UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Statistical investigation of the ocean charter market Proctor, Irving Leroy 1970

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

Item Metadata

Download

Media
831-UBC_1970_A4_5 P76.pdf [ 7.25MB ]
Metadata
JSON: 831-1.0102070.json
JSON-LD: 831-1.0102070-ld.json
RDF/XML (Pretty): 831-1.0102070-rdf.xml
RDF/JSON: 831-1.0102070-rdf.json
Turtle: 831-1.0102070-turtle.txt
N-Triples: 831-1.0102070-rdf-ntriples.txt
Original Record: 831-1.0102070-source.json
Full Text
831-1.0102070-fulltext.txt
Citation
831-1.0102070.ris

Full Text

A STATISTICAL INVESTIGATION OF THE OCEAN CHARTER MARKET by IRVING LEROY PROCTOR B. Comm., University of British Columbia, 1968 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF BUSINESS ADMINISTRATION in the Department of COMMERCE AND BUSINESS ADMINISTRATION We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA APRIL, 1970. In presenting th i s thes i s in pa r t i a l f u l f i lment of the requirements for an advanced degree at the Un ivers i ty of B r i t i s h Columbia, I agree that the L ibrary sha l l make i t f r ee l y ava i l ab le for reference and study. I fu r ther agree tha permission for extensive copying of th i s thes i s for scho lar ly purposes may be granted by the Head of my Department or by his representat ives. It is understood that copying or pub l i ca t ion of th i s thes i s fo r f i nanc ia l gain sha l l not be allowed without my wr i t ten permission. Department of Commerce and Business Administration. The Un ivers i ty of B r i t i s h Columbia Vancouver 8, Canada D a t e A p r i l , 30, 1970. ABSTRACT Most studies i n the area of ocean shipping are descriptive. Certain aspects of tramp shipping have been subjected to empirical analysis, but few authors have been concerned with an objective study of the behaviour of charter rates. The major purpose of this thesis i s to analyse the behaviour of tramp shipping rates over the years 1960 -1968, and to discover what impact the forces of supply and demand had on voyage and time charter rates during those years. To accomplish this objective, the thesis i s divided into two distinct parts: the f i r s t half of the text i s confined to identifying the various markets that exist i n the shipping industry today. Incorporated with this dis-cussion are pertinent facts and figures that exemplify the changing pattern of vessel ownership within the industry, as well as the impressive growth and diversification of the various facets of ocean shipping. The second h a l f of the text i s concerned with a s t a t i s t i c a l analysis of tramp charter rates, i e . , voyage and time charters. Monthly data were gathered on several variables of supply and demand i n the shipping industry. The r e l a t i o n s h i p s between these variables and charter rates were examined i n four d i s t i n c t categories: 1. between the various categories of rates, i e . , voyage, time and tanker rates. 2. the r e l a t i o n s h i p between l a i d up tonnage and charter rates. 3. the r e l a t i o n s h i p between charter rates and the various stages of a c t i v i t y i n the shipyards, i e . , ship ordering, ship launching and ship completions. 4. the r e l a t i o n s h i p between the demand for shipping space, as indicated by world sea trade, and charter r a t e s . A number of hypotheses concerning the economic behaviour of charter rates with respect to these variables were formulated and tested by means of a series of multiple regression models to determine whether these hypotheses could be accepted or rejected. