o r ma-34 t u r i t i e s of two years or more quotations are very rare. 3 3See Katz, pp. 38-42. 34 J Cf. Egon Sohmen, The Theory of Forward Exchange (Prince-ton Studies i n International Finance No. 17? Princeton, N.J.t International Finance Section, Department of Economics, Prince-ton University, 1966), p. 32, or Herbert G. Grubel, Forward Ex-change Speculation and the International Flow of C a p i t a l (Stan-ford University Press, 1966), pp. 5 2 , 116. -40-The theory of forward exchange i s only concerned with short-term phenomena and analyses, i n t e r a l i a , the influence a given term structure of i n t e r e s t rates has on the forward exchange rates established i n the market. Here, by concentrating on long-term phenomena and assuming that no forward exchange mar-kets e x i s t i t w i l l be shown how an expression f o r the e q u i l i -brium international term structure of i n t e r e s t rates can be derived as determined by i n d i v i d u a l traders* expectations about int e r e s t rates and exchange rate changes, and by t h e i r degree of r i s k aversion. I f we assume that investors only consume goods f o r which they have to pay i n t h e i r own currency, then any lending or borrowing a c t i v i t y i n the other country's capi-t a l market does not only involve i n t e r e s t arbitrage but also an element of speculation because the foreign exchange r i s k involved i n such transactions cannot be covered. F i r s t l y , the measurement of exchange rate r i s k w i l l be discussed. This i s followed by an extension of R o l l ' s model of equilibrium i n t e r e s t rates to open economies. F i n a l l y , the basic ideas of our international c a p i t a l market theory w i l l be elucidated by graphical means. - 4 1 -2.421 THE MEASUREMENT OF EXCHANGE RISK It i s usually argued that f i n a n c i a l transactions involv-ing foreign currencies are more r i s k y than purely domestic r i s k i s measured, t h i s need not necessarily be the case. As pointed out e a r l i e r , the r i s k i n e s s perceived by a trader specu-l a t i n g i n the forward market depends on the uncertainties sur-rounding his expectations of future spot i n t e r e s t rates. When speculating i n a foreign forward market, the expected return i n terms of the trader's home currency w i l l depend on uncer-t a i n t i e s surrounding his expectations of future spot i n t e r e s t rates and uncertainties about exchange rate changes. I f one were to measure r i s k by the variance i n the expected e f f e c t i v e Assuming that the amount and kind of information available to each investor i s s i m i l a r f o r both c a p i t a l markets, 3'' i t i s ''-'For example, see David K. Eiteman and A. I. S t o n e h i l l , Multinational Business Finance, (Reading, Mass.j Addison-Wesley), 1973• Chapter 11. 3 ^ I f we accepted the c a p i t a l asset p r i c i n g model as an em-p i r i c a l l y v a l i d description of how investors value f i x e d income s e c u r i t i e s , then the covariance between the e f f e c t i v e rate of re-turn on foreign investment and the "market'' return would be the correct r i s k measure. However, the question of whether the CAPM constitutes an appropriate theory of bond valuation i s beyond the scope of t h i s study. For some a n a l y t i c a l work extending the CAPM to the valuation of equities i n an i n t e r n a t i o n a l c a p i t a l market see Solnik. 3 ^ I n a l e s s than perfect world, we may expect information on foreign economies and t h e i r c a p i t a l markets to be more d i f f i c u l t to obtain than information on the domestic economy, to be more cos t l y and to be more d i f f i c u l t to i n t e r p r e t leading to more un-certainty with regard to foreign markets and/or to higher inform-atio n costs. transactions. 35 However, depending on how foreign exchange l,n» Cl,n - 4 2 -reasonable to expect the uncertainty surrounding expected i n t e r -est rates i n both markets to be equal, or var(R d ) = var(R:f ). ± ,n j. ,n Whether foreign investment i s perceived as more r i s k y than domestic investment or not w i l l then depend on whether v a r ( C l f n ) £ - 2 c o v ( R ^ n , C l f n ) . I t follows that i f investors perceive a negative r e l a t i o n s h i p between foreign i n t e r e s t rates and exchange rate changes t h i s could mean that foreign investments are regarded as l e s s r i s k y . 3 8 But only i f the negative covariance i s s u f f i c i e n t l y large to o f f s e t the increase i n variance due to var(C, _ J w i l l the t o t a l l , n variance on foreign investments be l e s s . However, i t seems questionable whether foreign exchange r i s k should be measured by the variance of the rate of change i n the exchange rate or by i t s covariance with a "market" rate of return. Subjective p r o b a b i l i t y density functions of t h i s variable are often highly skewed so that semi-variance or skew-ness would constitute better measures of r i s k because i n t e r -national traders seem to be preoccupied with p o t e n t i a l losses due to unfavourable exchange rate changes. Discussions i n the l i t e r a t u r e as well as the information gained from our interviews with Canadian corporations and underwriters indicate that borrow-ers as well as lenders i n international c a p i t a l markets regard O Q J A p r i o r i we might expect the covariance to be negative. During times of balance of payments problems countries tend to increase interest,rates to a t t r a c t f o r e i g n . c a p i t a l and to exert a deflationary pressure on the economy. But at the same time the uncertainty about the exchange rate usually increases con-siderably. Note that a devaluation of a foreign currency i s equivalent to an appreciation of the domestic currency, i . e . , C, w i l l be negative. -43-foreign exchange r i s k as an additive r i s k , not as a r i s k that can be d i v e r s i f i e d away. 7 This i s a major methodological j u s t i f i c a t i o n f o r our using Roll's model as a basis f o r extending term structure theory to open economies rather than Bierwag and Grove's p o r t f o l i o model. Though R o l l has shown that a s p e c i a l solution to his model " i s exactly the same as the r e s u l t s derived by Bierwag and Grove (1967) under a s t r i c t e r set of assumptions"^ 0 with-i n a two-period framework, the two theories are fundamentally d i f f e r e n t . P o r t f o l i o theory i s normative i n character and pre-scribes how investors should select t h e i r p o r t f o l i o s , given that t h e i r u t i l i t y increases with return and decreases with r i s k as measured by the p o r t f o l i o ' s variance. C l e a r l y , these are appeal-ing c h a r a c t e r i s t i c s . On the other hand, models based on port-f o l i o theory may not be able to explain observed investor be-haviour i f investors do not behave as prescribed by such models. R o l l also assumes that investors prefer more wealth to less and are risk-averse, but he leaves open the questions of how -''Tor a comprehensive treatment of exchange r i s k see Katz. Sidney M. Robbins and Robert B. Stobaugh i n t h e i r book Money i n  the Multinational Enterprise (New York! Basic Books, 1973)» p. 27, observe that, "anxious to avoid incurring losses through changes i n exchange rates, some f i n a n c i a l managers simply ignore the p o s s i b i l i t y of exploiting interest-rate d i f f e r e n t i a l s by moving money across boundaries." Our discussions with Canadian managers often revealed s i m i l a r a t t i t u d e s . Likewise, an Ameri-can underwriter remarked that U.S. f i n a n c i a l i n s t i t u t i o n s , which predominantly invest i n Canadian U.S.-pay bonds, " f i n d i t inap-propriate to assume exchange r i s k . " ^°Roll, p. 29. _44-to measure p o r t f o l i o r i s k and how investors should arrange t h e i r (bond) p o r t f o l i o s . Therefore his model must be regarded as more general and as p o s i t i v e i n character. Another a t t r a c t i v e feature of his theory i s that a l l i n t e r e s t rates are determined simultaneously as a function of borrowers' demand f o r and lenders* supply of funds. In a p o r t f o l i o model i t i s necessary to assume that the r i s k - f r e e i n t e r e s t rate i s exogenously de-termined. This i s not a very s a t i s f y i n g assumption i n an i n t e r -national context because of the i n t e r r e l a t i o n s h i p between i n t e r -est rates and c a p i t a l flows. In our t h e o r e t i c a l discussions that follow we s h a l l always regard foreign investments as more ri s k y than transactions i n -volving s e c u r i t i e s denominated i n the traders' home currency 41 only. It w i l l not be necessary f o r our purposes to be speci-f i c about how to measure foreign exchange r i s k . However, i n Appendix 1 i t w i l l be .demonstrated how under very r e s t r i c t i v e assumptions a mean-variance approach could be employed to devel-op a model of the determinants of the international term struc-ture of inter e s t rates. I n e f f i c i e n t or non-existing domestic c a p i t a l markets and p o l i t i c a l factors may lead people to regard foreign s e c u r i t i e s as safer than domestic investments. Our assumptions with regard to a perfect international c a p i t a l market exclude these possi-b i l i t i e s from the t h e o r e t i c a l analysis. -45-2.422 AN EXTENSION OP ROLL'S MODEL TO OPEN ECONOMIES WITH EXCHANGE RATE RISK In our two-country model world every trader i s faced with borrowing and lending opportunities i n the two national sub-markets of the international c a p i t a l market. For every one of the N periods there w i l l be two forward markets, one f o r funds denominated i n country D's currency and one f o r funds denomin-ated i n country F's currency. Consequently every trader i s as-sumed to have no longer N but 2N excess supply functions f o r forward loans. Again r e s t r i c t i n g our analysis to a two-period framework, these supply functions w i l l be given by -h _ »h, d d f f d d f fv (2-8) i q j " i f j ( r l " i rl» r l " l rl» r2 * i r 2 * r2 " i r 2 ) f o r j = i ; 2 and h = d, f . Assuming that the i t h investor i s a resident of country D, trading i n bonds denominated i n his country's numeraire does not involve any r i s k d i f f e r e n t from that discussed e a r l i e r f o r a closed economy. Investments i n foreign bonds, on the other hand, w i l l be more r i s k y because of the exchange rate r i s k involved. The personal forward rates .r_ and . r of any trader i r l n I n J are again chosen such that i q j = i f J ( ° ' °' °' 0 ) = °* The personal domestic forward rates ^ r n w i l l be given by expres-42 sion (2-3) i f trader i i s a resident of country D. His f o r -f eign forward rates ^ r n w i l l also depend on the four quantities 42 See p.23 above. -46-mentioned above plus an additional f i f t h factor» 5. The addi t i o n a l uncertainty surrounding his expectation of the rate of return r e a l i z a b l e i n the foreign market due to exchange r i s k . It i s assumed that these f i v e quantities can be expressed i n an additive form by the t y p i c a l equation i r n = Ei<*l.n> + i L n + X where ^Mn i s trader i ' s addi t i o n a l r i s k premium due to foreign exchange r i s k when trading i n country F*s forward market. In general, any trader's forward rate can be written as i r n " E i < l n > + A' where = -L^ + jM^» The foreign exchange r i s k premium, ^IV^, w i l l be equal to zero i f trader i invests i n his home mark-et. It w i l l be p o s i t i v e i f he intends to borrow abroad at the expected foreign spot rate, and i t w i l l be negative i f he ex-pects to lend at that rate. I t must be remembered that the ab-h • solute value of ^ L n w i l l be larger the l e s s confidence a trader has i n his expectation with regard to a future ( e f f e c t i v e ) spot rate. The basic c h a r a c t e r i s t i c s of the supply functions repres-ented by (2-8) are the same as those discussed e a r l i e r f o r a closed economy. However, the i t h investor's forward supply function i n a market foreign to him w i l l be steeper than his supply function i n h i s home market f o r the same period f o r two reasons 1 (1) The forward supply function i s derived from the supply function f o r corresponding future spot loans, and t h i s foreign spot supply curve w i l l have a steeper slope than the - i n -comparable domestic supply curve because of exchange r i s k . Whereas the y i e l d on an investment i n the domestic spot market i s certain a f t e r a tatonnement process has determined the e q u i l -ibrium spot rate at the beginning of the period, the y i e l d on a foreign spot investment i s also influenced by the rate of change i n the exchange rate during the period. (2 ) Because the investor i s assumed to be less confident about his expectation of the f u t u r e . e f f e c t i v e y i e l d on a foreign investment than on a domestic one, the angle 9 between his foreign spot and f o r -ward demand functions w i l l also be smaller than that between his domestic demand curves f o r the same period, making the f o r -ward function even steeper. Looking at the second period, i t w i l l be i n t e r e s t i n g to consider under what conditions the i t h investor, who may be a resident of country D, does neither enter the domestic nor the foreign forward market. To simplify the analysis, l e t us assume that the f i r s t - p e r i o d rates are such that he does not invest i n the domestic nor the foreign c a p i t a l market, that i s , we can neglect the influence trading i n the f i r s t period would have on investment plans during the second period. Furthermore, at the expected domestic second period spot rate he w i l l be a borrower. By d e f i n i t i o n , he w i l l not trade i n country D*s market i f r 2 = i r 2 ~ E i ^ R l 2^ + i L 2 * L i k e w i s e » he w i l l not enter the f f ** f f ' foreign forward market i f r g = i r 2 = E i(R ] L 2 ) + ^ 2 • N o t e A "P that ^ r 2 - - r 2 = E^C^ ; 2),that i s , the trader's two personal forward rates imply his expectation about the anticipated rate of change i n the exchange rate during period two. I f t h i s did -48-not hold, he would have an incentive to speculate i n at le a s t one of the two forward markets. A small example w i l l help to i l l u s t r a t e t h i s very important point. J Given the investor's expectations concerning future spot rates and h i s r i s k premia by, say, E f R ^ g ) = » 0 5 , Lg = L | = . 0 1 , Mg = . 0 0 , E ( R £ 2 ) = . 0 3 , M2 = . 0 1 , then ~ , 44 ' E ^ C 1 2) ~ * 0 1 because E ( R ^ 2 ) + L 2 + M2 = E ( R ^ 2 ) + l | + + E f C ^ ) or ( . 0 5 + . 0 1 + . 0 0 ) = ( . 0 3 + . 01 + . 0 1 ) + . 0 1 . Consider the domestic forward market f i r s t . Our investor ex-pects the future spot rate to be . 0 5 . At t h i s rate, he plans to borrow i n his home spot market, and whenever the forward rate i s below . 0 6 he w i l l s t a r t borrowing part of these funds i n the forward market i n order to reduce r i s k . At a forward rate r = . 0 6 the expected gain i n u t i l i t y from borrowing spot at a lower rate and the loss i n u t i l i t y due to r i s k inherent i n borrowing l a t e r rather than now just balance so that he w i l l not p a r t i c i p a t e i n the domestic forward market. S i m i l a r l y , he expects the foreign spot rate to equal . 0 3 and, because of an 43 ^To simplify the notation, reference to the i t h investor w i l l be dropped whenever possible without causing ambiguity. 44 As we have assumed that complete information on a l l markets i s available to a l l investors, we also assume that -49-expected devaluation of his home currency by one per cent or . 0 1 , he anticipates an e f f e c t i v e borrowing rate of .04 i n the foreign country. As t h i s i s below the expected domestic bor-rowing rate, he w i l l have an incentive to borrow part of h i s funds abroad i n spite of the r i s k of foreign borrowing. In-deed, given a domestic forward rate of . 0 6 , he w i l l s t a r t bor-rowing i n the foreign forward market whenever the foreign f o r -ward rate i s le s s than . 0 5 or whenever his perceived e f f e c t i v e borrowing rate i s le s s than . 0 6 . Only i f the foreign forward rate i s . 0 5 and therefore the investor's expected e f f e c t i v e borrowing rate abroad . 0 6 w i l l he have no incentive to enter the foreign forward market. I f , on the other hand, our invest-or expected no change i n the exchange rate, he would have a strong incentive to borrow i n the foreign forward market even i f his personal and the actual forward rate were both equal to . 0 5 . But t h i s i s inconsistent with our assumption that, when a l l personal forward rates equal actual forward rates, no supply or demand f o r forward funds w i l l m aterialize. This shows that an investor's personal forward rates imply his expectations about rates of change i n the exchange rate during respective periods and therefore need not be taken into account e x p l i c i t -l y i n his excess demand functions. In order to obtain an i n t u i t i v e notion of how i n t e r e s t rates are determined i n our model, l e t us continue with our example arid assume that a second investor, a resident of count-ry F, operates i n the market whose expectations are i d e n t i c a l to those held by the resident of country D. Note that what -50-i s a p o s i t i v e expected rate of change i n the exchange rate from country D*s point of view i s a negative change from the other country's point of view. But the absolute value of the r e l a -t i v e rate of change i s the same i f measured as an instantane-ous rate. Assume also that the second investor does not plan to borrow or lend at his country's expected spot rate of . 0 3 , that i s , his time and r i s k premium, ^Lg equals . 0 0 implying that his time preference d i f f e r s from that of the f i r s t invest-45 or who planned to borrow at t h i s rate. J As the expected spot rate i n country D i s . 0 5 , country F's resident plans to lend abroad because t h i s i s equivalent to an expected e f f e c t i v e y i e l d of .04 fef(Rj 2 ) + E f ( ^ i 2 ) = , 0 5 " ' P 1 ! * a r a t e a t which he would lend at home too. Of course, because of the higher r i s k involved he w i l l invest a smaller amount abroad than he would at home at a rate of .04 i f he i s risk-averse. Assuming his time and r i s k premium with respect to the foreign market, ^ L g ' t o e c l u a l f^z = ^ i s exchange r i s k premium, d 1 fM 2 equals - . 0 1 because E f ( R * t 2 ) + + fM^ + E f ( C 1 ( 2 ) = B f ( R * i 2 > + f l f + fMf or ( . 0 5 + . 0 0 - . 0 1 ) - . 01 = ( . 0 3 + . 0 0 + . 0 0 ) . When regarding our two investors as representative of the "average" trader of the respective countries, some i n t e r e s t i n g observations can be made. In closed economies, the domestic -'The subscript f r e f e r s to the resident of country F, the subscript d to the resident of country D. -51-forward rate f o r the second period, r d , w i l l equal . 0 6 because at t h i s rate no excess supply of loanable funds would be f o r t h -coming, and the foreign forward rate, r 2 , w i l l equal . 0 3 . In open economies, the supply of funds by traders from country P i n country D*s forward market tends to drive r d below . 0 6 , and s i m i l a r l y the demand f o r funds by borrowers from country D i n f country F's market forces r 2 up. The new i n t e r e s t rate d i f f e r -e n t i a l w i l l l i e somewhere i n the i n t e r v a l .OK r d - r f < . 0 3 depending on the exchange r i s k perceived, the traders* r i s k aversion, and t h e i r wealth. The higher the foreign exchange r i s k perceived and the more risk-averse they are, the fewer funds w i l l be lent or borrowed i n t e r n a t i o n a l l y ; the wealthier a person the more funds he w i l l move between national c a p i t a l markets. The i n t e r e s t rate d i f f e r e n t i a l cannot be equal to the expected rate of change i n the exchange rate because then it,would not pay to invest internatio>nally as exchange r i s k would have to be accepted at no expected gain. In other words, as long as c a p i t a l moves between two countries, the i n t e r e s t rate d i f f e r e n t i a l w i l l not only r e f l e c t exchange rate expecta-tions but w i l l be l a r g e r by a c e r t a i n r i s k premium. In a world dominated by risk-averse traders, expression ( 2 - 6 ) w i l l no longer hold and must be replaced by rt•- rt 8 8 c « = J + (2-9) n n n l»n n where denotes a foreign exchange r i s k premium. From country D*s point of view, i f the i n t e r e s t rate d i f f e r e n t i a l i s p o s i t i v e , -52-the domestic currency i s expected to lose i n value v i s - a - v i s the foreign currency, and both the expected rate of change i n the exchange rate and the exchange r i s k premium are p o s i t i v e . Un-less t h i s holds residents i n country D w i l l have no incentive to borrow abroad. S i m i l a r l y , i f the i n t e r e s t rate d i f f e r e n t i a l i s negative, both the expected exchange rate change and the ex-change r i s k premium are negative. Otherwise investors i n count-ry D w i l l not lend i n country F. More generally, from country D's point of view, the exchange r i s k premium w i l l always be Hrn = r d - r f - E,(C, ) (2-9a) d n n n d l , n and from the foreign country's viewpoint f m n = r n " 4 - V 8 ! . ^ <2-*>> where dm n = -fmn and ) = - E f ( C 1 > n ) . Expression (2-9a) and (2-9b) allow f o r the fact that the e f f e c t i v e i n t e r e s t rate d i f f e r e n t i a l , that i s , the i n t e r e s t rate d i f f e r e n t i a l a f t e r tak<-ing exchange rate expectations into account, may be p o s i t i v e even i f the nominal d i f f e r e n t i a l i s negative and vice versa. In our example i t was assumed that f o r both investors E ^ l , 2 ^ " E ^ l , 2 ^ ^ E ^ l , 2 ^ * I f t h e r e e x i s t systematic d i f f e r e n -ces i n time preferences between two countries, t h i s i s a reason-able assumption. C l e a r l y , our model does not presuppose that i n -ternational c a p i t a l w i l l only flow i n one d i r e c t i o n . For ex-ample, a resident of country D with a set of expectations d i f f -erent from those mentioned e a r l i e r may well lend abroad at the same time as other domestic traders borrow i n F. -53-We are now ready to derive an approximate solu t i o n f o r the 46 equilibrium i nternational term structure of i n t e r e s t rates. Each trader's excess supply functions as given by expression (2-8) w i l l be expanded around the points where there i s zero excess supply of funds, that i s , where r n - . r n = 0 f o r a l l n and h. This r e s u l t s i n the Maclaurin series .q*| = Z T ( r * - ,rJ})(J.fV3rJj +higher order terms i n r . (2-9) 1 3 n=l h=d n 1 n 1 J n With two countries, each trader w i l l have 2N or four excess supply functions. International c a p i t a l market equilibrium w i l l be obtained i f a l l forward rates are such that t o t a l ex-cess supply by a l l traders f o r a l l periods and a l l countries i s zero, or ' I I Z = 0. i j h 1 3 Denote by _ / „d _ f _d „f\, the column vector of excess supply of funds by the i t h investor i n each forward market; r = ( r j . r f . r f . r f ) -the column vector of forward rates i n each marketj As mentioned e a r l i e r (see p.27 above), our derivation pro-ceeds along s i m i l a r l i n e s to that of R o l l . However, because we dropped the subscript j from i n d i v i d u a l forward rates (see foot-note 19 above), the algebra i s l e s s involved. 47 fBy l e t t i n g j vary over 1, 2, n periods and h over 1, 2, k countries our r e s u l t s can e a s i l y be generalized to more than two periods and more than two countries. -54-£ i ~ ( i r l » i r l , i r 2 , i r 2 ) ' t h e c o l u m n v e c t o r o f t h e i t h i n v e s t o r ' s p e r s o n a l f o r w a r d r a t e s a n d by V i / 3 r i • • • • hfi/drz his matrix of p a r t i a l derivatives. Then the system of supply functions ( 2 - 9 ) may be written i n matrix notation as 2^ = F^r - Z^—i + higher-order terms. Summing over a l l traders, international c a p i t a l market eq u i l i b r i a urn w i l l be obtained i f t o t a l excess supply i s zero, that i s , i f S a i = o. i Neglecting higher-order terms, the set of equilibrium forward rates can be derived by solving 5- ai = £ - 2i£i> = 0 i 1 i 1 i i -f o r r = ( Z. F. ) - 1X F.r. . ( 2 - 1 0 ) i 1 i This important r e s u l t indicates that, as a f i r s t approximation, i n t e r e s t rates i n open economies depend on each trader's matrix of own and cross e l a s t i c i t i e s of supply, his expectations con-cerning future one-period spot rates at home and abroad, his -55-r i s k premia and, i m p l i c i t l y , his exchange rate expectations. A r e l a t i v e l y simple s o l u t i o n for, say, the second period domestic forward rate can be derived i f we assume that ( 1 ) a l l i n t e r -temporal cross e l a s t i c i t i e s are zero, that i s , that the F^ matrices are block-diagonal (only the cross e l a s t i c i t i e s between the domestic and foreign forward market f o r the same period may d i f f e r from zero), and that (2) the sum of the own e l a s t i c i t i e s i n the domestic market equals the sum of the own e l a s t i c i t i e s i n the foreign market and that the same holds f o r the cross e l a s t i c i t i e s . 48 Then 2 " l ^ i ^ i +vS/*£> d-TJ (E.(Rdt2) + i L d t 2 + i M d f 2) ? (E. (if 9) + ,L? 0 + . M ? 0 ) d/a-ix i 1,2' 1 1 , 2 I 1,2' Under these s p e c i a l conditions which may hold approximately i f two countries are quite s i m i l a r , the domestic forward rate i s seen to be a weighted-average of expectations of in t e r e s t rates i n country D and F over both domestic and foreign traders act-ive i n the market. The weights with respect to each investor's 48 For example, the second block of Z. *d/ _d\ _ _ T *d / _ f x _ F. w i l l be I <vf/ 4> -« I f a = d and b = c, the inverse of t h i s block i s 1 f d -cl = 1 f a -b| ad-bc |-b a j a^-b^ [-b a l This then leads to the special solution given i n the text. Note that a block-diagonal matrix can be inverted by inverting each block separately. -56-expectation of the domestic in t e r e s t rate depend on his i n c l i n -a t i o n to change his investment i n domestic bonds as the domestic rate changes which i s related to his wealth and his confidence i n his expectation of the (effective) y i e l d r e a l i z a b l e i n the future spot market i n country D coupled with his degree of r i s k aversion and his time preference. The weights with respect to each investor's expectation of the foreign spot rate f o r the same period are a function of h i s marginal propensity to s h i f t funds from or to country D as the rate i n country F changes which i s again related to his resources and his confidence i n his expectation of the (effective) future spot rate i n country F. Because of exchange r i s k , one should expect cross e l a s t i c i -t i e s to be r e l a t i v e l y small compared to the aggregate own el a s t -i c i t i e s so that r 2 w i l l be influenced by E(R d 2 ) m u c n more than by E ( R ^ 2 ) . However, i n general intertemporal e l a s t i c i t i e s w i l l not equal zero though we may suppose that they approach zero as we move away from the main diagonal. Also, e l a s t i c i t i e s of aggre-gate market excess supply functions w i l l d i f f e r i f there are differences i n wealth, r i s k aversion, and so on between nations. A more general s o l u t i o n f o r , say, r 2 can be derived from ( 2 - 1 0 ) by extracting an in d i v i d u a l equation f o r t h i s rate. The sys-tem of equations ( 2 - 1 0 ) can be rewritten as r =Z + L i ) ( 2 - 1 0 a ) where the matrix ^ = ( ^ F i ) ~ 1 F i and E^ and L i are column vectors of the i t h trader's expected spot rates and h i s (exchange) r i s k and time premia. Denoting the jth row of'V- by .v. and .y.'s -57-kth element by ^ L ^ * a n d remembering that r 2 i s the (2n-l)th or t h i r d element i n r, we obtain r2 = + £ i ^ i ° r 4 = ^ i ^ 3 3 E i ( R l , 2 ) + L f where L d ' = L* + M£ = I ^ v ^ + - ^ E ^ R ^ 2)J . Thus any forward rate r h = E(R h ) + L h* (2-12) n v l,n' n v ' can be seen to be a weighted average expectation of the future spot rate over a l l traders i n a l l c a p i t a l markets plus a function of t h e i r expectations of a l l other future one-period rates i n a l l markets, t h e i r degree of r i s k aversion, t h e i r wealth, t h e i r time preferences and, i m p l i c i t l y , t h e i r exchange rate expecta-t i o n s . 2.423 A GRAPHICAL ANALYSIS The basic ideas of our international c a p i t a l market theory may be best elucidated by graphical means. Let us aggregate the domestie excess forward supply functions of a l l residents of country D into one function f o r each period, and do the same with those functions of country F's investors. Also, the same holds f o r country F's markets. That i s , there w i l l be only two excess supply functions f o r each one-period forward market i n each country, one by residents of country D and one by traders from country F. Assuming a l l forward i n t e r e s t rates except those f o r the market under study as given, that i s , the other -58-FIGURE 2-2a INTERNATIONAL CAPITAL MARKET EQUILIBRIUM'FOR FORWARD LOANSs MARKET IN COUNTRY D -59-2 N - 1 markets are i n equilibrium, such excess supply functions f o r the nth period are drawn i n Figures 2 - 2 a and 2 - 2 b f o r both the domestic and the foreign markets. To simplify the graphs, i t i s assumed that, i n the aggregate, a l l traders have i d e n t i -c a l expectations with regard to exchange rate changes and future one-period spot rates. Looking f i r s t at the market i n country D, Figure 2 - 2 a , the excess forward supply function of domestic traders ( d S d d S d ) crosses the v e r t i c a l axis at r d * * . I f D were a closed economy t h i s would be the equilibrium forward rate because excess supply i s zero at t h i s point. However, i n an open economy, foreigners w i l l enter the market. They are assumed to have a lower time preference than l o c a l traders, and therefore they would supply funds i n country D at r ^ . Their supply function ( f S f ^S^) i s drawn with a steeper slope as an i n d i c a t i o n of the additional uncertainty about the ex-change rate on t h e i r part. I t can be seen from the graph that due to the supply of funds by foreigners the new equilibrium d* forward rate, r n , w i l l be lower and be established where the excess supply of funds by foreigners, q n 2 equals the excess demand f o r funds by l o c a l traders, q ^ . The expected spot rate f o r period n i n country F i s lower than the spot rate expected f o r the same period i n country D even i f e f f e c t i v e rates, that i s , rates taking into account the expected rate of change i n the exchange rate are considered. This follows from our assumption that investors i n country D have a higher time preference f o r period n. Looking at Figure 2 - 2 b , and supposing that traders i n both countries are s i m i l a r -60-FIGURE 2-2b INTERNATIONAL CAPITAL MARKET EQUILIBRIUM FOR FORV/ARD LOANS 8 MARKET IN COUNTRY F -61-i n t h e i r degree of risk-aversion and that uncertainties per-ceived with respect to future spot rates are also the same, the r e l a t i v e l y f l a t excess supply curve of residents i n country F f f i n t h e i r home market ( f S f .^S^ .) indicates that t h e i r aggregate wealth i s greater than that of traders l i v i n g i n country D whose domestic excess supply curve i s drawn with a steeper slope. The excess supply curve of country D fs residents i n the foreign market ( d S f d S f ) i s drawn les s steep than ( f s £ .^S^ .) i n Figure 2-2a as an i n d i c a t i o n that country D's traders are les s uncert-a i n about the expected exchange rate change than investors from country F, that i s , we assume that there are systematic d i f f e r -ences i n the exchange r i s k perceived between the two groups of ILQ f * investors. 7 The equilibrium forward rate i n country F, r n , i s higher than r ^ , the closed-economy equilibrium rate. Only at the higher rate does excess demand f o r funds by country D f residents, q n l i equal excess supply of funds by country F r e s i -f dents, ° A n 2 * 2.5 INFORMATION COSTS, TRANSACTION COSTS, AND EQUILIBRIUM IN-TEREST RATES IN AN INTERNATIONAL CAPITAL MARKET One of the assumptions made when developing our model of an international c a p i t a l market f o r long-term debt instruments was that information i s costless and that transaction costs are 7 T h i s assumption i s not c r u c i a l to our argument. Rather i t has been introduced because, f o r example, American i n s t i -t u t i o n a l investors, the main suppliers of funds to foreigners i n the U.S. market, seem to shy away from exchange r i s k to a much higher extent than foreign borrowers. - 6 2 -zero. Now t h i s assumption-*0 w i l l he relaxed and replaced by the following assertions! 2 a . Information and transaction costs i n a trader's home market are zero. 2 b . When trading i n a market foreign to him, an investor incurs variable information and transaction costs. They are measured as a percentage of the gross amount involved i n a transaction, are a decreasing function of traders' present wealth positions, and are known f o r each period with ce r t a i n t y . These assumptions allow us to concentrate on the difference i n transaction costs between trading i n the home market and trad-ing i n a foreign market.^ Transacting business i n a foreign environment usually involves l e g a l , taxation, language, and other problems not encountered at home. Longer communication channels, higher t r a v e l expenses, a r e l a t i v e s c a r c i t y of i n -formation on foreign markets and d i f f i c u l t i e s i n i n t e r p r e t i n g foreign data a l l w i l l contribute to t h i s difference i n trans-action costs. Many of these costs w i l l be f i x e d rather than v a r i a b l e . However, within a reasonable range, t o t a l costs may be regarded as a constant percentage of the amount involved i n a transaction though, more generally, they w i l l decrease r e l a -t i v e l y as transaction size increases. An investor's wealth and the average size of h i s transactions are usually highly correlated which j u s t i f i e s our assumption made. I m p l i c i t l y i t i s also assumed that, f o r example, a domestic resident borrow-ing i n a foreign market has to reimburse the foreigner f o r the extra costs incurred by the foreigner as a r e s u l t of lending to -*°See assumption 2 , p.21 above. ^Henceforth, the term transaction costs w i l l be understood to include information costs as w e l l . -63-him rather than to l o c a l borrowers. Usually i t w i l l be more d i f f i c u l t to obtain information on a foreign borrower and to establ i s h his credit-worthiness than that of a l o c a l fund seeker.-*2 Let -Z_ be the i t h investor's transaction costs f o r trans-1 n actions i n country h i n the nth period forward market. They are measured i n units of (percentage per period ) / l 0 0 , and are zero i f h i s the trader's home country. These transaction costs enter each investor's excess supply function as an argu-ment, that i s h _ .ph/ d d r»d._f „f. rif _d „d_ 7 d . f f „ f v i q j " i f j ( r l " i r l J i Z l j r l ' i r l ' i Z l , r 2 " i r 2 } i Z 2 j r 2 " i r 2 « i Z 2 ) ' ( 2 - 1 1 ) The nature of these modified excess supply functions can be best elucidated by deriving the excess forward supply of foreign funds by a t y p i c a l domestic trader f o r period n from h i s corres-ponding planned spot supply of funds i n country F at the st a r t of period n. In Figure 2 - 3 , these two supply functions are graphed. I f the future spot rate i n the foreign market were f R , the domestic trader would neither borrow nor lend abroad. f f Only at a rate R + Z, would he s t a r t lending i n the foreign x ,n market. I f the rate at which he can lend does not make up f o r the extra transaction costs incurred by trading i n the foreign  market, he i s better o f f by not p a r t i c i p a t i n g i n that market. I t follows that there i s a discontinuity i n his supply function 7 In Chapter 3 such transaction and information costs d i f f e r -e n t i a l s w i l l be discussed i n more d e t a i l . -64-FIGURE 2-3 A DOMESTIC INVESTOR'S EXCESS SUPPLY FUNCTIONS FOR SPOT AND FORWARD FUNDS IN A FOREIGN CAPITAL MARKET WITH VARIABLE TRANSACTION COSTS DIFFERENTIALS -65-f' f ' f ( d S s d S g ) f o r future spot loans around R . S i m i l a r l y , his f' f • supply function f o r forward funds ( d S j dS£ ) w i l l have two kinks. These d i s c o n t i n u i t i e s occur where the investor changes from being a borrower i n the forward market to being a lender i n the forward market, that i s , his forward supply function w i l l f also s h i f t upwards by 2Z, at the point where i t crosses the ±, n v e r t i c a l a x i s . Because of these d i s c o n t i n u i t i e s i n the supply functions, no straightforward mathematical solution f o r the int e r n a t i o n a l equilibrium term structure of in t e r e s t rates can be derived. But the jumps i n the forward supply functions w i l l be smaller the wealthier an investor i s and i n a l l p r o b a b i l i t y w i l l approach 53 zero f o r very large transactions. J Also, because of differences i n expectations and time preferences, d i s c o n t i n u i t i e s i n i n d i -vidual traders' curves w i l l occur at d i f f e r e n t i n t e r e s t rate l e v e l s . By aggregating i n d i v i d u a l excess supply functions a-cross a l l traders of a p a r t i c u l a r country a continuous supply function w i l l be obtained and at lea s t a graphical demonstra-t i o n of the e f f e c t s transaction costs d i f f e r e n t i a l s have on international i n t e r e s t rate d i f f e r e n t i a l s can be given. Consider Figure 2-4. The same assumptions as those made e a r l i e r with regard to Figures 2-2a and 2-2b a p p l y . ^ In addi-t i o n , transaction costs d i f f e r e n t i a l s due to trading i n a f o r -53 -^Indeed, when considering t o t a l transaction costs rather than only cost d i f f e r e n t i a l s , t o t a l costs i n a foreign market may be lower i f the foreign market i s very e f f i c i e n t with re-gard to huge transactions. 54 J See pages 57 -59 above. -66-FIGURE 2-4 INTERNATIONAL CAPITAL MARKET EQUILIBRIUM FOR FORWARD LOANSo WITH VARIABLE TRANSACTION COSTS DIFFERENTIALSs MARKET IN COUNTRY D -6?-eign market are taken into account. Whenever foreign investors would be lenders at a given i n t e r e s t rate, t h e i r aggregate f o r -d' d' ward supply function i n country D's market (^ S^ . .^S^  ) l i e s to the l e f t of the excess demand function which would p r e v a i l i f there were no d i f f e r e n t i a l i n transaction costs. The new function w i l l be to the r i g h t of the former demand curve when-ever foreigners would borrow i n D at a given rate. This follows from the f a c t that the aggregate excess supply function of f o r -eigners becomes very i n e l a s t i c around the point where i t crosses the v e r t i c a l a x i s . The same considerations would ap-ply to the excess supply function of domestic traders i n count-ry F. Note that the equilibrium one-period forward rate i n d*' the domestic market, r n , i s higher than i n an i n t e r n a t i o n a l c a p i t a l market without d i f f e r e n t i a l s i n transaction costs be-tween trading i n the home and trading i n a foreign market. The foreign forward rate i s lower than i t would be otherwise. It follows that the size and d i r e c t i o n of international i n t e r e s t rate d i f f e r e n t i a l s does not only r e f l e c t exchange rate expectations and differences i n time preferences but i s also influenced by imperfections due to transaction costs which, f o r a given amount, are assumed to be higher when a domestic and a foreign trader deal with each other than when two trad-ers of the same n a t i o n a l i t y are involved i n a transaction. Our model also shows that, even i f no changes i n exchange rates are expected and i f time preferences are i d e n t i c a l i n two countries, a Central Bank's monetary p o l i c y may cause s l i g h t i n t e r e s t rate d i f f e r e n t i a l s without inducing considerable i n -ternational c a p i t a l flows because of these differences i n trans--68-action costs. But a Central Bank's continuing attempts to s h i f t a country's aggregate excess supply functions f o r for -ward funds i n the domestic market upward w i l l meet increasing resistance as foreigners s h i f t t h e i r funds i n increasing amounts from t h e i r home market to the country i n question, given that foreign i n t e r e s t rates do not increase as well, and domestic residents w i l l "be tempted to borrow abroad rather than at home. Of course, other imperfections l i k e discriminatory taxes against f o r e i g n e r s , ^ special reserve requirements f o r deposits by foreigners as introduced by some European Central Banks i n 1 9 7 2 , ^ or other exchange controls a l l w i l l impede the free market mechanism and lead to distorted international i n t e r e s t rate d i f f e r e n t i a l s . Indeed, the Eurobond market, which i s re-garded by many as a t r u l y international c a p i t a l market, and i n which foreign investors deal with each other i n a t h i r d curren-cy cy, owes i t s existence mainly to such imperfections.-" A f t e r the recent a b o l i t i o n of U.S. c a p i t a l r e s t r a i n t programmes we may see a re-emergence of New York as a leading international c a p i t a l market. -^Witness the Canadian withholding tax, American i n t e r e s t equalization tax or German coupon tax. ^See, e.g., the London Economist, January 2 7 , 1973 i "Inter-national Banking Survey", p. 14 on exchange controls introduced by European countries i n 1972 . -"See Mendelson. 2 . 6 SUMMARY In summary, several c r u c i a l conclusions can be derived from our analysis! 1 . In a world dominated by risk-averse investors, i n t e r -national differences i n i n t e r e s t rates do not only re-f l e c t exchange rate expectations but also r i s k premia necessary to reimburse in t e r n a t i o n a l borrowers and lenders f o r accepting exchange r i s k . 2 . The persistence of such i n t e r e s t d i f f e r e n t i a l s i s com-pati b l e with the assumption that the international c a p i t a l market f o r long-term debt c a p i t a l i s perfect. 3. In countries whose residents have a r e l a t i v e l y high time preference f o r a certa i n period and where there-fore the intere s t rate l e v e l i s r e l a t i v e l y high com-pared to the "world" l e v e l of i n t e r e s t rates f o r that period, the respective domestic forward i n t e r e s t rate w i l l be lower a f t e r the country opens up f o r i n t e r -national c a p i t a l investment. S i m i l a r l y , i n t e r e s t rates w i l l r i s e i n countries whose residents have a r e l a t i v e l y low time preference.^ 8 k. On a net basis, there w i l l be a one-way long-term debt c a p i t a l flow from the country with the r e l a t i v e l y low int e r e s t rate l e v e l f o r a given period to the country -*8In a world with growing economies, the d i r e c t i o n of i n t e r -est rate d i f f e r e n t i a l s w i l l also r e f l e c t differences i n the re-turn on r e a l c a p i t a l and differences i n the demand f o r funds r e s u l t i n g herefrom. -70-with a higher interest rate l e v e l . This net one-way c a p i t a l flow w i l l be due to a) residents of the high-interest l e v e l country borrow-ing abroad i n the low-interest market, and b) residents of the low-interest l e v e l country lending abroad i n the high-interest market. On a gross basis, two-way c a p i t a l flows may be caused by differences i n expectations among investors. As there does not exist a world currency, the internation-a l c a p i t a l market ac t u a l l y consists of national sub-markets i n which domestic residents transact with foreigners.-^ 5. The amount lent or borrowed i n a foreign c a p i t a l mar-ket w i l l not only depend on a country's wealth posi-t i o n but also on i t s residents' degree of r i s k aver-sion and the exchange r i s k perceived. 6. I f information and transaction costs are higher when a domestic and a foreign trader deal with each other than when two traders of the same n a t i o n a l i t y are i n -volved i n a transaction, then i n t e r e s t rate d i f f e r e n -t i a l s are lar g e r than they would be i f the internation-a l c a p i t a l market were perfect. -^The fact that foreigners transact with foreigners i n a t h i r d c a p i t a l market can be explained by p o l i t i c a l uncertainties at home, i n e f f i c i e n t domestic c a p i t a l markets and other imper-fections not considered i n our theory. CHAPTER 3 CORPORATE BORROWING IN INTERNATIONAL CAPITAL MARKETSs THE CANADIAN - UNITED STATES CASE 3.1 INTRODUCTION Canada i s a net importer of long-term debt c a p i t a l . The data i n Table 3-1 show that t h i s inflow of funds i s almost ex-c l u s i v e l y due to the sale of new issues abroad by Candian borrow-ers. Besides corporations, p r o v i n c i a l and municipal governments obtain considerable amounts of funds i n foreign bond markets. The p o l i c y of the Canadian federal goverment i s to approach foreign c a p i t a l markets only during exchange crise s l i k e those i n 1962 and 19&8 i n order to replenish foreign exchange re-serves. 1 As the data on outstanding bonds indicate, a c t i v i t i e s of international investors i n secondary markets are of only minor importance. Since 1966, t h i s has resulted i n an outflow of funds from Canada. Both the fact that the inflow of long-term debt c a p i t a l i s predominantly due to the sale of new bonds abroad by borrowers other than the federal government and the absence of an active "'"See Bank of Canada, Annual Report f o r the years 1962 and 1968. -71-TABLE 3-1 PORTFOLIO TRANSACTIONS IN CANADIAN AND UNITED STATES BONDS BETWEEN CANADA AND THE UNITED STATES, 1960 - 1972 (In m i l l i o n s of do l l a r s ) Item 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 Canadian Bonds Corporate bonds New issues 143 350 323 382 410 803 746 306 537 526 519 Trade i n outstanding bonds -25 10 5 2 2 -12 -15 -17 -14 -3 -9 Retirements -94 -119 -127 -94 -114 -208 -127 -135 -229 -169 -127 Net flow 24 241 201 290 298 583 604 154 294 354 383 (Memo: New issues Can.-pay ) (67) ( 46) ( 54) ( 75) (123) (172) ( 93) ( 80) ( 94) ( 80) ( 27) ( 48) (132) ( sold abroad foreign-pay) (87) (308) (277) (317) (291) (635) (658) (235) (492) (517) (513) (315) (263) A l l Canadian bonds New issues 370 455 676 922 1028 1200 1409 1239 1391 1502 1024 868 1023 Trade i n outstanding bonds -9 74 84 35 38 21 -72 -63 -67 -27 -69 -72 -4 Retirements -206 -189 -228 -268 -259 -324 -458 -300 -371 -382 -325 -581 -410 Net flow 155 440 532 689 807 897 879 876 953 1093 630 215 609 (Memo: New issues Can.-pay ) (123) (135) (126) (148) (226) (242) ( 168) ( 137) (. 135) ( no) ( 77) ( 114) ( 225) ( sold abroad foreign-pay) (299) (369) (583) (822) (852) (974) (1240) (1333) (1715) (1767) (1078) (1029) (1546) United States Bonds New issues -6 -13 -10 -32 -8 -9 -18 -17 -22 -9 -10 Trade i n outstanding bonds -8 10 1 23 -5 12 -35 -39 -18 1 -86 9 -2 Retirements 12 5 17 17 5 5 6 8 11 7 5 Net flow -2 2 8 8 -8 8 -47 -48 -29 -1 -91 -a i Source: S t a t i s t i c s Canada, Quarterly Estimates of the Canadian Balance of International Payments (Catalogue No. 67-001), various issues; Security  Transactions with Non-Residents (Catalogue No. 67-002), various issues; The Canadian Balance of International Payments, 1946-1965 (Catalogue No. 67-505). Missing data for 1971 and 1972 are not yet a v a i l a b l e -73-international trade i n outstanding issues make a d i r e c t test of our model through regression analysis rather d i f f i c u l t . Furthermore, the Bank of Canada has attempted to i n t e r f e r e with the free play of economic forces by manipulating int e r e s t rate d i f f e r e n t i a l s and through moral suasion. Therefore we s h a l l r e l y on an i n d i r e c t test of the basic features of our theory. In the preceding chapter i t was shown that the p e r s i s t -ence of in t e r e s t rate d i f f e r e n t i a l s between national long-term c a p i t a l markets may be consistent with equilibrium i n the i n t e r -national c a p i t a l market, given that c e r t a i n conditions hold. The three conditions which are c r u c i a l f o r the derivation of our results are that (1) because of differences i n time prefer-ences between nations and r e s u l t i n g differences i n the demand fo r funds i n t e r e s t rate l e v e l s tend to d i f f e r among countries, that (2) expectations of exchange rate changes and exchange r i s k cause inte r e s t rate d i f f e r e n t i a l s to p e r s i s t , even i f the i n -ternational c a p i t a l market i s perfect, and that (3) d i f f e r -ences i n transaction and information costs between operating i n the domestic and a foreign market w i l l further reduce the i n t e r e s t - e q u i l i b r a t i n g influence of i n t e r n a t i o n a l long-term c a p i t a l flows. Now we w i l l turn to a more detailed discussion of these three points. Data taken from Canadian-United States experi-ence w i l l be presented as preliminary evidence i n support of See our discussion i n Sections 3«24 and 5'7 below. -7k-our assertions. Their implications for corporate borrowing i n foreign c a p i t a l markets, i n p a r t i c u l a r f o r Canadian corporate debt issues i n the United States, w i l l be discussed. Hypotheses w i l l be derived that allow us to test the theory developed. Information on Canadian corporate bonds sold to United States investors w i l l be used i n the empirical tests to follow i n Chapter 5' We believe that corporations are more sensitive to economic forces than p r o v i n c i a l or municipal governments which borrow heavily i n foreign c a p i t a l markets as well. P o l i t i c a l p r i o r i t i e s and budget requirements may have a considerable i n -fluence on the borrowing behaviour of the two l a t t e r groups. In t h i s chapter, we s h a l l f i r s t present some evidence on the relationship between interest d i f f e r e n t i a l s and exchange rate expectations and on the r e l a t i v e l y higher demand f o r funds i n Canada. The implications of differences i n time preferences between the two countries f o r corporate borrowing i n the United States are discussed. It w i l l be shown that, at times, the Bank of Canada has exerted considerable influence on Canadian-United States interest d i f f e r e n t i a l s . Next, ways to reduce exchange r i s k i n international long-term borrowing are analyzed. F i n a l l y , the impact of information and transaction costs on international c a p i t a l flows w i l l be explored. -75-3.2 THE CANADIAN-UNITED STATES INTEREST RATE DIFFERENTIAL, EXCHANGE RATE EXPECTATIONS AND DIFFERENCES IN THE AGGRE-GATE DEMAND FOR FUNDS BETWEEN CANADA AND THE UNITED STATES 3.21 CANADIAN-UNITED STATES INTEREST RATE DIFFERENTIALS AND EXCHANGE RATE BEHAVIOUR In Chapter 2 i t was shown that observed inte r e s t rate d i f -f e r e n t i a l s between two countries should r e f l e c t exchange rate expectations held by market p a r t i c i p a n t s . Porter, i n his ar-t i c l e on "The Term Structure of Exchange Rate Expectations" reports empirical tests of t h i s hypothesis which were not very 3 successful. He regressed exchange rate r a t i o s on y i e l d r a t i o s of Canadian and U.S. government s e c u r i t i e s f o r the period 1953 to I960. We experimented with s i m i l a r data f o r the period 1962 to 1973 and also obtained i n s i g n i f i c a n t estimates or c o e f f i c i e n t s with wrong signs. One explanation f o r these discouraging results could be that observed exchange rate changes do not t r u l y r e f l e c t expect-ed exchange rate changes. As i s well known, exchange rate move-ments often r e f l e c t p o l i t i c a l factors rather than purely eco-nomic forces. Another reason could be that the i n t e r e s t rate d i f f e r e n t i a l on federal government s e c u r i t i e s i s not a good i n -dicator of the y i e l d d i f f e r e n t i a l to which c a p i t a l flows re-spond. The data i n Table 3-1 showed that long-term debt c a p i t a l flows between Canada and the United States are almost exclus-i v e l y due to the sale of new issues to Americans, and the y i e l d Porter, pp. 633-636. -76-d i f f e r e n t i a l on new issues rather than that on seasoned bonds may be the relevant v a r i a b l e . For the United States, Moody's Investors' Service computes a composite average of y i e l d s on newly issued corporate bonds. For Canada, no d i r e c t l y comparable index i s a v a i l a b l e . But, as Peters reports, there seems not to exist any systematic f a c t -or influencing the i n t e r e s t d i f f e r e n t i a l between newly issued and seasoned Canadian corporate bonds.^ Consequently the d i f -f e r e n t i a l between Moody's index and McLeod, Young, Weir and Co.'s "10 I n d u s t r i a l Bonds Y i e l d Average"^ should provide us with an adequate proxy f o r the relevant i n t e r e s t d i f f e r e n t i a l . We regressed quarterly rates of change i n the exchange rate, C^ ., on past values of t h i s y i e l d d i f f e r e n t i a l , D^^, and ob-tained the following r e s u l t f o r the period fourth quarter 1963 7 to second quarter 1973« See Moody's In d u s t r i a l Manual or Moody's Bond Survey. ~*J. Ross Peters, Economics of the Canadian Corporate Bond  Market (Montreal! McGill-Queen's University Press, 1971), pp. B8-95. ^See McLeod, Young, Weir & Co., 40 Bond Monthly Average or the Bank of Canada Review. 'The quarterly y i e l d d i f f e r e n t i a l i s an average of month-end y i e l d d i f f e r e n t i a l s . When using monthly data almost i d e n t i -c a l r e s u l t s are obtained but estimates of the d i s t r i b u t e d l a g structure are less precise. The time period was chosen such that none of the lagged observations f a l l s into the second quarter of 1962 or e a r l i e r when the Canadian government forced a devaluation of the Canadian d o l l a r . Values i n brackets are standard errors. 0.215 2. Q.60 1.012 1.83 b P = 0.59 (0.61) . (0.20) - 0 . 2 1 OJ, = -0.5© (0.24) * (0 .32) - 0 . 5 1 (0.25) The di s t r i b u t e d l ag structure was constrained to a second degree polynomial and we imposed the r e s t r i c t i o n that i t assume a zero value at a lag of six periods. Experiments with higher degree polynomials and longer lags did not lead to improved estimates. Estimates covering the period from 1951 "to 1973 or only those per-iods when the Canadian d o l l a r was free to f l o a t , that i s , 1951 to 1962 and 1970 to 1973, were not very s i g n i f i c a n t though the signs on the c o e f f i c i e n t s were as expected. For the most recent period of a f l o a t i n g exchange rate, t h i r d quarter 1970 to second quarter 1973, the following r e s u l t was obtained: C, = -1 .75 + 4.1? D. , - 1.8 7 D. , R 2 = 0.434 t (0 .90) (1.62) t _ 1 (1 .63) t _ 2 DWS = 1.547 SER = 1.070 These re s u l t s are surpr i s i n g l y good and may be taken as evid-ence that the international c a p i t a l market has become more ef-f i c i e n t and sophisticated since the early ' s i x t i e s . The nega-Q For the period t h i r d quarter 1951 "to second quarter 1973, the only r e s u l t s i g n i f i c a n t by usual standards was as follows: C, = -1.40 + 1.5 8 D. ,, R 2 = 0.051 DWS = 1.730 SER = 1 .624. z (0.63) (0.73) t _ i -77-C. = - 1 .20 + b .D, . t (0 .62) o=l J " ^ J DWS SER l b . = 1.12 0=1 J (0 .68) b 3 = b. = t i v e constant r e f l e c t s the exchange r i s k premium required by international investors. Even i f the dependent variable, the (expected) exchange rate change, i s zero the i n t e r e s t rate d i f f e r e n t i a l i s p o s i t i v e . Our s t a t i s t i c a l r e s u l t s indicate that i n t e r e s t d i f f e r e n -t i a l s on long-term bonds r e f l e c t short-term and medium-term exchange rate expectations rather than long-run a n t i c i p a t i o n s . This i s not very surprising because (1) the impact which a given exchange rate change has on the p r o f i t a b i l i t y of an i n -ternational long-term c a p i t a l transaction i s the greater the Q e a r l i e r i t occurs; 7 (2) predicting exchange rate changes, say, f i f t e e n years hence i s extremely d i f f i c u l t ; and (3) i t i s widely believed that i n the long run the Canadian-U.S. d o l l a r exchange rate w i l l fluctuate around an equilibrium value of p a r i t y . 1 0 H i s t o r i c a l exchange rate data as shown i n Table 3-2 f o r the period 192? to 1972 lend strong support to such a b e l i e f . 7 C f . our discussion i n Section 3.3 below. 1 0 F u l l e r t o n remarks that "when the Canadian d o l l a r was worth substantially more than the U.S. d o l l a r , some U.S. buyers used par as a basis f o r exchange reserve calculations, with the gen-e r a l idea that p a r i t y i s a long term average f o r the rate". Douglas H. F u l l e r t o n , The Bond Market i n Canada (Torontoi Cars-well Co., 1962), p. 53" And Kindleberger notes that long-term investors "held the view that the Canadian d o l l a r could not get very f a r from the United States d o l l a r over the l i f e t i m e of a 15 to 20-year investment, so that a one per cent d i f f e r e n t i a l i n interest rates could not be discouraged by exchange r i s k " . Charles P. Kindleberger, Balance-of-Payments D e f i c i t s and the  International Market fo r L i q u i d i t y (Essays i n International Finance, No. 46; Princeton, New Jerseyt International Finance Section, Department of Economics, Princeton University, May 1965), p. 18. See also the discussion i n Freedman, pp. 117-118. -79-TABLE 3-2 UNITED STATES DOLLAR EXCHANGE RATE IN CANADA, 1927-1972 Year High Low Average 1927 100 3/16?? 99 13/16 100 1/8 1928 100 7/169S 99 3/4 100 5 / 6 4 1929 103 4 100 1/8 100 4 7 / 6 4 1930 101 9/3296 99 27/32 100 5/32 1931 1 2 4 7/8 4 99 31/32 1 0 4 1 3 / 6 4 1932 119 1/2 4 106 5/8 113 3 3 / 6 4 1933 - 123 4 95 1/2 109 2 7 / 6 4 1934 101 5/8 4 96 7/16 98 31/32 1 9 3 5 102 5/8 4 99 1/2 100 3 1 / 6 4 1936 100 11/164 99 100 1/16 1 9 3 7 100 5/164 99 3 / 4 99 6 3 / 6 4 1938 103 1/2 4 99 5 9 / 6 4 103 7 / 6 4 1939 112 % 99 6 3 / 6 4 — 1 9 4 0 - 4 5 111 4 110 — 1 9 4 6 110 1/2 4 100 — 1947 100 1/2 4 100 — 1 9 4 8 100 1/2 4 100 --1 9 4 9 110 1/2 4 100 1 / 4 --1 9 5 0 110 1/2 4 103 --1951 107 5/164 101 3/16 105 1 / 4 1952 101 1/8 4 95 7/8 9 7 7/8 1 9 5 3 . 99 25/324 96 3 / 4 98 5/16 1 9 5 ^ 98 3/4 H> 96 11/32 9 7 5/16 1 9 5 5 100 1/164 96 15/32 98 5/8 1956 99 31/324 9£ 21/32 98 13/32 1 9 5 7 98 5/8 % 94 7/32 9 5 7/8 1958 99 5/324 9 £ 3 / 4 9 7 1/16 1959 98 3/164 13/164 94 9/16 95 29/32 I960 99 9 4 15/16 96 31/32 1961 1 0 4 3/8 4 4 98 1 / 4 101 5/16 1962 109 1 0 4 11/32 106 7/8 1963 1 0 8 9/164 107 1 9 / 3 2 107 27/32 1 9 6 4 1 0 8 1 / 4 % 107 1 / 4 107 7/8 1 9 6 5 1 0 8 1/2 4 1 0 7 5/16 1 0 7 13/16 1 9 6 6 1 0 8 "13/324 107 11/32 107 3 / 4 1 9 6 7 1 0 8 11/324 107 1 / 4 107 7/8 1 9 6 8 109 4 7/324 107 1 / 4 107 3 / 4 1 9 6 9 1 0 8 107 1 / 4 107 11/16 1970 107 15/324 100 5/16 1 0 4 13/32 1971 102 17/324 15/164 99 5/16 100 31/32 1972 100 97 13/32 99 1/32 Note 1 From September l 6 t h , 1939, to September 30th, 1950, fixed rates set by the Foreign Exchange Control Board were in effect in Canada. Free market trading was resumed on October 2nd, 1950. Source 1 Bank of Montreal, Foreign Exchange Rates 1972 (Montreal, 1973). - 8 0 -TABLE 3-3 EXCHANGE RATE CHANGES AND MEAN ANNUAL YIELD SPREADS BETWEEN CANADA AND THE UNITED STATES, I960 - 1972 YEAR ANNUAL EXCHANGE RATE CHANGE IN PER CENT MEAN OF MONTH-END YIELD SPREADS GOVERNMENT SECURITIES LONG-TERM CORPORATE BONDS 3-MONTH BILLS LONG-TERM BONDS I960 4.47 0.32 1.28 0.84 1961 4.75 0.42 1.18 0.84 1962 3-33 1.25 1.20 0.99 1963 0.29 0.35 1.08 O.96 1964 -0.63 0.14 1.03 0.95 1965 0.09 -0.04 1.00 1.00 1966 0.82 0.08 1.04 0.84 1967 -0.26 0.23 1.09 1.12 1968 -0.75 0.84 1.49 1.21 1969 0.03 0.35 1.47 0.87 1970 -5.96 -0.38 1.33 0.3^ 1971 -0.78 -0.89 1.25 0.58 1972 -0.60 -0.68 1.18 0.84 Note: The y i e l d spread on long-term corporate bonds i s computed as the difference between McLeod, Young, Weir & Co.'s ten i n -d u s t r i a l bond series and Moody's composite average of y i e l d s on newly issued corporate bonds. Sources: Bank of Canada, Bank of Canada Review, and Bank of Canada  S t a t i s t i c a l Summary, various issues? Moody's Investors' Service, Municipal & Government Manual 1973, and Indu s t r i a l Manual 1973. -81-P a r t i c u l a r l y because of t h i s l a s t point one might expect y i e l d d i f f e r e n t i a l s on long-term s e c u r i t i e s to r e f l e c t a general ex-change r i s k premium rather than s p e c i f i c expectations about ex-change rate changes many years hence. This leads us to the question of what i s the basic cause f o r the continuous flow of long-term c a p i t a l into Canada^"1" (cf. Table 3-1) and a p o s i t i v e interest rate d i f f e r e n t i a l (cf. Table 3-3) independent of whether the Canadian d o l l a r i s generally ex-pected to appreciate or depreciate. From 1962 to 1968, when the American d o l l a r was at a considerable premium i n Canada, the interest rate d i f f e r e n t i a l should have been smaller than during e a r l i e r or l a t e r years or even negative i f i t were sole-l y a function of exchange rate expectations. 3.22 DIFFERENCES IN THE RELATIVE DEMAND FOR FUNDS BETWEEN CANADA AND THE UNITED STATES In Chapter 2 i t was shown that a trader's time preference determines whether he plans to borrow or lend at a given future one-period spot interest rate. The higher an investor's time preference f o r a p a r t i c u l a r period, the more he w i l l tend to be a debtor rather than a creditor during that period and the more his excess supply functions f o r both spot and forward loans 12 as drawn i n Figure 2-1 w i l l move upwards. S i m i l a r l y , i f one "^Complete data on long-term debt .capital flows be.t-we en .Cana-da and the United States are only available f o r 1952 and l a t e r years. 1955 w a s "the only year when there was a s l i g h t outflow of long-term debt c a p i t a l into the United States. For data sources see Table 3-1. 1 2See Section 2.21 above, pp. 24-26. -82-country's time preference i s higher than that of another country, t h i s w i l l cause her aggregate supply functions to be at a higher l e v e l than those of the other country as demonstrated i n Figures 2-2a and 2-2b. The reason f o r t h i s i s that at any given interest rate the former country's ( r e l a t i v e ) excess de-mand f o r funds i s higher (her excess supply of funds i s lower) 1 3 than that of the l a t t e r . J If investors are risk-averse and ex-change rates can change, then e f f e c t i v e i n t e r e s t rates i n one country w i l l consistently be higher than i n the other, and long-term c a p i t a l w i l l tend to flow i n only one d i r e c t i o n even i f the international c a p i t a l market i s i n equilibrium. In growing economies, differences i n e f f e c t i v e i n t e r e s t rates w i l l not only r e f l e c t differences i n l i q u i d i t y preferences but also d i f -14 ferences i n the expected return on new c a p i t a l investments. The t o t a l demand f o r funds i n a country w i l l then be determined by both these factors. Assuming that i n d i v i d u a l investors' time preferences f o r funds are a function of both t h e i r con-sumption plans and the expected return on investment i n r e a l assets, i t i s not necessary i n our t h e o r e t i c a l model to i n t r o -duce corporations as separate traders i n the c a p i t a l market. In perfect c a p i t a l markets, because optimal production decisions 1 3See Section 2.423 above, pp . 57 - 6 l . 1 4 Differences i n the e f f i c i e n c y of f i n a n c i a l intermediaries may also be of influence. We s h a l l not pursue t h i s point here because Neufeld has quite f o r c e f u l l y argued that the Canadian c a p i t a l market i s as e f f i c i e n t as the American market. See Edward P. Neufeld, "The Relative E f f i c i e n c y of the Canadian Capital Marketi The Consequences f o r Canadian-United States Fin a n c i a l Relations", i n Canadian-United States F i n a n c i a l Re-lationships (The Federal Reserve Bank of Boston Conference Series No. 6; Boston, Mass.t Federal Reserve Bank of Boston, 1 9 7 1 ) , PP. 1 0 0 - 1 1 5 . - 8 3 -are independent of owner tastes, corporate demand f o r funds can be regarded as demand f o r funds by i n d i v i d u a l owners even though they may not be a c t i v e l y involved i n the decision-making process. 1-* Kindleberger i s perhaps the best-known proponent of the view that inte r e s t rate d i f f e r e n t i a l s and international c a p i t a l move-ments are caused by differences i n l i q u i d i t y preferences. He has asserted that "much, perhaps most, of the lending by the United States to Europe, and perhaps a t h i r d to a h a l f of Uni-ted States lending to Canada and Japan, serve . . . i n an over-a l l economic sense to provide l i q u i d i t y . " 1 ^ A l i b e r has c r i t i -cized t h i s view because i t neglects the influence anticipated exchange rate changes have on the rate of return investors expect to r e a l i z e on s e c u r i t i e s denominated i n d i f f e r e n t num-eraires . He suggests that "the spread between y i e l d curves denominated i n various currencies r e f l e c t s the market's apprais-1 7 a l of exchange r i s k " . As our t h e o r e t i c a l analysis of the de-terminants of the international term structure of intere s t rates has demonstrated, both these factors are important f o r an ex-planation of y i e l d d i f f e r e n t i a l s and international c a p i t a l flows. In 1961 the Bank of Canada has argued that "the basic reason why intere s t rates are lower i n the United States than ^On the separation of production decisions and ownership see Fama and M i l l e r , Chapters 2 and 4 . ^ C h a r l e s P. Kindleberger, "Balance-of-Payments D e f i c i t s and the International Market f o r L i q u i d i t y " , p. 7 . See also Emile Despres, Charles P. Kindleberger, and Walter S. Salant, "The Dollar and World L i q u i d i t y " , The Economist, February 5 , 1 9 6 6 , pp. 5 2 6 - 5 2 9 . •^Robert Z. A l i b e r , "Exchange Risk, Y i e l d Curves, and the Pattern of Cap i t a l Flows", Journal of Finance, XXIV (May, 1 9 6 9 ) , pp. 3 6 1 - 7 0 . See also his paper on "Uncertainty, Currency Areas -84-i n Canada i s that the t o t a l l e v e l of borrowing, the aggregate demand f o r funds by governments, business and indiv i d u a l s com-bined, i s less (proportionately) i n the United States than i n 18 Canada." To check on t h i s claim we calculated the t o t a l 19 20 amount of funds 7 raised by non-financial sectors as a per-centage of gross national product i n Canada and the United States f o r the period 1962 to 1 9 7 2 . These data are presented i n Table 3 - 4 . They show that the r e l a t i v e demand f o r funds has indeed been consistently higher i n Canada than i n the Uni-ted States. Whereas the mean value f o r Canada was 1 3 « 9 5 during those eleven years, i t was only 1 0 . 8 5 f o r the United States. In other words, the r e l a t i v e demand f o r funds was, on average, almost 30 per cent higher i n Canada. We also calculated the net amount of Canadian bonds acquired by American investors as a percentage of the t o t a l funds raised by Canadian non-financial sectors during the same period. These figures are shown i n Table 3 - 5 * On average, inflows of long-term debt c a p i t a l from the United States accounted f o r approximately ten per cent of the t o t a l amount of funds raised by Canadians during most of t h i s period though a clear tendency towards lower values devel-and the Exchange System", Economica, XXIX (N.S.) (November, 1 9 7 2 ) , pp. 432-441. 18 Bank of Canada, Annual Report i 9 6 0 , p. 1 9 • 19 ^Consumer c r e d i t , bank and other loans, short-term paper, mortgages, bonds, and stocks. 20 Households, non-financial business, federal, p r o v i n c i a l (state) and municipal governments, and rest of the world. Be-cause the United States i s a net lender to the r e s t of the world, the U.S. data s l i g h t l y overstate the demand f o r funds by domestic sectors. But i n t e r e s t rates are a function of the t o t a l demand f o r funds, not domestic demand. TABLE 3-4 TOTAL FUNDS RAISED BY NON-FINANCIAL SECTORS AS A PERCENTAGE OF GROSS NATIONAL PRODUCT IN CANADA AND THE UNITED STATES, 1962 - 1972 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 Canada U n i t e d S t a t e s 11.9 9.7 12.3 9.9 13-5 10.7 13.7 10.3 13-8 9.2 15.2 10.5 15-1 11.0 13.0 9.8 11.2 10.0 • 17.0 13.9 16.8 14.4 Source» Based on data i n S t a t i s t i c s Canada, F i n a n c i a l Flow Accounts (Catalogue No. 13-002) and Income and Expenditure Accounts (Catalogue No. 13-001), v a r i o u s i s s u e s ; Board of Governors, F e d e r a l Reserve System, F e d e r a l Reserve B u l l e t i n , v a r i o u s i s s u e s . The Canadian data f o r 1969-1972 are taken d i r e c t l y from F i n a n c i a l Flow Accounts. For 196l and e a r l i e r y e a r s , no Canadian data were a v a i l a b l e . TABLE 3 - 5 NET AMOUNT OP CANADIAN BONDS ACQUIRED BY U.S. INVESTORS AS A PERCENTAGE OF TOTAL FUNDS RAISED BY CANADIAN NON-FINANCIAL SECTORS, 1 9 6 2 - 1 9 7 2 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1 0 . 5 - 1 2 . 5 . 1 2 . 1 1 1 . 9 . 1 0 . 4 8 . 8 . 8 . 8 1 0 . 5 • 6 . 6 I . 2 5 • 3 . 5 Sourcet See Tables 3 -1 and 3 - 5 . -87-oped i n recent years. Only the future can t e l l whether t h i s i s a fundamental trend i n d i c a t i n g a diminished reliance by Canadi-ans on American funds or whether i t i s only a temporary pheno-menon that w i l l reverse i t s e l f once the United States has solved i t s balance of payments problems. These data strongly support our assertion that differences i n time preferences between nations and r e s u l t i n g differences i n the demand f o r funds have a considerable influence on i n t e r -national c a p i t a l flows. However, whereas our theory predicts that, f o r example, c a p i t a l inflows into Canada should be due to both Americans acquiring Canadian-pay bonds i n Canada and Canadians s e l l i n g U.S.-pay bonds i n the United States, the data i n Table 3 -1 suggest that long-term debt c a p i t a l inflows are mostly due to Canadians borrowing abroad by issuing foreign-21 pay bonds. Several factors which have not been incorporated into our model can explain t h i s . (1) P r i o r to 1961, U.S.-pay issues were not subject to withholding taxes. This made them more appealing to United States investors and may have estab-22 23 l i s h e d a pattern. ' J (2) Canadian-pay issues "do not q u a l i f y 21 Data on the amount of foreign-pay bonds acquired by Ameri-can investors are not a v a i l a b l e . However, whereas Canadian pro-vinces have borrowed recently large amounts i n Europe and also i n Japan, we were unable to i d e n t i f y a single corporate issue that was denominated i n a foreign currency other than U.S. dol-l a r s or a U.S.-pay issue that was sold outside North America.. Consequently i t i s highly probable that the foreign-pay corpor-ate bonds mentioned i n Table 3 -1 as sold abroad were indeed placed almost exclusively i n the United States. 2 2Investment Dealers' Association of Canada, B r i e f to  Royal Commission on Banking and Finance (Toronto: Investment Dealers' Association of Canada, June 1962), Appendix M, "Non-Resident Investment". 2 3A- usually excellent discussion of American and Canadian - 8 8 -as l e g a l investments under the investment laws of many of the states" . This makes them less a t t r a c t i v e to American invest-ors and reduces t h e i r marketability at any point i n time. (3) A given inte r e s t rate d i f f e r e n t i a l provides Canadian com-panies paying income tax with a higher protection against un-favourable exchange rate changes when issuing U.S.-pay bonds than tax-free American i n s t i t u t i o n a l investors investing i n Canadian-pay bonds.2-* (4) Canadian issuers receiving revenues denominated i n U.S. do l l a r s can hedge the exchange rate r i s k whereas most United States investors have no debts denominated i n Canadian d o l l a r s . (5) United States insurance companies have to revalue foreign-pay bonds annually and must write o f f any 26 exchange losses immediately against surplus or other p r o f i t s whereas Canadian companies can write o f f exchange losses as they are r e a l i z e d over the l i f e of the bonds. tax regulations and United States investment guidelines as they apply to U.S. investment i n Canadian bonds can be found i n any public prospectus f o r a Canadian U.S.-pay issue. American i n -vestors can usually o f f s e t payments of Canadian withholding taxes against U.S. income tax l i a b i l i t i e s , and tax-free i n s t i t u -tions can request to be exempted from the Canadian withholding tax.. Douglas H. Full e r t o n , The Bond Market i n Canada (Torontot Carswell, 1962), p. 133. A summary of regulations of American f i n a n c i a l i n s t i t u t i o n s can be found i n U.S., 88th Congress, House of Representatives, Committee on Banking and Currency, Comparative Regulations of Fin a n c i a l I n s t i t u t i o n s (Washington, D.C.i U.S. Government Printing O f f i c e , 1963). Regulations con-cerning l i f e insurance companies* investments i n Canadian secur-i t i e s can be found on pp. 336-37• concerning investments i n other foreign countries on pp. 338-39. U.S. l i f e insurance companies are the main foreign investors i n Canadian corporate bonds. 2-*See our discussion i n Section 3«3;below. 2 6 F u l l e r t o n , p. 133. - 8 9 -In addition, there may exist systematic differences i n the evaluation of exchange rate r i s k between Canadian and American 27 traders, a phenomenon which can be incorporated into our 28 model. For example, i f U.S. investors' expectations of future exchange rate changes are more diffuse and uncertain than those of Canadians, the American aggregate excess supply function i n the Canadian market w i l l be steeper than the Canadian supply function i n the U.S. market ind i c a t i n g a r e l a t i v e l y higher de-mand by Canadians f o r U.S. d o l l a r loans than supply of Canadian d o l l a r loans by Americans. 3.23 DIFFERENCES IN TIME PREFERENCES AND CORPORATE BOND ISSUES IN THE UNITED STATES Hicks has reasoned that lenders, because of l i q u i d i t y pre-ference, prefer to invest i n shorter-term s e c u r i t i e s thereby causing a " c o n s t i t u t i o n a l weakness" on the long side of the 29 c a p i t a l market. 7 Following him i t i s sometimes argued that differences i n terms to maturity available on bonds i n two countries can be taken as an i n d i c a t i o n of differences i n time preferences. For Canada and the United States Neufeld has com-pared the average term to maturity of interest-bearing market-27 'For example, A l i b e r argues that "the explanation f o r an international bond market - why firms based i n one country issue debt denominated i n a foreign currency (why Canadian firms issue debt denominated i n the U.S. dollar) i s that these firms believe that lenders have over-priced exchange r i s k . " Robert Z. A l i b e r , "The Multinational Enterprise i n a Multiple Currency World", i n John H. Dunning (ed.), The Multinational Enterprise (Londont A l l e n & Unwin, 1971), p. 52. 28 See our discussion i n Section 2.42 above, p.46 and p.6l . 2Q 7Hicks, Chapters 11 and 13. -90-able federal government debt outstanding? 0 From 1946 to 1970 the average term to maturity has always been higher i n Canada with an average difference of about two years. However, such a comparison overlooks two things: (1) In Canada, federal govern-ment debt includes long-term obligations of government comp-anies l i k e Canadian National Railway and thereby increases the average maturity from what i t would be otherwise. (2 ) In the United States, the term to maturity of federal bond issues has been a r t i f i c i a l l y r e s t r i c t e d by a l e g a l c e i l i n g on inte r e s t rates the government i s allowed to pay and thus r e f l e c t s market forces at most very i n d i r e c t l y . Consequently Neufeld's r e s u l t s may not t r u l y r e f l e c t investors' preferences. A comparison of terms to maturity on corporate bonds i n the F i n a n c i a l Post's Record of New Issues and i n Moody's Bond Survey indeed strongly suggests that terms to maturity a v a i l -3 1 able i n the United States are longer than those i n Canada. J This may, however, r e f l e c t differences i n business character-i s t i c s p a r t i c u l a r l y with regard to r i s k rather than differences 32 i n time preferences. A more meaningful comparison would be 3 0 N e u f e l d , pp. 105-106. 3 1 J Soldofsky reports that i n 1955 "the term to maturity on private placements i n the United States varied from eight to twenty-seven years on most issues but that maturities of twenty-eight to thirty-two years were not uncommon, and maturities of up to forty-nine years were ava i l a b l e . A few issues had terms of one hundred years. Robert M. Soldofsky, "The Size and Matur-i t y of Direct Placement Loans", Journal of Finance, XV (March, I960), pp. 3 2 - 4 4 . More recent data or comparable Canadian data are not av a i l a b l e . ^ 2The r i s k i e r an enterprise i s the l e s s l i k e l y lenders w i l l concede long terms to maturity. -91-to compare the longest maturities obtainable by a Canadian firm i n the Canadian c a p i t a l market with those available to the same firm i n the United States market. However, i t i s d i f -f i c u l t to make such a comparison because most firms do not ap-proach both markets at the same time and firm c h a r a c t e r i s t i c s and market conditions change over time. S t i l l , i f differences i n time preferences between the two markets exemplify themselves i n differences i n terms to maturity, then we should expect that the following hypothesis holds t H - ^ J Consider only long-term bonds which may be defined as bonds having terms to maturity of more than twelve years. And compare only bonds that have been issued by Canadian corporations that have borrowed i n both Canada and the United States. Then the term to matur-i t y of U.S.-pay bonds w i l l be, on average, longer than the term to maturity of Canadian-pay bonds. Here H stands as an abbreviation f o r hypothesis. Hypotheses des-ignated i n t h i s manner w i l l be employed f o r additional empirical t e s t s of our international c a p i t a l market model. The r e s u l t s of these tests w i l l be reported i n Chapter 5> Companies presumably prefer longer to shorter maturities at least on part of t h e i r debt i f long-term l i a b i l i t i e s are re-garded as a permanent component of the c a p i t a l structure be-cause long maturities w i l l tend to minimize fi x e d transaction 33 costs and also the p o s s i b i l i t y of a " c r i s i s at maturity". ^ __ ^ F o r a discussion of t h i s phenomenon see Ramon E. Johnson, "Term Structure of Corporate Bond Yields as a Function of Risk of Default", Journal of Finance, XXII (May, 1 9 6 ? ) , pp. 313-^5. -92-In general, long-term debt i s considered as less r i s k y than short-term debt because the necessity to r o l l over large amounts of funds can put enormous strains on a firm's cash flow. As Gordon observes t h i s becomes e s p e c i a l l y important during times of f i n a n c i a l distresst "When i f ever a firm i n f i n a n c i a l d i s -tress goes into receivership depends i n part on the maturity 34 structure of i t s debt..." Consequently one would expect Canadian corporations to make some use of the longest terms to maturity available to them i n the American bond market i f these terms are indeed longer than the longest ones obtainable i n Canada. The a v a i l a b i l i t y of a private placement market may also provide us with some ind i c a t i o n as to whether Canadian invest-ors have higher time preferences than American investors. Direct placements have certain advantages over public issues f o r corporate borrowers because terms and provisions of the loan agreement can be t a i l o r e d to meet the firm's p a r t i c u l a r needs, because they provide considerably more f l e x i b i l i t y i n case tr u s t deed or other changes become necessary i n l a t e r years, and because they can often be arranged very f a s t and without any public exposure. In addition, transaction costs are usual-l y s i g n i f i c a n t l y lower. On the other hand, lenders may shy away from private placements as t h e i r marketability i s usually very l i m i t e d . A comparison of the percentage of new corporate M. J. Gordon, "Towards a Theory of F i n a n c i a l Distress", Journal of Finance, XXVI (May, 1 9 7 1 ) , pp. 3 4 7 - 5 6 . - 9 3 -bonds placed d i r e c t l y i n Canada and the United States shows that i n both countries t h i s figure declined from about f i f t y per cent and more during 196l to 19&6 to about t h i r t y per cent or less during the late ' s i x t i e s . J These data would not sug-gest any basic differences between the two markets. However, whereas Peters traces the decreased market share of private placements i n Canada to the increased "interest of i n s t i t u -t i o n a l investors i n the more marketable public o f f e r i n g s " 3 ^ Shapiro and Wolf argue that i n the United States "the percent-age of corporate debt sold i n the private market has declined dur-ing periods of rapid growth i n corporate borrowing and increas-ed during periods of modest growth" because the supply of funds to the private market has grown at a r e l a t i v e l y steady rate 37 without reacting to changes on the demand side. In f a c t , large American l i f e insurance corporations, the dominant buy-ers i n the dir e c t placement market, have progressively con-centrated t h e i r bond purchases i n t h i s market, but they had to c u r t a i l t h e i r acquisitions of new bonds because of sharp i n -creases i n p o l i c y loans during the late ' s i x t i e s . 3 8 To further inquire into such pot e n t i a l differences between the investment behaviour of Canadian and United States f i n a n c i a l i n s t i t u t i o n s 3^Peters, p. 16 and E l i Shapiro and Charles R. Wolf, The  Role of Private Placements i n Corporate Finance (Bostons Har-vard University, 1972), p. 112. •^Peters, p. 38. The e f f o r t s of Canadian i n s t i t u t i o n s to increase t h e i r p r o f i t through bond trading and t h e i r r e s u l t i n g interest i n l i q u i d bonds should also be noted. Cf. Fullerton, pp. 218-219, and Peters, pp. 100-101. 3 7 S h a p i r o and Wolf, p. 157. 3 8 I b i d . , p. 54. -94-i t i s therefore hypothesized that H 2s The comparative preference of Canadian versus Ameri-can f i n a n c i a l i n s t i t u t i o n s f o r bonds with a high degree of marketability has induced Canadian corporations to s e l l s e c u r i t i e s i n the U.S. direc t placement market. The a v a i l a b i l i t y of forward commitments can be a major fact o r a t t r a c t i n g corporate borrowers to the private placement 39 market. A forward commitment i s a firm agreement by a f i n -a n c i a l i n s t i t u t i o n to provide a s p e c i f i e d amount of funds at a sp e c i f i e d future date. Sometimes such agreements involve several closings that extend over many years. The intere s t rate i s usually agreed upon at the date the contract i s signed "and r e f l e c t s p r e v a i l i n g rather than anticipated future market condi-,.40 tions. In fa c t , forward funds are sometimes cheaper than 41 funds f o r immediate delivery. Forward commitments allow corporations to formulate investment plans with greater certain-ty and to arrange f o r funds to be available when they are actu-a l l y needed. U.S. l i f e insurance corporations usually commit more than f i f t y per cent and sometimes up to almost ninety per cent of t h e i r expected cash flows i n advance. About two-thirds •^However, since the middle ' s i x t i e s a few public offerings i n the United States have also contained delayed delivery pro-vi s i o n s . Ibid., p. 20 40 Ibid., pp. 20-21. This i s not surp r i s i n g i n view of the problems involved i n accurately forecasting int e r e s t rates. See Telser, pp. 562-564, or Michael J. P r e l l , "How Well do Experts Forecast Interest Rates?", Federal Reserve Bank of Kansas City Monthly Review, September-October, 1973, pp. 3-13. -Shapiro and Wolf, p. 25- See also Lawrence D. Jones, In-vestment P o l i c i e s of L i f e Insurance Companies (Boston: Harvard University, 1968), p. 328. -95-of the funds commited forward are taken down one to six months l a t e r , but time lags of up to twenty months and more have been 42 observed. For the Canadian market s i m i l a r data are not av a i l -able. Peters only notes that "the delayed take-down i s not a 43 c h a r a c t e r i s t i c of the majority of private placements." J If indeed Canadian investors are very liquidity-conscious and pre-f e r s t a b i l i t y of p r i n c i p a l to s t a b i l i t y of income, then i t i s not surprising that forward commitments are le s s prevalent i n the Canadian than i n the American bond market, and i t i s there-fore hypothesized that H ^ J The a v a i l a b i l i t y of forward commitments has been a major fac t o r a t t r a c t i n g Canadian corporations to bor-row i n the United States. 3.24 THE BANK OF CANADA'S INFLUENCE OJN INTEREST RATE DIFFERENTIALS When we developed our theory of the determinants of the i n -ternational term structure of inte r e s t rates, i t was assumed that a l l traders are price takers. However, at least one trader i n the Canadian market, the Bank of Canada, can influence "prices" to a s i g n i f i c a n t extent. Therefore i t s a c t i v i t i e s and i t s i n -fluence on inte r e s t rate l e v e l s i n Canada and on the time pre-ferences of Canadians as expressed by thev term structure of in t e r e s t rates should not be overlooked. During the ' f i f t i e s when the Canadian exchange rate was allowed to fluctuate f r e e l y Ibid., pp. 356, 380 and Shapiro and Wolf, pp. 24, 168-170. ^ P e t e r s , p. 3 1 . For only one year, 1966, he estimates that 15$ of a l l private placements accounting f o r 37.8$ of the volume, of such issues involved delayed take-downs. -96-there did not exist any p a r t i c u l a r need to influence the d i f -f e r e n t i a l between Canadian and U.S. i n t e r e s t rates f o r balance-of-payments purposes. In 1962 the Canadian d o l l a r was devalued and pegged. This caused an exchange c r i s i s which was accompan-ied by huge outflows of c a p i t a l . In order to restore confidence i n the Canadian d o l l a r , "central bank operations were according-l y directed toward promoting and maintaining a l e v e l of interest rates i n Canadian f i n a n c i a l markets which would help i n establish-ing a net inflow of c a p i t a l large enough to cover the current account d e f i c i t i n the balance of international payments and r e b u i l d the depleted foreign exchange reserves." In l a t e r years the Bank of Canada commented also on the need "to ensure that the d i f f e r e n t i a l of bond y i e l d s between Canada and the United States was adequate". J Only i n 1970, a f t e r the current account had s h i f t e d to a p o s i t i o n of substantial surplus and the Canadian d o l l a r was again f l o a t i n g , d i d "the need which had ex-i s t e d f o r many years to maintain interest rate l e v e l s high enough to a t t r a c t a net inflow of c a p i t a l to cover the current 4 6 account d e f i c i t " cease to e x i s t . C l e a r l y then the higher i n -t e r e s t rate d i f f e r e n t i a l s p r e v a i l i n g during the middle "sixties were at least p a r t l y due to the Bank of Canada's p o l i c y to main-t a i n interest rate d i f f e r e n t i a l s adequate f o r balance-of-pay-ments equilibrium. ^Bank of Canada, Annual Report 1962, p. 4. ^Bank of Canada, Annual Report 1966, p. 43. See also Bank of Canada, Annual Report 1968, pp. 35-37. Bank of Canada, Annual Report 1970, p. 6. -97-3.3 EXCHANGE RATE CHANGES AND THE COST OF FOREIGN BORROWING A unique c h a r a c t e r i s t i c of i n t e r n a t i o n a l f i n a n c i a l trans-actions i s that at least one of the two traders involved i s dealing with s e c u r i t i e s denominated i n a currency d i f f e r e n t from the one i n use i n his country. In order to evaluate the p r o f i t a b i l i t y of such a deal i t i s therefore necessary f o r him to take the e f f e c t of potential exchange rate changes into ac-count. On the following pages we s h a l l analyze the influence an adverse change i n the exchange rate w i l l have on the cost of foreign borrowing. This allows us to derive hypotheses con-cerning the f i n a n c i a l behaviour of Canadian corporations when s e l l i n g bonds i n the American c a p i t a l market. 3.31 HEDGING EXPORT EARNINGS THROUGH INTERNATIONAL BORROWING Before embarking on such a discussion, i t i s important to r e a l i z e that foreign borrowing does not e n t a i l an increase i n f i n a n c i a l r i s k f o r a l l corporations. Whenever a firm s e l l s part or a l l of i t s production abroad f o r foreign currencies i n a market where i t has no appreciable influence on the p r i c e , then l i a b i l i t i e s denominated i n these currencies w i l l decrease the company's exposure to exchange r i s k . Though in t e r e s t and 47 sinking fund payments occur only semi-annually, short-term investments of foreign currency revenue or swap agreements could bring about an almost complete elimination of exchange r i s k on part or a l l export income. It i s therefore hypothesized that 'These are the usual terms. There ex i s t cases where such payments are made on a monthly basis. -98-H^: Canadian corporations with U.S. d o l l a r revenue w i l l exhibit a higher tendency to issue U.S.-pay bonds than those firms not engaged i n export a c t i v i t i e s . Indeed, one might expect that firms which obtain a r e l a t i v e l y large and stable proportion of t h e i r t o t a l revenue from exports borrow abroad even i f the expected e f f e c t i v e cost of debt capi-t a l i s s l i g h t l y higher i n the foreign c a p i t a l market. The re-duction i n exchange r i s k should be a b i g enough incentive f o r such a transaction. In a sense the company would provide i t -s e l f with a forward exchange market extending over the l i f e of the loan i f we regard the semi-annual debt payments as forward sales of anticipated export income. 3 . 3 2 EXCHANGE RATE RISK AND THE COST ADVANTAGE OF LOWER INTEREST RATES ABROAD When comparing a domestic with a foreign borrowing oppor-tunity, a firm i s faced with the problem of how to evaluate two future payment streams which are denominated i n di f f e r e n t currencies, that i s , which cannot be compared d i r e c t l y . To arr i v e at a r a t i o n a l decision, i t i s necessary to form expect-ations about future exchange rates, to translate foreign currency debt payments into domestic currency at these rates and to com-pare the present value of a l l future payments to domestic and foreign lenders. For example, l e t us assume that a domestic firm needs X domestic do l l a r s f o r N periods. At a domestic i n t e r e s t rate of R^ the firm can s e l l enough bonds to raise the required a--99-mount. To obtain the same amount abroad, X = a X , i t would o f have to pay a periodic coupon of R^. Here a Q i s the present spot exchange rate, the value of one unit of foreign currency expressed i n terms of domestic money. Letting E(a n) denote the exchange rate expected to p r e v a i l at the end of the nth p e r i -od when the nth coupon becomes due, the present value of the foreign loan i n terms of domestic currency i s PV f = R * X f £ E ( a n ) / ( 1 + R^) n + [ E ( a N ) X f ] / ( l + R d ) N (3-1) Payments denominated i n foreign currency are converted into domestic money at the expected exchange rate and discounted at 48 the domestic cost of debt c a p i t a l . Assuming f o r the moment the firm to be r i s k - n e u t r a l , i t should borrow abroad whenever f f a QX > PV , that i s , when the present value of the payments to be made to foreign creditors i s lower than the value of the amount of domestic currency received. Cle a r l y , to a c t u a l l y perform such an analysis would be ex-tremely d i f f i c u l t . For example, i f the foreign-pay bond has a term to maturity o;f twenty years and ca r r i e s a semi-annual coupon, f o r t y future exchange rates would have to be estimated. A considerable s i m p l i f i c a t i o n can be achieved by assuming that Whether foreign payments are converted into domestic money and discounted at the domestic i n t e r e s t rate or whether domestic payments are translated into foreign currency and d i s -counted at the foreign int e r e s t rate i s immaterial f o r the out-come of the anal y s i s . One must not discount.cash flows denom-inated i n one currency at a discount rate applicable to c a p i t a l denominated i n another currency. -100-the exchange rate changes only once during the l i f e time of ILQ the loan. 7 Furthermore, rather than forming expectations about exchange rates d i r e c t l y , i t seems to be customary to evaluate the protection a given i n t e r e s t rate d i f f e r e n t i a l pro-vides against unfavourable exchange rate changes.^° If,the exchange rate remains unchanged from the present t i l l the end of the (M-l)th period and assumes a new value, a.jyj, thereafter, then expression (3-1) becomes-*1 PV f = a QR fX f X ^ l + R d ) ~ n + a MR fX f J: (1 + R d ) " n + a M X f ( l + R d ) _ N (3-2) Setting PV equal to X = a QX , equation (3-2) can be solved f o r a^, the new exchange rate necessary to eliminate the cost advan-tage of borrowing abroad at a lower i n t e r e s t rate. The most un-favourable case would be a depreciation of the domestic currency 49 7kt the other extreme, one may expect a more or less con-tinuous change i n the exchange rate i n the same d i r e c t i o n by X per cent per period. In t h i s case a comparison of domestic and foreign i n t e r e s t rates i s rather simple. Taking exchange losses (or gains) on i n t e r e s t payments and p r i n c i p a l into ac-count, a foreign i n t e r e s t rate R* i s equivalent to a domestic int e r e s t rate of R f + (1 + R f)E(Ct) where E(C-t) i s the (same) expected rate of change i n the exchange rate during each future period. This i s s i m i l a r to expression (2-4) i n Section 2.41 above. Obviously, i t i s not very r e a l i s t i c to assume a con-tinuous appreciation or depreciation of the Canadian d o l l a r v i s - a - v i s the U.S. d o l l a r . -*°See footnote 10 above. As noted e a r l i e r i n Section 2.421, international investors seem to be preoccupied with p o t e n t i a l exchange losses. Caves and Reuber, p. 40, have attempted to compute the "depreciation required to eliminate the advantage of foreign borrowing i n r e l a t i o n to i n t e r e s t - r a t e l e v e l and term to maturity." However, by defining the exchange rate as the "U.S. d o l l a r price of the Canadian d o l l a r " they a c t u a l l y com-puted the appreciation of the U.S. d o l l a r . ^ T o simplify the notation, the subscript N w i l l be dropped henceforth. -101-or an appreciation of the foreign currency immediately a f t e r the funds have been taken down. Then the f i r s t term on the f right-hand side of expression (3-2) drops out and, as PV = a QX , the new exchange rate that would n u l l i f y the in t e r e s t advantage of foreign borrowing i s N , L f • a M a /[ R fz ( i + R d r n + ( i + R d r N ] . 0 1 n=l Summing the geometric progression^ 2 and simplifying we obtain = a Q [(R f/R d) + [(R d - R f ) / R d ] ( l + RV 1*]"" 1. (3-3) Obviously the value of a M increases as the domestic interest rate, Rd, increases, and i t decreases as the foreign interest rate, R , increases. In other words, the larger the interest rate d i f f e r e n t i a l , R - R , the higher the value of a^ has to be i n order f o r the bond issuer to become i n d i f f e r e n t between domestic and foreign borrowing. Remember that an increase i n the exchange rate, a^, means a depreciation of the domestic cur-rency. A t y p i c a l graph of the relat i o n s h i p between the intere s t rate d i f f e r e n t i a l and the devaluation percentage that would eliminate the cost advantage of foreign borrowing i s presented i n Figure 3-1. I t should be noted that as the intere s t rate d i f f e r e n t i a l doubles, the protection against unfavourable ex-change rate changes more than doubles. D i f f e r e n t i a t i n g equation (3-3) with respect to R and 52 aM r 1 l = a r t (R f/R d) [i - i / d + R d ) N ] + i / d + R d ) N " / a o J 53 . [, . The percentage change i n the exchange rate i s [ ( a M - a.Q) 100. -102-F I G U R E 3-1 DEVALUATION PERCENTAGE REQUIRED TO OFFSET THE INTEREST ADVANTAGE OF FOREIGN BORROWING AS A FUNCTION OF THE INTEREST RATE DIFFERENTIAL •50 1 . 0 0 1 . 5 0 2 . 0 0 I n t e r e s t D i f f e r e n t i a l Note: The domestic i n t e r e s t r a t e i s assumed to be 4 . 5 r>cr cent, the term to _ m a t u r i t y to be twenty years, and the d e v a l u a t i o n to occur immediately a f t e r the bonds have been i s s u e d . -103--jFIGURE 3 - 2 DEVALUATION .PERCENTAGE.REQUIRED TO OFFSET A FIFTY BASIS POINTS INTEREST ADVANTAGE OF FOREIGN BORROWING AS A FUNCTION OF THE INTEREST RATE LEVEL Devaluation Percentage 3 ^ Note: J_ 1 0 Domestic I n t e r e s t Rate I t i s assumed that the term to mat u r i t y i s tv/entv years and that the devaluation occurs immediately a f t e r the bonds have been sued. -104-assuming the i n t e r e s t rate d i f f e r e n t i a l to remain constant, i t can be shown that f o r a p o s i t i v e i n t e r e s t rate d i f f e r e n t i a l d 54 a.„ w i l l decrease as R increases. This means that i f the i n -M terest rate d i f f e r e n t i a l remains constant, i t w i l l be less at-t r a c t i v e to borrow abroad the higher the general i n t e r e s t rate l e v e l i s because the less protection against a depreciation i s provided. This r e s u l t i s the same which was derived e a r l i e r i n Section 2.41 f o r one-period s e c u r i t i e s . ^ However, as Figure 3-2 demonstrates, the influence of the i n t e r e s t rate l e v e l on the protection provided by a given i n t e r e s t rate d i f f e r e n t i a l i s quite dramatic f o r a long-term bond. For example, as the interest rate doubles from 4.5 per cent to 9.0 per cent the de-preciation percentage required to eliminate a f i f t y basis points cost advantage of foreign borrowing decreases from 7.01 to 4.82 or by about 31 per cent. Of p a r t i c u l a r interest i s the influence the term to maturi-ty has on the r i s k involved i n foreign borrowing. Whereas the inte r e s t rate l e v e l and the d i f f e r e n t i a l are exogenous variables, a firm has considerable influence on the term of i t s bond issue. The p a r t i a l d i f f e r e n t i a l of equation (3-3) with respect to maturity, N, i s p o s i t i v e , i n d i c a t i n g that the unfavourable 5 \ s i t i s assumed that dR f/dR d = 1, da^/dRl = -a (R d - R f) [[(1 + R d ) N + 1 < - (1 + R d + N R d ) ] / [ R d 2 ( l + R d j N + 1 ] J ( A ) 2 where A denotes the right-hand side of equation (3-3) divided by a 0 . For R d - R f>0, da M/dR d< 0 because (1 + R d ) N + 1 > 1 + R d + NRd ^^See p. 35 above. 5 6 9 a M / a N = [ a Q ( R d - R f ) l n ( l + R d)] /JR d(l + R d ) N ( A ) ~ 2 ] > 0 i f (R d - R^)> 0. Again A denotes the right-hand side of equation (3-3) divided by a Q . -105-exchange rate change has to be l a r g e r the longer the term to maturity. As the term to maturity approaches i n f i n i t y , a^ approaches a constant valuei N ^ a M = a o ( R d / R f ) -For example, i f the foreign interest rate i s .05 and the domes-t i c i n t e r e s t rate .06, that i s , twenty per cent higher, then an immediate depreciation: of twenty per cent would eliminate the cost advantage on a perpetual foreign-pay bond. In figure 3-3 t h i s r e l a t i o n s h i p between maturity and protection against exchange r i s k i s i l l u s t r a t e d . So f a r we have assumed the worst case possible, an adverse change i n the exchange rate immediately a f t e r the foreign loan has been taken down. The same basis relationships demonstrated above hold i f the domestic currency r e t a i n s - i t s i n t ernational value f o r some time and depreciates only a few years a f t e r the bonds have been placed abroad. Of course, the l a t e r the ex-change rate change occurs, the higher the devaluation has to be to make foreign borrowing more expensive than a domestic loan. Most corporate bonds are sinking-fund bonds. Sinking fund requirements vary considerably, and whereas i n most cases f i f -ty per cent or more of a bond's p r i n c i p a l i s repayed through a sinking fund before f i n a l maturity, the balloon payment amounts to eighty per cent or more of the o r i g i n a l loan i n some instances. For sinking fund bonds i t holds also that the exchange r i s k i s lower the larger the inte r e s t rate d i f f e r e n t i a l , the lower the -106-FIGURE 3-3 DEVALUATION PERCENTAGE REQUIRED TO OFFSET A FIFTY BASIS POINTS INTEREST ADVANTAGE OF FOREIGN BORROWING AS A FUNCTION OF THE TERM TO MATURITY Devaluation Percentage In Years Note j I t i s assumed th a t the domestic i n t e r e s t r a t e i s 4.5 per cent and that the devaluation occurs immediately a f t e r the bonds have been issued. -107-FIGURE 3-4 DEVALUATION PERCENTAGE REQUIRED TO OFFSET A FIFTY BASIS POINTS INTEREST ADVANTAGE OF FOREIGN BORROWING AS A FUNCTION OF THE YEAR IN WHICH THE EXCHANGE RATE CHANGE OCCURS I I I I I o 5 1 0 1 5 Year I n Which The Exchange Rate Change Occurs Note J I t i s assumed t h a t s i n k i n g fund payments on S i n k i n g Fund  Bond A s t a r t s i x months a f t e r the date of issue and t h a t an equal amount i s paid semi-annually so as to r e t i r e the i s s u e w i t h the l a s t s i n k i n g fund payment. For S i n k i n g Fund Bond B payments s t a r t during the s i x t h year of i t s term, they are assumed to be equal i n s i z e and to r e t i r e 75.3 of the p r i n c i p a l before f i n a l m a t u r i t y . The domestic i n t e r e s t r a t e i s assumed to be 4 . 5 per cent and the bonds' term to maturity to be twenty y e a r s . -108-general l e v e l of inte r e s t rates, and the longer the term to maturity. However, the e a r l i e r sinking fund payments s t a r t and the la r g e r they are r e l a t i v e to the t o t a l amount borrowed, the lower i s the protection against unfavourable exchange rate changes during the early years of the term and the higher during l a t e r years. In Figure 3 - 4 we have graphed t h i s r e l a t i o n s h i p between exchange r i s k and year i n which the exchange rate change occurs f o r three d i f f e r e n t types of twenty year bondst a bond without sinking fund provisions, a bond where sinking fund pay-ments star t a f t e r f i v e years and seventy-five per cent of the p r i n c i p a l i s r e t i r e d before f i n a l maturity, and a bond where sinking fund payments s t a r t immediately and the f i n a l payment equals the semi-annual sinking fund payment. To enable us to graph the Figures presented above, we de-veloped a computer program f o r c a l c u l a t i n g the devaluation per-centages required to o f f s e t the inte r e s t advantage of foreign borrowing. In Table 3 - 6 such data are presented f o r sinking fund and non-sinking fund bonds, maturities of ten, twenty and t h i r t y years, and domestic interest rate l e v e l s of 9 . 0 and 4 . 5 per cent. The inte r e s t d i f f e r e n t i a l i s taken to be f i f t y basis p o i n t s . ^ Furthermore, d i f f e r e n t assumptions are made about the year i n which an adverse exchange rate change occurs. For example, i f a Canadian borrower has the option to issue a Cana-dian-pay bond with a 9 . 0 per cent coupon and 20 years to maturi-57 •^ 'On Canadian corporate bonds, a term of twenty years i s f a i r l y common. A domestic interest rate of nine per cent was chosen because i n late 1973 the Canadian corporate bond rate approached t h i s l e v e l . Casual observations on some Canadian-pay and U.S.-pay bonds issued by the same Canadian firm at the same point i n time strongly suggests that the relevant inte r e s t rate d i f f e r e n t i a l i s about f i f t y basis points. TABLE 3-6 DEVALUATION PERCENTAGES REQUIRED TO OFFSET A FIFTY BASIS POINTS INTEREST ADVANTAGE OF FOREIGN BORROWING TYPE OF MATURITY YEAR AT THE START OF WHICH DEVALUATION OCCURS BOND IN YEARS 10 15 20 25 1 3 5 30 DOMESTIC INTEREST RATE 9.0 PER CENT NO S.F. 10 3.36 3.99 4.73 7.22 WITH S.F. 10 1.98 2.94 4.67 43-05 NO S.F. 20 4.82 5-74 6.83 10.52 16.15 24.62 WITH S.F. 20 3.09 4.09 5.48 12.30 34.76 320.68 NO S.F. 30 5.44 6.48 7.72 11.95 18.47 28.46 43.67 66.59 WITH S.F. 30 3.79 4.83 6.20 11.86 24.03 53.79 150.64 1265.23 DOMESTIC INTEREST RATE 4.5 PER CENT NO S.F. 10 4.16 4.51 4.90 5-99 WITH S.F. 10 2.29 3.12 4.53 33-59 NO S.F. 20 7.01 7.62 8.29 10.21 12.54 15.33 WITH S.F. 20 3.99 4.83 5-92 10.67 24.19 179.31 NO S.F. 30 8.92 9.72 10.59 13.19 16.18 19.93 24.64 29.90 WITH S.F. 30 5-33 6.22 7-31 11.22 18.22 32.72 73.54 496.18 Notesj With r e s p e c t t o s i n k i n g fund (S.F.) bonds i t i s assumed that payments s t a r t a f t e r s i x months and t h a t an equal amount i s p a i d semi-annually so as to r e t i r e the bonds with the l a s t s i n k i n g fund payment. A l l i n t e r e s t and p r i n c i p a l payments are discounted semi-annually at the domestic i n t e r -e s t r a t e . -110-ty or to issue an 8 . 5 per cent bond denominated i n U.S. dollars with an i d e n t i c a l maturity, he should opt f o r the Canadian issue whenever he expects a devaluation of the Canadian d o l l a r by 6 . 8 3 per cent or more f i v e years hence or e a r l i e r . For a corporation paying a f i f t y per cent tax on i t s i n -come, the after-tax cost of debt c a p i t a l i s only 4 . 5 per cent i f i t can borrow at home at 9 « 0 per cent. Consequently the ex-change r i s k should be evaluated by using the after-tax cost as a discount rate.-* 8 Looking again at Table 3 - 6 , i f the choice i s between 30-year sinking fund bonds, a U.S.-pay issue i s pref-erable whenever i t i s expected that the Canadian d o l l a r w i l l be devalued only ten years from now or l a t e r and by less than 11.22 per cent. Obviously, tables l i k e the one presented above cannot answer the question whether a Canadian corporation should borrow i n the United States or not at any p a r t i c u l a r point i n time. Factors other than exchange r i s k may be of paramount importance. Never-theless, i f exchange r i s k i s mainly perceived as a po t e n t i a l loss due to an unfavourable change i n the exchange rate, data l i k e those i n Table 3 - 6 help to focus on the exchange r i s k i n -volved i n foreign borrowing. For example, a firm contemplating to f l o a t a foreign-pay issue should form some expectations about future exchange rates and can then use such data to derive an estimate of the p r o b a b i l i t y that foreign borrowing w i l l be more -^Though a f i f t y basis points i n t e r e s t advantage reduces to twenty-five points on an after-tax basis, the before-tax i n t e r -est rate d i f f e r e n t i a l i s s t i l l relevant f o r an evaluation of ex-change r i s k . For example, on a foreign-pay perpetual bond a twen-ty percent devaluation implies only a ten percent increase i n (interest) costs on an after-tax basis. - i n -expensive than a domestic loan. This p r o b a b i l i t y should become one of the major variables influencing the decision as to where to s e l l a new bond issue. What are the implications of our discussion f o r Canadian corporate borrowing i n the United States? Clearly, corpora-tions can influence two factors that have a bearing on exchange r i s k : the term to maturity and the schedule of sinking fund payments. It seems that Canadian borrowers (and American lend-ers) are very much concerned about the r i s k of an adverse change i n the exchange rate early during a bond's term.-^ If t h i s i s indeed true, then Canadian corporations should show a prefer-ence f o r U.S.-pay bonds which have long terms to maturity. As our e a r l i e r analysis has shown, exchange r i s k decreases as a bond's term to maturity increases. I f firms need medium-term funds, a Canadian-pay issue should be sold. To test whether Canadian corporations a c t u a l l y behave i n such a manner we hy-pothesize that H^: On average, the term to maturity of U.S.-pay bonds w i l l be longer than the term to maturity of Canadian-pay bonds. Whereas hypothesis above concerns only long-term bonds issued by those corporations that have sold debt instruments i n both -^See discussions i n the l i t e r a t u r e c i t e d above i n footnotes 10 and 50. -112-the Canadian and United States market, hypothesis i s aimed at a comparison of the term to maturity of a l l bonds issued by a l l Canadian corporations. Of course, i f corporations are r e l a t i v e l y c e r t a i n about future exchange rates and do not expect an adverse change during, say, the next ten years, then they should s e l l medium-term bonds i n the United States rather than long-term issues. However, we doubt that corporate managers would have much con-fidence i n long-range forecasts. Usually corporate bonds have sinking fund provisions. But the time elapsed between sale of the bonds and the f i r s t sinking fund payment and the size of i n d i v i d u a l payments r e l a -t i v e to the p r i n c i p a l vary considerably among issues of i d e n t i -c a l maturity. Consequently variations i n sinking fund pro-vis i o n s could also be used as a means to reduce the exchange r i s k involved i n borrowing abroad. It i s widely assumed that the Canadian-U.S. d o l l a r exchange rate w i l l fluctuate around a long-run value of one Canadian to one American d o l l a r . I f exchange rate expectations are diffuse and i f borrowers f i n d i t d i f f i c u l t to forecast whether an adverse exchange rate change w i l l be followed by a favourable one, whether the favour-able change w i l l materialize f i r s t , or whether no s i g n i f i c a n t change w i l l occur, then evenly d i s t r i b u t e d sinking fund pay-ments coupled with a long maturity should lead to an optimal -113-protection against exchange r i s k . Not only should exchange losses and gains offset each other to some extent over the term of the issue, but perhaps more importantly, a firm would avoid the considerable s t r a i n on i t s l i q u i d i t y that might be caused by an unexpected devaluation of the domestic currency shortly before a large balloon payment becomes due. To test f o r t h i s , we hypothesize that Hg» Sinking fund payments on U.S.-pay bonds w i l l be more evenly distributed over the term of the issue than those on Canadian-pay bonds. Again, i f corporate managers are r e l a t i v e l y c e r t a i n about t h e i r exchange rate expectations, than a d i f f e r e n t pattern of sinking fund payments may appear more favourable. It should be noted that sinking fund payments reduce the average maturity of an issue. But, as the data i n Table 3-6 showed, the protection against an early unfavourable exchange rate change increases f o r sinking fund bonds, too, as the term to maturity increases. -114-3.4 INFORMATION AND TRANSACTION COSTS IN AN INTERNATIONAL CAPITAL MARKET E a r l i e r i n Chapter 2 we argued that transaction and inform-atio n costs on international f i n a n c i a l transactions are higher than on comparable domestic deals. Now we s h a l l turn to a more detailed discussion of these costs as they a f f e c t borrowers and lenders and of the possible implications such transaction costs d i f f e r e n t i a l s may have f o r int e r n a t i o n a l long-term debt c a p i t a l flows. 3.41 TRANSACTION COSTS TO BORROWERS The term transaction costs w i l l be used as r e f e r r i n g to a l l costs incurred by a firm when issuing and serv i c i n g i t s bonds except f o r intere s t c o s t s . ^ Most of these costs are nonrecurring expenses incurred during the period when a new issue i s planned, negotiated, and delivered. The net present value of the funds borrowed, X can be defined as the d i f f e r -ence between gross proceeds, ^ ^ r o s s * a n ( * the present value of a l l transaction costs, T, or X . = - PV(T) net gross N where PV(T) = T / ( l + R) . More commonly, "net proceeds", n=0 n ^°See Section 2.5 above, pp.6l -69. ^ - I f bonds are not issued at par, the amortization of bond discounts or premia i s regarded as a f f e c t i n g i n t e r e s t costs. Ex-change losses are not taken to be part of transaction costs. They should be included into a comparative evaluation of domestic and foreign f i n a n c i a l opportunities through an exchange r i s k analysis. -115-Xj^p, are computed as ANP Agross -"-o where T Q includes only those expenses which have been incurred at time of issue, which can r e a d i l y be i d e n t i f i e d as being re-l a t e d to a p a r t i c u l a r issue, and f o r which actual disbursements have been made to persons outside the issuing corporation. Usu-a l l y T Q does not include any amount taking into account manage-ment time or other company resources spent on planning, negoti-ating, and s e l l i n g an issue and on disseminating information about the firm to p o t e n t i a l investors. Because of t h i s and because future transaction costs are neglected, net proceeds do not t r u l y r e f l e c t the present value of the funds obtained through a new issue but overstate i t s l i g h t l y . Because of transaction costs, the actual t o t a l cost of the funds to the borrower, when expressed i n percentage terms, i s always higher than the y i e l d return to lenders. It can be calculated as the i n t e r n a l rate of return at which the present value of future 62 inte r e s t and p r i n c i p a l payments equals net proceeds. The largest single expense item i s usually the compensa-t i o n paid to investment bankers f o r underwriting an issue or /To f o r a s s i s t i n g i n negotiating private placements. ^ Other trans-62 Of course, the use of X n e t would lead to a "correcter" re-s u l t f o r t h i s i n t e r n a l rate. Then a problem arises as to which rate to use f o r discounting future transaction costs. This ques-t i o n i s not of material importance to our ensuing discussion and w i l l not be pursued here. ^ E x t e n s i v e data on f l o t a t i o n expenses i n the U.S. can be found i n S e c u r i t i e s and Exchange Commission, Cost of F l o t a t i o n  of Corporate Securities 1951-1955 (Washington, D.C.: U.S. Gov-ernment P r i n t i n g O f f i c e , June, 1957). For some Canadian data see Peters, Chapter 4. I t seems that private issues i n Canada -116-action costs include some or a l l of the following itemst (1) P r i n t i n g and engraving of the bonds; (2) p r i n t i n g of r e g i s t r a -t i o n statements and prospectuses; (3) r e g i s t r a t i o n and stock exchange l i s t i n g fees; (4) taxes; (5) accounting fees; (6) l e -gal fees; (7) trustee's fees f o r handling i n t e r e s t payments and transfers, and f o r administering sinking funds or the redemption of the issue. Trustees' fees, p r i n t i n g and engraving costs, and l e g a l fees account f o r the major part of these "other ex-penses" on public issues, whereas on private placements l e g a l fees alone usually amount to f i f t y per cent or more of t o t a l 64 other expenses. In the United States, federal revenue stamps have also been a major cost item on large issues because they have been l e v i e d as a constant percentage of the amount borrow-ed. On the other hand, underwriting costs, trustee's fees, p r i n t i n g costs and some other expenses increase l e s s than pro-p o r t i o n a l l y with issue si z e , and l e g a l and auditors' fees are f i x e d rather than variable costs. Consequently t o t a l f l o t a t i o n costs generally decrease with issue size when expressed as a percentage of gross proceeds, and t h e i r impact on a firm's actual cost of debt c a p i t a l diminishes accordingly. Assuming f o r the moment that transaction costs are i d e n t i -c a l f o r domestic and foreign traders i n t h e i r respective home markets, there are s e v e r a l reasons why one would expect trans-are usually placed through an agent whereas i n the U.S. most private placements are sold d i r e c t l y without the a i d of an i n -termediary; c f . Shapiro and Wolf, p. 90. ^ C f . S e c u r i t i e s and Exchange Commission, pp. 551 69. Com-parable data f o r Canada are not a v a i l a b l e . -117-action costs to be higher when they operate i n a market which i s foreign to them. F i r s t of a l l , underwriting spreads may be higher because i t w i l l be more d i f f i c u l t to secure a market as lenders are less well informed about foreign economic, l e g a l . and p o l i t i c a l conditions and about p a r t i c u l a r foreign companies.^ Also, l e g a l fees, trustee's fees, and taxes may be higher be-cause of regulations p e c u l i a r to international transactions which 66 are of no relevance f o r domestic trading. For example, U.S. lenders and Canadian borrowers have to be concerned with the United States "Guidelines f o r Banks and Nonbank Fi n a n c i a l In-s t i t u t i o n s " with regard to t h e i r foreign lending a c t i v i t i e s as set out by the Board of Governers of the Federal Reserve System,^ the U.S. Internal Revenue Code's s t i p u l a t i o n s with regard to the inte r e s t equalization tax, Canadian withholding tax regulations, the Canada-United States Tax Convention, Cana-dian estate tax regulations, and the Estate Tax Treaty between the United States and Canada. In addition, when converting funds from one currency into another, transaction costs a r i s e . Having foreign debt outstanding can cause p e c u l i a r accounting 6 5 ^In Section 3.42 below we w i l l discuss the problem of i n -creased transaction costs to lenders. 66 A discussion of l e g a l and other aspects of foreign borrow-ing i n the United States during the l a t e ' f i f t i e s and early 'six-t i e s can be found i n Nathaniel Samuels, "The Investment Banking Background of Issuing and Marketing Foreign Securities i n the United States", International Financing and Investment, ed. John F. McDaniels (Dobbs Ferry, N.Y.t Published f o r the World Community Association of the Yale Law School by Oceana Publ., 1964), pp. 411-29, and i n "Legal Problems of Issuing and Market-ing Foreign Securities i n the United States", panel discussion, International Financing and Investment, pp. 4-30-60. 6 7 F e d e r a l Reserve B u l l e t i n , LI (March, 1965), pp. 371-376, and LVI (Jan., 1970), pp. 11-22. As of early 197^, these guide-l i n e s as well as the Interest Equalization Tax have been removed. -118-problems and makes i t necessary to account s p e c i f i c a l l y f o r 68 exchange gains and losses. The amount of time and money spent when negotiating a foreign issue i s probably higher because of longer t r a v e l and communication distances. I t may be more expensive to keep foreign investors informed about the firm's a c t i v i t i e s and performance than domestic investors. C l e a r l y then i t seems reasonable to expect transaction costs to be higher when traders of two d i f f e r e n t countries deal with each other. No data on underwriters' discounts and other transaction costs with regard to foreign corporate bond issues placed i n the U.S. market have been published. However, some information on the costs of f l o a t i n g foreign government bond issues i n the New York market i s av a i l a b l e . Nevin reports that besides underwriters, 0 discounts of 2.5 per cent to 3*5 per cent on issues varying i n size from U.S. $6 m i l l i o n to $35 m i l l i o n the issuer had to reimburse the underwriter f o r his expenses incur-red when "sounding out and securing a market" which usually amounted to $25,000 to $50,000. Other issuing expenses varied between $33i000 and $60,000, not counting expenses incurred on the borrowers*s side which "can be very considerable i n terms 69 of time and labour of senior o f f i c i a l s as well as of cash". 7 These data suggest that f l o t a t i o n costs on foreign bonds are 6R On such accounting problems see, f o r example, R. MacDon-ald Parkinson, Translation of Foreign Currencies, (Torontos The Canadian In s t i t u t e of Chartered Accountants, 1972). ^ E . Nevin, "Some Reflections on the New York New Issue Market", Qxford Economic Papers, XIII (N.S.) (February, 196l), pp. 84-102. His data on other expenses seem not to include trustees' or l e g a l fees, see p. 101. -119-perhaps twice as large as those incurred by domestic borrowers 70 i n the United States. If transaction costs are higher i n a foreign c a p i t a l mark-et, then the actual i n t e r e s t savings on foreign borrowing w i l l be higher the l a r g e r the bond issue at a given i n t e r e s t rate d i f f e r e n t i a l . To see t h i s , l e t us assume that t o t a l transaction costs f o r a domestic bond issue of $2 m i l l i o n amount to $100,000 or 5«0 per cent of gross proceeds. For a s i m i l a r foreign-pay issue, these costs are, say, s i x t y per cent higher or 8.0 per cent of gross proceeds. In order to eliminate t h i s disadvant-age of higher f l o t a t i o n costs abroad f o r a twenty-year bond and a domestic i n t e r e s t rate l e v e l of 8 per cent, the i n t e r e s t rate d i f f e r e n t i a l has to be approximately 31 basis points i n favour of the foreign market. This figure can be derived from Table 3-7 where we computed the interest rate d i f f e r e n t i a l required to offset the disadvantage of higher f l o t a t i o n costs abroad on a twenty year bond as a function of the domestic i n t e r e s t rate l e v e l and the d i f f e r e n t i a l i n f l o t a t i o n expenses expressed as a percentage of gross proceeds. Obviously, the required i n t e r -est rate d i f f e r e n t i a l increases as the i n t e r e s t l e v e l increases. On the other hand, i t w i l l be lower the longer the (average) maturity of an issue."''1 To continue with our example, l e t us 70 ' No comparable data are available f o r bonds issued by Ameri-can governments. However, f l o t a t i o n costs of a $6 m i l l i o n issue of Rhodesia & Nyasaland i n June, 1958 amounted to 4.8 per cent and those of a $35 m i l l i o n issue by the European Coal and Steel Community i n the same month to 2.8 per cent compared to about 2.0 per cent and 1.3 per cent respectively f o r an average domestic corporate bond issue of comparable s i z e . See Nevin, p. 87 and Securities and Exchange Commission, p. 37. 71 For an extensive analysis of transaction costs i n r e l a t i o n to maturity i n a closed economy see Burton G. Malkiel, The Term  Structure of Interest Rates (Princeton, N.J.jPrinceton University press, 1966), Chapter 5. -120-TABLE INTEREST RATE DIFFERENTIALS IN MARKET REQUIRED TO ELIMINATE FLOTATION COSTS ABROAD 3 - 7 FAVOUR OF THE FOREIGN CAPITAL THE DISADVANTAGE OF HIGHER ON A TWENTY-YEAR BOND D i f f e r e n t i a l In Total F l o t a t i o n Ex-Domestic Interest Rate penses As A Percen-tage of Proceeds 2 . 0 4 . 0 6 . 0 8 . 0 1 0 . 0 5 . 0 0 . 3 1 5 . 3 7 8 .446 . 5 4 2 . 608 4 . 0 0 . 2 5 0 . 3 0 1 . 3 5 5 . 4 1 5 .482 3 . 0 0 . 1 8 6 . 2 2 5 . 2 6 5 . 3 0 9 .'358 2 . 0 0 . 1 2 3 .149 . 1 7 6 . 2 0 5 . 238 1 . 0 0 . 061 . 0 7 4 . 0 8 7 . 1 0 2 .118 . 5 0 . 0 3 0 . 0 3 7 .043 . 0 5 0 . 058 . 2 5 . 0 1 5 . 0 1 8 . 0 2 1 . 0 2 5 . 0 2 9 Sources Derived from Bond Values Tables by int e r p o l a t i o n . -121-assume that on a domestic bond issue of $20 m i l l i o n t o t a l trans-action costs amount to $400,000 or 2.0 per cent of gross pro-ceeds. I f on a comparable foreign-pay bond issue expenses are f i f t y per cent higher or 3.0 per cent of gross proceeds, then already a 10 basis points i n t e r e s t d i f f e r e n t i a l w i l l elim-inate t h i s disadvantage of higher f l o t a t i o n costs i n the foreign market. If our assertion that transaction costs i n foreign markets are higher i s correct, then loan size becomes a decisive f a c t o r as to whether a domestic or a foreign issue i s cheaper, given the i n t e r e s t rate d i f f e r e n t i a l . I t i s therefore hypothesized that Hr,: The average size of U.S.-pay bond issues w i l l be larger than that of Canadian-pay issues. 72 As loan size and size of company are c l o s e l y related, our above analysis implies that large Canadian corporations should f i n d i t more a t t r a c t i v e to borrow i n the United States than smaller ones. Also, we may expect that those corporations which have ready access to both markets s e l l only some of t h e i r larger issues abroad and a l l others, p a r t i c u l a r l y smaller ones, at home. Because of exchange r i s k i t i s unreasonable to expect them to borrow s o l e l y i n the United States. E a r l i e r we assumed that f l o t a t i o n costs on bonds sold i n the home market are i d e n t i c a l f o r both the domestic and the Cf. Securities and Exchange Commission, p. 41. -122-foreign market. In the Canadian-United States case t h i s does not hold f o r public offerings of corporate bonds. The data re-73 ported i n Table 3-8 suggest that f o r smaller issues up to a size of about $5 m i l l i o n transaction costs are lower i n 74 Canada.' P a r t i c u l a r l y the differences i n underwriting spreads seem to indicate that the Canadian market i s more e f f i c i e n t with regard to small bond issues and the American market with '7*5 regard to large issues. J The differences i n other expenses 73 '^Note that i n Table 3-8 underwriting spreads reported f o r United States corporate bonds are an average of three categor-ies only, "Manufacturing", "Mining", and "Other" industries. Two other c l a s s i f i c a t i o n s , " E l e c t r i c i t y , Gas and Water" and "Communication" have been excluded. In Canada, major public u t i l i t i e s are government owned. Therefore Canadian data are only to a minor extent influenced by bonds issued by public u t i l i t i e s whereas the usually reported U.S. " A l l Industries" underwriting spreads and other expenses are a weighted average of two-thirds public u t i l i t y and one-third other industry bonds. U.S. underwriting spreads on public u t i l i t y bonds usually amount to only one-half to one-third of the compensa-t i o n paid on s i m i l a r bonds issued by other in d u s t r i e s . For comparisons with Canadian data, t h i s introduces a considerable downward bias into the U.S. " A l l Industries" data. 74 Similar observations have been made by The Investment Dealers* Association of Canada, Appendix E, Part I - Corpora-t i o n Finance, pp. 14-15, and by Peters, p. 53* Comparable data on private placements are not a v a i l a b l e . Peters, p. 47, reports very rough estimates f o r Canada which, i f correct,would suggest that f o r a l l issue sizes transaction costs on private placements are s l i g h t l y higher i n Canada. ^-*Our re i n t e r p r e t a t i o n of the data on underwriting spreads i n the United States as compared to Canadian data may shed some additi o n a l l i g h t on the ongoing discussion as to whether the Canadian c a p i t a l market i s less e f f i c i e n t than the United States market. Both F u l l e r t o n , pp. 311-313, and Peters, pp. 83-84 have argued that the Canadian market i s l e s s e f f i c i e n t than the U.S. market though taking notice of the obvious differences i n volume and i n s t i t u t i o n a l investors' resources between the two markets which should lead to the expectation that the absolute e f f i c i e n c y with regard to large bond issues i s higher i n the United States. Recently Neufeld, p. 105, asserted that the r e l a t i v e (compared to market size) e f f i c i e n c y of the Canadian c a p i t a l market i s i d e n t i c a l to that of the U.S. marketi "There i s no evidence here [ i n his data] that the market f o r long-term c r e d i t instruments i s less developed i n Canada than i n the United States". We may add that there i s some evidence i n d i -TABLE 3 - 8 AVERAGE FLOTATION COSTS OF PUBLIC CORPORATE BOND ISSUES AS A PERCENTAGE OF GROSS PROCEEDS IN CANADA AND THE UNITED STATES AND COST DIFFERENTIALS BETWEEN THE TWO COUNTRIES Size of Issues ( M i l l , of $$) .Underwriting Spread Other Expenses Total Expenses Canada u s a D i f f . Canada u s a D i f f . Canada u s a D i f f . • 5 - . 9 5 - 7 4 7 . 5 3 - 1 . 7 9 1 . 9 6 3 . 9 6 - 2 . 0 0 7 . 7 0 1 1 . 4 9 - 3 . 7 9 1 . 0 - 1 . 9 5 - 3 0 6 . 3 1 - 1 . 0 1 1 . 5 3 2.40 - . 8 7 6 . 8 3 8 . 7 1 - 1 . 8 8 2 . 0 - 4 . 9 3.84 4.46 - . 6 2 • 75 1 . 7 5 - 1 . 0 0 ^ . 5 9 6 . 2 1 - 1 . 6 2 5 . 0 - 9 . 9 2 . 9 0 2 . 6 2 + .28 .71 .92 - . 21 3 . 6 1 3 . 5 ^ + . 0 7 1 0 . 0 - 1 9 . 9 2 . 5 8 1 . 6 2 + . 9 6 . 3 8 . 6 0 - .22 2 . 9 6 2 . 2 2 + . 7 ^ 2 0 . 0 - 4-9.9 2.08 1 . 3 1 + . 7 7 . 1 9 .40 - .21 2 . 2 7 1 . 7 1 + . 5 6 5 0 . 0 & over 1 . 5 0 . 9 6 + . 5 4 . 06 . 2 6 - . 20 •< 1 . 5 6 1 . 2 2 + . 3 4 aAverage of "Manufacturing", "Mining", and "Other" industries ( i . e . , public u t i l i -t i e s have been excluded) weighted by number of issues i n each size class. Sources« Peters, p. 46 and Se c u r i t i e s and Exchange Commission, p. 3 8 . -124-are probably i n part due to differences i n l e g a l and other requirements. These differences i n f l o t a t i o n costs between Canada and the United States w i l l reinforce the postulated tendency that only r e l a t i v e l y large Canadian issues w i l l be flo a t e d i n the United States. One of the features of private placements that i s p a r t i -c u l a r l y a t t r a c t i v e to corporations i s lower transaction c o s t s . 7 ^ Whether t h i s leads to lower o v e r a l l debt costs i s not clear. Peters has argued that i n s t i t u t i o n a l investors i n Canada t r y to reap the benefits of lower transaction costs by taking "up to one-quarter of 1% more on private placements than 77 they w i l l on public offerings of roughly parable q u a l i t y . " ' For the United States, i t has even been suggested that "the t o t a l cost i s l i k e l y to be somewhat higher f o r a private place-ment than f o r a public o f f e r i n g . " 7 8 On the other hand, Shapiro and Wolf present data which indicate that i n the United States the y i e l d on private placements i s only marginally higher or even lower than on public issues of comparable q u a l i t y except f o r bonds of highest quality. 7-^ As Canadian bonds never obtain the highest c r e d i t r a t i n g , lower transaction costs should provide a strong incentive f o r Canadian corporations to place eating that the absolute e f f i c i e n c y of the Canadian market with regard to small public bond issues seems to be higher than that of the U.S. market. 7 6 S h a p i r o and Wolf, p. 46. 7 7 P e t e r s , p. 6 2 . 7 8James C. Van Home, Fi n a n c i a l Management and Policy (Engle-wood C l i f f s , N . J . J Prentice-Hall, 1968), p. 2 3 7 . 7^Shapiro and Wolf, p. 28. See also p. 46 there. 8 0 0 n t h i s see Peters, p. 98 and Ripley, "United States In-vestment i n Canadian Securities 1958-1965", p. -125-t h e i r U.S.-pay bonds p r i v a t e l y rather than to have them under-written. Note that t h i s argument reinforces our e a r l i e r assert-81 ion that U.S.-pay bonds w i l l tend to be private placements. If the flow of information i s imperfect and i f i t i s cost-l y to obtain new information, companies which have subsidiaries i n the United States or which are (partly) owned by Americans may f i n d i t easier to approach the U.S. market f o r long-term c a p i t a l . Through t h e i r dealings with t h e i r foreign a f f i l i a t e or parent, these firms are more f a m i l i a r with the foreign f i n a n c i a l system. In addition, American shareholders or d i r e c t -ors may be h e l p f u l i n establishing good relationships to finan-c i a l i n s t i t u t i o n s i n the United States and i n gathering informa-t i o n on the foreign c a p i t a l market. I t i s therefore asserted that Hgt Canadian corporations which have a f f i l i a t e s i n the United States or which have American stockholders are more l i k e l y to approach the U.S. bond market than those firms that have no connections abroad. See hypothesis H 2, p.94 above. But i f indeed the major-i t y of U.S.-pay bonds i s placed d i r e c t l y , i t may be d i f f i c u l t to discover whether t h i s i s mainly due to differences i n l i q u i d i -ty preferences of i n s t i t u t i o n a l investors between the two count-r i e s or a function of transaction costs. -126-3.42 TRANSACTION COSTS AND LENDING BEHAVIOUR So f a r , we have mainly dealt with the impact which higher information and transaction costs i n international financing are expected to have on borrowing behaviour. But international and foreign tax laws, regulations concerning c a p i t a l flows, lack of r e a d i l y available information on foreign corporations and t h e i r home countries and s i m i l a r imperfections i n international c a p i t a l markets w i l l a l l lead to higher transaction costs f o r lenders as well. Consequently one would expect them to ask f o r a higher rate of interest on loans to foreigners than on other-wise comparable loans to domestic borrowers. At least part of lenders' transaction costs are fixed and independent of investment size so that t h e i r impact on the y i e l d required by suppliers of 82 funds should diminish as loan size increases. S i m i l a r l y , high-er information and other costs involved i n lending to foreigners w i l l lose i n r e l a t i v e importance as the amount of funds supplied increases. It i s therefore to be expected that the difference between interest rates requested on loans to foreigners and rates demanded on loans to domestic borrowers declines as issue size increases. This influence of transaction costs on lending behaviour should reinforce our e a r l i e r conclusion that U.S.-pay issues w i l l , on average, be larger than Canadian-pay issues. Because of the impact of transaction costs the actual interest 82 For an a n a l y t i c a l treatment of the relationship between i n -formation and transaction costs and lending behaviour see D.J. Aigner and CM. Sprenkle, "A Simple Model of Information and Lending Behaviour", Journal of Finance, XXI (March, 1 9 6 8 ) , pp. 1 5 1 - 6 6 . Peters, p. 48, presents data which indeed show that the " y i e l d return to lenders" decreases with issue size as one would expect i f transaction costs of lenders are f i x e d or only semi-variable . -127-rate d i f f e r e n t i a l between domestic and foreign c a p i t a l markets should be an increasing function of loan s i z e . In general, quite apart from the fact that size of issue and size of issuing corporation are correlated, large companies may f i n d i t easier and more a t t r a c t i v e to approach a foreign bond market than smaller firms. The higher an investor^ sub-jective p r o b a b i l i t y estimate of the borrower's default on the loan, the higher w i l l be the rate of return demanded by him. The r i s k of default as perceived by lenders i s a function of the amount and qu a l i t y of information a v a i l a b l e . Information on leading Canadian corporations i s more re a d i l y available to 84 U.S. investors than data on smaller Canadian firms. Also, large companies are more l i k e l y to be widely known abroad, p a r t i c u l a r l y i f t h e i r business a c t i v i t i e s extend to foreign countries. 8-* However, because of the high interdependence be-tween size of issue and size of corporation, i t seems not very meaningful to introduce an independent hypothesis asserting that mainly large Canadian corporations w i l l borrow i n the United States. Note that the above considerations also suggest that 8 3 G f . Aigner and Sprenkle, pp. 153-156. 84 For example, Standard & Poor's and Moody's manuals report on leading Canadian corporations and sometimes give, even more exhaustive information than i s available from the F i n a n c i a l Post's publications. 8-*With respect to the Eurobond market, Mendelson, p. 125, note 37• remarks that "the a c c e p t a b i l i t y of a name i s not neces-s a r i l y a function of s i z e . It i s rather a function of the famil-i a r i t y of the bankers or the investors with the firm." -128-H^t A company which has already issued a U.S.-pay bond before or which has i t s stocks l i s t e d on a U.S. stock exchange w i l l have easier access to the United States c a p i t a l market. Through an e a r l i e r bond issue, a firm has established a c r e d i t r a t i n g and investors already possess some information about i t . 86 Because of the lower r i s k and costs involved, investment bank-ers w i l l be more i n c l i n e d to underwrite the issue of such a 87 company. ' A stock l i s t i n g w i l l s i m i l a r l y f a m i l i a r i z e investors with a c e r t a i n corporation and lead to the constant dissemina-t i o n of information on t h i s firm. With respect to bonds issued by American corporations, i t has been observed that a f t e r a f i r s t private placement has been arranged successfully, " i t i s common for the borrower and lender to continue the r e l a t i o n -ship and f o r the lender to accommodate the borrower i n prefer-88 ence to others when the money supply i s t i g h t . " The same may hold f o r Canadian corporate borrowers i n the United States. 86 Costs l i k e educating salesmen and c l i e n t s . 8 7Nevin, p. 99. note 1, reports that "Australia enjoys unusually favourable terms i n New York, mainly because i t i s a very f a m i l i a r borrower i n that market. " qq Shapiro and Wolf, p. 3. - 1 2 9 -3 . 5 SUMMARY In t h i s chapter we turned to a more detailed discussion of the influence which (1) differences i n time preferences between countries, (2) exchange r i s k , and (3) transaction costs are ex-pected to have on international i n t e r e s t rate d i f f e r e n t i a l s and r e s u l t i n g c a p i t a l flows, p a r t i c u l a r l y with regard to Canadian corporate borrowing i n the United States. Some evidence was presented which indicates that interest rate d i f f e r e n t i a l s are p a r t l y but not s o l e l y a r e f l e c t i o n of ex-change rate expectations. By c a l c u l a t i n g the demand for funds i n Canada and the United States as a percentage of gross national product i n the respective country we showed that the r e l a t i v e demand f o r funds seems to be considerably higher i n Canada. We suggested that t h i s may be a major cause of the continuing i n -flow of long-term debt c a p i t a l into Canada. To further inquire into a possible difference i n time preferences between the two nations and the impact t h i s may have on c a p i t a l flows between them,the following hypotheses were advanced f o r empirical t e s t -ings H^: Consider only long-term bonds which may be defined as bonds having terms to maturity of more than twelve years. And compare only bonds that have been issued by Canadian corporations that have borrowed i n both Canada and the United States. Then the term to matur-i t y of U.S.-pay bonds w i l l be, on average, longer than the term to maturity of Canadian-pay bonds. -130-HgJ The comparative preference of Canadian versus Ameri-can f i n a n c i a l i n s t i t u t i o n s f o r bonds with a high de-gree of marketability has induced Canadian corpora-tions to s e l l s e c u r i t i e s i n the U.S. dir e c t placement market. Hy The a v a i l a b i l i t y of forward commitments has been a major fac t o r a t t r a c t i n g Canadian corporations to borrow i n the United States. Next we analyzed the exchange r i s k involved i n borrowing i n the American c a p i t a l market. The issuance of U.S.-pay bonds tends to reduce a corporation's exposure to exchange r i s k i f i t receives U.S. d o l l a r income. If Canadian borrowers want to protect themselves against an early adverse change i n the ex-change rate, then they should issue only bonds with long terms to maturity. Furthermore, i t i s l i k e l y that sinking fund pay-ments provide a better protection against exchange r i s k the more evenly they are distributed over the term of an issue. Therefore we asserted that H ^ J Canadian corporations with U.S. d o l l a r revenue w i l l exhibit a higher tendency to issue U.S.-pay bonds than those firms not engaged i n export a c t i v i t i e s . H^: On average, the term to maturity of U.S.-pay bonds w i l l be longer than the term to maturity of Canadian-pay bonds. -131-Hgt Sinking fund payments on U.S.-pay bonds w i l l be more evenly d i s t r i b u t e d over the term of the issue than those on Canadian-pay bonds. F i n a l l y , we b r i e f l y discussed those factors that l e d us to believe that information and transaction costs are higher on international transactions than on comparable domestic ones. From an analysis of the impact of higher transaction costs on borrower and lender behaviour we concluded that H?s The average size of U.S.-pay bond issues w i l l be larger than that of Canadian-pay issues. Hgj Canadian corporations which have a f f i l i a t e s i n the United States or which have American stockholders are more l i k e l y to approach the U.S. bond market than those firms that have no connections abroad. H^t A company which has already issued a U.S.-pay bond before or which has i t s stocks l i s t e d on a U.S.stock exchange w i l l have easier access to the United States c a p i t a l market. We also suggested that large Canadian corporations should f i n d i t easier and more a t t r a c t i v e to s e l l bonds abroad than smaller firms. However, because of the expected high c o r r e l a t i o n be-tween size of issue and size of corporation such a hypothesis could probably not be tested independently from H 9 above. CHAPTER 4 DATA COLLECTION AND SAMPLE DESCRIPTION In t h i s chapter we b r i e f l y discuss the data c o l l e c t i o n pro-cess, and a description of our data samples w i l l be given. Only information essential f o r a better understanding of the empiri-c a l results presented i n Chapter 5 w i l l be provided. For a more extensive discussion the reader should r e f e r to Appendix 2 where also the data and interview questionnaires used f o r t h i s study are reproduced. 4 . 1 DATA COLLECTION In order to test the hypotheses advanced i n Chapter 3 , extensive information on c h a r a c t e r i s t i c s of Canadian-pay and U.S.-pay bonds and on c h a r a c t e r i s t i c s of the issuing Canadian corp-orations was needed. As such data are not rea d i l y available, i t was necessary to c o l l e c t most of the data f o r t h i s study from o r i g i n a l sources l i k e public prospectuses, private place-ment memoranda and annual reports. In many instances the re-spective corporations supplied us with these documents or f i l l e d out a data questionnaire we sent them. Investment bank-ers were also very h e l p f u l . For the period January i 9 6 0 through May 1973 information was gathered on Canadian-pay and U.S.-pay bonds issued by corp-orations located i n Canada and the proceeds of which were i n -tended f o r use i n Canada. Up to May 1973 Canadian corporations, whether Canadian-controlled or foreign-controlled, had not yet - 1 3 2 --133-issued bonds denominated i n any other currency though several firms had borrowed short-term funds denominated i n Swiss Franks, Deutsche Marks and so on. For 1968 and l a t e r years random samples of Canadian-pay corporate bonds were drawn from the annual editions of the Fin -a n c i a l Post's Record of New Issues. Issues contained i n these samples have been c l a s s i f i e d according to whether or not they have been sold by corporations that have issued U.S.-pay bonds at least once since i 9 6 0 . The sample of bonds issued by corp-orations that have sold long-term s e c u r i t i e s i n Canada only i s c a l l e d "Sample 1" and the acronym "CAonly" w i l l be used to iden-t i f y t h i s sample. The other bonds contained i n our random samples were grouped into "Sample 2 " ; they have been issued by firms that approached both the Canadian and U.S. c a p i t a l mark-ets, and the acronym "CAandUS" w i l l be used. In addition, we attempted to c o l l e c t information on Canadi-an-pay bonds not contained i n these random samples but issued since 1968 by corporations that borrowed at least once i n the U.S. market since i 9 6 0 . This "Sample 3", also i d e n t i f i e d by the acronym "CAandUS", together with " S a m p l e 2 " provides us with almost complete information on a l l Canadian-pay bonds sold by those Canadian companies that placed t h e i r s e c u r i t i e s both i n Canada and the United States. For the period i 9 6 0 through 1 9 6 7 , data on Canadian-pay bonds were gathered only f o r those bonds that had been issued by corporations which approached the U.S. market at least once TABLE 4 - 1 DEFINITION OF BOND SAMPLES Item 1 2 Sample 3 4 5 Type of Issue Can.-pay Can.-pay Can.-pay Can.-pay U.S.-pay Time Period Covered 68 - 73 68 - 73 68 - 73 60 - 67 60 - 73 Randomly Selected yes yes no no no Issuer Placed Bonds i n the U.S. no yes yes yes -Acronym Used "CAonly" "CAandUS" "CAandUS" "CAandUS" "USpay" a F o r a detailed explanation of the random and non-random selection process see Appendix 2 . -135-since i960. This proved less d i f f i c u l t than c o l l e c t i n g data on a random sample "because most firms f a l l i n g into t h i s cate-gory turned out to be among the larger and better known Canadi-an companies. These bonds have been grouped into "Sample 4" which i s also i d e n t i f i e d by the acronym "CAandUS". The number of U.S. d o l l a r bonds issued by Canadian corp-orations during a given time period i s usually small r e l a t i v e to the t o t a l number of new corporate bonds sold during the same period. Considerable e f f o r t s were therefore made to c o l l e c t complete data on a l l U.S.-pay bonds issued since January i960. But i n a few instances where corporations placed more than three bond issues i n the United States during the period covered by t h i s study a random number table was used to select at least three of them or f i f t y per cent of a l l U.S. d o l l a r issues, whichever led to the higher number of issues selected. This seemed advisable to avoid "Sample 5"'s being dominated by and biased towards a few large corporations which are or used to be very active borrowers i n the United States and which usually provided us with a l l the information requested. For "Sample 5" the acronym "USpay" i s used. In Table 4-1 summary information on the d e f i n i t i o n of samples 1 to 5 i s presented f o r easier reference. Some descriptive data f o r the four samples covering the period January 1968 through May 1973 can be found i n Table 4-2. Of course, i t was not possible to obtain complete i n -formation on a l l bond issues i n a l l of our samples. But when TABLE 4-2 DESCRIPTION OF SAMPLES COVERING PERIOD 1968 TO MID-1973i COMPARISON OF NUMBER OF ISSUES ON WHICH INFORMATION HAS BEEN OBTAINED WITH TOTAL NUMBER OF ISSUES IN EACH SAMPLE ACCORDING TO SELECTED ISSUE CHARACTERISTICS Sample 1 Samole 2 Sample 3 Sample 5 Issue "CAonly" "CAandUS" "CAandUS" "USpay" C h a r a c t e r i s t i c s Tot. Obt. M i s s i n g Tot. )bt. M i s s i n g M i s s i n g M i s s i n g Nos. % Nos. % Tot. Dbt. Nos. % Tot. Dbt. 1NOS . 1 m (1) (2) ( 4 ) = (8) = (12) = (16) = (3) (2)-(3) (5) (6) 17) (6)-(7) (9) (10) (11) (10)-(11) (13) ( 1 4 ) (15) : i 4 ) - ( l 5 ) (17) Year of Issue 1968 16 11 5 31J5 6 6 0 — 7 7 0 9 7 2 2 2 4 1969 19 1 9 0 - 12 12 0 - 0 0 0 - 11 10 1 9 4 1970 2 4 23 1 4 4 12 12 0 — 8 8 0 - 4 3 1 254 1 9 7 1 25 2 4 1 44 7 7 0 12 12 0 _ 4 3 1 254 1972 26 2 4 2 84 6 6 0 - 7 7 0 — 6 5 1 174 1973 12 11 1 84 5 5 0 - 2 2 0 - 3 3 • 0 -T o t a l 122 112 10 8 4 48 43 0 - 36 36 0 - 37 31 6 164 S i z e of Issue 0 ( M i l l , o f $$) l e s s than 1.00 2 2 0 — 0 0 0 _ 0 0 0 0 0- 0 1.00 - 2 . 4 9 9 6 3 33% 0 0 0 _ 0 0 0 1 1 0 2.50 - 4 . 4 9 22 17 5 23% 2 2 0 - 3 3 0 - 3 3 0 4.50 - 6 . 9 9 19 1 8 l 5% 3 3 0 - 1 1 0 _ 2 2 0 7.00 - 1 1 . 4 9 31 31 0 9 9 0 - 4 4 0 — 5 4 1 2 0 4 11.50 - 1 7 . 4 9 13 12 l 8 4 8 8 0 - 4 4 0 - 7 4 3 434 17.50 - 2 7 . 4 9 1 7 17 0 - 8 8 0 - 9 9 0 - 6 4 2 33?? 27.50 - 4 2 . 4 9 7 7 0 - 5 5 0 - 3 3 0 _ 3 3 0 42.50 - 6 2 . 4 9 2 2 0 - 9 9 0 - 9 9 0 — 3 3 0 62.50 & over 0 0 0 - 4 4 0 - 3 3 0 - 7 ! 7 0 -TABLE 4-2 - Continued Issue Sample 1 Samole 2 Sample 3 Sample 5 "CAonly" "CAandUS" "CAandUS" "US-Dav" V*\ ^ "4- f~\ V* M ^* A m una r s c x e r i s x i c s Tot. Obt. . a s s i n g M i s s i n g M i s s i n g M i s s i n g Nos. % Tot Obt. Nos. % Tot. Obt. Nos. % Tot. Obt. Nos. $ (1) (2) (3) (4) = (5) (8) = (12) = (16) = (2)-(3) (6) (7) (6)-(7) (9) (10) (11) (10)-(11) (13) (14) (15) (14)-(15) (17) Placement of Issue P u b l i c O f f e r -ings 90 90 0 - 43 43 0 - 29 29 0 - 5 4 1 20$ P r i v a t e P l a c e -ment 32 22 10 31$ 5 5 0 - 7 7 0 - 32 27 5 In d u s t r y C l a s s i -f i c a t i o n Banking & F i n . 21 21 0 - 14 14 0 _ 2 2 0 _ 6 3 3 50$ Mining 0 0 0 - 0 0 0 - 0 0 0 — 0 0 0 _ Non-ferrous Metals 1 1 0 - 2 2 0 - 5 5 0 — 3 3 0 _ O i l s 6 6 0 - 2 2 0 - 3 3 0 _ 5 5 0 _ P r o p e r t y Dev. 10 9 1 10$ 1 1 0 - 3 3 0 — 2 2 0 _ P u b l i c U t i l . 25 24 1 4$ 17 17 0 - 10 10 0 — 11 11 0 _ Pulp & Paper 2 2 0 - 3 3 0 - 3 3 0 - 4 2 2 50$ Transporta-t i o n 3 3 0 - 2 2 0 - 3 3 0 - 4 3 1 25$ Trust & Loan Companies 13 13 0 - 0 0 0 - 0 0 0 — 0 0 0 — Pipe Lines 1 1 0 - 6 6 0 - 7 7 0 _ 2 2 0 _ Others 40 32 8 20$ 1 1 0 - 0 0 0 - 0 0 0 --138-comparing the number of issues on which information has been obtained with the t o t a l number of issues i n each sample ac-cording to year of issue, size of issue, placement of bonds and industry c l a s s i f i c a t i o n of issuing company i t can be seen that the missing data are f a i r l y evenly d i s t r i b u t e d and do not introduce any serious biases into our samples. Not surpris-ingly, information obtained on private placements i s less complete than that on public offerings, and data on U.S.-pay bonds were more d i f f i c u l t to secure. Similar information on the two samples covering the period i960 through 1967 i s reported i n Table 4-3. In general, the missing information i s again f a i r l y evenly d i s t r i b u t e d . Table 4-3 provides also summary s t a t i s t i c s f o r the t o t a l number of U.S.-pay bonds which have been i d e n t i f i e d as having been issued from i 9 6 0 through I 9 6 7 . Not a l l of these bonds were included i n our "Sample 5" f o r t h i s period because (1) i n some instances no company address could be secured nor the place-ment agent i d e n t i f i e d , that i s , i t was impossible to even at-tempt to c o l l e c t data f o r these issues; (2 ) some of these bonds were i d e n t i f i e d only a f t e r the data c o l l e c t i o n process had been terminated; (3) several bonds, mostly issued by f i n a n c i a l i n -s t i t u t i o n s and public u t i l i t i e s , were excluded through the random selection process described e a r l i e r . -139-TABLE 4-3 DESCRIPTION OF SAMPLES COVERING PERIOD I960 TO 196?» COMPARISON OF NUMBER OF ISSUES ON WHICH INFORMATION HAS BEEN OBTAINED WITH TOTAL NUMBER OF ISSUES IN EACH SAMPLE ACCORDING TO SELECTED ISSUE CHARACTERISTICS Issue Samole 4 "CAandUS" Sample 5 "USpay" >\ « l C h a r a c t e r i s t i c s M i s s i n g M i s s i n e '. 1—1 (1) T o t a l (2) Obt. (3) Nos. (4) = (2)-(3) $ (5) T o t a l (6) Obt. (7) Nos. (8) = (6)-(7) <$> ( 9 ) • w 3 CD T3 r H CO . H r H CO < M . H Year of Issue I960 1961 1962 1963 1964 1965 1966 1967 3 2 5 11 6 4 12 9 3 2 5 . 9 6 2 10 8 0 0 0 2 0 2 2 1 18$ 50$ 17$ 11$ 5 7 6 8 11 17 17 15 4 6 5 5 6 14 10 10 1 1 1 3 5 3 7 5 20$ 14$ 17$ 37^ 46$ 18$ 4 l $ 33$ 6 10 7 12 16 24 21 20 T o t a l 52 45 7 13$ 86 60 26 30$ 116 S i z e of Issue ( m i l l , of $$) l e s s than 1.00 1.00 - 2 . 4 9 2.50 - 4 . 4 9 4.50 - 6 . 9 9 7.00 - 11 . 4 9 11.50 - 17 . 4 9 17.50 - 27 . 4 9 27.50 - 42 . 4 9 42.50 - 62 . 4 9 62.50 & over 3 6 4 6 11 11 5 4 2 0 3 i 5 11 8 3 4 2 0 0 1 0 1 0 3 2 0 0 0 17$ 17$ 27$ 40$ 2 7 9 9 20 8 11 12 3 5 1 4 8 8 11 5 ? 8 3 5 1 3 1 1 9 3 4 4 0 0 5 0 $ 43$ 11$ 11$ 45$ 37$ 36$ 33$ 4 11 14 9 24 13 14 14 7 5 - 1 4 0 -T A B L E 4-3 - Continued Issue Characteristics ( 1 ) Sample 4 "CAandUS" Sample 5 "USpay" All U.S.-Pay Issues Ident-ified Total ( 2 ) Obt ( 3 ) Missing Obt. ( 7 ) Kissing All U.S.-Pay Issues Ident-ified Nos. ( 4 ) = ( 2 ) - ( 3 ) ( 5 ) Total ( 6 ) Nos. ( 8 ) = ( 6 ) - ( 7 ) 1 TIT 7° ( 9 ) All U.S.-Pay Issues Ident-ified Placement of Issue Public Offer- 23 2 1 2 9 $ 5 5 0 5 ing Private Place- 2 9 • 2 4 5 1 7 $ 8 1 5 5 2 6 3 2 $ 1 1 1 ment Industry Classic fication Banking and 1 0 9 1 1 0 $ 1 1 8 3 2 7 $ 2 4 Finance Mining 0 0 0 3 3 0 3 Non-ferrous 0 0 0 _ 2 2 0 2 Metals Oils 1 0 8 2 2 0 $ 1 4 1 0 4 2 9 $ 1 5 Property Devel- 3 3 0 - 4 3 1 2 5 $ 8 opment Public U t i l i - 1 1 9 2 1 8 $ 9 9 0 _ 1 4 ties Pulp and Paper 4 4 0 _ 2 0 1 0 1 0 5 0 $ 2 1 Transportation 3 3 0 _ 2 2 0 • a 6 Trust and Loan 1 1 0 _ 1 1 0 1 Companies Pipe Lines 3 3 0 — 7 7 0 8 Others 7 5 2 2 9 $ 1 3 5 8 6 1 $ 1 4 4.2 INTERVIEWS -141-In order to gain a better understanding of the f i n a n c i a l behaviour of Canadian corporations and of the functioning of the Canadian corporate bond market, interviews were conducted with s i x Canadian investment bankers, four l i f e insurance i n -vestment o f f i c e r s , and twenty-one corporate executives. A l l interviews took place during July and August 1973 when the author paid v i s i t s to Montreal, Toronto, Calgary and Vancouver. In addition, three written questionnaires were received, one from a Canadian corporation and two from American underwriters. A l l the interviews sought with investment houses and l i f e insurance companie . were successfully completed. Of the twenty-eight personal interviews sought with corporate f i n a n c i a l o f f i c e r s , twenty-one did materialize, and one questionnaire from a public u t i l i t y was received by mail. As can be seen from Table 4-4, the companies chosen covered firms i n a l l those industries that are p a r t i c u l a r l y active borrowers i n the United States. The interview questionnaires used and a concise summary of the answers obtained from corporate managers can be found i n Appendix 2. -142-TABLE 4-4 COMPARISON OF CORPORATE INTERVIEWS SOUGHT WITH INTERVIEWS GRANTED BY INDUSTRY CLASSIFICATION Industry C l a s s i f i c a t i o n Number of Sought Interviews Granted Banking and Finance 2 2 Mining 1 0 Non-ferrous Metals 3 2 O i l and Gas 5 4 Property Development 2 1 Public U t i l i t i e s 5 4 Pulp and Paper 3 3 Transportation 2 0 Pipe Lines 4 4 Others 1 1 Total | 28 21 CHAPTER 5 A COMPARATIVE ANALYSIS OF CANADIAN CORPORATE DEBT ISSUES IN THE UNITED STATES AND CANADIAN BOND MARKETS 5.1 INTRODUCTION In t h i s chapter s t a t i s t i c a l tests of the hypotheses de-veloped i n Chapter 3 w i l l be presented. The information obtained through interviews w i l l be summarized and used to interpret the s t a t i s t i c a l r e s u l t s . Our hypotheses concentrate on differences between p a r t i c u l a r c h a r a c t e r i s t i c s of bonds denominated i n U.S. doll a r s and of bonds denominated i n Canadian d o l l a r s and on d i f -ferences between firms that have issued bonds abroad and those that have not. Discriminant analysis w i l l be used as the main s t a t i s t i c a l technique. This method allows us to construct a rule from sample observations which can be employed to assign a new observation to one of two or more mutually exclusive popu-l a t i o n s . Whenever a certa i n c h a r a c t e r i s t i c as asserted i n our hypotheses indeed contributes s i g n i f i c a n t l y to the discriminatory power of the derived decision rule t h i s w i l l be judged as support f o r that p a r t i c u l a r hypothesis. A f t e r a short exposition of the discriminant analysis technique the influence of interest rate d i f f e r e n t i a l s and of differences i n the demand f o r funds between Canada and the United States on Canadian corporate borrowing abroad w i l l be analyzed. This i s followed by a discussion of exchange rate r i s k and transactions costs differences and t h e i r impact on -143--144-U.S. d o l l a r debt issues. Based on these r e s u l t s , summary dis-criminant functions w i l l be derived which lead to the best ex post and ex ante c l a s s i f i c a t i o n of Canadian corporate bond issues into those denominated i n U.S. d o l l a r s and those denom-inated i n Canadian d o l l a r s . F i n a l l y , the influence of non-economic factors on corporate borrowing behaviour w i l l be dis-cussed. Usually empirical results w i l l be presented f o r three time periods: 1. The period i960 to 196? f o r which we c o l l e c t e d data only on those bonds that were issued by corporations which have ap-proached the U.S. bond market at l e a s t once since i960. That i s , we compare the Canadian-pay bonds i n "Sample 4" with the U.S.-pay bonds i n "Sample 5" f o r t h i s period. Con-sequently r e s u l t s f o r the period i960 to 196? should be re-garded as representative only with respect to Canadian corp-orations that have borrowed i n the United States and are not in d i c a t i v e f o r the f i n a n c i a l behaviour of a l l Canadian firms during t h i s time. 2. The period January 1968 to May 1970 during which the Canadian exchange rate was f i x e d . 3. The period June 1970 to May 1973 when no par value f o r the Canadian d o l l a r was i n e f f e c t . A l l discriminant analysis re s u l t s for the period January 1968 to May 1973 are based on a comparison of the random sample of Canadian-pay bonds ("Sample 1" and "Sample 2" combined) with the sample of U.S.-pay bonds ("Sample 5") f o r t h i s period. -145-As a bond's c h a r a c t e r i s t i c s l i k e coupon, issue p r i c e , maturity, and so on are determined before or at the date a prospectus i s issued or the placement agreement i s signed the "date of con-t r a c t " rather than the issue or delivery date of the bonds has been ohosen to c l a s s i f y bonds i n our samples into these three time periods. 5.2 STATISTICAL METHODS The objective of discriminant analysis i s to derive from sample sets of observations on members of two or more mutually exclusive populations l i n e a r combinations of these measurements that allow to predict the group to which a new member belongs. Let us concentrate on the case of two populations, and l e t x' = ( x ^ . X g , . . •»xjc) D e the vector of measurements on k char-a c t e r i s t i c s of an in d i v i d u a l drawing, b a (kxl) vector of di s -criminant c o e f f i c i e n t s , and z the discriminant score.^ The general problem i s to f i n d a discriminant function which d i f f e r e n t i a t e s between the two populations as much as poss i b l e by minimizing the overlap i n discriminant scores between the two groups. This i s achieved by maximizing the squared difference between the mean discriminant score f o r each group, — — 2 (z., - z0) , r e l a t i v e to the pooled variance within samples, where n. i s the number of observations on population is The case of several populations i s discussed i n Theodore W. Anderson, An Introduction to Multivariate S t a t i s t i c a l Analysis (New Yorkt Wiley, 1958), Chapter 6. Here we follow Gerhard Tintner, Econometrics (New Yorkt Wiley, 1952), Chapter 6. See z = x'b (5-D 2 -146-(5-2) Here M s a Lagrange m u l t i p l i e r . Substituting from expression (5-1) into equation (5-2) we obtain x^ i s the vector of mean scores on the k variables f o r sample i , and where i s an (n^xk) matrix of deviations of in d i v i d u a l measure-ment vectors from the mean scores f o r each group. S i s an e s t i -mate of the common covariance matrix f o r the k variables. Set-t i n g the derivatives of F with respect to b equal to zero does not y i e l d a unique solution f o r the vector of discriminant coef-2 — — f i c i e n t s . But when a r b i t r a r i l y l e t t i n g A = (x-^  - .x^'b- the following solution i s obtainedi Multiplying b by any scalar d i f f e r e n t from zero does not change the discriminatory power of the discriminant function. This feature can be used f o r transforming the vector of discriminant c o e f f i c i e n t s i n any convenient manner. When the discriminant score f o r an observation i s greater than the value of the dis-also J. Johnston, Econometrics .(2nd ed.; New York: McGraw-Hill, 1972), pp. 334-340. p George W. Ladd, i n "Linear P r o b a b i l i t y Functions and Dis-criminant Functions", Econometrica, Vol. 34, No. 4 (October 1966), pp. 873-85, discusses alternative estimation methods which lead to discriminant c o e f f i c i e n t s that d i f f e r only by a factor of pro-p o r t i o n a l i t y . The discriminant functions to be presented l a t e r were estimated using equation (5-*0« F = b'(x x - x 2 ) ( x ^ - x 2)'b - (b'Sb). (5-3) S = ( X ^ i + X ^ 2 ) / , ( n l + n2 " 2 ) b = S" 1(x 1- x 2 ) • -147-criminant function evaluated at the average of the two mean vec-tors, x , b > f ( x 1 + x 2)'b, then t h i s observation w i l l be assumed to have come from population 1 and vice versa. An a r b i t r a r y choice has to be made whenever an in d i v i d u a l score i s equal to i?(x^ + X g J ' b . When l a t e r reporting estimates of discriminant functions, the value -§-(x + x) 'b w i l l be included as a constant. This causes the mean discriminant scores f o r the two groups of bonds to be equi-distant from zero except f o r rounding errors. The functions are always presented such that scores above zero indicate a bond issue denominated i n U.S. dollars and that negative values would cause an issue to be c l a s s i f i e d as Canadian-pay. The s t a t i s t i c a l significance of an estimated discriminant function can be determined by using a standard F-test where F i s a function of Hotelling's T^. Like most tests i n multivariate analysis t h i s t e s t presupposes that the underlying d i s t r i b u t i o n s are normal and that the covariance matrices are equal. But a v i o l a t i o n of these assumptions usually leads to only a mild d i s -3 agreement between nominal and actual si g n i f i c a n c e l e v e l s . The 3 -\A.s Cooley and Lohnes point out, "many research workers pre-f e r to ignore the issue of the homogeneity of group dispersions on the grounds that the test [for the equality of group means] i s probably f a i r l y robust under departures from i t s assumptions". See William W. Cooley and Paul R. Lohnes, Multivariate Data Analy-s i s (New Yorkt Wiley, 1971), p. 228. For experimental support of th i s assertion see B a s i l P. Korin, "Some Comments on the Homo-sce d a s t i c i t y C r i t e r i o n M and the Multivariate Analysis of Variance Tests T 2, W and R", Biometrika, LIX ( A p r i l , 1972), pp. 2 1 5 - 1 6 , and the l i t e r a t u r e c i t e d there. On the question of normality, c f . H. J. Arnold, "Permutation Support f o r Multivariate Techniques", Biometrika, LI (1964), pp. 65-70. - I n -significance of an i n d i v i d u a l variable entered i n the discrim-inant function can be assessed by t e s t i n g f o r the equality of the means of the two conditional d i s t r i b u t i o n s of t h i s v a r i -able given the other variables included. Note that i n several instances logarithmic transformations of variables which approxir mated a lognormal rather than a normal d i s t r i b u t i o n were used. Ex post and ex ante c l a s s i f i c a t i o n r esults w i l l be presented as an additional means to assess the usefulness of our estimates f o r a c t u a l l y discriminating between Canadian-pay and U.S.-pay issues. Ex post c l a s s i f i c a t i o n r e s u l t s indicate how well the derived function a c t u a l l y separates the two types of bonds con-tained i n the sample that was used f o r estimating i t . Ex ante c l a s s i f i c a t i o n r e s u l t s show us how well the estimated discrim-inant function predicts the group membership of new sets of ob-servations . Note that discriminant analysis allows us only to test f o r systematic differences between the c h a r a c t e r i s t i c s of Canadian-pay and U.S.-pay bonds, and f o r differences between those Cana-dian corporations that s e l l new issues i n the United States and those that do not. It does not indicate whether such differences are caused by borrower behaviour or are a function of lender pre-ferences. In other words, we cannot t e s t d i r e c t l y f o r causal rela t i o n s h i p s . However, our interview r e s u l t s w i l l help us to overcome t h i s obstacle. j . . Because the logarithm of zero i s undefined, we added 1 to our observations i n those cases where the lower l i m i t of a variable equalled zero. -149-5.3 INTEREST RATE DIFFERENTIALS, DIFFERENCES IN TIME PREFER-ENCES AND CANADIAN CORPORATE BORROWING IN THE UNITED STATES As discussed e a r l i e r i n Chapter 3 there i s considerable evidence i n support of the assertion that the r e l a t i v e demand fo r loanable funds i s higher i n Canada than i n the United States. It i s to be expected that t h i s leads not only to differences i n e f f e c t i v e i n t e r e s t rates between the two countries but also to differences i n lender behaviour, p a r t i c u l a r l y with regard to time preferences exhibited. F i r s t i n t e r e s t rate d i f f e r e n t i a l s and t h e i r impact on corporate borrowing i n the United States w i l l be discussed. Then discriminant analysis r e s u l t s w i l l be presented as an i n d i c a t i o n f o r the strong influence differences i n time preferences between Canadian and U.S. i n s t i t u t i o n a l i n -vestors have on long-term c a p i t a l flows into Canada. Interest Rate D i f f e r e n t i a l s . - There cannot be any doubt that the difference i n long-term inte r e s t rates usually observed between Canada and the United States constitutes an incentive f o r Canadian corporations to borrow south. A l l but two of the company o f f i c i a l s interviewed indicated that lower inter e s t rates were one of the factors that attracted them to the U.S. market, and t h i s impression was reinforced by answers obtained from underwriters. But the actual inte r e s t advantage i s con-siderably less than a comparison of y i e l d indices between the two countries would suggest.^ A casual comparison of y i e l d ^See the data i n Table 3-3 above. Because of higher i n -formation and transaction costs and additional r i s k s ( l i k e po-t e n t i a l changes i n tax regulations or exchange controls) i n -volved i n lending to foreigners, American investors seem to de-mand a higher y i e l d on foreign U.S.-pay bonds than on otherwise comparable domestic bonds. - 1 5 0 -d i f f e r e n t i a l s between Canadian-pay and U.S.-pay bonds, par t i c u -l a r l y on bonds issued by the same firm on the same date or on dates that were close together, would suggest that the cost ad-vantage at time of issue i s usually between twenty-five and seventy-five basis points. In order to obtain a more precise notion of the i n t e r e s t rate d i f f e r e n t i a l , we computed the difference between the y i e l d on a l l bonds i n our samples and the ( r i s k - f r e e ) yield, on long-term Canadian government bonds at date of contract.^ The results are presented i n Table 5 - 1• Comparing these " r i s k premiums" on bonds placed p r i v a t e l y i n Canada and the United States by corpor-ations that have approached both markets (see columns 6 and 7 i n Table 5 - 1 ) suggests that the average actual i n t e r e s t advantage to Canadian corporations has decreased considerably from about seventy basis points i n e a r l i e r years to about twenty-five points during the most recent period. However, our data on bonds p r i -vately placed i n Canada since 1969 a r e biased insofar as these issues were usually either sold by public u t i l i t i e s and finance companies or had r e l a t i v e l y short terms to maturity. Because of t h i s and because many U.S.-pay bonds would, as an a l t e r n a t i v e , have to be sold to the public i n Canada, a comparison of columns _ Because of the wide swings i n i n t e r e s t rates i n recent years a dir e c t comparison of y i e l d s on Canadian-pay and U.S.-pay bonds i s meaningless. Subtracting the government rate i s an at-tempt to adjust f o r these changes i n i n t e r e s t l e v e l s over time. Commonly t h i s y i e l d d i f f e r e n t i a l i s c a l l e d a " r i s k premium" on corporate bonds. Cf. C e c i l R. Dichand, "The Determinants of Risk Premiums and Development of Rating Classes f o r P u b l i c l y Traded Canadian Corporate Bonds", Proceedings of the F i r s t Annual  Conference of the Canadian Association of Administrative Sciences (Kingston, Ont., 1 9 7 3 ) , pp. 2 / 1 3 3 - 2 / 1 6 5 . TABLE 5 - 1 MEAN DIFFERENCE BETWEEN YIELDS ON NEWLY ISSUED CANADIAN CORPORATE BONDS AND YIELDS ON LONG-TERM CANADIAN GOVERNMENT BONDS AT DATE OF CONTRACT3-P u b l i c O f f e r i n g s P r i v a t e Placements P e r i o d Can.-nay Bonds U.S.-pay Can.-pay Bonds U.S.-pay Bonds Sample 5 (7) (1) bamole 1 "CAonly" (2) Samples 2 ,3,4 "CAandUS" (3) Bonds Sample 5 (H) Samole 1 "CAonly" (5) Samples 2 ,3,4 "CAandUS" (6) I960 - 1967 - 1.012 . ( . 3 2 6 . 2 1 ) ' 3 . 7 1 5 (.392, 5 ) - 1 . 1 3 4 ( .672 , 24 ) . 4 2 2 ( . 4 8 8 , 5 5 ) 1968 . 9 7 1 ( . 4 6 5 , 6 ) 1.239 ( . 2 6 3,10) - .060 (.000, 1) 1 . 1 1 5 ( . 3 0 7 , 5 ) 1.358 (.169, 3 ) 1 . 0 4 4 -(.497, 6 ) 1/1969 - 5/1970 1 . 2 5 4 (.293 ,24) 1 . 2 8 6 ( . 5 5 5 , 1 9 ) . 6 5 5 ( . 1 3 4 , 2) 1 . 0 4 0 (.782, 5 ) .867 ( . 3 0 2 , 3 ) • 9 5 5 ( . 5 0 6,12) 6/1970 - 5/1973 1 . 4 8 3 (.617,60) 1 . 4 2 0 (.474 ,43) 1 - 5 5 0 (.000, 1) 1 . 6 2 5 ( . 8 4 8,12) 1 . 2 7 5 ( 1 . 1 3 1 , 6 ) 1 . 0 4 6 ( . 7 5 2 , 9) The y i e l d on long-term Canadian government bonds has been obtained from Bank of Canada S t a t i s t i c a l Summary and Bank of Canada' Review, v a r i o u s i s s u e s . : ^Values i n b r a c k e t s are standard d e v i a t i o n s and numbers of observations r e s p e c t i v e l y . -152-3 and 7 i n Table 5-1 seems to be more meaningful. This would indicate that, on average, the i n t e r e s t advantage of foreign borrowing has decreased only from .60 per cent to about .35 per cent with a s l i g h t tendency to increase again i n recent years. The improvement of the Canadian bond market and easier monetary p o l i c i e s i n Canada i n recent years probably have a l l contributed to narrowing the actual y i e l d d i f f e r e n t i a l . Because of the small number of public offerings • of U.S.-pay bonds by Canadian corporations a comparison of r e l a t i v e y i e l d s on private and public U.S. d o l l a r denominated bonds (columns 4 and <7) has to be made with great caution. However, when excluding two small public issues i n the early s i x t i e s , the general impression i s that the i n t e r e s t cost on public 7 8 offerings i s lower, a pattern also observable i n Canada. From these data i t must not be concluded that lower inter e s t costs are the only or the main fa c t o r inducing Canadian corpora-''This i s not surprising because usually only large, i n t e r -national corporations s e l l U.S.-pay bonds to the p u b l i c . Q C l e a r l y , the interest rate d i f f e r e n t i a l s that can be der-ived from the data i n Table 5-1 do not necessarily r e f l e c t the actual int e r e s t savings on U.S.-pay bonds. Shorter maturities on some Canadian d o l l a r issues and lower in t e r e s t costs on these medium-term bonds introduce a s l i g h t bias into the com-parisons f o r recent years. Perhaps more importantly, many U.S. d o l l a r issues could have been placed i n Canada only at a premi-um because of t h e i r size or because of some other characteris-t i c s . As was repeatedly pointed out by many corporate managers, investment dealers and representatives of l i f e insurance com-panies, funds as such have usually been avai l a b l e i n Canada i n s u f f i c i e n t amounts to meet corporate demand and, contrary to often found b e l i e f s , a s c a r c i t y of funds i n the domestic c a p i t a l market cannot be regarded as a major reason f o r corporate borrow-ing abroad. A possible exception was 1969 when an unforeseen heavy demand f o r comparatively cheap p o l i c y loans and forward commitments i n the mortgage market considerably reduced the l i f e insurance firms' uncommited funds available f o r investment i n corporate bonds. -153-tions to s e l l t h e i r bonds abroad. Rather the impression gained from our interviews i s that the in t e r e s t rate d i f f e r e n t i a l serves as a constraint's only i f the savings i n cost of debt c a p i t a l are judged large enough to make up f o r the perceived a d d i t i o n a l r i s k s of foreign borrowing w i l l Canadian corporations consider denominating bonds i n a foreign currency. However, most of them w i l l not place bonds i n the United States just f o r intere s t savings; as our re s u l t s w i l l show, ad d i t i o n a l factors are usual-l y necessary to lead to the decision to issue U.S.-pay bonds. On the other hand, when there i s no material i n t e r e s t advantage as during much of the period 1970 through 1972, Canadian corpor-ations tend to not borrow abroad. Though several issues were delivered during t h i s period to United States investors, our sample contains only two issues f o r which contracts were signed during June 1970 through May 1972. F i n a n c i a l o f f i c e r s of corpora-tions that sold bonds i n Canada during t h i s period indicated that they did not further consider issuing U.S. d o l l a r denomin-ated bonds once i t was discovered that the y i e l d on such bonds would have to be higher or would not be lower than on comparable Canadian-pay bonds. This r i s k premium, the difference between a bond's y i e l d and the y i e l d on Canadian government bonds, w i l l be l a t e r i n -cluded i n our discriminant analyses. I t w i l l be interpreted as a measure o.f the influence exchange r i s k has on Canadian corp-orate borrowing i n the United States. Of course, i t also measures the influence of lower inter e s t costs abroad, but these factors are clos e l y i n t e r r e l a t e d as our discussion has indicated. - 1 5 4 -Differences i n Time Preferences. - Our interviews revealed that the r e l a t i v e l y higher demand for funds i n Canada indeed exerts a considerable influence on the lending behaviour of Canadian f i n a n c i a l i n s t i t u t i o n s and thereby on the type of bond issue that can be sold i n Canada. A l l four l i f e insurance com-panies v i s i t e d indicated that they have a preference f o r r e l a -t i v e l y l i q u i d investments. This reduces t h e i r perceived r i s k , allows immediate p o r t f o l i o adjustments and makes i t feasible to attempt to improve p r o f i t s through bond trading. As a matter of fact, two o f f i c e r s indicated that t h e i r firms sometimes turn over t h e i r bond p o r t f o l i o s more than once per annum. Always facing an a t t r a c t i v e array of investment opportunities they prefer to buy p u b l i c l y offered bonds, avoid forward commitments and, depending on t h e i r market outlook, regard investments i n extendible or prepayable bonds 1 0 as p a r t i c u l a r l y a t t r a c t i v e be-cause of the usually short i n i t i a l term to maturity of these 11 bonds'. On the other hand, underwriters and corporation o f f i c e r s "We promised to a l l those who contributed information to t h i s study not to reveal t h e i r i d e n t i t y . This l e d to sometimes very frank and open discussions. When r e f e r i n g to p a r t i c u l a r interviews, i t i s therefore not possible to state place, time, or name of the interviewee. 1 0 E x t e n d i b l e or prepayable bonds are s e c u r i t i e s where a f t e r an i n i t i a l period of usually f i v e or s i x years the in d i v i d u a l lender can decide whether he wants to extend the term of the bonds owned by him or not. Extensions range usually from f i v e to f i f -teen years, sometimes with multiple options. When measuring the term to maturity of these bonds, always the i n i t i a l term was chosen. Bonds having such features seem not to be available i n the American market, but a few Canadian U.S.-pay bonds have been extendible. Unfortunately, the term prepayable i s sometimes a l -so used f o r borrower-callable bonds and can lead to confusion. 1 10ne insurance o f f i c e r indicated that his firm does not regard extendible bonds as p a r t i c u l a r l y a t t r a c t i v e s e c u r i t i e s . -155-characterized U.S. i n s t i t u t i o n a l investors as constantly i n search f o r good qua l i t y investments and consequently less con-cerned about the l i q u i d i t y of a p a r t i c u l a r bond issue. This preoccupation of Canadian investors with the l i q u i d -i t y of t h e i r investments has made i t very d i f f i c u l t f o r most Canadian corporations to place t h e i r bonds p r i v a t e l y i n Canada, p a r t i c u l a r l y larger issues with long maturities. Peters observed that during the early s i x t i e s private placements amounted to up to ?0 per cent of the d o l l a r amount of a l l gross new corporate issues i n Canada and that t h i s percentage declined steadily since 1966 to about 2k per cent i n 1968. Our data indicate that 13 t h i s tendency has continued, see Table 5-1• On the other hand, most corporations v i s i t e d indicated a preference f o r private placements because they are regarded as o v e r a l l cheaper, easier and f a s t e r to arrange, and as providing considerably more f l e x i b i l i t y i n case trust deed or other changes become necessary i n l a t e r years. To avoid public exposure, some p r i v a t e l y owned firms w i l l consider private placements only. Public issues are looked upon as a means to improve public r e l a t i o n s , to provide stockholders with additional investment opportunities i n t h e i r company, and to make future financing easier. When faced with a choice between a public issue i n Canada and a private place-ment i n the United States many Canadian corporations have se-lected the l a t t e r option. 1 2 P e t e r s , p. 38. 13 -'In the United States the r e l a t i v e volume of the private placement market increased again i n the 'seventies; see John D. Rea and Peggy Brockschmidt, "The Relationship Between P u b l i c l y Offered and P r i v a t e l y Placed Corporate Bonds", Federal Reserve  Bank of Kansas C i t y Monthly Review, November, 1973, pp. 11-20. -156-We did not f i n d any evidence i n d i c a t i n g that Canadian borrowers have encountered s i g n i f i c a n t problems i n placing bonds d i r e c t l y i n the United States. This holds even f o r the l a t e ' s i x t i e s when a large demand for p o l i c y loans reduced the amount of free investment funds available to American l i f e insurance companies. Many Canadian corporations entertain a continuing relationship with t h e i r American lenders and seem to belong to that group of borrowers which receives p r e f e r e n t i a l treatment 14 durmg a cr e d i t squeeze. Also, several loan agreements signed during that time period contain delayed delivery provisions. The s t a t i s t i c a l r e s u l t s presented i n Tables 5-2a, 5-2b and 5-2c tend to support our e a r l i e r hypothesis that the a v a i l a b i l i t y of a private placement market i n the United States has been a major factor inducing Canadian corporations to borrow abroad.^ In a l l discriminant function estimates the private placement binary variable i s very s i g n i f i c a n t . Of course, t h i s dummy: variable also measures the influence which lower transaction costs have on corporations' preferences f o r d i r e c t placements, a point to be discussed l a t e r . One other factor, which i s c l o s e l y related to the dearth of a private placement market i n Canada, strongly a t t r a c t s cer-t a i n Canadian corporations to the U.S. market« the f a c t u a l non-a v a i l a b i l i t y of long-term forward commitments from Canadian f i n -a n c i a l i n s t i t u t i o n s . Some corporations are able to plan t h e i r needs fo r long-term funds well into the future and w i l l under-14, 15! See Section 3.42 above. See hypothesis H?, p. 94 above. TABLE 5-2a DISCRIMINANT ANALYSIS RESULTS, 1960 - 1967: TIME PREFERENCE MEASURES Discriminant Function Means and St. D.s. Item C o e f f i c i e n t s F-Values F-Probab. Can.-pay Bonds U.S.-pay Bonds Term to Maturity i n Years .074 1.79 (.1799) 19.29 (4.09) 20.12 (3.94) P r i v a t e Placement Dummy 2.567 25.26 (.0000) .533 (.149) .917 (.427) Constant -3.322 - - -Function - 13.32 (.0000) - -Number of Observations - 105 45 60 Ex Post C l a s s i f i c a t i o n : Percentage of Issues C o r r e c t l y C l a s s i f i e d 72% - 49% 90% TABLE 5-2b DISCRIMINANT ANALYSIS RESULTS, JANUARY 196 8 - MAY 1970: TIME PREFERENCE MEASURES Discriminant Function A Discriminant Function B Means and St. D.s. Item C o e f f i c i e n t s F-Values F-Probab. C o e f f i c i e n t s F-Values F-Probab. Can.-pay Bonds U.S.-pay Bonds Term to Maturity i n Years .0098 2.69 (.1010) .151 7.11 (.0090) 11.56 (7.34) 21.36 (6.30) Several Closings Dummy 10.990 54. 82 (.0000) 6.987 31. 32 (.0000) .031 (.174) .619 (.498) Pr i v a t e Placement Dummy 4.3178 14.65 (.0003) 4.840 26.31 (.0000) .185 (.391) . 857 (.359) Log (Years Elapsed between Contract and Delivery+1.0) 207.000 23.34 (.0000) - - .0042 (.0059) .0160 (.0205) Constant -9.700 - -7.277 - - -Function - 49.43 (.0000) - 45.68 (.0000) - -Number of Observations - 86 - 86 65 21 Ex Post C l a s s i f i c a t i o n : Percentage of Issues C o r r e c t l y C l a s s i f i e d 92% - 89% - A B 95% 89% A B 81% 90% TABLE 5-2c DISCRIMINANT ANALYSIS RESULTS, JUNE 1970 - MAY 1973: TIME PREFERENCE MEASURES Discriminant Function A Discriminant Function B Means and St. D.s. Item C o e f f i c i e n t s F-Values F-Probab. C o e f f i c i e n t s F-Values F-Probab. Can.-pay Bonds U.S.-pay Bonds Term to Maturity i n Years .121 2. 71 (.0985) .130 3.29 (.0692) 15.36 (6.42) 21. 97 (3.38) Several Closings Dummy 11.430 22.44 (.0000) 11.042 21.60 (.0000) .011 (.103) . 300 (.483) Pri v a t e Placement Dummy 6.590 34.49 (.0000) 6.590 35.69 (.0000) .149 (.358) .900 (.316) Log (Years Elapsed between Contract and Delivery+1.0) 210.000 1.79 (.1810) - - .0038 (.0021) .0053 (.0073) Constant -6.400 - -7.600 - - -Function - 20.55 (.0000) - 26.59 (.0000) - -Number of Observations - 104 - 104 94 10 Ex Post C l a s s i f i c a t i o n : Percentage of Issues C o r r e c t l y C l a s s i f i e d 85% - 87% - A B 86% 85% A B 80% 100% - 1 6 0 -take c e r t a i n investment projects only i f financing can be ar-ranged well i n advance. Almost without exception these firms are forced to borrow i n the United States because Canadian l i f e insurance companies are reluctant to commit t h e i r funds even a few months into the future at a predetermined interest rate. Two of the four interviewed insurance o f f i c e r s said that t h e i r firms do not enter into forward contracts, and the two others indicated that they, might agree to such a deal, but usually only i n a mortgage-type s i t u a t i o n . The pot e n t i a l loss i n y i e l d i n case interest rates r i s e i s seen as a major r i s k involved i n forward commitments, p a r t i c u l a r l y as supposedly some Canadian corporations defaulted on forward agreements i n the s i x t i e s during times when intere s t rates declined. It was also pointed out that i n s t i t u t i o n a l investors have no need f o r forward con-tr a c t s because s u f f i c i e n t a t t r a c t i v e investment opportunities are available on a spot basis. American investors i n search f o r a t t r a c t i v e investments fo r t h e i r huge cash flows are, on the other hand, prepared to enter into negotiations with Canadian corporations two to three years before the scheduled ( f i r s t ) closing, and i t i s not un-common to sign the f i n a l contract twelve to twenty months before the bonds w i l l be delivered. The time elapsed between signing of the f i n a l agreement and the date of ( f i r s t ) delivery of the bonds or between date of issue of the f i n a l prospectus and the 16 This statement applies only to investments i n corporate bonds and seems not to hold f o r other types of investments, par-t i c u l a r l y mortgage funds. - 1 6 1 -f i r s t o f f e r date has been measured f o r each issue. For the period i 9 6 0 to 1 9 6 7 , no s i g n i f i c a n t difference between Canadian-pay and U.S.-pay bonds could be detected, and the data indicate l i t t l e demand f o r long-term forward commitments by Canadian corp-orations. However, rapidly r i s i n g i n t e r e s t rates i n the late • s i x t i e s changed t h i s as evidenced by discriminant functions-A reported i n Tables 5-2b and 5-2c. It i s i n t e r e s t i n g to note that contracts f o r three U.S. d o l l a r issues delivered i n late 1970 and 1971 were signed i n 1969 already and therefore included i n estimates of discriminant functions f o r the period 1968 to May 1 9 7 0 . Major investment projects usually involve expenditures stretched over months and years. E s p e c i a l l y i f large amounts are involved, some companies prefer to obtain loan contracts providing f o r several closings rather than to r e l y on short-term interim financing. Again such agreements usually cannot be arranged f o r i n Canada and firms are forced to resort to the U.S. market. Our sample includes several issues where bonds were delivered i n up to four closings stretching over one to four years. Indeed, one large issue (even f o r the American bond market) has been drawn down i n sixteen closings over f i v e 17 years. ' Public offerings of Canadian corporate bonds i n the United States sometimes also include delayed delivery provisions. Unfortunately exact information on the number and dates of closings and the amount involved each time was not available 17 In a singular e f f o r t Canadian i n s t i t u t i o n a l investors provided funds f o r an additional Canadian-pay issue amounting to one-tenth of the U.S.-pay issue at almost i d e n t i c a l conditions. -162 i n many instances even though firms made attempts to recon-struct t h i s information from t h e i r f i l e s . Therefore only a dummy variable f o r more than one closing could be included i n the s t a t i s t i c a l analysis. As can be seen i n Tables 5-2b and 5-2c, i t was very important i n discriminating between the two types of bonds f o r the period from 1968 through May 1973. This variable did not improve the discriminatory power of the function reported i n Table 5-2a f o r the period i960 to 1967. However, our data suggest that the p o s s i b i l i t y of arranging f o r several clos-ings did a t t r a c t several Canadian issues to the U.S. during that time period, and t h i s binary variable i s s i g n i f i c a n t i n discrim-inant function A reported below i n Table 5-8a. The significance both of the variable measuring the time elapsed between signing of a loan agreement and the data of ( f i r s t ) delivery and of the several closings binary variable lends s t a t i s t i c a l support to our hypothesis Kj that the a v a i l -a b i l i t y of forward commitments has been a major factor a t t r a c t -lfi ing Canadian corporations to the American c a p i t a l market. It also indicates that lower transaction costs were not the main or at least not the only reason why Canadian corporations placed t h e i r bonds p r i v a t e l y i n the United States. E a r l i e r i t was asserted that differences i n time prefer-ences between Canada and the United States may lead to longer maturities on U.S.-pay bonds. 1^ Five (or roughly 2%) of the See p. 95 above. See p. 91 above. -163-managers interviewed said that longer terms to maturity a v a i l -able i n the U.S. market were a major factor of a t t r a c t i o n . A close analysis of our data showed that indeed the term to matur-i t y on some U.S.-pay bonds i s twenty-five to t h i r t y years or up to f i v e years longer than on comparable Canadian-pay bonds issued by the same or other firms belonging to the same indust-20 ry. But the a v a i l a b i l i t y of longer maturities i n the United States i s neither a widely known fac t nor do most Canadian corp-orations seem interested i n longer terms than are presently a v a i l -able i n Canada. In Table 5-3 the mean maturities of Canadian-pay and U.S.-pay bonds are compared f o r bonds having terms to maturity of more than twelve years and which have been issued by 21 corporations that have approached the U.S. market. In the l i g h t of the foregoing discussion i t i s not su r p r i s i n g that these data do not indicate a s i g n i f i c a n t difference i n the mean maturi-t i e s on bonds having r e l a t i v e l y long terms to maturity. There-fore, from a s t a t i s t i c a l point of view, our hypothesis has not been confirmed, though a s l i g h t tendency of U.S.-pay bonds to have longer maturities i s obvious. Almost a l l Canadian corporate bonds issued during the early and middle ' s i x t i e s had terms to maturity close to twenty years. With the increased l i q u i d i t y consciousness of i n s t i t u t i o n a l 20 However, as one interviewee pointed out, terms of more than t h i r t y years offered to some American firms, i n p a r t i c u -l a r to public u t i l i t i e s , are not available on bonds issued by Canadian corporations. 21 A maturity of twelve years has been chosen as a outpoint because standard terms seem to be f i v e , s ix, ten, twenty and twenty-five years to maturity with a gap between twelve and f i f t e e n years. TABLE 5-3 MEAN TERMS TO MATURITY OF BONDS HAVING TERMS TO MATURITY OF MORE THAN TWELVE YEARS AND WHICH HAVE BEEN SOLD BY CORPORATIONS THAT HAVE BORROWED IN THE UNITED STATES Period Can.-Pay Bonds U.S.-Pay Bonds I960 - 1967 20.31 . 2 0.46 (2.42, 42) a (3.5^. 58) Jan. 1968 - May 1970 21.78 22.15 (5.17, 16) (5.28, 20) June 1970 - May 1973 20.36 21.97 ( .87, 36) (3.38, 10) Values i n brackets are standard deviations and numbers of observations respectively. -165-investors and the t a i l o r i n g of issues to meet the medium-term investment needs of hanks, trust companies and other investors the mean term to maturity on a l l bonds offered i n Canada has de-creased considerably i n recent years. Including the term to maturity as a variable when estimating discriminant functions usually led to marginally s i g n i f i c a n t r e s u l t s depending somewhat on the*presence of other measures intended to capture differences i n time preferences. These re s u l t s are reported i n Tables 5-2a, 5-2b, and 5-2c. In order to allow f o r an a d d i t i o n a l evaluation of the discriminatory power of the estimated functions, ex post c l a s s i -f i c a t i o n r e s u l t s are reported i n the bottom of each Table. These re s u l t s are very s a t i s f a c t o r y , p a r t i c u l a r l y f o r the two more re-cent periods. As a further check on our estimates, a discrim-inant function including the same variables as those mentioned i n Table 5-2a was estimated f o r the period i960 to mid-1965 and used f o r predicting the population of the bonds i n the "CAandUS" sample 4 and the "USpay" sample 5 f o r the.period mid-1965 to 22 1967. Seventy-four per cent of a l l bonds were co r r e c t l y c l a s s i -f i e d . S i m i l a r l y , using mid-1965 to 1967 estimates to predict the group membership of bonds i n the "CAandUS" samples 2 and 3 and the "USpay" sample 5 f o r 1968 to May 1970 led to a correct 23 p r e d i c t i o n i n ninety per cent of the cases. ^ When discriminant 22 Mid-1965 was a convenient cutpoint leading to a r e l a t i v e -l y even d i s t r i b u t i o n of Canadian-pay and U.S.-pay bonds i n both the estimation and the prediction samples. 23 J I t should be remembered that f o r the. period i960 to 1967 data have been coll e c t e d only on bonds issued by corporations that have borrowed i n the United States. Estimates based on these data should therefore not be used to predict the group membership of bonds issued by Canadian firms that never sold U.S. - 1 6 6 -function A reported i n Table 5-2b wa's employed to c l a s s i f y bonds i n samples 1 ("CAonly"), 2 ("CAandUS") and 5 ("USpay") f o r the -period June 1970 to May 1973» ninety-three per cent of a l l bonds but only f o r t y per cent of the U.S.-pay bonds were c o r r e c t l y c l a s s i f i e d . Function B led to a correct prediction f o r a l l U.S.-pay bonds, eighty-nine per cent of Canadian-pay bonds and ninety-one per cent of a l l bonds. Note that t h i s l a s t r e s u l t i s better than the ex post c l a s s i f i c a t i o n obtained from discrim-inant function B i n Table 5-2c. To summarize our analysis, there are strong indications that the r e l a t i v e l y higher demand f o r funds i n Canada than i n the United States has made i t possible f o r Canadian f i n a n c i a l i n s t i t u t i o n s to be selective and to invest only i n r e l a t i v e l y l i q u i d corporate bonds without facing a dearth of a t t r a c t i v e i n -24 vestment opportunities. This has resulted i n a rather t h i n private placement market i n Canada. Canadian investors are generally not w i l l i n g to commit funds f o r delivery several months or even years into the future or to permit a corporation to draw down a loan i n several closings. American i n s t i t u t i o n a l investors, p a r t i c u l a r l y the larger l i f e insurance companies, seem to be l e s s concerned with the l i q u i d i t y of t h e i r invest-d o l l a r denominated bonds, i . e . , bonds i n sample 1. On the other hand, 1968 to 1973 estimates are based on a comparison of our random sample of Canadian-pay bonds (samples l.and 2) with the sample of U.S.-pay bonds (sample 5). 24 One l i f e insurance o f f i c e r pointed out that recently, at times, Canadian l i f e insurance companies were r e l a t i v e l y l i q u i d and therefore prepared to accept private placements with interest rates comparable to those on public issues only i n order to not forego an investment i n an a t t r a c t i v e corporation. But these seem to have been rare instances. And, as one corporate manager said, the size of such issues i s l i m i t e d f o r c i n g companies either to resort to public issues or to place part or a l l of an issue i n the United States i f large amounts of funds are needed. -167-vents and eager to maintain a f u l l y invested p o s i t i o n . Their willingness to enter into direct placement agreements has been a major factor a t t r a c t i n g Canadian corporate borrowers to the United States c a p i t a l market. The discriminant analysis r e s u l t s reported i n Tables 5-2a, 5-2b and 5-2c tend to confirm t h i s ob-servation. They also support our hypothesis that the a v a i l a -b i l i t y of long-term forward commitments and the p o s s i b i l i t y of taking a loan down i n several closings stretched-' sometimes over several years leads Canadian firms to borrow abroad. The hy-pothesis that U.S.-pay bonds w i l l have, on average, longer terms to maturity than Canadian-pay bonds issued by the same group of firms when comparing only bonds that have r e l a t i v e l y long terms to maturity i s not confirmed by the data reported i n Table 5-3. But our interviews revealed that terms to maturity available i n the United States are sometimes longer than the longest terms offered to the same firms i n Canada and that t h i s i s a factor making the American market more a t t r a c t i v e to some f i n a n c i a l o f f i c e r s i n Canada. Mainly due to lender demand the mean ma-t u r i t y of Canadian-pay bonds has decreased considerably i n re-cent years whereas no such change could be detected f o r U.S. do l l a r bonds. -168-5.4 EXCHANGE RATE RISK AND INTERNATIONAL LONG-TERM FINANCING BY CANADIAN CORPORATIONS When issuing a "bond denominated i n a foreign currency, a firm faces two peculiar problems. ( 1 ) The proceeds of the issue have to be converted into Canadian d o l l a r s , and the exchange rate p r e v a i l i n g at the time of conversion determines the amount of Canadian do l l a r s a c t ually obtained. ( 2 ) Changes i n exchange rates during the bonds*lifetime may change the actual costs of foreign borrowing considerably. Through our interviews we at-tempted to gain a better understanding of how Canadian corpora-tions a c tually evaluate and manage these r i s k s involved i n bor-rowing i n the United States. We were very surprised to d i s -cover that many f i n a n c i a l o f f i c e r s of major Canadian corporations seemed not to regard the p o s s i b i l i t y of an adverse change i n the exchange rate as a major r i s k involved i n international f i n -a n c i a l transactions, at least not with respect to U.S. d o l l a r borrowing. The question "When you were contemplating the i s s u -ance of U.S. d o l l a r bonds, did you regard the p o s s i b i l i t y of a devaluation of the Canadian d o l l a r as a serious r i s k influencing your choice of where to f l o a t the new bonds?"was answered with "No" by t h i r t e e n (or 62%) of the managers interviewed. Seven of these firms did not have enough foreign exchange earnings to cover regular interest and sinking fund payments on outstanding U.S.-pay bonds. Only eight o f f i c e r s acknowledged that foreign exchange r i s k entered t h e i r decision as a major vari a b l e . Three managers defended t h e i r neglect of foreign exchange r i s k with the argument that, i n t h e i r opinion, the Canadian-U.S. d o l l a r exchange rate w i l l continue to fluctuate around par - 1 6 9 -so that exchange gains and losses w i l l o f f s e t each other i n the long run. This seems to he a widely held b e l i e f . Five execu-tives were more s p e c i f i c and pointed out that as long as the American d o l l a r was at a premium v i s - a - v i s the Canadian d o l l a r only an upward revaluation of the Canadian d o l l a r was expected. Consequently they did not think that borrowing i n the United States entailed any exchange r i s k at a l l . We also .encountered four instances where the issuing corp-orations have been able to cover part or a l l of the exchange r i s k through agreements with Canadian and American customers. As a rule, the firms involved i n such contracts are u t i l i t i e s or regu-lated corporations i n the o i l and gas industry. Two types of agreements seem to prevail» (1) Exchange losses and gains are f o r the customer's account. (2) The customer i s b i l l e d i n Cana-dian dollars but, on a one-to-one basis, must provide s u f f i c i e n t U.S. dollars to enable the bond issuer to discharge a l l his foreign currency interest and p r i n c i p a l payments which become due during a given period. Of those corporations which indicated that foreign exchange r i s k was a major variable entering t h e i r borrowing decision only one attempted to forecast the future trend of the exchange rate. A l l others r e l i e d on a more or l e s s subjective evaluation of the protection a given interest rate d i f f e r e n t i a l provides against unfavourable exchange rate changes. Two firms showed us studies i n which they had attempted to quantify the exchange rate change necessary to eliminate the cost advantage of lower inte r e s t rates abroad, but both did not use a present value approach. Other -170-firms based t h e i r r i s k evaluation on information obtained from underwriters 2^, Canadian and American banks and in-house econo-mists . Except f o r three instances where the U.S. d o l l a r proceeds were used to reduce short-term U.S. d o l l a r l i a b i l i t i e s or to import equipment into Canada, firms converted these foreign funds almost immediately into Canadian currency. Usually t h i s i s done through the spot exchange market over a time period which, de-pending on the amount involved, may stretch over up to two weeks i n order to avoid disruptions of the exchange market that might lead to unfavourable exchange rates. Only four corporations attempted to completely eliminate the exchange r i s k involved i n such transactions by negotiating forward exchange contracts that matured at the time of delivery of the bonds. Four other firms mentioned that they sold part of the proceeds of an issue i n the forward exchange market. During the term of the bonds, U.S. dol l a r s needed to meet interes t and sinking fund payments are usually obtained only a few days beforehand i n the spot exchange market unless company revenue denominated i n U.S. dollars i s av a i l a b l e . Only three of the larger corporations interviewed indicated that they have 2-'in general, investment bankers play an important role i n a firm's decision to place a,.new issue i n the United States. When they recommend the sale of U.S.-pay bonds, they are usually asked by t h e i r c l i e n t s f o r an opinion regarding the exchange r i s k i n -volved. The economist of one investment dealer demonstrated f o r us t h e i r computer program that calculates the protection against unfavourable exchange rate changes which a given interest rate d i f f e r e n t i a l provides. The output data seemed to be i d e n t i c a l to those calculated by our program and presented above i n Table 3 - 6 . -171-attempted to cover foreign exchange l i a b i l i t i e s i n the forward market f o r up to one and a half years. One other firm even ob-26 tained coverage f o r up to four years i n advance. However, i n the l a t t e r case the trust deed r e s t r i c t s that corporation's out-standing foreign currency debt which i s not protected against exchange losses through forward exchange contracts to twenty per cent of i t s t o t a l debt. Consequently t h i s firm covers as much of i t s foreign currency l i a b i l i t i e s as possible through for-ward exchange contracts i n order to be able to increase i t s U.S. d o l l a r borrowings. Even at times of an unsettled international monetary scene usage of the forward market by Canadian corpora-tions to reduce foreign exchange r i s k does not increase. Trustees and lenders also seem not too concerned with f o r -eign exchange r i s k . We came across only two instances where a t r u s t deed took foreign exchange r i s k e x p l i c i t l y into account. In one case, a l l U.S. d o l l a r revenue of a Canadian corporation has to be routed through an account with i t s New York trustee. Af-ter making monthly deductions f o r i n t e r e s t and sinking fund payments the remaining sum i s remitted to Canada. The other case has been discussed i n the preceding paragraph. In December 1973 one of the companies interviewed, also a regular borrower i n the United States, proposed to revise i t s indenture by includ-ing, amongst others, an a r t i c l e dealing with the valuation of foreign currency debt. This could become very important i n d i s -putes about whether or not a corporation has met c e r t a i n pro-Such long-term deals are not r e a d i l y a v a i l a b l e . However, they can be arranged through banks occasionally i f s u f f i c i e n t time i s permitted. - 1 7 2 -visions s tipulated i n i t s t r u s t deed. Four of the six Canadian investment "bankers interviewed did also not attach major importance to exchange rate r i s k . One of them even said that because of the close t i e s between the Canadian and American economies "we think that exchange r i s k i s not a relevant v a r i a b l e . " Two of them also mentioned that, i n t h e i r opinion, some Canadian firms are overly concerned with foreign exchange r i s k , an obvious reference to those major Cana-dian corporations that do not borrow i n the United States. How-ever, three of the six underwriters pointed out that, as a gener-a l rule, they prefer to recommend U.S.-pay issues only to corporations with s u f f i c i e n t U.S. d o l l a r income to cover the exchange r i s k . Given the p r e v a i l i n g expectation that the value of the Canadian currency v i s - a - v i s the U.S. d o l l a r w i l l fluctuate around a par value of one Canadian to one U.S. d o l l a r , we showed i n Section 3 « 3 2 above that foreign exchange r i s k w i l l decrease as 27 a foreign-pay bond's term to maturity increases. ' The r e s u l t s reported e a r l i e r i n Tables 5-2a, 5-2b and 5-2c and the discrim-inant functions shown i n Tables 5-4a and 5-4b would lend support to the a s s e r t i o n that Canadian corporations are aware of t h i s e f f e c t and prefer to s e l l U.S.-pay bonds with long terms to maturity i n order to reduce foreign exchange r i s k . However, only two of the twenty-one corporate managers interviewed had recognized t h i s r elationship between exchange r i s k and term to 27 See hypothesis H^, p. I l l above. TABLE 5-4a DISCRIMINANT ANALYSIS RESULTS, 1960-196?« EXCHANGE RISK MEASURES Item Discriminant Function Means and St. D.s Coe f f i c i e n t s F-Values F-Probab. Can.-pay Bonds U.S.-pay Bonds Term to Maturity i n Years .093 1.20 (.2756) 19-29 (4.09 ) 20.12 (3.9^ ) Average Maturity i n Years - .162 4.70 (.0307) 14.21 (*.55 ) 14.18 (4.64 ) Risk Premium -2.732 41.18 (.0000) 1.077 ( .537) .462 ( .478) Constant 2.564 - - -Function -14.83 (.0000) - -Number of Observations - 105 ^5 60 Ex Post C l a s s i f i c a t i o n Percentage of Issues Correctly C l a s s i f i e d 78% - 73% 82% TABLE 5-4b DISCRIMINANT ANALYSIS RESULTS, JANUARY 1968-MAY 1970» EXCHANGE RISK MEASURES Discriminant Function Means and St. D.s. Item Coe f f i c i e n t s F-Values F-Probab. Can.-pay Bonds U.S.-pay Bonds Term to Maturity i n Years .451 15.22 (.0003) 11.56 (7.3^ ) 21.36 (6.30 ) Average Maturity i n Years - .39^ 6.35 (.0132) 9.65 (5.44 ) 14.73 (5-58 ) Risk Premium -1.251 3.12 (.0776) 1.224 ( .445) .892 ( .512) Log (Percentage of Revenue Denominated i n US $+1.0) 1.211 4.01 ( .046) .130 ( .429) .701 ( .825) Constant -1.795 - - -Function -13.46 (.0000) - -Number of Observations - 86 65 21 Ex Post C l a s s i f i c a t i o n Percentage of Issues Correctly C l a s s i f i e d 80% - 83^ 71# -175-maturity. Most underwriters seemed not to be aware of the fact that the mean term to maturity on Canadian-pay bonds has decreased considerably since the early and middle ' s i x t i e s where-as t h i s does not hold f o r U.S.-pay bonds. As an explanation they offered the hypothesis that there may ex i s t a market i n the United States f o r medium-term bonds issued by Canadian corpora-tions but that placement agents had neglected to develop t h i s market. Our own explanation i s that, as a rule, Canadian corpor-ations prefer to issue long-term bonds. Whereas U.S. l i f e i n -surance companies are b a s i c a l l y long-term investors, Canadian i n s t i t u t i o n s have developed a ce r t a i n preference f o r medium-term issues which has led at times to lower i n t e r e s t rates on these bonds. As a consequence, corporate borrowers sometimes trade o f f interest costs and term to maturity on Canadian-pay bonds. This has not been necessary i n the United States because of a ready dir e c t placement market f o r Canadian corporate bonds with long terms to maturity. E a r l i e r we had also asserted that, i n order to reduce ex-change r i s k , sinking fund payments on U.S.-pay bonds w i l l be more 2 8 evenly d i s t r i b u t e d over time than those on Candian-pay bonds. Or, i n other words, given a bond's term to maturity, the average 29 term to maturity on U.S.-pay bonds w i l l be shorter. ' The s i g -28 See hypothesis H/-, p. 113 above. 29 ^Average maturity has been measured as the mean maturity of a l l sinking fund and balloon payments weighted by t h e i r s i z e . When computing a bond's average maturity, only mandatory sinking fund or purchase fund provisions have been taken into account. In-dividual payments have not been discounted as i s done when calcu-l a t i n g a bond's "duration". For a discussion of t h i s concept see Lawrence Fisher and Roman L. Weil, "Coping With the Risk of Fluc-tuations t Returns to Bondholders from Naive and Optimal Strate-gies," Journal of Business, LXVII (October, 1971), pp. 415-420. -176-n i f i c a n t negative c o e f f i c i e n t s of the average maturity variable i n the discriminant functions reported i n Tables 5-^a and could be interpreted as confirming t h i s hypothesis. However we are reluctant to give such an i n t e r p r e t a t i o n to our data. In our interviews we discussed t h i s point and discovered that underwriters and f i n a n c i a l managers do generally not regard sink-ing fund provisions as a potential instrument to reduce foreign exchange r i s k . I t seems that sinking fund payments are usually determined by U.S. lender preferences.-^ Whether exchange r i s k considerations enter into U.S. l i f e insurance companies' deliber-ations about sinking fund requirements on Canadian bonds i s not known. As an a d d i t i o n a l measure of the influence which exchange r i s k has on Canadian corporate borrowing i n the United States we included the r i s k premium on corporate bonds, the difference between.a bond's y i e l d to maturity and the y i e l d on long-term Canadian government bonds, into the discriminant analysis. As can be seen i n Table 5-^a, t h i s variable i s extremely s i g n i f i -cant f o r the early period, but i n Table 5-^0 i t i s no longer s i g n i f i c a n t by common standards. It w i l l be remembered that we also hypothesized that Cana-dian firms with U.S. d o l l a r revenue w i l l show a greater tendency -^In addition i t was pointed out during interviews that i n -s t i t u t i o n a l investors, p a r t i c u l a r l y American l i f e insurance com-panies, have a preference f o r pro-rata sinking funds. We did not inquire into why Canadian lenders are prepared to accept r e l a -t i v e l y longer average terms to maturity than U.S. lenders. We became aware of t h i s only through our s t a t i s t i c a l analysis which was performed a f t e r we had conducted the interviews. -177-to borrow i n the United States than companies not engaged i n ex-31 port a c t i v i t i e s . Unless information on the percentage of t o t a l revenue denominated i n U.S. dol l a r s during the year pre-ceding the date of issue was obtained from the firm i t s e l f , t h i s datum was estimated from information contained i n prospectuses and annual reports. For the period 1968 to May 1970 the discrim-inant function reported i n Table indicates that corporations with U.S. d o l l a r revenue indeed exhibited a greater preference to s e l l U.S.-pay bonds than those firms active only i n domestic 32 markets. But again we are reluctant to regard t h i s as a con-clusive confirmation of our hypothesis. For the period June 1970 to May 1973 no evidence f o r such a relat i o n s h i p between ex-port a c t i v i t i e s and foreign borrowing could be found i n spite of the fact that the Canadian d o l l a r had appreciated consider-ably i n 1970 and has been f l o a t i n g since June of that year. In-deed, none of our exchange r i s k measures l e d to a meaningful es-timate of a discriminant function f o r t h i s l a t t e r period. This may be pa r t l y due to a r e l a t i v e l y large proportion of public u t i l i t y bonds i n our sample of U.S.-pay bonds f o r that period. Through our interviews we gained the impression that Canadian f i n a n c i a l managers see only a loose r e l a t i o n s h i p between export earnings and foreign borrowing and do not regard the issuance of U.S.-pay bonds as a means to hedge future U.S. d o l l a r income. 31 See hypothesis Hj,, p. 98 above. 32 J As our estimates f o r the period i 9 6 0 to 1967 are based only on samples of bonds issued by corporations that approached both the Canadian and the American bond markets, the export revenue variable cannot be expected to be s i g n i f i c a n t f o r that period. - 1 7 8 -When te s t i n g the predictive power of the discriminant function reported i n Table 5-^ a t eighty-one per cent of the bonds i n our t e s t sample f o r the period mid -1965 to 1967 were corre c t l y c l a s s i f i e d and eighty per cent of the bonds f o r the period 1968 to May 1 9 7 0 . Using the function shown i n Table 5-^b, only seven-ty per cent of the bonds i n our test sample f o r the period June 1970 to May 1973 were correctly c l a s s i f i e d . These r e s u l t s , p a r t i c u l a r l y the l a s t one, are not very encouraging and seem to indicate that these relationships are not very stable over time. In summary, both the information obtained through i n t e r -views and our data suggest that foreign exchange r i s k had only a weak influence on the c h a r a c t e r i s t i c s of bonds placed i n the United States. S t a t i s t i c a l evidence i n support of our hypothe-ses that U.S.-pay bonds have r e l a t i v e l y long terms to maturity and provide f o r sinking fund payments to be evenly d i s t r i b u t e d over time has been presented. But our interviews led us to con-clude that these differences i n bond c h a r a c t e r i s t i c s are mainly due to differences i n lender preferences rather than due to deliberate attempts by Canadian corporations to reduce exchange r i s k . S i m i l a r l y , data to support the claim that there exists a re l a t i o n s h i p between a firm's U.S. d o l l a r revenue and i t s foreign borrowing have been reported. But again the evidence i s r e l a t i v e l y weak. Foreign currency borrowing i s not seen as a means to reduce the exchange r i s k connected with export earn-ings. Even corporations with large and stable U.S. d o l l a r reven-ues f e e l no incentive to s e l l U.S. d o l l a r bonds unless i t involves some expected interest cost savings. Except f o r such a trade-off - 1 7 9 -between exchange r i s k and lower interest costs abroad, Canadian corporations do usually not make deliberate e f f o r t s to reduce foreign exchange r i s k further through varying a bond's character-i s t i c s or through extensive use of the forward exchange market. The r i s k premium variable which intends to measure the advantage of lower inter e s t costs i n the United States was extremely s i g -n i f i c a n t i n our estimate of a discriminant function f o r the early and middle ' s i x t i e s but i s no longer s i g n i f i c a n t i n estimates f o r recent periods. This seems to indicate that factors other than exchange r i s k (or lower interest costs abroad) have become dom-inant i n Canadian corporations' decisions to issue U.S.-pay bonds. 5 . 5 TRANSACTION COSTS DIFFERENCES AND INTERNATIONAL DEBT ISSUES For the period 1968 to 1973 data regarding placement fees and other issuing expenses on p r i v a t e l y placed Canadian-pay and 33 U.S.-pay bonds are reported i n Table 5 - 5 a . Total f l o t a t i o n costs are s i m i l a r f o r both types of bonds with a s l i g h t tendency to be higher on U.S. d o l l a r issues. As expected, t h i s i s mainly due to "other expenses". These data confirm the opinion voiced by most of the interviewed managers that placement fees are about the same on Canadian-pay and U.S.-pay issues but that other trans-action costs are higher f o r foreign currency bonds. Our e f f o r t s to obtain at least very rough estimates of transaction costs incurred with regard to a firm's o f f i c i a l s ' time spent on nego-t i a t i n g a p a r t i c u l a r issue and t r a v e l and s i m i l a r expenses were a complete f a i l u r e . No corporation seems to c o l l e c t such data. 3 3 For the period i 9 6 0 to 1967 the information on f l o t a t i o n costs of d i r e c t l y placed Canadian-pay bonds was too l i m i t e d to allow a meaningful comparison. Even f o r 1968 to 1973 "the data are incomplete and can only be regarded as suggestive. TABLE 5-5a FLOTATION COSTS OF PRIVATELY PLACED CORPORATE BONDS: 1968-1973 AS A PERCENTAGE OF GROSS PROCEEDS Size of Issue Average Size No. of Issues Placement Fee Other Expenses Total Flotation Costs ( M i l l , of $$) of Issues Average Range Average Range Average Range Canadian--Pay Bonds up to 2.50 1.450 2 4.100 2.20-6.00 1.444 0.22-2.67 5.544 2.42-8.67 • 2.50- 6.99 4.611 9(6)a 0.901 0.25-1.75 0.715 0.24-1.33 1.615 0.84-2.83 7.00-11.49 9.556 9 0.798 0.37-1.25 0.255 0.12-0.50 1.053 0.53-1.53 11.50-27.A9 18.667 3 0.799 0.25-1.00 0.156 0.13-0.19 0.955 0.42-1.19 27.50-62.49 37.500 4 0.452 0.25-0.87 0.285 0.04-0.92 0.737 0.39-1.17 62.50 & over 75.000 1 0.250 - 0.067 - 0.317 -U.S.-Pay Bonds up to 2.49 - - - - - - - -2.50- 6.99 4.250 2 ( l ) a 0.779 0.25-1.25 0.444 - 1.223 -7.00-11.49 9.333 3 0.661 0.50-0.75 0.711 0.18-1.66 1.372 0.93-2.16 11.50-27.49 17.429 7 0.700 0.25-1.00 0.143 0.06-0.23 0.842 0.42-1.06 27.50-62.49 44.500 4 0.521 0.40-0.60 0.678 0.25-1.30 1.199 0.65-1.90 62.50 & over 189.000 4 0.567 0.40-1.10 0.210 0.09-0.31. 0.777 0.59-1.37 Number of issues for which data on other expenses were available. TABLE 5-5b FLOTATION COSTS OF PUBLICLY OFFERED CORPORATE BONDS: 1968-1973 AS A PERCENTAGE OF GROSS PROCEEDS Size of Issue Average Size No. of Issues Underwrit ing Spread Other Expenses Total Flotation Costs (.Mill, of ?S) of Issues Average Range Average Range Average Range Canadian-Pay Bonds up to 0.99 0.462 2 5.297 3.00-8.00 2.378 1.40-3.53 7.675 4.40-11.53 1.00- 2.49 1.333 3 2.250 1.50-2.50 0.950 0.25-2.00 3.200 2.75- 3.80 2.50- 4.49 3.206 16 2.597 1.00-6.00 0.836 0.17-1.67 3.433 2.00- 7.67 4.50- 6.99 5.394 17 2.337 0.50-4.00 0.536 0.14-1.25 2.873 0.90- 4.83 7.00-11.49 9.041 35 2.078 0.50-3.00 0.422 0.09-1.13 2.500 0.70- 3.71 11.50-17.49 14.526 19 2.311 1.50-4.00 0.346 0.17-1.00 2.657 1.70- 4.67 17.50-27.49 22.076 33 1.695 0.0 -2.50 0.246 0.05-0.75 1.940 0.09- 2.90 27.50-42.49 33.917 12 1.569 0.75-2.00 0.190 0.01-0.50 1.759 0.76- 2.50 42.50-62.49 52.368 19 1.705 1.30-2.40 0.185 0.05-0.33 1.890 1.45- 2.61 62.50 & over 92.500 6 1.637 1.30-2.00 0.199 0.10-0.30 1.836 1.40- 2.22 U.S.-Pay Bonds a up to 17.49 - - - - _ _ 17.50-27.49 25.000 2 1.060 1.00-1.12 0.550 0.38-0.72 1.610 1.38- 1.84 27.50-42.49 30.000 2 1.060 1.00-1.12 0.371 0.23-0.51 1.431 1.23- 1.63 42.50-62.49 50.000 1 1.050 - 1.130 - 2.180 _ 62.50 & over 150.000 1 0.875 - 0.147 - 1.022 Data include two issues sold during the second half of 1967. -182-But i t was pointed out that quite often the Vice President Finance spends the larger part of his time on securing long-term funds, and sometimes even the President i s extensively involved i n such a c t i v i t i e s . In one instance three employees of a large Canadian corporation spent about one year on preparing comparative analy-ses of Canadian and U.S. c a p i t a l markets. They also made an ex-tensive evaluation of the exchange r i s k before the firm decided to place i t s f i r s t large U.S.-pay issue. Of eleven managers that ventured a guess seven estimated that U.S. d o l l a r issues are more costly than Canadian d o l l a r issues i n terms of time spent on preparing and s e l l i n g an issue and i n terms of expenses usual-l y not allocated to "other issuing expenses". Four thought there i s no difference. In Table 5~5"b f l o t a t i o n costs on public bonds issued since 34 1968 are compared. There exists only a small number of public-l y sold U.S.-pay bonds ranging i n size from twenty-five to one hundred and f i f t y m i l l i o n U.S. d o l l a r s . For these large bond issues underwriting spreads are considerably lower i n the United States than i n Canada whereas "other issuing expenses" tend to be higher. Though t o t a l f l o t a t i o n expenses are generally lower i n the United States on these large issues, the dearth of public offerings of Canadian corporate bonds i n the United States seems to be due mainly to three factorst (1) The r e l a t i v e ease with 3 4 The data on U.S.-pay bonds include two issues sold during the second h a l f of 1967. For the period i960 to mid-1967 we could i d e n t i f y only three public offerings, two of which were below four m i l l i o n d o l l a r s i n size and which had f l o t a t i o n costs considerably i n excess of those on comparable bonds sold i n Can-ada. - 1 8 3 -which dire c t placements can be secured i n the United States; (2 ) the lengthy and, i n terms of management time, c o s t l y process of preparing a public U.S. d o l l a r o f f e r i n g f o r sale; and (3 ) the hesitance of most Canadian firms to r e g i s t e r with the Ameri-can Securities and Exchange Commission because of the information they have to disclose and the costs involved i n keeping t h e i r 35 f i l e i n New York up-to-date. J But Canadian managers and i n -vestment bankers think that some of the planned huge energy de-velopment projects can only be financed through public offerings i n the New York and perhaps the European and Japanese c a p i t a l • markets. Consequently we may see i n future an increase i n pub-l i c l y offered foreign currency bonds issued by Canadian corpora-tions. In view of the foregoing discussion i t i s not surprising that we found considerable s t a t i s t i c a l evidence (see Tables 5 - 6 a , 5 - 6 b , and 5 - 6 c ) i n support of our assertion that U.S.-pay issues w i l l , on average, be larger i n size than Canadian-pay issues and that only larger Canadian corporations place t h e i r bonds i n the 36 United States. The size of the issuing company has been measured by i t s assets. The negative sign of t h i s variable i n 35 -^ Some firms believe that they would have to change t h e i r accounting system i n order to produce a l l the information requir-ed by the SEC. At least one Canadian corporation that i s a regu-l a r borrower i n the United States but which has not yet issued a public bond denominated i n U.S. d o l l a r s has nevertheless r e g i s -tered with the SEC to provide ready access to information on i t -s e l f to the New York investment community and to be able to act f a s t e r i n case a public o f f e r i n g becomes desirable. •^See hypothesis H 7, p. 121 above. For a d d i t i o n a l evidence see Tables 5 - 8 a , 5 - 8 b , and 5 - 8 c below. - The c o r r e l a t i o n c o e f f i -cients between the logarithms of issue size and asset size of i s -suing corporation were . 6 3 0 f o r 1969 to 1 9 6 7 , .664 f o r 1968 to 1970 and . 8 0 7 f o r 1970 to 1973 f o r a l l bonds i n our samples. T A B L E 5-6a D I S C R I M I N A N T A N A L Y S I S R E S U L T S , 1960-1967: M E A S U R E S R E L A T E D T O T R A N S A C T I O N S C O S T S D i s c r i m i n a n t F u n c t i o n M e a n s a n d S t . D . s . I t e m ..- C o e f f i c i e n t s F - V a l u e s F - P r o b a b . C a n . - p a y B o n d s U . S . - p a y B o n d s P r i v a t e P l a c e m e n t Dummy 2.608 24.60 ( .0000) .533 (.505) .917 (.279) L o g ( I s s u e S i z e i n - M i l l , o f $ $ ) 1.446 6.70 ( .0107) .857 (.510) 1.050 (.514) L o g ( A s s e t s i n M i l l , o f $ $ ) -.781 2.87 ( .0894) 2.072 (.618) 2.058 (.641) C o n s t a n t -1.657 — — — F u n c t i o n — 10.83 ( .0000) — — N u m b e r o f O b s e r v a t i o n s . . 105 45 60 E x P o s t C l a s s i f i c a t i o n P e r c e n t a g e o f I s s u e s 72% — 55% 83% C o r r e c t l y C l a s s i f i e d T A B L E 5 - 6 b D I S C R I M I N A N T A N A L Y S I S R E S U L T S , J A N U A R Y 1 9 6 8 - M A Y 1 9 7 0 : M E A S U R E S R E L A T E D T O T R A N S A C T I O N S C O S T S D i s c r i m i n a n t F u n c t i o n M e a n s a n d S t . D . s . I t e m C o e f f i c i e n t s F - V a l u e s F - P r o b a b . C a n . - p a y B o n d s U . S . - p a y B o n d s L o g ( I s s u e S i z e i n M i l l o f $ $ ) 3 . 0 7 3 1 4 . 5 1 ( . 0 0 0 4 ) . 9 6 4 ( . 3 6 6 ) 1 . 3 8 8 ( . 6 3 0 ) L o g ( P e r c e n t a g e o f C o m p a n y O w n e d b y F o r e i g n e r s + 1 . 0 ) 1 . 0 6 3 5 . 0 2 3 ( . 0 2 6 3 ) . 4 2 7 ( . 7 3 7 ) 1 . 1 0 8 ( . 8 7 1 ) P r i v a t e P l a c e m e n t Dummy 5 . 5 5 1 4 7 . 7 6 ( . 0 0 0 0 ) . 1 8 5 ( . 3 9 1 ) . 8 5 7 ( - 3 5 9 ) E a r l i e r U . S . - p a y I s s u e s Dummy 1 . 4 3 1 3 . 4 6 (. . 0 6 3 4 ) . 3 5 4 ( . 4 8 2 ) . 6 6 7 ( . 4 8 3 ) C o n s t a n t - 8 . 0 5 3 — — — F u n c t i o n — 2 3 . 7 4 ( . 0 0 0 0 ) — — N u m b e r o f O b s e r v a t i o n s 86 — 6 5 2 1 E x P o s t C l a s s i f i c a t i o n P e r c e n t a g e o f I s s u e s C o r r e c t l y C l a s s i f i e d 87% — 88% 86% T A B L E 5 - 6 c D I S C R I M I N A N T A N A L Y S I S R E S U L T S , J U N E 1 9 7 0 - M A Y 1 9 7 3 : M E A S U R E S R E L A T E D TO T R A N S A C T I O N S C O S T S D i s c r i m i n a n t F u n c t i o n M e a n s a n d S t . D . s . I t e m C o e f f i c i e n t s F - V a l u e s F - P r o b a b . C a n . - p a y B o n d s U . S . - p a y B o n d s L o g ( P e r c e n t a g e o f C o m p a n y O w n e d b y F o r e i g n e r s O t h e r T h a n A m e r i c a n s + 1 . 0 ) 1 . 7 0 6 5 . 0 9 ( . 0 2 4 9 ) . 1 5 2 ( . 4 9 1 ) . 5 2 8 ( . 8 5 1 ) L o g ( A s s e t s i n M i l l , o f $ $ ) . 6 2 8 2 . 3 0 ( . 1 2 8 0 ) 2 . 1 7 4 ( . 7 6 2 2 . 4 9 2 ( . 6 0 1 ) P r i v a t e P l a c e m e n t Dummy 6 . 3 2 4 4 1 . 38 ( . 0 0 0 0 ) . 1 4 9 ( . 3 5 8 ) . 9 0 0 ( . 3 1 6 ) C o n s t a n t - 5 . 8 2 8 — — — F u n c t i o n 1 6 . 7 0 ( . 0 0 0 0 ) — — N u m b e r o f O b s e r v a t i o n s — 1 0 4 9 4 1 0 E x P o s t C l a s s i f i c a t i o n : P e r c e n t a g e o f I s s u e s C o r r e c t l y C l a s s i f i e d 86% — 85% 90% -187-the discriminant function reported i n Table 5-6a where we compare only bonds issued by those corporations that have borrowed i n the United States indicates that these corporations place t h e i r l a r g -er bonds abroad and smaller ones i n Canada. Given a firm's assets size, the p r o b a b i l i t y of a bond being c l a s s i f i e d as a U.S. d o l l a r issue increases with issue s i z e . The data i n Table 5-7 confirm t h i s , too, except f o r the period June 1970 to May 1973» As w i l l be discussed i n the l a s t section of t h i s chapter, p o l i t i -c a l pressure exerted upon Canadian underwriters and corporations i s at least p a r t l y to blame fo r the small number of issues and the modest amounts borrowed i n the United States during t h i s time i n t e r v a l . No evidence could be detected that Canadian corporations decreased the size of t h e i r U ,S .-pay issues i n order to reduce exchange r i s k when the Canadian d o l l a r was f l o a t -ing. This tendency that only larger Canadian corporations place only t h e i r larger bond issues i n the United States i s reinforced by the lending behaviour of U.S. l i f e insurance companies. It i s not uncommon f o r these i n s t i t u t i o n s to send t h e i r investment analysts to Canada f o r an extensive study of a p a r t i c u l a r corp-oration, and i t may take several weeks before they commit t h e i r funds. A l l underwriters said that U.S. lenders generally look only at medium to large corporations and are only interested i n 3 7 lending larger amounts. ' Otherwise i t would not be "worth t h e i r -''But some Canadian firms, mainly smaller finance companies, have been able to place now and then small U.S.-nay issues. T A B L E 5 - 7 M E A N I S S U E S I Z E O F B O N D S S O L D B Y C O R P O R A T I O N S W H I C H H A V E BORROWED I N T H E U N I T E D S T A T E S ( I n m i l l i o n s o f d o l l a r s ) P e r i o d C a n . -- P a y B o n d s U . S . - P a y B o n d s 1 9 6 0 -- 1 9 6 7 1 2 . 06 2 0 9 9 ( 1 1 . 4 4 , 4 5 ) a ( 25 . 1 3 , 6 0 ) J a n . 1 1 9 6 8 • - M a y 1 9 7 0 1 8 . 87 62 . 0 8 (14 6 0 , 3 5 ) ( 1 0 7 9 1 , 2 1 ) J u n e 1 9 7 0 • - M a y 1 9 7 3 3 8 . 67 2 1 1 5 ( 2 8 . 26 , 4 9 ) ( 1 5 4 0 , 1 0 ) Values in brackets are standard deviations and numbers of observations respectively. -189-time" to monitor the Canadian borrower. 3 8 They also pointed out that the lack of information with regard to Canada, differences i n l e g a l requirements between the two countries and the p o t e n t i a l 39 r i s k s involved i n foreign lending cause U.S. investors to de-mand a higher y i e l d on Canadian bonds denominated i n U.S. dollars than on American bonds of comparable q u a l i t y . Only eleven of the managers interviewed agreed with t h i s view. But the data shown e a r l i e r i n Table 5-1 strongly confirm t h i s assertion that the i n -terest advantage enjoyed by Canadian corporations i n the U.S. bond market i s considerably lower than a comparison of Canadian and U.S. bond y i e l d indices would suggest. Another deterrent f o r smaller Canadian corporations to borrow i n the United States are the detailed and sometimes rather r e s t r i c t i v e t rust deeds demanded by U.S. lenders. Larg-er corporations are i n a much better p o s i t i o n to provide the ex-pert s t a f f support necessary f o r negotiating and executing such contracts. 38 Note that t h i s i s not t y p i c a l f o r t h e i r lending behaviour. As Shapiro and Wolf, p. 2, remark, "the most important character-i s t i c of the private placement market i s that i t serves as the major source of long-term debt financing f o r smaller, less f i n -a n c i a l l y secure companies. One reason i s that small borrowers tend to s e l l small issues..." The average issue size of private placements i n the United States has been about f i v e m i l l i o n d o l l a r s , i b i d . , pp. 91-94. 39 J U.S. i n s t i t u t i o n s have been mainly concerned about changes i n withholding tax or American in t e r e s t equalization tax regula-ti o n s . Some trust deeds contain a r t i c l e s protecting the lender against such e v e n t u a l i t i e s . 40 There was much general agreement that t r u s t deeds fo r U.S.-pay bonds are much more detailed and usually more r e s t r i c t i v e than deeds f o r comparable Canadian-pay bonds. But one manager pointed out that t h e i r sole reason f o r borrowing i n the United States on two occasions was that Canadian i n s t i t u t i o n a l investors would not agree to cer t a i n a r t i c l e s i n the indenture which were acceptable to a U.S. l i f e insurance company. It was also repeatedly mention-ed that U.S. i n s t i t u t i o n s are more f l e x i b l e regarding l a t e r changes i n indentures. - 1 9 0 -Foreign ownership has had considerable impact on Canadian corporate borrowing i n the United States. The data i n Table 5 - 6 b show that during 1968 to May 1970 the p r o b a b i l i t y of a p a r t i c u l a r bond to be c l a s s i f i e d as a U.S.-pay issue increased as the per-centage of the company's equity owned by foreigners increased. Since June 1970 t h i s pattern has changed. Corporations owned by foreigners other than Americans s t i l l exhibit a tendency to borrow abroad (see Table 5 - 6 c ) whereas U.S.-owned firms have de-veloped a s l i g h t preference f o r the Canadian c a p i t a l market (see Table 5 - 8 c below). Growing nationalism, p a r t i c u l a r l y grow-ing resentment of U.S. investment i n Canada, i s probably respon-s i b l e f o r t h i s . Parent pressure because of U.S. balance of pay-ments problems may also have had some influence. In general, relationships between the board of directors and foreign interests seem to have had considerable influence on U.S. d o l l a r borrowing. One corporation mentions i n i t s prospect-uses f o r Canadian-pay public offerings under "Material Contracts" verbal agreements between the firm and one of i t s di r e c t o r s , a partner i n a New York investment house, concerning the placement of U.S.-pay issues. We encountered firms where executives i n i -t i a t e d relationships to New York underwriters whom they were fam-i l i a r with from e a r l i e r employment i n the United States. Ex-change of information on international c a p i t a l markets among subsidiaries of multinational corporations seems to be widespread. In at least one instance a European parent company used i t s New York connections to arrange financing f o r a Canadian sub-s i d i a r y without the l o c a l corporation having much influence on the p a r t i c u l a r s of the deal. -191-On the other hand, neither our interviews nor our data l e d to any evidence indicating a r e l a t i o n s h i p between foreign invest-ment by Canadian corporations and U.S. d o l l a r borrowing f o r dom-4 l e s t i c purposes. S i m i l a r l y , whether a corporation has i t s shares l i s t e d i n the United States or not seems not to influence i t s access to the U.S. bond market. But t h i s i s probably true only with respect to the private placment market. Because of the unexpectedly small number of public U.S.-pay offerings no general conclusions could be derived as to what determines a Canadian corporation's a c c e p t a b i l i t y i n the U.S. r e t a i l bond market. One American investment banker remarked i n his written questionnaire that "the U.S. market i s more name conscious and access i s generally reserved f o r only larger c r e d i t worthy com-panies . " Once a Canadian corporation has sold an issue to U.S. f i n -a n c i a l i n s t i t u t i o n s , the placement of future bonds with the same lenders i s usually simpler and cheaper. Both underwriters and managers mentioned that American l i f e insurance corporations l i k e to be informed about planned U.S. d o l l a r issues by former borrow-ers, and i t i s not uncommon that a Canadian firm places two or 42 three issues with the same small group of U.S. i n s t i t u t i o n s . Instead of drafting anew bulky t r u s t deeds the sale of even large issues i s sometimes arranged through an exchange of l e t t e r s , and we encountered several cases where no placement agent was involved i n the deal. There was general agreement among under-41 Consequently hypothesis Hg, p. 125 above can be regarded as only p a r t l y confirmed. ^Comparatively small issues are sometimes bought by only one or two i n s t i t u t i o n s . -192-writers and f i n a n c i a l managers that regular borrowers enjoy a s l i g h t interest advantage and that the Canadian firm's bargain-ing p o s ition i s the stronger the more f a m i l i a r the lender i s with i t from e a r l i e r contracts. A binary variable measuring whether a corporation had U.S.-pay issues outstanding or not at the date of issue of a p a r t i c u l a r bond contributes s i g n i f i c a n t l y to the discriminatory power of the functions reported i n Tables 5 - 6 b , 5 - 8 b and 5 - 8 c and thus lends s t a t i s t i c a l support to these 43 conclusions. J As already discussed e a r l i e r , U.S. d o l l a r bonds are usually placed d i r e c t l y because Canadian corporations have a preference f o r private placements. Part of the reason why d i r e c t placements are preferred i s that they are believed to provide o v e r a l l cheaper debt funds than public offerings. Savings i n placement fees, other issuing expenses, and i n company o f f i c i a l s ' time and other resources are assumed to more than o f f s e t the sometimes margin-a l l y higher i n t e r e s t costs. The i n c l u s i o n of a private place-ment dummy among the variables intended to measure the influence of transaction costs on borrowings i n the United States improved the significance of the discriminant functions presented i n Tables 5 - 6 a , 5 - 6 b and 5 - 6 c considerably. When using the same variables included i n the discriminant function reported i n Table 5 - 6 a to predict the population of bonds i n our samples f o r the time i n t e r v a l mid -1965 to 1 9 6 7 , 43 ^Cf. also hypothesis Hg, p. 12 8 above, which i n part i s sup-ported by these r e s u l t s . Similar s t a t i s t i c a l r e s u l t s were not obtained f o r i 9 6 0 to 1967 because our samples f o r that period i n -clude only bonds issued by corporations that have borrowed i n the United States. - 1 9 3 -sixty-eight per cent of a l l bonds were co r r e c t l y c l a s s i f i e d . For the period 1968 to May 1 9 7 0 , eighty-six per cent of a l l bonds were grouped r i g h t . The discriminant function estimated f o r 1968 to May 1970 (see Table 5 - 6 b ) predicted the correct population f o r eighty-seven per cent of a l l bonds f o r the period June 1970 to May 1 9 7 3 . Note that again t h i s r e s u l t i s better than the ex post c l a s s i f i c a t i o n reported i n Table 5 - 6 c . To summarize our discussion, we have found strong evidence that transaction and information costs have had a considerable influence on foreign borrowing by Canadian corporations. Be-cause of extra costs connected with international transactions f o r both borrowers and lenders, the average size of U.S.-pay bond issues i s considerably larger than that of Canadian-pay issues except f o r the most recent period when firms bowed to p o l i t i c a l pressure and sold only medium-sized issues abroad. These bonds are usually only issued by l a r g e r Canadian corpora-t i o n s . Companies (partly) owned by foreigners or which through t h e i r directors have connections to New York underwriters are more l i k e l y to borrow abroad than other Canadian corporations. Once a firm has gained access to the U.S. bond market and Ameri-can i n s t i t u t i o n s are f a m i l i a r with i t s business, the placement of future issues i s usually easier and sometimes can be arranged very fast and informally. Almost a l l U.S.-pay issues are not sold to the general public but are placed d i r e c t l y , p a r t l y because of savings i n transaction costs and p a r t l y because of the great-er f l e x i b i l i t y possible through arm's length negotiations with a very small number of investors and because of other advantages discussed e a r l i e r . -194-5 . 6 SUMMARY ANALYSIS OF FACTORS INFLUENCING CANADIAN CORPORATE BORROWING ABROAD In the preceding sections considerable evidence has been presented to confirm our t h e o r e t i c a l conclusions that (1) d i f f e r -ences i n time preferences between countries, (2) exchange rate r i s k and (3) transaction costs differences influence the i n t e r -national flov; of long-term c a p i t a l . However, each of these three factors has been analyzed i n i s o l a t i o n . We s h a l l now draw those in d i v i d u a l r e s u l t s together to show t h e i r combined e f f e c t on Canadian corporate borrowing i n the United States. In Table 5 - 8 a discriminant analysis r e s u l t s f o r the period i 9 6 0 to 196? are presented. Because of c o l l i n e a r i t y problems some variables are not s i g n i f i c a n t i n the presence of others, and therefore two d i f f e r e n t estimates are reported. The most s i g n i f i c a n t variable i n function A i s the r i s k premium measure which attempts to capture the influence which exchange r i s k and lower inte r e s t costs have on Canadian borrowing i n the United States. A private placement dummy measuring the influence of differences i n time preferences between the two countries and of transaction costs i s also very s i g n i f i c a n t . The several closings binary variable indicates that the a v a i l a b i l i t y of con-t r a c t s involving delayed takedowns was a f a c t o r a t t r a c t i n g Cana-dian corporations to the American market. In function B, besides the private placement dummy, two variables r e l a t i n g to informa-t i o n and transaction cost, size of issue and asset size of is s u -ing corporation, are marginally s i g n i f i c a n t . The U.S d o l l a r revenue measure i s not s i g n i f i c a n t by common standards but leads TABLE 5-8a DISCRIMINANT ANALYSIS RESULTS, 1960-1967' COMBINATION OF SEVERAL MEASURES D i s c r i m i n a n t F u n c t i o n A Di s c r i m i n a n t Function B Means and St.D.s Item C o e f f i c i e n t s F-Values F-Probab. C o e f f i c i e n t s F-Values F-Probab. Can.-pay Bonds U.S.-pay Bonds S e v e r a l C l o s i n g s Dummy- 1.757 5-35 (.0216) - -.022 ( .149) .233 ( .426) P r i v a t e Placement Dummy 2.667 19.02 (.0001) 2 .618 24.28 (.0000) • 533 ( .505) .917 ( .279) R i s k Premium -2 .293 24.20 ( .0000) - - 1.077 ( .537) .461 ( .478) Log (Percentage of Revenue Denominated i n US $$ + 1) - - • 331 • 99 (.3225) .368 ( .637) .637 ( .824) Log (Issue S i z e i n M i l l i o n s of $$) - - 1.240 4.20 (.0408) .857 ( .510) 1 .050 ( .514) Log (Assets i n M i l l , of $$) - - - .748 2.56 (.1085) 2.072 ( .618) 2.058 ( .641) Constant + ,4o4 - - 1 . 7 0 3 - - -F u n c t i o n - 23.58 ( .0000) - 8.37 (.0000) - -Number of Observations - 105 - 105 45 60 Post C l a s s i f i c a t i o n Percentage of Issues C o r r e c t l y C l a s s i f i e d 837S - 72^ • -A B 80% 56# A B W 85# - 1 9 6 -to a marginal improvement i n ex post and ex ante c l a s s i f i c a t i o n r e s u l t s . Estimates based on the variables mentioned i n Table 5 - 8 a f o r functions A and B corre c t l y predicted the c l a s s i f i c a t i o n of ninety-six and sixty-nine per cent, respectively, of a l l bonds i n our test samples f o r the time i n t e r v a l mid -1965 to 1967 and of eighty-six and eighty-six per cent, respectively, of a l l cases f o r the period 1968 to May 1 9 7 0 . When evaluating our res u l t s f o r the period i 9 6 0 to 1967 two things sould be borne i n mind: (1) these data r e f l e c t only the f i n a n c i a l behaviour of those corporations that have borrowed i n the United States during the time i n t e r v a l i 9 6 0 to 1 9 7 3 , and (2) information f o r t h i s early period i s less complete than f o r l a t e r periods. Nevertheless, a l l the data presented indicate that factors predicted by our theory influenced long-term c a p i t a l flows between Canada and the United States. The estimates f o r the time i n t e r v a l January 1968 to May 1970 (see Table 5 - 8 b ) show that combinations of several variables lead to an almost perfect ex post c l a s s i f i c a t i o n of the bonds i n our samples. The predictive power of these functions i s equal-l y impressive. For the period June 1970 to May 1 9 7 3 , function A and function B lead to a correct c l a s s i f i c a t i o n of ninety-six and of ninety-five per cent of a l l cases, respectively. Most s i g n i f i c a n t are three variables measuring the influence of d i f -ferences i n lenders' time preferences on foreign issues: the several closings and private placement binary variables and a measure of the time elapsed between contract and delivery of a bond issue. A maturity variable i s also s i g n i f i c a n t . Measures TABLE 5-8b DISCRIMINANT ANALYSIS RESULTS, JANUARY 1968-MY 1970« COMBINATION OP SEVERAL MEASURES Item D i s c r i m i n a n t F u n c t i o n A Dis c r i m i n a n t Function B Means and St.D.s C o e f f i c i e n t s F-Values F-Probab. C o e f f i c i e n t s F-Values F-Probab. Can.-pay 3onds U.S .-pay Bonds Term to M a t u r i t y i n Years .162 6.21 (.0142) .157 4.48 (.0354) 11.56 ( 7.34 ) 21.36 ( 6.30 ) Sev e r a l C l o s i n g s Dummy 7.999 31.24 (.0000) IO.58 38.61 ( .0000) .031 ( .174 ) .619 ( .498 ) Log (Years Elapsed between Co n t r a c t and D e l i v e r y + 1 - - 201.0 18.05 (.0001) .0042 ( .0059) .0160 ( .0205) P r i v a t e Placement Dummy 5-183 22.80 (.0000) 4.971 16.347 (.0002) .184 ( .391 ) .857 ( .358 ) R i s k Premium -I.450 1.84 (.1749) - -1.224 ( .445 ) .892 ( .512 ) Log (Percentage of Company Owned by Fo r e i g n e r s + 1) 1.416 5.46 (.0209) 1.318 4.06 (.0447) .42? ( .737 ) 1.108 ( .871 ) E a r l i e r U.S.-pay Issues Dummy 2.944 9.10 (.0035) - - .35^ ( .482 :) .66? ( .483 ) Log (Assets i n M i l l , of $$) - - 1.855 4.59 (.0334) 2.195 ( .620 ) 2.475 ( .626 ) Constant -9.026 - -6.00 - - -F u n c t i o n - 30.22 (.0000) - 37.02 (.0000) Number of Observations - 86 - 86 65 21 Ex Post C l a s s i f i c a t i o n 94?? - 92% - k95% B95'£ k 9 0 % B8l<£ -198-intended to capture the influence of differences i n information and transaction costs l i k e the percentage of a corporation owned by foreigners, an " e a r l i e r U.S.-pay issues" dummy and the asset size of the issuing firm are equally important variables. In the presence of these variables both measures d i r e c t l y related to exchange r i s k , the r i s k premium variable and the variable measuring the percentage of a firm's revenue denominated i n U.S. d o l l a r s , are i n s i g n i f i c a n t . Nevertheless, we included"the r i s k premium measure into discriminant function A i n Table 5-813 to demonstrate the remarkable change i n the importance of t h i s variable f o r discriminating between Canadian-pay and U.S.-pay bonds when compared to the e a r l i e r period. We would regard our 1968 to May 1970 data as most approp-r i a t e f o r testing our theory. During t h i s period p o l i t i c a l i n -terference with the economic forces influencing the f i n a n c i a l behaviour of Canadian corporations was almost absent, and we com-pare a sample of U.S.-pay bonds with a random sample of a l l Canadian-pay bonds issued during t h i s time i n t e r v a l . The high predictive power of the estimated functions indicates that the basic factors influencing Canadian corporate borrowing i n the United States have not materially changed i n recent years. The estimates i n Table 5-8c f o r the period June 1970 to May 1973 confirm t h i s because b a s i c a l l y the same variables are included as those mentioned i n Table 5-8b. However, some small-er changes are obvious. Whereas measures related to differences i n time preferences are s t i l l very s i g n i f i c a n t as i n estimates f o r e a r l i e r years, variables intended to capture the influence TABLE 5-8c DISCRIMINANT ANALYSIS RESULTS, JUNE 1970-MAY 1973i COMBINATION OF SERVERAL MEASURES Item D i s c r i m i n a n t F u n c t i o n A Disc r i m i n a n t Function B Means and St.D.s. C o e f f i c i e n t s F-Values F-Probab. C o e f f i c i e n t s F-Values F-Probab. Can.-pay Bonds U .S .-pay Bonds Term to M a t u r i t y i n Years - - .162 4.40 (.0365) 15.36 ( 6.42 ) 21.97 ( 3-38 ) Se v e r a l C l o s i n g s Dummy 12.528 25-91 (.0000) - 11.98? 21.97 (.0000) .011 ( .103) .300 ( .483) P r i v a t e Placement Dummy 7.28? 40.96 (.0000) 7.345 38-74 (.0000) .149 ( .358) .900 ( .316) E a r l i e r U.S.-pay Issues Dummy 2.805 7.35 (.0071) _ _ .266 ( .444) .600 ( .516) Log (Percentage of Company Owned by For e i g n e r s Other Than Americans + 1) 1.868 4.61 (.0324) 1.705 3.63 (.0567) .152 ( .491) .528 ( .851) Log (Percentage of Company Owned by Americans + 1 ) 0 - - 1.255 3.19 (.0736) .396 ( .716) .360 ( .68?) Log (Assets i n M i l l , of $$) - - 1.109 2.95 (.0853) 2.174 ( -762) 2.492 ( .601) Constant - 6.619 - -11.422 - -F u n c t i o n - 23.55 (.0000) - 15.89 (.0000) _ Number of Observations - 104 - 104 94 10 Ex Post C l a s s i f i c a t i o n 96% - 91?? - k97% B90% k9C% B100% - 2 0 0 -of differences i n information and transaction costs have l o s t somewhat i n importance except f o r the " e a r l i e r U.S.-pay issues" dummy. Indeed discriminant function B i n Table 5 - 8 c suggests that firms (partly) owned by Americans have developed a s l i g h t aversion against U.S.-pay bonds. As w i l l be discussed i n the following section, growing nationalism i n Canada and p o l i t i c a l pressure can explain these r e s u l t s . Somewhat puzzling i s the complete i n s i g n i f i c a n c e f o r t h i s period of any measure intended to capture the influence of foreign exchange r i s k on corporate borrowing i n the United States except perhaps f o r the maturity variable. Since June 1970 the Canadian d o l l a r has been f l o a t i n g and has fluctuated around par v i s - a - v i s the U.S. d o l l a r most of the time a f t e r appreciating i n value sharply i n 1 9 7 0 . We had expected that t h i s would focus attention on exchange r i s k but we were unable to detect any evidence of t h i s . For example, when including the r i s k premium as a variable into the discriminant functions reported . i n Table 5 - 8 c , i t was " s i g n i f i c a n t " only at the twenty-eight and nineteen per cent l e v e l , respectively. When comparing our results f o r the period i 9 6 0 to 1967 with those f o r 1968 to 1973» the increased significance of measures related to differences i n time preferences and the diminished importance of the r i s k premium variable (or the i n t e r -est rate d i f f e r e n t i a l ) are most noteworthy. V/e do not have a ready explanation f o r t h i s . Our interviews concentrated most-l y on recent financing a c t i v i t i e s and did not provide many clues. However, i t seems that the increased l i q u i d i t y conscious--201-ness of Canadian f i n a n c i a l i n s t i t u t i o n s since the middle ' s i x t i e s , the improved e f f i c i e n c y of the Canadian c a p i t a l market i n recent years and a greater f i n a n c i a l s o p h i s t i c a t i o n of Canadian corpor-ations have a l l contributed to these r e s u l t s . 5.7 NON-ECONOMIC INFLUENCES ON CORPORATE BORROWING BEHAVIOUR Our theory of long-term international debt c a p i t a l flows i s s o lely based on an analysis of economic factors that influence borrower and lender behaviour. Though no direc t s t a t i s t i c a l evidence can be presented f o r t h i s assertion, our discussion with underwriters and f i n a n c i a l managers strongly suggest that non-economic variables also have had a su r p r i s i n g l y great e f f e c t on Canadian corporate debt financing i n the United States. Two influences seem of p a r t i c u l a r importanceJ growing nationalism i n Canada and pressure exerted by f i n a n c i a l a u t h orities i n Ottawa. More than half of the corporations which we interviewed have not approached the U.S. bond market since 1970. When asked why they did not s e l l U.S. d o l l a r bonds i n recent years, several managers answered that, as a Canadian corporation, they wanted to maintain a Canadian image by financing i n Canada thereby keeping t h e i r name i n front of the Canadian public. Others said that major stockholders, out of national f e e l i n g s , requested Canadian financing. Mangers of subsidiaries of multinational corporations mentioned the changing p o l i t i c a l climate with regard 44 to American dir e c t investment i n Canada. Consequently they On t h i s see, f o r example, John Fayerweather, Foreign In-vestment i n Canadai Prospects f o r National Policy (New Yorkt International Arts and Sciences Press, 1973). -202-thought i t opportune to finance Canadian projects i n Canada.^ Requests by the Bank of Canada and the Federal Minister of Finance to r e s t r i c t financing a c t i v i t i e s to the Canadian market seem to have had even more influence on the borrowing be-haviour by Canadian corporations i n recent years. In i t s Annual Report f o r the year 1970 the Bank of Canada notes that " i n view of the continued u n d e s i r a b i l i t y of a large inflow of c a p i t a l , both the Minister of Finance and the Bank of Canada attempted to reinforce the e f f e c t of declining in t e r e s t rates i n Canada by asking Canadian borrowers to explore very c a r e f u l l y the possi-b i l i t i e s of doing t h e i r necessary financing i n Canadian markets before they had recourse to borrowing abroad. Since the f i r s t quarter of 1970 the use of external markets has f a l l e n to a low 46 l e v e l and t h i s has been a hel p f u l development." According to our information, the Bank of Canada also s o l i c i t e d the coopera-t i o n of Canadian investment dealers. Underwriters play an import-ant role i n foreign borrowing by corporations because they often have been the i n i t i a t i n g force that f i n a l l y l e d to U.S.-pay issues. This Bank of Canada request has put Canadian underwriters into a compromising p o s i t i o n . They f e e l that i t i s t h e i r o b l i -gation to provide t h e i r c l i e n t s with objective information on a l l c a p i t a l markets accessible to Canadians and to advise them on where to obtain the cheapest funds at the best conditions. 45 ^From a Canadian point of view i t probably would be more desirable to reduce foreign direct investment by approaching the Canadian c a p i t a l market f o r equity funds and perhaps even increase foreign borrowing. I t seems that a c o n t r o l l i n g interest by f o r -eigners rather than the use of foreign funds as such i s objection-able to Canadians. 46 Bank of Canada, Annual Report 1970» p. 7. - 2 0 3 -On the other hand, they attempt to he "good corporate c i t i z e n s " and, perhaps more importantly, want to protect t h e i r share i n the allotment of new government s e c u r i t i e s . General practice seems to he to inform corporate c l i e n t s about conditions i n the U.S. bond market but to no longer recommend foreign-pay issues. As a rule, investment dealers try to avoid o f f i c i a l contact with the Bank of Canada regarding any p a r t i c u l a r issue planned but may informally make the Bank aware of i t . Usually the c l i e n t i s expected to inform the Bank of Canada of a new U.S.-pay issue. He may also ask f o r the Bank's opinion i f he f e e l s t h i s i s advisable. The majority of corporations interviewed agreed that the Bank of Canada's p o s i t i o n on foreign borrowing influenced t h e i r f i n a n c i a l decisions i n recent years. Most of those that borrowed i n the United States since 1970 indicated that they consulted with the Bank before issuing foreign currency bonds. Information gathered through our interviews suggests that the Bank of Canada gives a "permission" to borrow abroad only i f an issue meets certai n c r i t e r i a , some of which seem to bes (1 ) Issue size must be r e l a t i v e l y small, not more than twenty to t h i r t y m i l l i o n dol-l a r s . (2) The borrower should present convincing reasons why he cannot obtain the funds i n Canada at an acceptable interest rate. Useful arguments are a need f o r a long-term forward com-mitment or f o r delayed d e l i v e r i e s . (3 ) I f the proceeds are used f o r the import of equipment or payment of other U.S. d o l l a r debt obligations, larger issue sizes are permissible. Most firms seem very reluctant to defy o f f i c i a l requests f o r changes i n t h e i r - 2 0 4 -financing p o l i c i e s because of fear of government r e t a l i a t i o n . In at least one case a corporation advised the Bank of Canada, as a matter of courtesy, about an impending large U.S.-pay issue; i t had to cancel the deal because of pressure exerted upon i t . CHAPTER 6 SUMMARY AND CONCLUSIONS This study has presented a new attempt to gain a better understanding of those forces that lead to the movement of funds from one country to another. We r e s t r i c t e d our attention to the international market f o r long-term debt c a p i t a l . The empirical analysis focused on c a p i t a l flows between Canada and the United States, p a r t i c u l a r l y on Canadian corporate borrowing i n the United States. F i r s t we developed a model of the international term structure of intere s t rates. Assuming that c a p i t a l markets are perfect, we showed that interest rate d i f f e r e n t i a l s between countries r e f l e c t exchange rate expectations as well as r i s k premia necessary to reimburse international investors f o r ac-cepting exchange r i s k . The demand f o r funds by a country's residents, which i s a function of t h e i r time preferences, de-termines whether a nation's e f f e c t i v e i n t e r e s t rate l e v e l i s higher or lower than that of other countries. I f information and transaction costs are higher when two traders from d i f f e r e n t countries deal with each other than when both investors are from the same country, then interest rate d i f f e r e n t i a l s w i l l tend to be larger than they would be otherwise. -205--206-The inflow of long-term debt c a p i t a l into Canada i s almost exclusively due to the sale of new bond issues abroad by Canadian borrowers. A c t i v i t i e s of in t e r n a t i o n a l investors i n secondary markets are only of minor importance. This makes a direct test of our model through regression analysis rather d i f f i c u l t . Furthermore, the Bank of Canada has attempted to in t e r f e r e with the free play of economic forces by manipulating interest rate d i f f e r e n t i a l s and through moral suasion. Conse-quently we had to re l y on an i n d i r e c t test of the basic fea-tures of our theory. We presented some evidence which suggested that the y i e l d d i f f e r e n t i a l between Canadian and American corporate bonds i n -deed r e f l e c t s exchange rate expectations to some extent, but only short-term and medium-term expectations. It i s widely be-liev e d that i n the long run the exchange rate w i l l fluctuate around an equilibrium value of one Canadian to one United States d o l l a r . Given such expectations, the high l e v e l of interest rates i n Canada and the continuous inflow of long-term debt c a p i t a l are probably due to higher time preferences of Canadians. We calculated the ( r e l a t i v e ) demand for funds i n Canada and i n the United States as a percentage of GNP and showed that i t i s considerably greater i n Canada. We analyzed the impact which such differences i n time preferences between the two countries might have on Canadian corporate borrowing behaviour. Dis-cussions i n the l i t e r a t u r e suggested that Canadian f i n a n c i a l i n s t i t u t i o n s have a comparative preference f o r bonds with a high degree of marketability. To test f o r t h i s , we hypothesized -207-that the a v a i l a b i l i t y of a well-functioning private placement market, of long-term forward commitments, and of longer maturi-t i e s should be factors a t t r a c t i n g Canadian corporations to the U.S. bond market. Both our discriminant analysis r e s u l t s and our interviews confirmed that the r e l a t i v e abundance of a t t r a c t i v e investment opportunities i n Canada has allowed Canadian f i n a n c i a l i n s t i t u -tions to concentrate t h e i r investments i n comparatively market-able s e c u r i t i e s . Corporations wanting to s e l l t h e i r bonds d i r e c t l y often had to resort to the U.S. market, and long-term forward commitments and contracts providing f o r several closings are extremely d i f f i c u l t to arrange f o r i n Canada. The longer maturities available i n the American bond market seem to be of lesser interest to Canadian firms. These findings raise some s i g n i f i c a n t questions! ( l ) How can we explain differences i n time preferences between countries? A comparative analysis of the age structure of d i f f e r e n t nations, of savings behaviour, of investment opportunities i n r e a l assets, and of i n f l a t i o n a r y experiences should be of considerable i n t e r -est. (2) How can the functioning of the Canadian private place-ment market be improved? Can Canadian f i n a n c i a l i n s t i t u t i o n s indeed consistently increase t h e i r p r o f i t s through bond trading? I f yes, i s the secondary market f o r Canadian corporate bonds i n -e f f i c i e n t ? (3) A comparative analysis of the d i f f e r e n t i n -vestment strategies pursued by Canadian and American l i f e insurance corporations seems warranted to determine whether they - 2 0 8 -r e s u l t i n marked differences i n benefits accruing to share-holders and policy holders. Our model indicated that exchange r i s k should have a con-siderable influence on borrower (and lender) behaviour. To test f o r t h i s , we analyzed how Canadian corporations could re-duce the exchange r i s k involved i n s e l l i n g U.S.-pay bonds, given that they want to protect themselves against an early adverse change i n the exchange rate. We showed that exchange r i s k de-creases as the term to maturity increases. Also, i f exchange rate expectations are rather d i f f u s e , then sinking fund payments evenly d i s t r i b u t e d over the term of an issue should be preferred. Corporations receiving income denominated i n U.S. do l l a r s can provide themselves with a long-term forward exchange market by s e l l i n g U.S. d o l l a r bonds. Our s t a t i s t i c a l r e s u l t s suggested that Canadian corpora-tions behave i n the predicted manner, but the evidence was r e l a t i v e l y weak. Through our interviews we found out that cor-porations do not deliberately a l t e r bond c h a r a c t e r i s t i c s to re-duce exchange r i s k . Rather, lender behaviour i s responsible for the observed differences i n maturities and sinking fund provisions. A s l i g h t l y lower interest rate seems to be a l l the protection against exchange r i s k firms require. The function and purpose of the forward exchange market i s often not well understood,^ - and few corporations can af f o r d to employ a foreign Forward exchange contracts are sometimes regarded as a means to speculate rather than as a means to protect future U.S. d o l l a r revenue or U.S. d o l l a r l i a b i l i t i e s against exchange r i s k . -209-exchange s p e c i a l i s t . Usually foreign borrowing i s not regarded as a means to s e l l future export earnings forward. A closer analysis of exchange r i s k management by Canadian corporations and by those i n other countries should prove valu-able. The development of r e l a t i v e l y simple, operational models to a s s i s t firms i n t h i s task i s needed. Also, there should be considerable opportunities for improved services by banks and underwriters i n t h i s respect. F i n a l l y , we looked at the influence which information and transaction costs have on international c a p i t a l flows. For both Canadian borrowers and American lenders these costs seem to be higher.when dealing with foreigners. Consequently i t i s not surprising that only larger Canadian corporations borrow i n the United States and that only r e l a t i v e l y large issues are sold abroad. Once a Canadian firm i s f a m i l i a r to American lenders, the placement of future issues can often be arranged r e l a t i v e l y e a s i l y and quickly. Connections to U.S. investors through fo r -eign stockholders or directors have also proven h e l p f u l . From a corporate viewpoint, material differences i n trust deeds of Canadian-pay and U.S.-pay issues seem to be a major concern. A comparative analysis and evaluation of the economic implications of such differences would be highly desirable. When considered together, both our interviews and our discriminant analysis r e s u l t s have provided strong support for our t h e o r e t i c a l conclusions that (1) differences i n time - • -Sophisticated models l i k e the one developed by Lietaer are beyond the f i n a n c i a l and technical c a p a b i l i t i e s of most Canadian corporations. See, Bernard A. Lietaer, F i n a n c i a l Management of  Foreign Exchange, (Cambridge, Mass.: MIT Press, 1971). -210-preferences, (2) exchange r i s k , and (3) transaction costs have a considerable influence on the d i r e c t i o n and volume of i n t e r -national c a p i t a l flows. A t y p i c a l U.S.-pay bond issued by a Canadian corporation i s a private placement, i s r e l a t i v e l y large i n size, often involves a long-term forward commitment by an American lender, and i s sometimes drawn down i n several closings. The inte r e s t rate i s s l i g h t l y lower than i n Canada. The issuing corporation i s r e l a t i v e l y large i n size, has often foreign stockholders, and has sold U.S.-pay bonds before. It seems to be engaged i n export a c t i v i t i e s to a greater extent than a t y p i c a l corporation s e l l i n g only Canadian-pay bonds. We did not compare our sample of Canadian U.S.-pay bonds with a sample of American U.S.-pay bonds. However, i t seems that the t y p i c a l American private placement i s small i n size 3 and i s issued by a "smaller, less f i n a n c i a l l y secure" firm. Also, features l i k e long-term forward commitments and several take-downs seem to be less prevalent. In addition, the interest rate i s usually lower than on Canadian U.S.-pay issues. The investment behaviour of American lenders with respect to foreign U.S.-pay bonds should be worth further exploration. In Chapter 1 we indicated that a major deficiency of Canadian-United States c a p i t a l flow data i s that information i s available only on the amount of new corporate issues de-l i v e r e d to American lenders. I f data based on the date of Cf. footnote 33 i n Chapter 5 above. -211-c o n t r a c t s h o u l d b e c o m e a v a i l a b l e , t h e n a m o r e r i g o r o u s r e g r e s -s i o n a n a l y s i s o f C a n a d i a n c o r p o r a t e b o r r o w i n g a b r o a d m a y b e -c o m e f e a s i b l e . T h e i n f o r m a t i o n p r o v i d e d b y o u r s t u d y s h o u l d c o n t r i b u t e t o t h e s u c c e s s o f s u c h r e s e a r c h . BIBLIOGRAPHY Aigner, D.J., and Sprenkle, CM. "A Simple Model of Information and Lending Behavior," Journal of Finance, XXI (March, 1968), 1 5 1 - 6 6 . A l i b e r , Robert Z. "The Interest P a r i t y Theorems A Reinterpreta-t i o n , " Journal of P o l i t i c a l Economy, LXXXI (November/Decem-ber, 1 9 7 3 ) , 1451-59. A l i b e r , Robert Z. "Uncertainty, Currency Areas and the Exchange System," Economica, XXIX (N.S.) (November, 1 9 7 2 ) , 432-441. A l i b e r , Robert Z. "The Multinational Enterprise i n a Multiple Currency World," i n John H. Dunning (ed.), The Multination-a l Enterprise. London* A l l e n & Unwin, 1 9 7 1 , 49-56. A l i b e r , Robert Z. "Exchange Risk, Y i e l d Curves, and the Pattern of Capital Flows," Journal of Finance, XXIV (May, 1 9 6 9 ) , 3 6 l -7 0 . Anderson, Theodore W. An Introduction to Multivariate S t a t i s t i -c a l Analysis. New Yorkt Wiley, 1 9 5 8 . Arnold, H.J. "Permutation Support f o r Multivariate Techniques," Biometrika, LI ( 1 9 6 4 ) , 6 5 - 7 0 . Arrow, Kenneth J . Aspects of the Theory of Risk-Bearing. Hel-s i n k i : Yrjo* Jahnssonin SSStio', 1 9 6 5 . Bank of Canada. Annual Report f o r the years i 9 6 0 to 1 9 7 2 . Bierwag, G.O., and Grove, M.A. "A Model of the Term Structure of Interest Rates," Review of Economics and S t a t i s t i c s , IL (February, 1 9 6 7 ) , 50~^E2~. Bohi, Douglas R. "The International Interdependence of Interest Rates," Kyklos, XXV (November, 1 9 7 2 ) , 5 9 7 - 6 0 0 . Branson, W i l l i a n H., and H i l l , Raymond D. J r . "Capital Movements Among Major OECD Countries: Some Preliminary Results," Journal of Finance, XXVI (May, 1 9 7 1 ) , 2 6 9 - 8 6 . Caves, Richard E., and Johnson, Harry G. Readings i n Internation-a l Economics. Homewood: Irwin, 1 9 6 8 . Caves, Richard E., and Reuber, Grant L. Capital Transfers and  Economic Policyt Canada, 1 9 5 1 - 1 9 6 2 . Cambridge, Mass.: Harvard University Press, 1 9 7 1 . - 2 1 2 --213-Cooley, William W., and Lohnes, Paul R. Multivariate Data  Analysis. New York: Wiley, 1 9 7 1 . Dipchand, C e c i l R. "The Determinants of Risk Premiums and De-velopment of Rating Classes f o r P u b l i c l y Traded Canadian Corporate Bonds," i n Proceedings of the F i r s t Annual Con-ference of the Canadian Association of Administrative  Sciences (Kingston, Ont., 1973), 2/133-2/165. Despres, Emile, Kindleberger, Charles P., and Salant, Walter S. "The Dollar and World L i q u i d i t y , " The Economist, February 5 , 1 9 6 5 , 5 2 6 - 5 2 9 . Dunn, Robert M. J r . Canada's Experience with Fixed and Flexible  Exchange Rates i n a North American C a p i t a l Market. Washing-ton and Montreal! Canadian-American Committee, 1 9 7 1 . Eiteman, David K., and S t o n e h i l l , A. I. Multinational Business  Finance. Reading, Mass.t Addison-Wesley, 1 9 7 3 . E l l i s , Howard S., and Metzler, L.A. (eds.) Readings i n the  Theory of International Trade. Homewoodl Irwin, 1 9 5 0 . Fama, Eugene F., and M i l l e r , Merton H. The Theory of Finance. New Yorki Holt, Rinehart and Winston, 1 9 7 2 . Fayerweather, John. Foreign Investment i n Canadat Prospects  for National P o l i c y . New Yorki International Arts and Sciences Press, 1973* Fisher, Lawrence, and Weil, Roman L. "Coping With the Risk of Fluctuations J Returns to Bondholders from Naive and Opti-mal Strategies," Journal of Business, LXVII (October, 1 9 7 1 ) . 408 -31 . Floyd, John E. "International Capital Movements and Monetary Equilibrium," American Economic Review, LIX (September, 1 9 6 9 ) , 472 -92 . Freedman, Charles. "Long-Term C a p i t a l Flows Between the United States and Canada." Unpublished Ph.D. thesis, Massachu-setts Institute of Technology, 1 9 7 0 . Fullerton, Douglas H. The Bond Market i n Canada. Torontot Carswell Co., 1 9 6 2 . Gordon, M.J. "Towards a Theory of F i n a n c i a l Distress," Journ-a l of Finance, XXVI (May, 1 9 7 1 ) , 34 -7 -56 . Grubel, Herbert G. "Internationally D i v e r s i f i e d P o r t f o l i o s ! Welfard Gains and Capital Flows," American Economic Re-view, LVIII (December, 1 9 6 8 ) , 1301-T4^ Grubel, Herbert G. Forward Exchange,Speculation and the Inter-national Flow of C a p i t a l . Stanford, Cal.j Stanford Univer-s i t y Press, 1966. -214-"Guidelines f o r Banks and Nonbank Finan .c i a l I n s t i t u t i o n s , " Federal Reserve B u l l e t i n , LI (March, 1 9 6 5 ) , 3 7 1 - 7 6 , and LVI (January, 1 9 7 0 ) , 1 1 - 2 2 . Haas, Richard D. "A P o r t f o l i o Model of International Capital Flows." Unpublished Ph.D. thesis, Department of Economics, Duke University, 1 9 7 1 . Helleiner, Gerald K. "Connections Between United States' and Canadian Ca p i t a l Markets, 1 9 5 2 - 1 9 6 0 , " Yale Economic Essays , II ( F a l l , 1 9 6 2 ) , 350-400. H e l l i w e l l , John F., et a l . The Structure of RDX2. Bank of Canada Research Studies No. 7 , Ottawa $ Bank of Canada, 1 9 7 1 . Hicks, J.R. Value and C a p i t a l . 2 n d ed.: London: Oxford Uni-v e r s i t y Press, 1946. Homer, Sidney, and Johannesen, Richard I. The Price of Money. An A n a l y t i c a l Study of U.S. and Foreign Interest Rates. New Brunswick, N.J.* Rutgers University Press, 1 9 6 9 . "International Banking Survey," Economist, January 2 7 , 1 9 7 3 -International Monetary Fund. Annual Report on Exchange R e s t r i c t -ions. Washington, D.C., 1 9 7 3 . Investment Dealers' Association of Canada. B r i e f to the Royal  Commission on Banking and Finance. Toronto: Investment Dealers' Association of Canada, June 1 9 6 2 . Johnson, Harry G. "Towards a General Theory of the Balance of Payments," i n Harry G. Johnson, International Trade and  Economic Growth: Studies i n Pure Theory (Cambridge, Mass.: Harvard University Press, 1961), 153-168. Johnston, J. Econometrics.2nd ed.; New York: McGraw-Hill, 1 9 7 2 . Johnson, Ramon E. "Term Structure of Corporate Bond Yields as a Function of Risk of Default," Journal of Finance, XXII (May, 1 9 6 7 ) , 3 1 3 - 4 5 . Jones, Lawrence D. Investment P o l i c i e s of L i f e Insurance Com-panies . Boston! Harvard University, 1968. Katz, Samuel I. "Exchange-Risk Under Fixed and F l e x i b l e Exchange Rates," The B u l l e t i n of the I n s t i t u t e of Finance, Graduate School of Business Administration, New York University, Nos. 83-84 (June, 1 9 7 2 ) . Kindleberger, Charles P. International Economics. 4th ed.; Homewood: Irwin, I96W. -215-Kindleberger, Charles P. Balance-of-Payments D e f i c i t s and the International Market f o r L i q u i d i t y . Essays i n Internation-a l Finance, No. 4 6 ; Princeton, New Jerseys International Finance Section, Department of Economics, Princeton Uni-v e r s i t y , May 1965. Korin, B a s i l P. "Some Comments on the Homoscedasticity C r i t -erion M and the Multivariate Analysis of Variance Tests T2, W and R," Biometrika, LIX ( A p r i l , 1972), 215-16. Krueger, Anne 0. "Balance of Payments Theory," Journal of Eco-nomic Literature, VII (March, 1969), 1-26. Krueger, Anne 0. "The Impact of Alter n a t i v e Government P o l i c i e s Under Varying Exchange Systems," Quarterly Journal of Eco-nomics, LXXIX (May, 1965), 195-20o\ Ladd, George W. "Linear P r o b a b i l i t y Functions and Discriminant Functions," Econometrica, XXXIV (October, 1966), 873-85. "Legal Problems of Issuing and Marketing Foreign S e c u r i t i e s i n the United States," panel discussion, i n John F. McDaniels (ed.), International Financing and Investment (Dobbs Ferry, N . Y . J Oceana Publ., 1 9 6 4 ) , 4 3 0 - 6 0 . Lietaer, Bernard A. Financial Management of Foreign Exchange. Cambridge, Mass.i MIT Press, 1971. McDaniels, John F. International Financing and Investment. Dobbs Ferry, N.Y.t Published f o r the World Community Association of the Yale Law School by Oceana Publ., 1964. Malkiel, Burton G. The Term_Structure of Interst Rates. Prince-ton, N.J.i Princton University Press, 1 9 6 6 . Mendelson, Morris. "The Eurobond and Ca p i t a l Market Integration," Journal of Finance, XXVII (March, 1972), 110-126. Metzler, Lloyd A. "The Process of International Adjustment. Under Conditions of F u l l Employment!, a Keynesian View," i n R.Ej Caves and H.G. Johnson (eds.), Readings i n Internation-a l Economics (Homewoods Irwin, 1968), 4 6 5 - 4 8 6 . M i l l e r , Merton H., and Modigliani, Franco. "Dividend Policy, Growth and the Valuation of Shares," Journal of Business, XXXIV (October, 1961), 411-33. M i l l e r , Norman C , and Whitman, Marina V.N. "Alternative Theories and Tests of U.S. Short-Term Foreign Investment," Journal of Finance. XXVII (December, 1973), 1131-1150. Mundell, Robert A. International Economics. New Yorki MacMillan, I968. -216-Neufeld, Edward P. "The Relative E f f i c i e n c y of the Canadian Cap i t a l Market» The Consequences f o r Canadian-United State Fin a n c i a l Relations," i n Canadian-United States Fi n a n c i a l  Relationships (The Federal Reserve Bank of Boston Confer-ence Series No. 6; Boston, Mass.j Federal Reserve Bank of Boston, 1971), 100-115. Nevin, E. "Some Reflections on the New York New Issue Market," Oxford Economic Papers, XIII (N.S.)(February, 1961), 84-ToT. Parkinson, R. MacDonald. Translation of Foreign Currencies. Torontot The Canadian Inst i t u t e of Chartered Accountants, 1972. Peters, J. Ross. Economics of the Canadian Corporate Bond Mark-et . Montreal*. McGill-Queen*s University Press, 1971. Porter, Michael G. "A Theoretical and Empirical Framework f o r Analysing the Term Structure of Exchange Rate Expectations, IMF Staff Papers, XVIII (November, 1971), 613-642. P r e l l , Michael J. "How Well do Experts Forecast Interest Rates? Federal Reserve Bank of Kansas C i t y Monthly Review, Septem-ber-October, 1973, 3-13. Ragazzi, Giorgio. "Theories of the Determinants of Direct For-eign Investment," IMF Staff Papers, XX (July, 1973), 471-498. Rea, John D., and Brockschmidt, Peggy. "The Relationship Be-tween P u b l i c l y Offered and Pr i v a t e l y Placed Corporate Bonds Federal Reserve Bank of Kansas C i t y Monthly Review, Novem-ber, 1973, 11-20. Ripley, Duncan M. "Some Determinants of Canadian Municipal and Pr o v i n c i a l Bond Flotations i n the United States," Review  of Economics and S t a t i s t i c s , LII (November, 1970), 417-26. Ripley, Eleanor D. "United States Investment i n Canadian Securi t i e s 1958-1965." Unpublished Ph.D. thesis, Department of Economics, Harvard University, 1969. Robbins, Sidney M., and Stobaugh, Robert B. Money i n the Multi-national Enterprise. New York! Basic Books, 1973. R o l l , Richard. The Behaviour of Interest Rates. New Yorki Basic Books, 1970. Samuels, Nathaniel. "The Investment Banking Background of Issu-ing and Marketing Foreign S e c u r i t i e s i n the United States," i n John F. McDaniels (ed.), International Financing and In-vestment (Dobbs Ferry, N.Y.i Oceana Publ., 1964), 411-29. -217-Securities and Exchange Commission, Cost of F l o t a t i o n of Corpor-ate Securities 1 9 5 1 - 1 9 5 5 . Washington, D.C.: U.S. Govern-ment P r i n t i n g Of f i c e , June, 1 9 5 ? • Shapiro, E l i , and Wolf, Charles R. The Role of Private Place-ments i n Corporate Finance. Bostoni Harvard University, 1 9 7 2 . Sohmen, Egon. The Theory of Forward Exchange. Princeton Studies i n International Finance No. 1 7 ; Princeton, N.J.: Inter-national Finance Section, Department of Economics, Prince-ton University, 1 9 6 6 . Soldofsky, Robert M. "The Size and Maturity of Direct Place-ment Loans," Journal of Finance, XV (March, i 9 6 0 ) , 3 2 - 4 4 . Solnik, Bruno H. European Capital Markets. Lexington, Mass.j Lexington Books, 1 9 7 3 . Spiegel, Jeremy J . "Risk, Interest Rates and the Forward Ex-change," Quarterly Journal of Economics, LXXXVI (May, 1 9 7 2 ) , 3 0 3 - 0 9 . S p i t a l l e r , E r i c h . "A Survey of Recent Quantitative Studies of Long-Term Ca p i t a l Movements," IMF S t a f f Papers, XVIII (March, 1 9 7 1 ) , 1 8 9 - 2 2 0 . Stern, Robert M. The Balance of Payments. Chicago 1 Aldine, 1 9 7 3 . Stern, Robert M. "Discussion," Journal of Finance, XXVI (May, 1 9 7 1 ) , 3 0 8 - 9 . Takayama, Akira. International Trade. New York? Holt, Rine-hart and Winston, 1 9 7 2 . Telser, L.G. "A Cri t i q u e of Some Recent Empirical Research on the Explanation of the Term Structure of Interest Rates," Journal of P o l i t i c a l Economy, LXXV (Supplement; August, 1967), 546-651. Tintner, Gerhard. Econometrics. New Yorkt Wiley, 1 9 5 2 . U.S., 8 8 t h . Congress, House of Representatives, Committee on Banking and Currency. Comparative Regulations of Financial  I n s t i t u t i o n s . Washington, D . C . J U.S. Government P r i n t i n g Off i c e , 1 9 6 3 . Van Home, James C. Financial Management and Po l i c y . Engle-wood C l i f f s , N.jTi Prentice-Hall, 1 9 6 8 . Wong, Stanley. "The 'F-Twist' and the Methodology of Paul Samuel-son," American Economic Review, LXIII (June, 1 9 7 3 ) , 312 -325 . DATA SOURCES Annual Reports of Various Canadian Corporations. Bank of Canada Review. Ottawa: Bank of Canada, various issues. Bank of Canada S t a t i s t i c a l Summary. Ottawa: Bank of Canada, various issues. Basic Corporations Cards (Yellow). Toronto: F i n a n c i a l Post Corporation Service, various years. The Canadian Balance of International Payments, 1946-1965. Catalogue No. 67-505; Ottawa: S t a t i s t i c s Canada. Data Questionnaires Received from Various Canadian Corporations Concerning Bonds Issued from i960 through 1 9 7 3 . Federal Reserve B u l l e t i n . Washington, D.C.: Board of Governors, Federal Reserve System, various issues. F i n a n c i a l Flow Accounts. Catalogue No. 13-002; Ottawa: S t a t i s -t i c s Canada, various issues. The F i n a n c i a l Post Survey of I n d u s t r i a l s . Toronto: McLean-Hunter, various years. The Fi n a n c i a l Post Survey of Mines. Toronto: McLean-Hunter, various years. The Fi n a n c i a l Post Survey of O i l s . Toronto: McLean-Hunter, various years. Foreign Exchange Rates 1972. Montreal: Bank of Montreal, 1973. 40 Bond Monthly Average. Toronto: McLeod, Young, Weir & Co., various issues. Income and Expenditure Accounts. Catalogue No. 13-001; Ottawa: S t a t i s t i c s Canada, various issues. Moody's Bond Survey. New York: Moody's Investors Service, various issues. Moody's In d u s t r i a l Manual. New York: Moody's Investors Service, various years. -218--219-Moody's Municipal & Government Manual 1973» New Yorks Moody's Investors Service. Private Placement Memoranda of Bonds Issued by Various Canadian Corporations from I960 through 1973. Public Prospectuses of Bonds Issued by Various Canadian Corpora-tions from I960 through 1973. Quarterly Estimates of the Canadian Balance of International Payments^ Catalogue No. 67-001; Ottawai S t a t i s t i c s Cana-da, various issues. Record of New Issues. Torontot F i n a n c i a l Post Corporation Ser-vice, various issues. Register of Corporations, Directors and Executives. New Yorki Standard & Poor's Corporation, 1973. Security Transactions with Non-Residents. Catalogue No. 67-002; Ottawaj S t a t i s t i c s Canada, various issues. Standard Corporation Descriptions. New Yorki Standard & Poor's Corporation, various years. APPENDIX 1 EQUILIBRIUM INTEREST RATES IN AN INTERNATIONAL CAPITAL MARKET: A PORTFOLIO APPROACH In Chapter 2 we indicated that because of methodological considerations we used Roll's model of the determinants of e q u i l -ibrium interest rates i n a closed economy as a basis f o r extend-ing term structure theory to open economies."'' We believe that t h i s leads to a more general model of the behaviour of in t e r -national lenders and borrowers of long-term funds than a port-f o l i o approach. On the other hand, Bierwag and Grove have shown that p o r t f o l i o theory can be applied to develop "a model of the 2 term structure of intere s t rates;" consequently i t should be of interest to extend t h e i r theory to open economies as well and to compare the res u l t s with those obtained e a r l i e r i n Chap-ter 2. Before we do that, the following caveats should be noteds 1. The p o r t f o l i o model i s inherently a one-period model. This necessitates rather r e s t r i c t i v e assumptions con-cerning the investment opportunities open to investors. 2. The r i s k - f r e e interest rate i s assumed to be exogen-ously given. In open economies, i t seems more reason-able to assume that a l l int e r e s t rates are determined simultaneously by the demand f o r and supply of funds 1See pp.42 -44 above. G. 0. Bierwag and M.A. Grove, "A Model of the Term Struc-ture of Interest Rates," Review of Economics and S t a t i s t i c s , IL -220--221-by investors i n a l l countries. 3. The r i s k of foreign investments.is assumed to be meas-ured by the variance of the expected rate of return on a foreign-currency denominated security. As pointed out i n Section 2.421 above, t h i s assumption seems to be questionable. F i r s t , we s h a l l review Bierwag and Grove's model. Then we dem-onstrate that t h e i r theory can be extended to open economies. Bierwag and Grove's Model. - In t h e i r multi-period model Bierwag and Grove assume that there exist n investors who share 3 a common time horizon of T periods. Attention i s r e s t r i c t e d to the case of simple i n t e r e s t . Then the relat i o n s h i p between observed inter e s t rates and implied on-period forward rates i s J _ JR. = >Z r. f o r j = 1, 2 T. (Al-1) J i =l 1 Investors are assumed to have only the following investment op-tion s . The f i r s t option i s to invest f o r T periods by purchasing a bond of maturity T. The (r i s k l e s s ) return on t h i s option i s TR T ~S\% T n e second option i s to invest f o r T - l periods by purchasing a bond of maturity T - l , and to reinvest f o r one p e r i -od at time T - l by purchasing a bond of maturity one. The return on t h i s option i s (T-l)R rp_ 1 + R^ ~J?2' Here the t i l d e de-notes a random variable. The t h i r d option i s to invest f o r T-2 periods, and to reinvest f o r two periods at T-2 for a t o t a l re-(February, 196?), pp. 50-62. ^Ibid., pp. 59-60. - 2 2 2 -return of (T - 2)R T_ 2 + 2 R 2 T _ 2 = ^ 3 , and so on. I f an investor allocates x. of his funds to investment op-t i o n j , then the expected return on his p o r t f o l i o i s _ T 3 = I x . E ( J . ) . j=l J J 2 To simplify the analysis, we s h a l l always assume that 2 1 x i = !• Investors' wealth can e a s i l y be introduced into the analysis by multiplying the x. by an i n d i v i d u a l investor's t o t a l funds. J The t o t a l variance of an investor's p o r t f o l i o i s 2 T T s = XL 5Z_x.x .s. . i= 2 j=2 1 J 1 J where s.. i s the expected covariance between the return on op-t i o n i and the return on option j . Assuming that the investor chooses an optimal p o r t f o l i o by — 2 maximizing a u t i l i t y function, U, over the moments f and S , we can solve f o r the optimal values of the x., (subject to the usu-a l p o r t f o l i o conservation constraint "/Vx . = 1) by d i f f e r e n t i a t -ing U p a r t i a l l y with respect to the x-. As necessary conditions J for a constraint maximum of U we obtain T  r *T x ? s i . = m[E(f.) - E(j?1)J f o r j = 2, 3, • • • , T. (Al-2) Here m = Qu/9j>)/-(28u/dS ), a measure of the investor's r i s k aversion. Note that expression (Al-1) may be employed i n equa-LL This rather r e s t r i c t i v e assumption can be relaxed somewhat by assuming that an investor assigns "the same expected return and variance of return at time t to a bond spanning the remain-ing T - l periods of his horizon as he does to any combination of short and long bonds spanning the same set of periods." Ibid., P. 59. - 2 2 3 -tions (Al-2) to substitute implied one-period forward rates f o r the differences between the observed market rates. For example, f o r j = 2, (T-1)R T_ 1 - TR T = - r T , and E ( ^ 2 ) - E{f±) = E ( R l j T _ 1 j Let S be the ( T - l ) ( T - l ) covariance matrix lis. . |[ and x = (Xg, x^» Xrp)' a column vector. Denote by R = ( r T , r T - 1 + r t , . . . , r 2 + r ^ + . . . + r T ) f a column vector of forward rates, and by E = [ E ( R l f T r l ) . 2 E ( R 2 t T _ 2 ) ( T - l ) E ( R T _ l f l ) ] ' a column vector of expected rates. Using these d e f i n i t i o n s , the so l u t i o n (Al-2) can be written as x* = mS'1(E - R). Market equilibrium i s obtained when excess demand by a l l investors f o r a l l investment options i s zero. Note that when n-1 markets are i n equilibrium, then the nth market must be i n equilibrium as well. Let the subscript k r e f e r to the kth investor. Then the condition f o r market equilibrium i s _n -1 1_ mk S k ( E k - R) = 0. (Al - 3 ) k 1 The system of equations (Al - 3 ) can e a s i l y be solved f o r the vect-or of equilibrium forward rates R* = ( Z mk S ^ 1 ) " 1 Z mk S k " 1 ^ . -224-An Extension to Open Economies. - When extending Bierwag and Grove's model to open economies we r e t a i n t h e i r basic assumptions as outlined above. However, we w i l l r e s t r i c t our analysis to a two-period model and to two "representative" investors from two countries, the domestic country D and a foreign country F. An extension to T periods and n investors should be obvious from the e a r l i e r discussion. It i s also assumed that there exist two c a p i t a l markets, one i n country D and one i n country F. In such a model world, an investor from country D i s faced with four investment options from which he may choose» 1. A two-period bond denominated i n country D's currency and y i e l d i n g a r i s k l e s s rate of return of Rg per period and 2Rg over the investor's time horizon. 2. A two-period bond denominated i n country F's currency f ~ ,~ and y i e l d i n g 2Rg + c^ + c 2» where c^ . denotes the rate of change i n the exchange rate during period j . 3. A one-period bond denominated i n country D's currency followed by another one-period bond also denominated i n country D's currency. The t o t a l return on t h i s d ~d option i s + R^ 2' 4. A one-period bond denominated i n country F's currency followed by another one-period bond also denominated i n country F's currency. The t o t a l return on t h i s op-f f ~ ~ ~ t i o n 1 S R]_ + R]_ 2 + c i + c2' Note that our investor must not switch investments from country D to country F or vice versa before the end of his investment horizon. -225-Using our e a r l i e r notation, the domestic investor's expected rate of return on his t o t a l p o r t f o l i o i s and the t o t a l variance i s S d = ^ X x i x i s i i * a i=2 j=2 1 J 1 J D i f f e r e n t i a t i n g his u t i l i t y function p a r t i a l l y with respect to the x., employing our e a r l i e r matrix notation and defining Q, as the column vector of excess rates of return expected by the domes-t i c investor on the three ri s k y investment options, we obtain the following optimal solution: Note that x d = ( xd2' xd3 , xd^' Of p a r t i c u l a r interest i s the vector of excess rates of return, Q d < The f i r s t element, the expected excess rate of re-turn on a two-period bond denominated i n country F*s currency, i s 2Rg + E d ( c 1 ) + E^Cc'g) - 2R 2» Taking note of the d e f i n i t i o n of forward rates i n equation (Al-1) and defining o. - r* - (Al-5) as the one-period rate of change i n the exchange rate implied by the term structure of interest rates i n the two countries, then f f f the excess rate of return on option two becomes 2R 2 - R ^ + R2. - (2R* - R* + H « ) + B d ( S 1 ) + E a ( c 2 ) = r | - rg + i f - r{ + B „ ( ^ ) + E d ( c 2 ) = E d(c' 1) + E^Cc'g) - c 1 - c 2 . In other words, the ex--226-pected excess rate of return on option two equals the difference between the sum of the expected rates of change i n the exchange rate during those two periods and the sum of the market implied forward rates of change i n the exchange rate. Defining these as E(C T) = E(C 2) and C T = C 2, respectively, the excess rate of re-turn on investment option two becomes ^d ~" Q o = E j ( C o ) - C 9 . (Al-6a) S i m i l a r l y , the expected excess return on option three i s ^3 = E d ( ^ l , 2 ) " r2' (Al-6b) Making use of the d e f i n i t i o n s ( A l - l ) and (Al-5), i t can be shown that the expected excess return on option four i s Q £ = E d ( R ^ 2 ) - + E d ( C 2 ) - C 2 . (Al-6c) Note that because of equation (Al-6c) an expression l i k e (Al-4) cannot be solved immediately f o r the vector of forward inter e s t rates and the forward rate of change i n the exchange rate over the investor's time horizon. But we want to obtain a solution f o r the market forward rates as a function of investors' ex-pectations and t h e i r degree of r i s k aversion. Fortunately, equa-t i o n (Al-4) can e a s i l y be transformed into the desired form. Consider an i d e n t i t y matrix 1^ = AB where the matrices A and B are defined as r A = 1 0 1 0 1 0 and B 1 0 0 0 1 0 • 1 0 1 Premultiplying Q d i n equation (Al-4) by 1^  we obtain -22?-4 = " ^ d ^ d o r = m ^ 1 ^ - R) (Al-7) where Z_d = S d A, the column vector of expected rates E_d = [ E d ( C 2 ) , E d(R^ 2 ) , E d(R^ 2 ) ] ', and the column vector of forward rates R = (C 2, r 2 , r 2 ) ' . Our next task i s to derive the foreign investor's demand f o r investment options two, three, and four as a function of the vector (E - R). However, before we proceed with t h i s , one pec u l i a r problem should be noted. In equation (Al-5) we defined c. = r- - r . as the forward rate of change i n the exchange rate J J J implied by the term structure of in t e r e s t rates i n the two coun-t r i e s , But., as w i l l be remembered from Chapter 2, t h i s i s not quite correct unless the interest rates are defined as continu-a l l y compounded rates of return and the rate of change i n the exchange rate as an instantaneous r e l a t i v e rate of change. Therefore E(c'-) must be redefined as being only a function of J the expected rate of change i n the exchange rate, that i s , as representing an expected inter e s t rate d i f f e r e n t i a l that would eliminate the advantage (disadvantage) of higher (lower) nom-i n a l interest rates abroad, given the investor's exchange rate expectations. Of course, c. must be reinterpreted i n a s i m i l a r manner.^ For low inter e s t rate l e v e l s and small rates of change i n exchange rates t h i s difference between inter e s t rate d i f -f e r e n t i a l s and implied rates of change i n exchange rates i s ^Instead of using simple i n t e r e s t we could also redefine the intere s t rates as continually compounded rates. For periodic compounding the problem i s not immediately amenable to a solu-t i o n . -228-n e g l i g i b l e , and therefore we s h a l l continue to use our somewhat inaccurate terminology to avoid unnecessary confusion. Also, as a posi t i v e i n t e r e s t rate d i f f e r e n t i a l from one country's point of view i s a negative interest rate d i f f e r e n t i a l from the other country's point of view, l e t us introduce the following d e f i n i -tions i c c •r* -r* = -o]= -(r!s - i> (A1-8) and E(c'j) = -E(cJ) . (Al-9) Let us now return to our foreign investor. For him invest-ment option one, a two-period bond denominated i n country D's d ~' ~' currency, becomes a r i s k y investment y i e l d i n g 2R 2 + c j + c2 * Option two i s a r i s k l e s s investment f o r him returning 2R2, and d ~ d ~» ~' his return on option three i s R^ + ^ 2 + °1 + c2* F i n a l l y , f ~ f option four y i e l d s R^ + R^ 2 . The foreign investor's expected rate of return on his t o t a l p o r t f o l i o becomes J and the variance of t h i s p o r t f o l i o i s = Y. Z x . x . s f . f o r i , j = 1 , 3 , 4 . 1 j 1 J 1 J Remember that f o r the foreign investor option two i s a r i s k l e s s investment. D i f f e r e n t i a t i n g the foreign investor's u t i l i t y func-t i o n p a r t i a l l y with respect to the x. and proceeding as e a r l i e r , the following solution f o r the optimal a l l o c a t i o n of his funds -229-among the three ris k y investments i s obtained! 2£f = roS"1 Q f. (Al-10) Here f l ^*f3' ^ f4 * Of p a r t i c u l a r interest i s the vector of expected excess rates of return on the foreign investor's three r i s k y invest-ment options, Q^. The expected excess rate of return on option one i s 2R2 + E f(c^) + E f ( c 2 ) - 2R 2 • Defining E(C 2) = E(c.[) + ~' * • • E ( c 2 ) and C2 = c i + c2* a n d " t a k i n § note of expressions (Al-1), (Al-8) and (Al-9) we obtain = ^^.(Cg) - C 2 = C 2 — E^(C 2)-as the foreign investor's expected excess rate of return on op-t i o n one. S i m i l a r l y , i t can be shown that the expected excess return on option three i s Q 3 = Ef(^1,2) " r2 + C2 " E f ( £ 2 ) ' F i n a l l y , on option four the excess return becomes % = V«l,2> " r2« In order to obtain x^ as a function of (E^ - R), where - f = L Ef^2)' E f ^ i 2^' E f ^ i 2^] ' a n d - i s d e f i n e d a s above, l e t the i d e n t i t y matrix lj e c L u a l GG where G i s defined as G = - 1 0 0 - 1 1 0 0 0 1 -230-Premultiplying Q f i n equation (Al-10) by 1^ = GG we get x f = " i f V ^ 1 ( E f - R) (Al-11) where = S^G. Though expression (Al-11) looks very s i m i l a r to equations (Al-7)» we cannot yet solve f o r the vector R be-cause (Al-7) gives us the demand of the domestic investor f o r investment options two, three and four whereas (Al-11) provides us with an expression f o r the foreign investor's demand f o r i n -vestment options one, three and four. Therefore, as the next step i n our derivation, we have to determine the foreign investor's demand f o r investment option two. Actually, a n a l y t i c a l l y i t i s more convenient to derive a solution f o r one minus his demand f o r option two, or 1 - x.^. F i r s t , rewrite expression (Al-11) as * — — x f l Z f l # x f 3 - m^  z f 3 x f 4 Z f 4 The v„. denote the three row vectors i n V f -1 Because of the p o r t f o l i o conservation constraint ^ x. = 1, one minus the f o r -eign i n v e s t o r ^ demand f o r option two equals x ^ + x^j + x.^, or 1 - x f 2 = m f ( v f l + v f 3 + v f / j >) ( E f - R). To simplify t h i s expression, set v ^ 2 = v f l + v ^ + v f^. Then 1 - x f2 m f v f 2 ( E f - R). The foreign investor's demand f o r investment options two, three - 2 3 1 -and four can now be written as followst r m. •f ( E f - R) . Or, when simplifying the above expression, we get x- = m +.Z„ x(E 4 ? - R) . (Al - 1 2 ) Expression (Al - 1 2 ) f i n a l l y provides us with the foreign invest-or's demand f o r investment opportunities two, three and four as a function of his expectations and market observed forward rates. Market equilibrium i s obtained when the excess demand f o r investment opportunities two, three and four i s zero, or when the domestic investor's demand f o r a p a r t i c u l a r option equals the foreign investor's supply of that option. Employing expres-sions ( A l - 7 ) and ( A l - 1 2 ) , market equilibrium implies that Let i denote a column vector i = ( 1 , 0 , 0 ) ' . A solution f o r the vector R of equilibrium forward rates i s then In equilibrium, the international term structure of interest rates i s a quite elaborate function of i n d i v i d u a l investors' expectations of future spot interest rates, of t h e i r expectations 1 0 = m d Z d 1 ( E d - R) + m fZ~ 1(E f - R ) . 0 (Al - 1 3 ) -232-of interest rate d i f f e r e n t i a l s which are functions of exchange rate expectations, of a complicated structure of elements of the inverse of t h e i r variance-covariance matrices and of t h e i r degree of r i s k aversion, a l l weighted by the sum of a l l investors' measures of r i s k aversion multiplied by the respective elements of the inverse of t h e i r covariance matrices. As the r i s k - f r e e rate i n each country i s assumed to be exogenously determined, exchange rate expectations enter e x p l i c i t l y into the solution. When extending t h i s analysis to T periods, a solution f o r 2T-2 forward i n t e r e s t rates and C T, the forward rate of change i n the exchange rate over the investors' time horizon, can be derived. A simpler r e s u l t can be obtained by assuming that a l l traders i n the c a p i t a l markets have i d e n t i c a l expectations. In that case expression (Al-13) becomes R* = E - ( X ^ 1 ) - 1 ! , ( A 1 . 1 4 ) k Here E i s the vector of commonly shared expectations. As only the f i r s t element i n _i i s di f f e r e n t from zero and the f i r s t element i n R i s C 2» the exchange rate change over the investors' time horizon implied by the term structure of intere s t rates, i t follows from expression (Al-14) that C 2 does not represent an unbiased estimate of investors' exchange rate expectations. On the other hand, the forward rates c o r r e c t l y r e f l e c t investors' expectations of intere s t rates i n both countries, given common expectations. This l a t t e r r e s u l t i s i d e n t i c a l to the one derived -233-by Bierwag and Grove f o r a closed economy. E a r l i e r i n Chapter 2 we assumed that investors lend or bor-row i n a foreign c a p i t a l market only i f the inter e s t rate d i f f e r -e n t i a l i s larger than the expected rate of change i n the exchange rate that would tend to o f f s e t the advantage of higher (lower) intere s t rates abroad. I f t h i s i s true, then observed inter e s t rate d i f f e r e n t i a l s overestimate the rate of change i n the exchange rate a c t u a l l y expected by investors. The same holds i n our pre-sent model only i f the f i r s t element of the right-hand term i n expression (Al-14) i s negative. For risk-averse investors the m^  w i l l be p o s i t i v e . However, whether the whole element i s posi-t i v e or negative depends then on the i n d i v i d u a l investors' co-variance matrices. I f an investment i n foreign bonds i s regarded as a means to reduce o v e r a l l p o r t f o l i o r i s k because of expected low or negative correlations between domestic and foreign invest-ment opportunities, then i t i s conceivable that indeed observed interest rate d i f f e r e n t i a l s underestimate rather than overesti-mate expected rates of change i n exchange rates. It becomes an empirical question as to what act u a l l y holds i n a p a r t i c u l a r case. Because of t h i s l a s t r e s u l t one may be i n c l i n e d to regard the p o r t f o l i o model developed i n t h i s appendix as a more general theory than the one presented above i n Chapter 2. However, the very r e s t r i c t i v e assumptions that had to be made i n order to de-ri v e a solution to our problem render the p o r t f o l i o model a less appealing and le s s general t h e o r e t i c a l basis than the model de-veloped by R o l l . 2T Ibid., p. 60 "^ The empirical r e s u l t s reported on pages - above lent support to t h i s view. APPENDIX 2 SAMPLE SELECTION, DATA SOURCES, AND QUESTIONNAIRES Almost a l l of the data employed f o r the empirical tests presented above i n Chapter 5 had to be c o l l e c t e d from o r i g i n a l sources l i k e public prospectuses, private placement memoranda, and corporate f i l e s . On the following pages we s h a l l provide information on how we selected our bond samples and the time periods covered by these samples, which sources we used to ident-i f y Canadian-pay and U.S.-pay bonds to be included i n these samples, and where we obtained the data on bond c h a r a c t e r i s t i c s needed f o r our analysis. Also, the interview questionnaires that served as a guideline f o r our discussion with Canadian f i n a n c i a l managers, investment bankers, and l i f e insurance of-f i c e r s are reproduced i n t h e i r o r i g i n a l form. Sample Selection. - Information has been gathered only on Canadian-pay and U.S.-pay bonds issued by corporations located i n Canada and the proceeds of which were intended f o r use i n Canada. Of course, we did not exclude an issue i f i t s proceeds were used f o r the repayment of U.S. d o l l a r debt or f o r the pay-ment of imported goods. But the following bonds were excluded from our samples: (1) Canadian-pay bonds issued i n Canada by firms located abroad, (2) U.S.-pay bonds issued i n the United -234-- 2 3 5 -States or i n the Eurobond market by foreign subsidiaries of Canadian companies, and (3 ) bonds issued by Canadian chartered banks because they are not allowed to s e l l t h e i r bonds abroad. In other words, our analysis involves only the f i n a n c i a l beha-viour of such Canadian firms which, at lea s t i n p r i n c i p l e , can borrow abroad to finance domestic investments. We are not i n -terested i n Canadian corporate borrowing abroad f o r investment i n foreign countries because t h i s does not lead to an inflow of long-term c a p i t a l into Canada and because the managerial problems involved are quite d i f f e r e n t . The empirical part of our study i s mainly concerned with analyzing and comparing corporate f i n a n c i a l a c t i v i t i e s regard-ing straight debt issues sold i n the bond market. Consequently convertible bonds and bonds issued together with stocks and sold i n so-called "units" have not been included into our bond samples. Whereas such issues are quite common i n the Canadian market we could i d e n t i f y only two convertible U.S.-pay bonds. U.S. d o l l a r issues also rarely carry warrants, but such bonds have been included. Two foreign customer loans made to a Cana-dian mining company by buying two U.S.-pay bond issues, two U.S.-pay bonds not sold but issued as c o l l a t e r a l , bonds guar-anteed by p r o v i n c i a l governments, parent company loans and long-term U.S. d o l l a r loans granted by the American Export-Import Bank to Canadian companies have a l l been excluded from our analysis. -236-From October 1950 through A p r i l 1962 and from June 1970 onwards the Canadian d o l l a r has been allowed to f l o a t f r e e l y i n foreign exchange markets. It was pegged to the United States d o l l a r only during the period from May 1962 through May 1970. As a f i n a n c i a l manager's perception of exchange r i s k may be a function of the p r e v a i l i n g exchange system, i t seemed desirable to compare Canadian corporate financing behaviour during times of a f l e x i b l e exchange rate with that during times of a fixed rate. However, when undertaking preliminary i n -q u i r i e s concerning the a v a i l a b i l i t y of data needed f o r t h i s study, i t soon became obvious that even basic information on "When did which corporation issue what kind of bond" was not r e a d i l y a v a i l a b l e , p a r t i c u l a r l y with respect to United States d o l l a r bonds and private placements i n Canada. In order to re-duce the workload to a manageable l e v e l and increase the prob-a b i l i t y of a c t u a l l y obtaining the desireed data, i t was there-fore decided to gather data only on U.S.-pay and Canadian-pay bonds issued from i960 through May 1973. The number of U.S. d o l l a r bonds issued by Canadian corp-orations during a given time period i s small r e l a t i v e to the t o t a l number of new corporate bond issues flo a t e d during the same period. Considerable e f f o r t s were therefore made to c o l -l e c t complete data on a l l U.S.-pay bonds issued since January i960. However, as discussed i n Chapter 4, f o r the period i960 to 1967 a few U.S.-pay issues were excluded from our sample through a random sel e c t i o n process."'" This has been done to "*"A company's U.S.-pay issues were numbered consecutively, and a random number table was used to select at least three of -237-avoid a pot e n t i a l bias which might have arisen because the larger, more regular borrowers i n the United States usually provided us with a l l the information requested whereas data on other U.S.-pay issues were sometimes d i f f i c u l t to obtain. A l l U.S.-pay bonds have been grouped into "Sample 5". Table 4-1 (see p. 134 above) provides a summary of the c h a r a c t e r i s t i c s of a l l our bond samples. In order to allow a comparison of Canadian corporate f i n -ancing behaviour i n the United States with corporate a c t i v i -t i e s i n the domestic bond market, i t was necessary to obtain information on the c h a r a c t e r i s t i c s of Canadian-pay bonds. Our o r i g i n a l plan c a l l e d f o r the random s e l e c t i o n of a sample of Canadian d o l l a r issues covering the period from i960 onwards. However, p a r t i c u l a r l y on many small issues and private place-ments, the needed data are simply not available f o r early years. Consequently only f o r 1968 and l a t e r years random samples of Canadian-pay corporate bonds were drawn. 1969 was an unusual year f o r the Canadian bond market because of the very r e s t r i c t -ive monetary p o l i c y pursued by the Bank of Canada, and there-fore 1968 was included though some problems i n gathering data f o r these two early years were anticipated. These samples of Canadian-pay bonds were drawn from the respective annual Record them or f i f t y per cent, whichever led to the higher number of issues selected. -238-of New Issues.^ The 1973 sample was obtained from a cumula-t i v e record of t h i s publication covering the period January to May 1973. Bonds i n these random samples have been grouped into "Sample 1" ("CAonly") i f they have been issued by corporations that have sold long-term s e c u r i t i e s i n Canada only. A l l other bonds have been included i n "Sample 2" ("CAandUS"). In addition, information has been col l e c t e d on Canadian-pay bonds not contained i n these random samples but issued since 1968 by corporations that borrowed at least once i n the U.S. market since i 9 6 0 . This has been done to obtain more complete information on domestic bonds sold by t h i s group of corporations and to allow a better comparison between bonds floated by them i n the Canadian and U.S. bond markets. "Sample 3" ("CAandUS") contains these issues. Because of the data c o l l e c t i o n problems mentioned e a r l i e r , we gathered f o r the period from i 9 6 0 through 1967 data only on Canadian-pay bonds issued by corporations that approached the U.S. market at l e a s t once since i 9 6 0 . This proved less d i f f i -cult than c o l l e c t i n g data on a random sample because most firms f a l l i n g into t h i s category are among the larger and better known Canadian companies. Some corporations i n t h i s group have p ^Published by the F i n a n c i a l Post Corporation Service, Toron-to. The number of straight bond issues i d e n t i f i e d f o r each year and the number of issues i n our random samples are as follows 1 1968, 44 and 2 5 ; 1971, 76 and 35; 1 9 6 9 , 29 and 29; 1972; 62 and 35; 1970, 57 and 35; 1973, 34 and 18 ( u n t i l May). Bonds issued by chartered banks or non-Canadian corporations were deleted from the random samples. In a few instances bonds being members of the sample f o r a p a r t i c u l a r year turned out to have i n fact been issued during December of the prece ding or January of -239-been borrowing almost continually i n the Canadian market. In order to avoid them dominating t h i s "Sample 4" ("CAandUS"), only at most four of t h e i r issues were selected. If they had issued Canadian-pay bonds close to the date of t h e i r U.S.-pay bonds included i n the analysis, then these Canadian d o l l a r bonds were selected. Otherwise a random number table was used. Data Sources. - One of the most d i f f i c u l t tasks encountered during the empirical part of t h i s study was the i d e n t i f i c a t i o n of i n d i v i d u a l U.S.-pay bond issues. To s t a r t with, we obtained from the Bank of Canada a l i s t of si x t y Canadian corporations which have issued debt i n the United States since I960. Though t h i s l i s t was incomplete, i t proved to be very h e l p f u l . From the following year and were l a t e r c l a s s i f i e d accordingly. Unfortunately, the Bank of Canada refused to provide us with more detailed information. Two researchers at an American i n s t i t u t i o n reported that the Bank of Canada made available to them pr i v a t e l y the "date of issue and delivery, as well as the amount of each issue" with respect to new Canadian bonds pay-able i n United States currency. See Caves and Reuber, p. 250. ^In i t s l e t t e r of A p r i l 13, 1973 to Professor W. Wini-ata, the Bank of Canada mentions the following! "The enclosed l i s t of over s i x t y corporations was prepared from our records of i n d i v i d u a l issues by Canadian borrowers i n the domestic and foreign c a p i t a l markets. There are names which do not appear i n the l i s t f o r a variety of reasons. F i r s t , i n the case of a private placement not p u b l i c l y announced, we do not release information, Second, we have excluded issues which were exchanges f o r assets rather than f o r the r a i s i n g of funds. F i n a l l y , we have omitted the occasional case i n which a small corporation made minor and tentative use of the U.S. market. The three exclusions together do not involve more than twenty names and do not represent a s i g n i f i c a n t c r e d i t flow when compared to the a c t i v i t i e s of the l i s t e d corporations." -240-t h i s l i s t , four firms had to be deleted. Two were bankrupt and i t was v i r t u a l l y impossible to obtain information on t h e i r bond issues. One company issued only a convertible bond i n the United States. And a fourth firm obtained only parent company loans. An item by item search of every annual ed i t i o n of the Record of New Issues published since 1 9 6 0 , ^ of various annual reports and prospectuses, and information obtained from Canadian and American investment houses led to the i d e n t i f i c a -t i o n of thirty-one additional Canadian corporations that sold U.S. d o l l a r bonds since i 9 6 0 . Some of these companies have been quite regular borrowers i n the United States. With the help of t h i s l i s t of eighty-seven Canadian corp-orations that have sold one or more U.S.-pay issues since i 9 6 0 an attempt was made to i d e n t i f y i n d i v i d u a l U.S. d o l l a r debt 7 issues from public sources. For many private companies th i s was impossible, and the information available on public comp-anies proved to be incomplete with regard to U.S.-pay debt. ^Peters, p. 3, notes that "apart from the incomplete data published by the F i n a n c i a l Post on new corporate bond issues (private placements and public o f f e r ings), l i t t l e data exist respecting new corporate debt financing i n Canada". Our research indicates that t h i s i s s t i l l true. But i n view of the fact that some new issues are never p u b l i c l y announced the data seem to be as complete as one can expect them to be. ^Whether t h i s indicates that the Bank of Canada's i n f o r -mation on c a p i t a l inflows i s incomplete i s d i f f i c u l t to judge. A^ complete l i s t of data sources used f o r t h i s study can be found on pages 218 -219 above. -241-Consequently every firm was approached d i r e c t l y and asked f o r Q basic information on i t s foreign bonds l i k e date of issue, amount involved, and maturity. In general, the response was very favourable though a few corporations either outrightly refused the requested information or did not respond i n spite of two l e t t e r s and, i n several cases, a dir e c t phone c a l l . Therefore f o r four companies not a single U.S.-pay issue could be i d e n t i f i e d though i t i s known that they have foreign debt outstanding. The i d e n t i f i c a t i o n of Canadian-pay bonds not included i n our random samples but sold by these eighty-seven companies since i 9 6 0 was less troublesome as published records are f a i r l y complete i n t h i s respect. But i n several instances we also had to e n l i s t the help of the firms concerned. A f t e r a l l bonds had been i d e n t i f i e d on which detailed i n -formation was to be collected, considerable e f f o r t s were made to obtain public prospectuses or private placement memoranda fo r the respective issues. Such documents usually contain a l l the information needed f o r t h i s research. Local o f f i c e s of Canadian investment dealers were very h e l p f u l i n securing pros-pectuses on recent public issues by larger firms. However, information on smaller issues or bonds offered several years ago and on private placements was not r e a d i l y a v a i l a b l e . g It was s u r p r i s i n g l y d i f f i c u l t to secure the addresses of a l l eighty-seven firms, p a r t i c u l a r l y of privately-owned compan-ies and of those that had changed t h e i r names. For three firms addresses are s t i l l missing but at l e a s t one issue f o r each of these three companies has been i d e n t i f i e d . -242-This made i t necessary to approach the head o f f i c e s of fo r t y -nine Canadian and American underwriters i n an attempt to se-cure documents on p a r t i c u l a r issues underwritten or placed by these investment bankers. Many were very cooperative and sent a l l the information on f i l e . Others, as a matter of po l i c y , did not release documents older than one year. Also, private placement memoranda are generally not available through invest-ment dealers. So i t turned out that i n many instances, p a r t i c u l a r l y with regard to private placements i n Canada and the United States, the corporations themselves remained as the only source of information. Every e f f o r t was made to explain to them the importance of t h i s research f o r a better understanding of the Canadian and inte r n a t i o n a l c a p i t a l markets and s t r i c t c o n f i -d e n t i a l i t y of any information provided was assured. Again many companies responded very favourably whereas others did not respond at a l l . Several private corporations gave detailed data on bond c h a r a c t e r i s t i c s but omitted balance sheet and i n -come statement data. In Exhibit A2-1 (see pp.24 5 -250 below) the data questionnaire used f o r obtaining t h i s information i s shown. Besides data needed f o r te s t i n g our theory some addi-t i o n a l information has been c o l l e c t e d which, we hope, v / i l l prove useful f o r further research on the Canadian corporate bond market. Public sources have sometimes been used f o r c o l l e c t i n g a d d i t i o n a l information. But i n general these sources do not provide the detailed data on bonds required for t h i s study. -243-Interviews. - When reviewing the l i t e r a t u r e on long-term c a p i t a l flows between the United States and Canada we discover-ed that the decision-process that leads to the issuance of U.S.-pay bonds by Canadian corporations has apparently never been analyzed i n d e t a i l . Therefore we thought i t advisable not to r e l y exclusively on an interpretation of s t a t i s t i c a l r e s u l t s but to obtain additional information through discussions with finan-c i a l managers who have been involved i n such decisions. The questionnaire on which these discussions were based i s reproduced i n i t s o r i g i n a l form i n Exhibit A 2 - 2 (see pp . 2 5 1 - 266 below. A concise summary of the answers obtained can be found on pp. 2 6 7 - 2 7 2. ) A preliminary version of t h i s questionnaire had been tested i n exploratory interviews with three firms. A l l i n t e r -views took place during July and August of 1973 when we paid v i s i t s to Montreal, Toronto, Calgary, and Vancouver. Given these geographical constraints, those corporations were selected that either had several U.S. d o l l a r bonds outstanding or had at least one such bond issued since 1 9 6 9 . The interviewee was usually the Vice President Finance or the Treasurer of the firm. In general, interviews lasted between f o r t y - f i v e and seventy-five minutes though sometimes discussions went on f o r two hours and more. Only twenty-one of the twenty-eight i n t e r -views asked f o r with corporations did come about because of problems caused by vacation time, d i s i n t e r e s t , and s t r i k e s . The advice of investment bankers plays an important role i n a corporation's decision to s e l l bonds abroad. Therefore we also interviewed six Canadian underwriters. The question-naire we used i s shown i n Exhibit A2-4 (see pp. 2 7 3 - 2 8 5 below). -244-The investment houses v i s i t e d were those most active i n the placement of U.S.-pay corporate s e c u r i t i e s . The relevant i n -formation was derived from the Record of New Issues which usu-a l l y l i s t s the Banking Group manager or the p r i n c i p a l placement agent f o r each issue recorded. These six investment dealers are also among the leading underwriters of Canadian-pay "bonds. The inverviewees were vice-presidents or partners active i n the corporate underwriting f i e l d . In three instances, two members of the respective house par t i c i p a t e d i n the interview. In addition, questionnaires were mailed to f i v e American under-writers. Unfortunately, only two responded. L i f e insurance companies are the most important i n s t i t u -Q t i o n a l investors i n Canadian corporate bonds. Therefore i n -vestment o f f i c e r s i n charge of the corporate bond p o r t f o l i o i n four companies were interviewed regarding t h e i r invest-ment p o l i c i e s and strategies. The questionnaire i s reproduced i n Exhibit A2-5- (see pp.286 - 2 9 6 below). The companies select-ed were those most active i n the corporate bond market, given the geographical constraints imposed by our research budget. Interviews with insurance companies were sought because dis-cussions i n the l i t e r a t u r e suggested that i n s t i t u t i o n a l invest-ment behaviour may be one of several factors leading to the placement of bonds with American i n v e s t o r s . ^ 9See Bank of Canada Review, March 1973, p. S129, Table A7: "Estimated D i s t r i b u t i o n of Holdings of P r o v i n c i a l , Municipal, Corporate and Other Bonds." "^On t h i s point see also Fullerton, p. 3 1 2 , who remarked with respect to the f i f t i e s : "The small group of i n s t i t u t i o n a l buyers of new corporate issues i n Canada increases underwriting r i s k s , because i f the majority decide against a new issue there i s no other place to turn to except the r e t a i l market or to vola-t i l e demand i n the United States." -245-EXHIBIT A2-1 DATA QUESTIONNAIRE -246-Questionnaire Concerning Characteristics of U.S. Dollar and  Selected Domestic Debt Issues of Canadian Corporations Where the prospectus (or a statement in lieu of the prospectus) is available to the researcher, most of the following questions (with the possible exception of questions 5, 11, 12, and 20) can be answered therefrom. Consequently, this questionnaire i s being used only where neither of the aforementioned statements is available to us. Responses to the following questions are needed for s t a t i s t i c a l analyses only and w i l l remain confidential. Any information provided by your company w i l l be used in such a way as not to reveal information about your firm. In order to make your task as easy as possible, answers have been drawn up whenever possible in such a manner that you have only to insert the data requested or to indicate by a check (/) which part of the answer applies to your case or whether the answer i s YES or NO. An OR means we need only one of the two pieces of information mentioned. If the questionnaire is to be mailed please send i t to Dr. W. Winiata Faculty of Commerce University of B r i t i s h Columbia Vancouver 8, B. C. Canada 1. Name of firm: 2. Month and Year of issue: (Or date of f i r s t delivery i f issued in several closings) 3. Date prospectus was issued OR contract was finalized: 4. Total amount issued: 5. Has the issue been delivered in several closings? YES NO If YES: Please indicate dates and amounts of drawdowns. Date Amount 1st drawdown: 2nd drawdown: 3rd drawdown: 4th drawdown: -247-6. Coupon: % (If not issued at par , indicate yield % OR issue price_ 7. Are these sinking fund or ser i a l bonds? YES NO (If YES, please provide repayment schedule on page 4, see question 21) 8. Month and year of (final) maturity: 9. Are these bonds extendable or prepayable? YES NO If YES: Please indicate date(s) to which bonds are extendable/ at which bonds are prepayable. Bonds are extendable to/prepayable at: 10. Was this a public offering or a private placement or both 11. Underwriting discount (placement fee): % (OR: Total compensation payed to underwriter(s)/placement agent(s): $ 12. Other issuing expenses: $ ^ . 13. Are the bonds unsecured or secured If secured, are they secured by a f i r s t mortgage 14. Are the bonds redeemable before maturity? YES NO If YES: a) For the f i r s t time redeemable during the year ending at %. b) Redeemable at par (100.00%) for the f i r s t time i n 19 . c) Do the bonds have a non-call feature? YES NO If YES: Bonds are not redeemable prior to 19 i f refunded by debt yielding less than present issue. 15. Have warrants been attached to the bonds? YES NO -248-16. Are your company's stocks lis t e d at a stock exchange? YES NO If YES: At a Canadian stock exchange , at a U.S. stock exchange , at another foreign stock exchange ? 17. What was the intended use of the proceeds of this issue? a) For capital expenditures, working capital, and so on b) For funding of shorter-term Canadian debt or U.S. debt c) For refunding of long-term debt denominated in Canadian dollars or denominated in U.S. dollars . Please comment. 18. To what extent is your firm owned by a) U.S. shareholders? % b) Other foreign shareholders? % 19. Does your firm have branch plants or subsidiaries in a) the U.S.? YES NO b) other foreign countries? YES NO 20. Does your firm engage in exports or imports of any kind? YES NO If YES: a) Can you provide us with a rough estimate of that part of your total operating revenue ( gross income) that was denominated in U.S. dollars in 19 ? (The year preceeding the date of issue). About U.S. $ . (OR: About % of total revenue). b) Can you provide us with a rough estimate of that part of your total expenses that was denominated in U.S. dollars during the same year? About U.S. $ . (OR: About % of total expenses) -249-When these bonds are sinking fund or seri a l bonds, please provide retirement schedule. Give data only for mandatory retirements. If the firm may opt to increase the sinking fund payments at i t s discretion, please indicate . -250-We realize that the information which i s requested xn the following mixtions i s typically not available from private corporations. However, tor ?he aggregative'analyses which we intend to perform, this information i s very important as i t would enhance the quality of our results considerably. We sincerely hope that i t w i l l be made available. 22. (a) What were your firms total assets at the end of your f i s c a l year immediately preceeding this issue? 19 $ OR (b) If this issue was intended to finance your company's i n i t i a l investment and to allow you to commence operations would you provide pro-forma balance sheet data? (If this i s the case, skip questions 24 and 25). 23. Please provide the following information on your firm's l i a b i l i t i e s and shareholder 1s equity at the end of the same year: a) current and accrued l i a b i l i t i e s : $ b) Medium-term and Long-term debt: $ . c) Deferred credits, including deferred income taxes, and other items: $ d) Total shareholders' equity (including common and preferred equity paid i n , retained earnings and surpluses): $ 24. For the five f i s c a l years preceeding the date of this issue, what was your firm's total operating revenue (gross income/gross earnings) during each year? 19_: 19 : 19 : 19 : 19_: 25. What were your firm's net earnings (net income) during the same years? 19 : 19 : 19 : 19 : 19 : 26. If this issue i s guaranteed by another corporation , please provide in questions 22 to 25 in separate columns the same information for the guarantor. If the issue has several guarantors , please provide data only for principal or largest guarantor. The guarantying firm's name i s -251-EXHIBIT A2-2 INTERVIEW QUESTIONNAIRE, CANADIAN CORPORATIONS - 2 5 2 -Introduction to an Interview with a Canadian  Issuer of U.S. Dollar Bonds As you already know from your correspondence with Dr. Winiata, I am conducting a study on international bond financing by Canadian cor-porations. The purpose of this research is to gain a better understand-ing of the financial behaviour of Canadian corporations, particularly with regard to foreign borrowing a c t i v i t i e s . It is hoped that these efforts w i l l provide new insights into the determinants of international capital flows and the relationships between Canadian and international capital markets. As part of i t s cultural exchange program with Germany the Canadian Government, through the Canada Council, i s supporting my studies and pro-viding most of the financing for this research. Interviews w i l l be conducted with about twenty-five industrial com-panies. In addition, data w i l l be assembled from another seventy companies, from investment dealers and from major institutional investors. Any infor-mation provided by a company w i l l be used in such a way so as not to reveal information about that company. The analyses w i l l be in terms of Canadian corporate behaviour in general. The sequence of the questions to follow attempts to roughly trace in chronological order the considerations and decisions involved from planning a new issue, to selling the bonds and to servicing the debt outstanding. You may give multiple answers to these questions whenever you think that this better reflects a l l the factors you took into consideration. Whenever possible, please answer these questions with respect to your most recent U.S. dollar issue. K. Stroetmann, Ph.D. Candidate, Faculty of Commerce and Business Administration University of British Columbia Vancouver, Canada -253-I. WHY DID YOU APPROACH THE UNITED STATES BOND MARKET? In the main this group of questions w i l l explore the reasons behind your decision to float bonds denominated in United States dollars. As a pre-face I should like to ask you a question on the kind of long-term financing instruments you took into consideration when f i n a l l y deciding for a U.S.-pay issue. 1. A firm in need of long-term outside funds usually has a choice between new equity financing and the issuance of bonds. In addition, a Canadian firm intending to float a new bond issue often can choose between bonds de-nominated in Canadian dollars and bonds denominated in U.S. dollars. Many companies, however, consider only one alternative. -When your firm was searching for long-term funds and f i n a l l y issued U.S.-pay bonds, did you make a choice between (or did you choose a package of) a) A new stock issue, a bond issue i n Canada, and a bond YES NO issue in the United States? b) A Canadian bond issue and a U.S. bond issue? YES NO c) A stock issue and a U.S. bond issue? YES NO d) Or did you only consider the U.S. bond market? YES NO 2. Quite different reasons are often mentioned as to why Canadian firms approach the U.S. bond market. In your particular case, did your company take into consideration the flotation of an issue denominated in U.S. dollars a) Because you thought the Canadian market would not be willing to absorb another issue of your firm because of earlier issues by your company? YES NO b) Because the Canadian market was depressed and would not absorb a straight bond issue without an equity sweetener? YES NO c) Because your underwriter advised you to consider the U.S. market? YES NO d) Because U.S. investors approached your firm with the intention of lending money to your firm? YES NO e) Because of earlier successful U.S.-pay bond issues by your company? YES NO f) (Please see next page.) - 2 5 4 -f) Because you considered the size of the bond issue as too large for the Canadian market? YES NO If YES: Why did you not float part of the issue i n Canada and part of i t in the United States? If you did this, what determined the size of the U.S.-p a y issue relative to the size of the Canadian-pay issue? Or, are there any other reasons why you considered flotation of a U.S.-pay issue? 3. Typically, a firm would l i k e to be free to specify the characteristics of a bond issue (including maturity, c a l l features, and so on) according to i t s individual needs. Often, however, these characteristics are influenced by capital market conditions and lender behavior. When you made the decision for a U.S. dollar issue, were you particularly attracted by a) longer terms to maturing available in the U.S. market compared to maturities in the Canadian market? YES NO b) Less restrictive c a l l features? YES NO c) More suitable sinking fund requirements? YES NO d) Less restrictive indentures? YES NO e) Lower underwriters' discounts and issuing expenses? YES NO f) Lower interest rates? YES NO -255-4. Speaking about the characteristics of your U.S. dollar issue, what determined the maturity of the bonds: a) The longest maturity available to your firm at that point i n time in the U.S. market? YES NO b) The lenders' preferences with regard to maturity? YES NO c) The estimated l i f e time of the investment project to be financed through that issue? YES NO d) Your firm's preferences with regard to maturity? YES NO Please comment. CDoes, for example, the exchange rate risk influence the length of the term to maturity you prefer?) 5. Becoming familiar with a foreign capital market is some-times not an easy task. Through your parent company or subsidi-aries in the United States or through other sources, did your firm already have contacts with American financial institutions which made access to the U.S. bond market easier? YES NO If YES: please specify source of earlier contact: Parent company? YES NO U.S. subsidiary? YES NO Exports to the U.S.? YES NO Earlier U.S. issues? YES NO Other: (Please comment.) -256-II. INTEREST AND TRANSACTION COSTS We should expect a firm to try to minimize i t s cost of capital (percen-tage-wise) . The yield required by investors w i l l be the major factor deter-mining the cost of debt capital to a company. However, underwriters' compen-sations and other transaction costs may also be of considerable influence on the cost of capital. The next group of questions w i l l explore some of the differences between Canadian and U.S. bond markets with respect to interest and other costs. A) Interest Costs 6. Interest rate differentials observable between the Canadian and the U.S. capital markets do not truly reflect the actual d i f -ference i n interest costs as far as borrowing by a Canadian corp-oration i s concerned. In many cases, American investors seem to demand a higher yield on Canadian bonds denominated in U.S. dollars than on American bonds of comparable quality. Has this been your= experience? YES NO Do you think i t makes any difference in interest costs whether you approach the U.S. bond market regularly or not? YES NO Please comment: 7. Forecasting capital market conditions is a very d i f f i c u l t task. When you were choosing the exact date of issue for your U.S. dollar bonds, was your timing influenced (in the sense of advancing or postponing the date of issue by a month or more) by a) your expectations with regard to potential changes in long-term interest rates in the U.S.? YES NO b) Your expectations with regard to short-term interest rates at home? YES NO -257-B) Underwriters' Discounts 8. Almost no evidence exists on underwriting compensations payed by Canad-ian firms issuing U.S.-pay bonds. In your opinion, are underwriters' discounts -on issues of Canadian firms usually higher i n the U.S. than in the Canadian market a) On public offerings? (higher in the U.S. market) YES NO, THE SAME NO, LOWER b) On private placements? (higher in the U.S. market) YES NO, THE SAME NO, LOWER 9. Sometimes U.S.-pay issues by Canadian firms are underwritten by New York investment bankers. Was the investment banker who was the principal underwriter of your U.S. dollar issue the same who usually underwrites your Canadian bond issues? YES NO If YES: Why did you not choose a U.S. underwriter? If NO: Why did you choose a U.S. underwriter? -258-10. Often a firm can choose between a public offering and a private place-ment when issuing new bonds. What determines your choice between a public offering and a private placement when issuing bonds in Canada? Please comment. Have you ever considered a public bond offering in the United States? YES NO In your opinion, what are the advantages and/or disadvan-tages of a public offering in the U.S. bond market compared to a public bond offering in the Canadian bond market? C) Other Transaction Costs 11. Apart from underwriters' discounts, do you think that other flotation costs l i k e printing of bonds and prospectuses, trustee's fees, legal fees, and so on are higher for a U.S.-pay issue than for a comparable issue denomin-ated in Canadian dollars when the issue is a) a public offering? (higher in the U.S.) YES b) a private placement? (higher in the U.S.) YES NO, THE SAME NO, LOWER NO, THE SAME NO, LOWER 12. When issuing the U.S.-pay bond in question, did you pre-pare an estimate of the costs your company incurred with regard to your company's o f f i c i a l s ' time, travelling costs, and other company resources? YES NO Can you provide me with a rough estimate of these costs? YES NO $ Were these costs higher than for a comparable Canadian -pay issue? YES NO -259-I I I . EXCHANGE RATE RISK When i s s u i n g a bond denominated i n a f o r e i g n currency, a f i r m faces two p e c u l i a r problems. (1) The proceeds of the i s s u e have to be converted i n t o Canadian d o l l a r s , and the exchange r a t e p r e v a i l i n g at the time of conversion determines the amount of Canadian d o l l a r s a c t u a l l y obtained. (2) Changes i n exchange r a t e s during the bonds' l i f e t i m e may change the a c t u a l costs of f o r e i g n borrowing c o n s i d e r a b l y . I would l i k e to explore these two problem areas. A) Exchange Rate R i s k At Time of Issue 13. The timing of a bond i s s u e i s u s u a l l y a f f e c t e d by a host of f a c t o r s . When you chose the exact date of i s s u e f o r the U.S.-pay bond, was your t i m i n g i n f l u e n c e d ( i n the sense of advancing or postponing the date of i s s u e by a month or more) by your expec-t a t i o n s w i t h regard t o ' p o t e n t i a l exchange r a t e changes? YES NO I f YES: a) Were these expectations w i t h regard to short-term f l u c t u a t i o n s around the exchange r a t e l e v e l p r e v a i l i n g d u r i n g t h a t time period? YES NO b) Were these expectations w i t h regard to changes i n the p a r i t y of the Canadian d o l l a r (during times of a f i x e d exchange r a t e ) or long-term changes i n the exchange r a t e l e v e l p r e v a i l i n g before that p a r t i c u l a r time p e r i o d (during times of a f l o a t i n g exchange r a t e ? ) YES NO 14. Did you d e l i v e r the U.S.-pay bonds to your underwriter (or the lender) a) Against payment i n U.S. d o l l a r s ? YES NO b) Against Canadian funds r e p r e s e n t i n g the Canadian d o l l a r e q u i v a l e n t of the proceeds? YES NO I f you obtained U.S. d o l l a r s , d i d you convert p a r t or a l l of those U.S. funds i n t o Canadian Funds? YES NO I f YES: a) Immediately i n the spot exchange market? YES NO b) Immediately i n the forward exchange market? YES NO c) Or d i d you n e g o t i a t e at an e a r l i e r date a forward exchange c o n t r a c t maturing at the time of d e l i v e r y of the bonds? YES NO -260-15. Sometimes the p o s s i b i l i t y of a d e v a l u a t i o n of the Canadian d o l l a r i s regarded as a s e r i o u s r i s k which may lead to higher c o s t s of f o r e i g n borrowing than a n t i c i p a t e d . On the other hand, i t i s argued that f o r e i g n borrowing w i l l u s u a l l y t u r n out to be cheaper than borrowing i n Canada. When you were contemplating the issuance of U.S. d o l l a r bonds, d i d you regard the p o s s i b i l i t y of a devalua-t i o n of the Canadian d o l l a r as a s e r i o u s r i s k i n f l u e n c i n g your choice of where to f l o a t the new bonds? Please comment. YES If YES: F o r e c a s t i n g f u t u r e exchange r a t e s i s very d i f f i c u l t and few people attempt to make long-range f o r e c a s t s . Did you attempt a q u a n t i t a t i v e e v a l u a t i o n of the exchange r a t e r i s k by a) F o r e c a s t i n g the f u t u r e trend of the exchange rat e ? YES b) Or, by e s t i m a t i n g f u t u r e exchange r a t e s and c a l c u l a t i n g whether they would e l i m i n a t e any cost advantage d e r i v -a b l e from lower i n t e r e s t r a t e s abroad? YES What i n f o r m a t i o n d i d you use when attempting to evaluate the exchange r a t e r i s k or to f o r e c a s t exchange r a t e s ? Please comment. Did the exchange r a t e r i s k i n f l u e n c e your choice between s i n k i n g fund bonds and bonds where the p r i n c i p a l becomes due only at maturity? YES Or are other c o n s i d e r a t i o n s r e l e v a n t f o r such a choice? Please comment. -261-16. Some people argue that the larger the size of a U.S. dollar issue the greater is the firm's exposure to exchange rate risk. Was the size of your U.S.-pay bond issue solely determined by your financial needs, or did your view of the exchange rate r i s k influence the size? The size of the issue was solely determined by our financial needs. YES NO If NO: a) If, in your view, the exchange rate risk had been lower, would you have issued more U.S.-pay bonds? YES NO b) If, in your view, the exchange rate risk had been higher, would you have issued i ) fewer U.S.-pay bonds and substituted Canadian-pay bonds? YES NO i i ) Or would you have issued no U.S.-pay bonds and only Canadian-pay bonds? YES NO Please comment. -262-B) Exchange Rate Risk Management During Term of the Issue 17. As long as U.S. dollar bonds are outstanding, your firm has a continu-ous need for U.S. dollars to meet interest (and sinking fund) payments. Where and when do you usually obtain the necessary U.S. funds? a) Usually in the spot exchange market approximately days before payments become due. YES NO b) Usually in the forward market by negotiating a forward contract for delivery of U.S. dollars. Please comment on the length and. other matters including a v a i l a b i l i t y . YEg NO c) Usually Company revenue denominated in U.S. dollars i s available to meet these payments. YES NO 18. When the international monetary scene is unsettled and/or when you expect an international monetary c r i s i s , do you cover your U.S. dollar commitments at an earlier date than usual by buying the needed U.S. dollars in advance? YES NO If YES: Do you buy those U.S. dollars in a) the spot exchange market approximately days before payments become due? YES NO b) Or, in the forward market by negotiating a day forward contract for delivery of U.S. funds? YES NO (Please indicate the number of days in each case.) - 2 6 3 -19. I f most or a l l o f t h e p r i n c i p a l o f an i s s u e were to become due a t m a t u r i t y ( t h a t i s , i f t h e r e were a l a r g e b a l l o o n payment), would t h i s i n f l u e n c e y o u r e v a l u a t i o n o f t h e exchange r a t e r i s k i n v o l v e d i n b o r r o w i n g i n the U.S. market compared to a s i n k i n g f und bond? I f YES: a) How would i t i n f l u e n c e your e v a l u a t i o n o f t h e exchange r a t e r i s k i n v o l v e d ? P l e a s e comment: b) Would you e x p e c t to be a b l e t o r e f u n d the bond i s s u e by a n o t h e r U.S.-pay i s s u e ? 20. Have you e v e r c o n s i d e r e d d i v e r s i f y i n g t h e exchange r a t e r i s k by d e n o m i n a t i n g a new bond i s s u e i n a c u r r e n c y o t h e r t h a n the C a n adian o r U.S.? P l e a s e comment: 20a. Does t h e exchange r a t e r i s k i n f l u e n c e t h e c a l l f e a t u r e s you n e g o t i a t e f o r U.S.-pay bonds? I f YES, p l e a s e comment. I f NO: What u s u a l l y d e t e r m i n e s the c a l l f e a t u r e s on y o u r bonds? -264-IV. MISCELLANEOUS QUESTIONS F i n a l l y , I have some miscellaneous questions d e a l i n g w i t h the a t t i t u d e of shareholders toward f o r e i g n borrowing, the i n f l u e n c e of the Bank of Canada on f o r e i g n borrowing and w i t h the frequency (or infrequency) at which your company approached the U.S. bond market. 2 1 . Do you t h i n k that your company's stockholders/owners are YES, NO i n d i f f e r e n t w i t h regard to f o r e i g n borrowing by your firm? i n d i f f e r e n t I f NO: Or, do you assume that they are g e n e r a l l y not i n favour -of f o r e i g n borrowing? YES, NO, not i n favour i n favour How i s your stock p r i c e a f f e c t e d , i f at a l l ? 2 2 . Now and then, the Bank of Canada announces that i t i s i n favor or against f o r e i g n borrowing by Canadian governments and c o r p o r a t i o n s . Did the Bank of Canada's p o s i t i o n on f o r e i g n borrowing i n f l u e n c e your d e c i s i o n on where to f l o a t the bond i s s u e i n question? YES NO Did you c o n s u l t w i t h the Bank of Canada before i s s u i n g your U.S. d o l l a r denominated bonds? YES NO IF YOUR FIRM WENT "REGULARLY" TO THE U.S. MARKET DURING THE LAST FIVE YEARS, THAT I S , ISSUED TWO OR MORE U.S.-PAY BONDS SINCE 1968, SKIP QUESTION 23; OTHERWISE, SKIP QUESTION 24. 2 3 . (ANSWER THIS IF YOUR COMPANY HAS ISSUED NO BONDS/ONLY ONE ISSUE IN THE UNITED STATES SINCE 1968.) Why d i d you not approach the U.S. bond market more r e g u l a r l y ? a) There was no need f o r long-term debt c a p i t a l . YES NO b) Because i n t e r e s t r a t e s payable by your Company i n the U.S. were not s u f f i c i e n t l y lower ( r e l a t i v e to i n t e r e s t r a t e s on your bonds i n the Canadian market). YES NO c) Because f l o t a t i o n costs ( u n d e r w r i t e r s ' discounts and other expenses) are too high f o r your f i r m i n the U.S. market. YES NO Continued .... -265-23. Cont'd. d) Because conditions with respect to the following are, for your company, less attractive i n the United States than in Canada? i) maturity YES NO ii) c a l l features YES NO i i i ) sinking fund payments YES NO iv) other indentures? YES NO e) Because you considered the exchange rate risk to be too high? YES NO f) Because of other reasons? Please comment. 24. (ANSWER THIS QUESTION IF YOUR FIRM APPROACHED THE U.S. BOND MARKET MORE OR LESS REGULARLY DURING RECENT TIMES.) Why is the United States capital market attractive to your Firm? a) Call features are less restrictive. YES NO b) Underwriters' discounts and issuing expenses are lower. YES NO c) Interest rates are lower. YES NO d) Maturities are longer. YES NO e) Sinking fund requirements are less restrictive. YES NO f) Indentures are less restrictive in the U.S. than the Canadian market. in YES NO Please comment on these and any other reasons which made you approach the U.S. bond market regularly. -266-V. POSITION OF INTERVIEWEE Finally, may I ask some questions on your personal background? 25. a) F i r s t , what is your present position in this company? b) For how long have you been in this position? years c) For how long have you been associated with this firm? years d) Are you usually involved in decisions regarding the issuance of Canadian-pay bonds? YES NO e) Were you close to any U.S.-pay issue of your company? YES NO f) How old are you? years g) Would you briefly describe your experience and education to date? -26?-EXHIBIT A2-3 SUMMARY OF ANSWERS OBTAINED FROM CANADIAN CORPORATIONS Here we present a concise summary of the answers obtained from corporate managers with respect to the questions contained i n the prece ding questionnaire. As interviewees usually responded to a question only i f they could answer i t with YES, only the number of YES answers w i l l be given except where s p e c i f i c a l l y noted. Not a l l managers answered a l l questions, and sometimes multiple answers were given. In some instances we summarize the r e s u l t s of our dis ussions only because quantitative data would be rather meaningless. It must also be borne i n mind that i n many cases simple "YES" or "NO" answers were q u a l i f i e d through additional comments. Questions Answers 1. a) YESs 5 b) YESs 13 c) YESs 0 d) YESs 3 2. a) YESs 2 b) YESs 2 c) YESs 10 d) YESs 0 e) YES $ 8 f) YESs 9, but i t was generally agreed that "size i s no longer a major consideration." Why did you not Transaction costs increase; "U.S. investors s p l i t the issue? s t i c k to interest rate agreed upon even i f contract has not yet been signed;" arrange-ment of private placements i n the U.S. i s fast e r than of public offerings i n Canada. Why did you s p l i t the issue? P o l i t i c a l considerations; a v a i l a b i l i t y of funds i n Canada. - 2 6 8 -Question: Answer: Other reasons? A v a i l a b i l i t y of private placements and for -ward commitments: 5 (See also question 10) U.S. do l l a r income: 1 3. a) b) c) d) e) f) YES: YES: YES: YES: YES: YES: 5 o l 2 3 19 4. a) b) c) d) YES: YES: YES: YES: 5 3 7 7; nobody indicated that exchange r i s k considerations influenced decisions regard-ing the term to maturity; several managers said that, depending on market conditions, they attempt to trade o f f interest costs and maturities i n Canada. 5- parent comp.: U.S. subsidiary: Exports to U.S.: E a r l i e r U.S. Issues: Others« YES: 18 4 2 2 14 Connections through shareholders or direc-tors: 5 6. a) b) YES: YES: 11 14 NO: NO: 10 3 Don't know: a) b) YES: 11 NO: 8 YES: 5 NO: 10 (Our general impression was that the need for funds rather than inte r e s t rate expec-tations i s the main variable influencing the timing of a bond issue). 8. a) b) YES: 0 NO, THE SAME; 4 Don't know: 10 YES: 0 NO, THE SAME: 11 Don't know: 4 NO, LOWER: 4 NO, LOWER: 1 - 2 6 9 -Question: Answeri Why not U.S. underwriter? Why U.S. under-writer? YES: 7 NO: 9 Not required f o r private placement, U.S. investors know us: 5. "As a Canadian firm we use a Canadian underwriter, he has enough placement power;" "Our Canadian underwriter has a U.S. o f f i c e , and the costs are the same." Better placement power: 7; parent company decision: 2 . 10. Major reasons f o r choosing a public issue i n Canada are: 1. "That's where money i s a v a i l a b l e . " There i s a "continuous market" fo r such issues, even very large issues can be sold public-l y , and future financing i s easier. 2 . It provides desired public exposure and offers an additional investment opportu-n i t y to stockholders. 3. Trust deeds are less stringent, and the borrower i s not "at the mercy of a few lenders." Major reasons f o r choosing a private place-ment are: 1. They are regarded as o v e r a l l cheaper, and as f a s t e r and easier to arrange. 2 . Because of the small number of bond hol-ders, they provide more f l e x i b i l i t y , and t r u s t deeds are easier to change. 3. Public exposure i s avoided. Disadvantages of public offerings i n the U.S.: 1. SEC.requirements. 2 . Bank of Canada discourages them. 3. The Canadian market fo r public offerings provides s u f f i c i e n t funds. Advantages of private placements i n the U.S.: 1. A v a i l a b i l i t y of forward commitments and delayed d e l i v e r i e s . 2 . American lenders f e e l bound by verbal agreements. 3. Even very large amounts are usually r e a d i l y a v a i l a b l e . 4 . Small number of lenders. 5. Trust deeds are more stringent, but American lenders are more f l e x i b l e and t r u s t deeds are easier to change. -270-Question: 11. a) b) Answers YES: 13 NO, THE SAME: Don't know: 4 YES: 10 NO, The SAME: Don't know: 1 1 NO, LOWER: 1 6 NO, LOWER: 0 12. Rough estimate? Costs higher than i n Canada? YES: 0 NO: 21 YES: 0 NO: 21 (But one firm indicated that one o f f i c e r spent three months f u l l time on the preparation of the issue. Another manager said that three s t a f f mem-bers spent almost a year on analyzing the two c a p i t a l markets). YES i NO: 13. I f YES: a) b) YES: 3 NOs 18 YES: 2 NO: 1 YES: 1 NO: 2 (Those who answered "YES" were managers i n the banking and finance industry. Usu-a l l y they also borrow U.S. doll a r s on a short-term basis. In gerneral, managers pointed out that they f i n d i t d i f f i c u l t to forecast interest rates, and that forecasting ex-change rates i s almost impossible.) 14. a) b) Funds converted into Can. $$ a) b) c) YES: YES: 20 1 NO: 3 (Total amount: 3) YES: 18 YES: 12 the funds: YES: 0 YES: 8 (Total amount: of the funds: 4) (Sometimes funds not needed immediately are invested i n short-term U.S. d o l l a r papers. In one case, the tr u s t deed required t h i s e x p l i c i t l y . ) 9 ; only part of 4; only f o r part 15. Evaluation of exchange r i s k : a) b) Information used: YES: 8 NO: 13 YES: 1 YES: 6 Canadian and U.S. house economists, rates. banks, underwriters, i n -past history of exchange -271-Questiont Answers Did exchange r i s k influence sinking funds? YESs NO: 16. YES: 10 NO: 17. a) b) c) YES: hand) YES: month YES: 9 (approximately 1 to 10 days before-4 (length of contract i s between 1 and four years). 10 (Two companies obtain U.S. dollars by contract to cover part of t h e i r U.S. debt obligations). 18. If YES: a) b) YES: 3 YES: 1 YES: 2 forward). NO: 18 (approximately 15 days beforehand), (approximately 6 to 16 months 19. If YES: a) b) YES: 8 NO: 6 Don't know: 2 A l l associated a higher r i s k with a large balloon payment, but only two managers were s p e c i f i c about the expected r i s k by indica-t i n g that sinking fund payments tend to spread exchange r i s k ing out that a large depressing effect on YES: 2 (5 others f i e d YES: "It situation.") over time and by point-payment may have a the exchange market, answered with a q u a l i -depends on the c a p i t a l market 20. YES: 5 NO: l6 (Several firms have looked at European c a p i t a l markets, but interest rates turned out to be higher there. Also, the perceived exchange r i s k i s considerably higher.) 20a. YES: 0 NO: 14 (Because lenders i n s i s t on f i n a n c i a l non-c a l l clauses, other c a l l features are of l i t t l e i n t e r e s t and determined by standard market practice.) -272-Question: Answers 21. I f NOs YES: 12 NO s 9 YES, not i n favours 5 NO, i n favours 4 (Widely held public corporations indicated that t h e i r stockholders are i n d i f f e r e n t between domestic and foreign borrowing, and no e f f e c t on stock prices i s expected. Managers of other firms said that major stockholders on the Board of Directors favoured domestic financing. In two cases, foreign parent companies preferred U.S. financing.) 22. Consultation with Bank of Canada? YESs 13 NO s 8 YESs 8 NO s 11 (Several companies that did not consult with the Bank of Canada indicated that, under the changed circumstances, they would now ask f o r the Bank's advice.) 23. a) b) c) d) e) f) i ) - i v ) YESs YESs YESs YESs YESs 5 8 1 0 1 Policy of the Bank of Canadas 3 "As a Canadian company we borrow i n Canada," changed p o l i t i c a l climates 3 Canadian bond market has improved, issues are easier to place nows 3 2k. a) b) c) d) e) f) Other reasonss YESs 0 YESs 0 YESs 5 YESs 0 YESs 0 YESs 2 (One manager said that more concerned about the p o s s i b i l changing indentures i f i t became He pointed out that, as a matter one Canadian i n s t i t u t i o n refuses in t r ust deeds.) Private placements, forward commi available s 3 Increase f i n a n c i a l f l e x i b i l i t y by i n contact with both the Canadian U.S. c a p i t a l markets 1 they were i t y of opportune. of p olicy, changes tments keeping and the - 2 7 3 -EXHIBIT A2-4 INTERVIEW QUESTIONNAIRE , UNDERWRITERS -274-Introduction to an Interview with a Canadian Underwriter of Canadian  Corporate Bonds Denominated in U.S. Dollars As you already know from your correspondence with Dr. Winiata, I am conducting a study on international bond financing by Canadian cor-porations. The purpose of this research is to gain a better understanding of the financial behavior of Canadian corporations, particularly with regard to foreign borrowing a c t i v i t i e s . It i s hoped that these efforts w i l l provide new insights into the determinants of international capital flows and the relationships between Canadian and international capital markets. As part of i t s cultural exchange program with Germany the Canadian Government, through the Canada Council, is supporting my studies and providing most of the financing for this research. Interviews w i l l be conducted with several Canadian investment dealers who have underwritten U.S. dollar bonds issued by Canadian cor-porations. The underwriter's advice seems to play an important role in a firm's decisions as to where to float a bond and as to what the character-i s t i c s of the issue should be. Any information provided w i l l be used in such a way so as not to reveal information about any particular company. The analyses w i l l be in terms of Canadian corporate behavior in general. The sequence of the questions to follow attempts to roughly trace i n chronological order the considerations and decisions involved from planing a new issue to placing the bonds. You may give multiple answers to these questions whenever you think that this better highlights the factors that are relevant with respect to new U.S.-pay bond issues of Canadian corporations. K. Stroetmann, Ph.D. Candidate, Faculty of Commerce and Business Administration, University of British Columbia, Vancouver, Canada. -275-I . WHY DO CANADIAN CORPORATIONS APPROACH THE UNITED STATES BOND MARKET? In the main t h i s group of questions w i l l explore the reasons behind the d e c i s i o n by Canadian Corporations to f l o a t bonds denominated i n United States d o l l a r s (and the underwriter's i n f l u e n c e on these d e c i s i o n s ) . 1 . When one of the fi r m s w i t h which you have u n d e r w r i t i n g r e l a t i o n s h i p s approaches the United States bond market, does i t u s u a l l y go because you have recommended t h i s or i s i t u s u a l l y the f i r m ' s management that f i r s t suggests that an i s s u e be placed i n the United States? a) U s u a l l y we advise the f i r m to approach the U.S. market. YES NO b) U s u a l l y the f i r m ' s management f i r s t suggests the U.S. market. Please comment. YES NO 2. When, or under what circumstances do you recommend to your c l i e n t s to p l a c e a U.S.-pay i s s u e i n the Un i t e d States? Please comment b r i e f l y as subsequent questions w i l l enquire i n t o s p e c i f i c aspects. 3. Does a Canadian company have to meet c e r t a i n c r i t e r i a b efore you would, recommend the U.S. market f o r a bond iss u e ? Or, i n your o p i n i o n , are the U.S. and Canadian markets e q u a l l y a c c e s s i b l e ? Please comment. -276-4. Quite d i f f e r e n t reasons are o f t e n mentioned as to why Canadian firms approach the U.S. bond market. In your 1 o p i n i o n , what are the main reasons why Canadian companies take i n t o c o n s i d e r a t i o n the f l o a t a t i o n of an i s s u e denominated i n U.S. d o l l a r s ? a) Because the co r p o r a t i o n ' s management th i n k s the Canadian market would not be w i l l i n g to absorb another i s s u e of t h e i r f i r m because of e a r l i e r i s s u e s by the same company? YES NO b) Because the Canadian market i s depressed and they t h i n k i t would not absorb a s t r a i g h t bond i s s u e without an e q u i t y sweetener? YES NO c) Because t h e i r underwriter advised them to consider the U.S. market? YES NO d) Because U.S. i n v e s t o r s approached the f i r m w i t h the i n t e n t i o n of le n d i n g money to the firm? YES NO e) Because of e a r l i e r s u c c e s s f u l U.S.-pay i s s u e s by the same company? YES NO f ) Because the f i r m ' s management considers the s i z e of the bond i s s u e as too l a r g e f o r the Canadian market? YES NO I f YES: Why do they not choose to f l o a t p a r t of the i s s u e i n Canada and p a r t of i t i n the United States? Indeed, some firm s f l o a t p a r t of an i s s u e i n Canada and p a r t of i t i n the U n i t e d S t a t e s . In your o p i n i o n , what determines the s i z e of the U.S.-pay i s s u e r e l a t i v e to the s i z e of the Canadian-pay issue? - 2 7 7 -Or do you think there are other reasons why Canadian corporations consider flotation of a U.S.-pay issue? Typically, a firm would like to be free to specify the characteristics of a bond issue (including maturity, c a l l features, and so on) according to it s individual needs. Often, however, these characteristics are influenced by capital market conditions and lender behavior. When Canadian firms decide in favour of a U.S. dollar issue, do you think they are particularly attracted by a) Longer terms to maturing available i n the U.S. market compared to maturities in the Canadian market? YES b) Less restrictive c a l l features? YES c) More suitable sinking fund requirements? YES d) Less restrictive indentures? YES e) Lower underwriters' discounts and issuing expenses? YES f) Lower interest rates? YES Speaking about the characteristics of U.S. dollar issues, what usually determines the maturity of these bonds: a) The longest maturities available i n the U.S. market? YES b) The lenders' preferences with regard to maturity? YES c) The estimated l i f e time of the investment project to be financed through that issue? YES d) The firm's preferences with regard to maturity? YES Pleas e commen t. -278-e) Do you t h i n k t h a t the exchange r a t e r i s k i n f l u e n c e s your c l i e n t s ' choice of m a t u r i t i e s , that i s , do they p r e f e r longer o r s h o r t e r m a t u r i t i e s than they would otherwise because of p o s s i b l e exchange r a t e changes? YES NO Do they ask you f o r advise i n t h i s respect? YES NO Please comment. Are m a t u r i t i e s on Canadian-pay bonds u s u a l l y determined by the same f a c t o r s , or are there s i g n i f i c a n t d i f f e r e n c e s between the two markets w i t h respect to m a t u r i t i e s a v a i l a b l e to Canadian corporations? Please comment. 0 - 2 7 9 -I I . INTEREST AND TRANSACTIONS COSTS We should expect a f i r m to t r y to minimize i t s cost of c a p i t a l (percentage-wise). The y i e l d r e q u i r e d by i n v e s t o r s w i l l be the major f a c t o r determining the cost of debt c a p i t a l to a company. However, under-w r i t e r s ' compensations and other t r a n s a c t i o n costs may a l s o be c o n s i d e r a b l e i n f l u e n c e on the cost of c a p i t a l . The next group of questions w i l l e x plore some of the d i f f e r e n c e s between Canadian and U.S. bond markets w i t h respect to i n t e r e s t and other c o s t s . A) I n t e r e s t Costs 7. I n t e r e s t r a t e d i f f e r e n t i a l s observable between the Canadian and the U.S. c a p i t a l markets do not t r u l y r e f l e c t the a c t u a l d i f f e r e n c e i n i n t e r e s t costs as f a r as borrowing by a Canadian c o r p o r a t i o n i s concerned. In many cases, American i n v e s t o r s seem to demand a h i g h e r y i e l d on Canadian bonds de-nominated i n U.S. d o l l a r s than on American bonds of comparable q u a l i t y . Has t h i s been your experience? YES NO Do you t h i n k i t makes any d i f f e r e n c e i n i n t e r e s t costs whether Canadian c o r p o r a t i o n s approach the U.S. bond market r e g u l a r l y or not? YES NO Please comment. 8. F o r e c a s t i n g c a p i t a l market c o n d i t i o n s i s a very d i f f i c u l t task. In your experience, when Canadian firms choose the exact date of i s s u e f o r U.S. d o l l a r bonds, i s t h e i r t i m i n g u s u a l l y i n f l u e n c e d ( i n the sense of advancing or post-poning the date of i s s u e by a month or more), by a) T h e i r expectations w i t h regard to p o t e n t i a l changes i n long-term i n t e r e s t r a t e s i n the U.S.? YES NO b) By t h e i r e xpectations w i t h regard to s h o r t -term i n t e r e s t rates at home? YES NO -280-Do you usually make particular recommendations i n this respect? YES Please comment. B) Underwriters' Discounts Almost no evidence exists on underwriting compensations payed by Canadian firms issuing U.S.-pay bonds. In your opinion, are underwriter's discounts on issues  of Canadian firms usually higher in the U.S. than in the Canadian market? a) On public offerings? (higher in the U.S. market)YES b) On private placements? (higher in the U.S. market) YES If, i n your opinion, there are differences i n underwriting costs, why do you think these differences exist? Please comment. For foreign-pay issues, underwriters sometimes demand reimbursement for additional expenses incurred besides the usual underwriters' discount. Do you sometimes charge your clients for additional expenses peculiar to U.S.-pay issues? YES If YES: Could you itemize these costs? -281-10. Sometimes U.S.-pay i s s u e s by Canadian firms are underwritten by New York investment bankers. Why, i n your o p i n i o n , do some Canadian companies p r e f e r U.S. underwriters? Please comment. 11. Often fi r m s can choose between a p u b l i c o f f e r i n g and a p r i v a t e placement when i s s u i n g new bonds. What u s u a l l y determines t h e i r choice between a p u b l i c o f f e r i n g and a p r i v a t e placement when i s s u i n g bonds i n Canada? Please comment. In your o p i n i o n , what are the advantages and/or . disadvantages to Canadian cor p o r a t i o n s of p u b l i c o f f e r i n g s i n the U.S. bond market compared to p u b l i c bond o f f e r i n g s i n the Canadian bond market? C) Other T r a n s a c t i o n Costs 12. Apart from u n d e r w r i t e r s ' d i s c o u n t s , do you t h i n k t h a t other f l o t a t i o n costs l i k e p r i n t i n g of bonds and prospectuses, t r u s t e e ' s f e e s , l e g a l f e e s , and so on are higher f o r a U.S.-pay i s s u e than f o r a comparable i s s u e denominated i n Canadian d o l l a r s when the i s s u e i s a) A p u b l i c o f f e r i n g (higher i n the U.S.) YES NO, the same NO, lower b) A p r i v a t e placement? (higher i n the U.S.) YES NO, the same NO, lower -282-I I I . EXCHANGE RATE RISK When i s s u i n g a bond denominated i n a f o r e i g n currency, a f i r m faces two p e c u l i a r problems. ( 1 ) The proceeds of the i s s u e have to be converted i n t o Canadian d o l l a r s , and the exchange r a t e p r e v a i l i n g at the time of conversion determines the amount of Canadian d o l l a r s a c t u a l l y obtained. (2) Changes i n exchange ra t e s during the bonds' l i f e t i m e may change the a c t u a l costs of f o r e i g n borrowing c o n s i d e r a b l y . I would l i k e to explore these two problem areas. 13. The tim i n g of a bond i s s u e i s u s u a l l y a f f e c t e d by a host of f a c t o r s . When Canadian cor p o r a t i o n s choose the exact date of i s s u e f o r t h e i r U.S.-pay bonds, do you t h i n k that t h e i r t i m i n g i s i n -fluenced ( i n the sense of advancing or postponing the date of i s s u e by a month or more) by expectations w i t h regard to p o t e n t i a l exchange r a t e changes? YES NO I f YES: a) Are these u s u a l l y expectations w i t h regard t o short-term f l u c t u a t i o n s around the exchange r a t e l e v e l p r e v a i l i n g during t h a t time period? YES NO b) Or are these u s u a l l y expectations w i t h regard to changes i n the p a r i t y of the Canadian d o l l a r (during times of a f i x e d exchange r a t e ) or long-term changes i n the exchange r a t e l e v e l p r e v a i l i n g b e f o r e t h a t p a r t i c u l a r time p e r i o d (during times of a f l o a t i n g exchange r a t e ) ? YES NO 14., Do your c l i e n t s sometimes ask you f o r an e v a l u a t i o n of the exchange r a t e r i s k i n v o l v e d i n f o r e i g n borrowing? YES NO I f YES: Fo r e c a s t i n g f u t u r e exchange ra t e s i s very d i f f i c u l t and few people attempt to make long-range f o r e c a s t s . Do you attempt a q u a n t i t a t i v e e v a l u a t i o n of the exchange r a t e r i s k by a) F o r e c a s t i n g the f u t u r e trend of the exchange rate? YES NO b) Or, by e s t i m a t i n g f u t u r e exchange r a t e s and c a l c u l a t i n g whether they would e l i m i n a t e any cost advantage d e r i v a b l e from lower i n t e r e s t r a t e s abroad? YES NO -283-What i n f o r m a t i o n do you use when attempting t o evaluate the exchange r a t e r i s k or to f o r e c a s t exchange rate s ? Please comment. 15. The l a r g e r the s i z e of a U.S. d o l l a r i s s u e the g r e a t e r i s the f i r m ' s exposure to exchange r a t e r i s k . Is the s i z e of U.S.-pay bond i s s u e s u s u a l l y determined by the f i n a n c i a l needs of the i s s u i n g c o r p o r a t i o n , or i s i t your experience that the exchange r a t e r i s k i n f l u e n c e s the s i z e ? Do your c l i e n t s ask you f o r s p e c i f i c recommendations i n t h i s respect? Please comment. 16. In your o p i n i o n , do Canadian cor p o r a t i o n s e x h i b i t a d i s t i n c t p r e f e r e n c e f o r s i n k i n g fund bonds r a t h e r than f o r bonds where the p r i n c i p a l becomes due only at m a t u r i t y because of the exchange r a t e r i s k ? YES NO 17. What u s u a l l y determines the c a l l f e a t u r e s of U.S.-pay bonds? Please comment. 18. Have you ever recommended to a Canadian cor-p o r a t i o n that i t d i v e r s i f y the exchange r a t e r i s k by denominating a new bond i s s u e i n a currency other than the Canadian or U.S.? YES NO Please comment. -284-IV. MISCELLANEOUS QUESTIONS F i n a l l y , I have two questions d e a l i n g w i t h the Bank of Canada' i n f l u e n c e on f o r e i g n borrowing and w i t h American buyers of Canadian U.S pay bonds. 19. Now and then, the Bank of Canada announces that i t i s i n f a v o r or against f o r e i g n borrowing by Canadian governments and c o r p o r a t i o n s . Judging from your experience, does the Bank of Canada's p o s i t i o n on f o r e i g n borrowing u s u a l l y i n f l u e n c e corporate d e c i s i o n s on whether t o f l o a t U.S.-pay bonds? YES Do c o r p o r a t i o n s or do you sometimes consult w i t h the Bank of Canada before i s s u i n g U.S. d o l l a r denominated bonds? YES 20. Judging from your experience i n p l a c i n g Canadian-pay and U.S.-pay bonds i s s u e d by Canadian c o r p o r a t i o n s , why do American i n v e s t o r s e x h i b i t such a d i s t i n c t preference f o r U.S.-pay bonds i n s p i t e of the f a c t that u s u a l l y the y i e l d on Canadian-pay bonds i s s u e d by the same corporations i s s i g n i f i c a n t l y higher? Please comment. -285-V. POSITION OF INTERVIEWEEE Finally, may I ask some questions on your personal background? 2 1 . a) F i r s t , what is your present position in this firm? b) For how long have you been in this position years. c) For how long have you been associated with this firm? years. d) Are you usually involved i n the underwriting of Canadian-pay corporate bonds? YES NO e) Are you sometimes involved in the under-writing of U.S.-pay issues of Canadian governments? YES NO f) How old are you? years g) Would you b r i e f l y describe your experience and education to date? -286-EXHIBIT A 2 - 5 INTERVIEW QUESTIONNAIRE, CANADIAN LIFE INSURANCE CORPORATIONS -287-I n t r o d u c t i o n to an Interview w i t h an Investment O f f i c e r  of a Canadian L i f e Insurance Company As you already know from your correspondence w i t h Dr. W i n i a t a , I am conducting a study on i n t e r n a t i o n a l bond f i n a n c i n g by Canadian c o r p o r a t i o n s . The purpose of t h i s research i s to gain a b e t t e r understanding of the f i n a n c i a l behavior of Canadian c o r p o r a t i o n s , p a r t i c u l a r l y w i t h regard to f o r e i g n borrow-in g a c t i v i t i e s , and of the f u n c t i o n i n g of the Canadian c a p i t a l market. I t i s hoped that these e f f o r t s w i l l a l s o provide new i n s i g h t s i n t o the determinants of i n t e r n a t i o n a l c a p i t a l flows and the r e l a t i o n s h i p s between Canadian and i n t e r n a t i o n a l c a p i t a l markets. As p a r t of i t s c u l t u r a l exchange program w i t h Germany the Canadian Government, through the Canada C o u n c i l , i s supporting my s t u d i e s and p r o v i d i n g most of the f i n a n c i n g of t h i s research. There f o l l o w f i f t e e n questions e x p l o r i n g your f i r m ' s investment p o l i c y w i t h regard to corporate bond i s s u e s , the i n f l u e n c e lender behavior has on the c h a r a c t e r i s t i c s of such bond i s s u e s , and the a l l e g e d r e l a t i o n s h i p that e x i s t s between Canadian lender behavior and corporate borrowing i n the U n i t e d S t a t e s . These questions are only intended to guide our d i s c u s s i o n . Please f e e l f r e e to concentrate on any other problem area which you t h i n k w i l l l e a d to a b e t t e r understanding and i l l u m i n a t i o n of your f i r m ' s investment behavior and the Can-adian c a p i t a l market's f u n c t i o n i n g i n general. K a r l Stroetmann Ph.D. Candidate F a c u l t y of Commerce and Business A d m i n i s t r a t i o n U n i v e r s i t y of B r i t i s h Columbia Vancouver, Canada -288-I . PORTFOLIO OBJECTIVES AND CONSTRAINTS In t h i s f i r s t group of q u e s t i o n s , I would l i k e to explore the o b j e c t i v e s and b a s i c p r i n c i p l e s behind your company's investment p o l i c y w i t h regard to corporate bond i s s u e s . A) Investment O b j e c t i v e s 1. To begin w i t h , would you b r i e f l y d e s c r i b e your Company's o b j e c t i v e s w i t h respect to your investments i n corporate bonds? B) Government I n f l u e n c e s : Legal and P u b l i c P o l i c y C o n s t r a i n t s 2. With respect to the s i z e and q u a l i t y of i s s u i n g f i r m , the v a r i e t y of c o r -porate bonds a v a i l a b l e f o r investment i n the Canadian market i s q u i t e impres-s i v e . I understand that there are l e g a l requirements a f i r m has to meet before you are allowed to i n v e s t i n i t s bonds. Could you please s h o r t l y o u t l i n e what the main l e g a l p r o v i s i o n s are which r e s t r i c t your investments i n corporate bonds? Besides those l e g a l requirements, are there other industry-wide (or by your company self-imposed) r e s t r i c t i o n s on investments i n corporate bonds? -289-Could you (and would you be prepared to) i n v e s t i n bonds i s s u e d by a newly e s t a b l i s h e d company l i k e a new p i p e l i n e company or a new pulp m i l l i f these bonds were guaranteed by a w e l l - e s t a -b l i s h e d parent c o r p o r a t i o n ? YES NO Would i t make any d i f f e r e n c e whether the parent i s a Canadian, an American, or a f i r m of another n a t i o n a l i t y ? YES NO Please comment: 3. Now and then, the Bank of Canada pursues a very r e s t r i c t i v e monetary p o l i c y by l i m i t i n g the c r e d i t f a c i l i t i e s a v a i l a b l e to c h a r t e r e d banks and other f i n a n c i a l i n s t i t u t i o n s and thereby reducing the c r e d i t a v a i l a b l e to Canadian c o r p o r a t i o n s . Does the Bank of Canada's p o l i c y i n f l u e n c e your investment behavior i n the sense t h a t you do not buy corporate bonds which, given d i f f e r e n t c a p i t a l market c o n d i t i o n s , you would have i n v e s t e d i n ? YES NO Please comment; During those times, do you i n v e s t incoming funds i n short-term paper? Please comment: YES NO - 2 9 0 -C) Your Company's Investment P o l i c i e s and C o n s t r a i n t s 4. Assuming you have a wide s e l e c t i o n of corporate bonds a v a i l a b l e f o r investment, which f a c t o r s determine your company's choice of bonds to be acquired? 5. What determines how much of a s i n g l e c o r p o r a t i o n s ' s t o t a l debt you are prepared to i n v e s t i n ? Do there e x i s t l i m i t s as to the percentage of a s i n g l e c o r p o r a t i o n ' s debt you may acquire? ygg NO Please comment: To what extent are you prepared to i n v e s t above your u s u a l l i m i t s i n a c e r t a i n c o r p o r a t i o n ' s bonds i f they o f f e r an i n t e r e s t r a t e above the market y i e l d on other bonds of comparable q u a l i t y ? YES NO Please comment: -291-6. What determines how much of a s i n g l e corporate bond i s s u e you would a c q u i r e f o r your f i r m ' s p r o t f o l i o ? Do you sometimes take up s m a l l e r or even l a r g e r bond i s s u e s i n t h e i r e n t i r e t y ? YES NO Please comment: 7. Some c o r p o r a t i o n s are a b l e to p l a n t h e i r needs f o r l o n g -term funds w e l l i n t o the f u t u r e and enter i n t o forward agreements w i t h i n s t i t u t i o n a l i n v e s t o r s . A l s o , c o r p o r a t i o n s sometimes i s s u e bonds i n s e v e r a l c l o s i n g s which may s t r e t c h over a p e r i o d from a few months to s e v e r a l years. Do you ever enter i n t o such forward agreements w i t h c o r p o r a t i o n s ? YES NO I f YES: a) What f a c t o r s determine how f a r i n advance you w i l l commit your funds? b) Would you i n d i c a t e what your commitment fee u s u a l l y i s ? C) With respect to number and time p e r i o d over which draw-downs occured, would you comment on your Company's experience w i t h i s s u e s i n which s e v e r a l c l o s i n g s were involved? - 2 9 2 -8. Extendable and prepayable bonds instruments i n the Canadian market, a t t r a c t i v e to you? Please comment: have become q u i t e common Are those bonds p a r t i c u l a r l y YES NO 9. Do you p r e f e r secured r a t h e r than unsecured corporate bonds f o r investment? YES NO Please comment: II. BORROWER PREFERENCES AND MARKET PRACTICES T y p i c a l l y , a c o r p o r a t i o n would l i k e to be f r e e to s p e c i f y the c h a r a c t e r i s -t i c s of a bond i s s u e ( i n c l u d i n g m a t u r i t y , c a l l f e a t u r e s , and so on) according to i t s i n d i v i d u a l needs. Often, however, these c h a r a c t e r i s t i c s are i n f l u e n c e d by c a p i t a l market c o n d i t i o n s and lender behavior. The next group of fou r questions i s intended to ga i n a b e t t e r understanding of the f a c t o r s i n f l u e n c i n g the c h a r a c t e r i s t i c s of a corporate bond i s s u e . 10. Judging from your e x p e r i e n c e , what u s u a l l y determines the term to m a t u r i t y of corporate bonds: a) The estimated l i f e time of the p r o j e c t to be fina n c e d by a p a r t i c u l a r i s s u e ? YES NO b) The i s s u i n g f i r m ' s preferences w i t h regard to ma t u r i t y ? YES NO c) The l e n d e r s ' preferences w i t h regard to ma t u r i t y ? YES NO P.T.O. -293-Question 10. cont'd. When you acq u i r e corporate bonds, does the term to m a t u r i t y i n f l u e n c e your investment d e c i s i o n ? YES NO Please comment: In your o p i n i o n , would Canadian c o r p o r a t i o n s be able to p l a c e bonds w i t h r a t h e r long terms to m a t u r i t y o f , say, 30 years i n the Canadian market i f they wanted to do so? YES NO Please comment: 11. Can a c o r p o r a t i o n u s u a l l y f r e e l y choose between s i n k i n g fund bonds and a bond i s s u e where the p r i n c i p a l i s to be repayed only at m a t u r i t y ? YES NO Does the e x i s t e n c e of a s i n k i n g fund p r o v i s i o n i n f l u e n c e your investment d e c i s i o n s ? YES NO Please comment: 12. Judging from your e x p e r i e n c e , are the c a l l f e a t u r e s of corporate bonds u s u a l l y : a) Open to n e g o t i a t i o n between the i s s u i n g c o r p o r a t i o n s and the lenders? YES NO b) Are they u s u a l l y determined by the lender? YES NO c) Or are they u s u a l l y determined by standard market p r a c t i c e ? YES NO Are there other f a c t o r s that determine the c a l l f e a t u r e s of corporate bonds? YES NO Please comment: 13. .In your o p i n i o n , are there i d e n t i f i a b l e f a c t o r s which determine a Canadian c o r p o r a t i o n ' s preferences w i t h respect to p u b l i c o f f e r i n g s versus p r i v a t e p l a c e -ment? Do you p r e f e r to a c q u i r e corporate bonds through the market or through p r i v a t e placements? Are there d i f f e r e n c e s i n net y i e l d r e a l i z a b l e by your firm? -295-III. LENDER BEHAVIOR AND CORPORATE BORROWING IN THE U.S. F i n a l l y , I would l i k e to concentrate on the r e l a t i o n s h i p t h a t , a c c o r d i n g to some people, e x i s t s between Canadian corporate borrowing i n the United States and the investment behavior of Canadian i n s t i t u t i o n a l i n v e s t o r s . 14. Again, j u d g i n g from your experience and knowledge of the Canadian c a p i t a l market, do you t h i n k that Canadian c o r p o r a t i o n s (or t h e i r u n d e r w r i t e r s ) approach the U.S. market only a f t e r they have convinced themselves that there i s no "market" i n Canada f o r the p a r t i c u l a r bond is s u e ? YES NO I f YES: Are you aware of an i n s t a n c e where you turned down a Canadian c o r p o r a t i o n which subsequently f l o a t e d a bond i s s u e i n the U.S.? YES NO 15. With regard to the a l l e g e d i n a b i l i t y of the Canadian market to provide s u f f i c i e n t long-term funds to c e r t a i n c o r p o r a t i o n s , i n your o p i n i o n , i s i t i n f a c t t r u e t h a t : a) there.are not s u f f i c i e n t funds a v a i l a b l e i n Canada to w e l l e s t a b l i s h e d and c r e d i t - w o r t h y c o r p o r a t i o n s ? YES NO b) Or, i n your o p i n i o n , i s i t tru e that Canadian c o r p o r a t i o n s are a t t r a c t e d to the U.S. market by more l i b e r a l c o n d i t i o n s w i t h r e s p e c t to m a t u r i t i e s , s i n k i n g fund requirements, and so on a v a i l a b l e there r a t h e r than being " f o r c e d " i n t o the U.S. market by a s c a r c i t y of funds a v a i l a b l e i n Canada? YES NO c) Or, do you t h i n k both of the f a c t o r s mentioned above cause Canadian f i r m s to approach the U.S. market? YES NO I f YES: Why would American i n v e s t o r s demand l e s s r e s t r i c t i v e c o n d i t i o n s f o r a Canadian corporate bond i s s u e than Canadian lenders? Please comment: IV. POSITION OF INTERVIEWEE Before f i n i s h i n g t h i s i n t e r v i e w , may I ask. some questions on your p e r s o n a l background? 15. a) F i r s t , what i s your present p o s i t i o n i n t h i s firm? b) For how long have you been i n t h i s p o s i t i o n ? y e a r s . c) For how long have you been a s s o c i a t e d w i t h t h i s f i r m ? y e a r s . d) Are you u s u a l l y i n v o l v e d i n d e c i s i o n s r e g a r d i n g investments i n bonds i n g e n e r a l , or do you s p e c i a l i z e i n investments i n bonds i s s u e d by a c e r t a i n group of borrowers ? Can you pl e a s e i d e n t i f y t h i s group? e) f) How o l d are you? _years. Would you b r i e f l y d e s c r i b e your experience and education to date? 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0100013/manifest

Comment

Related Items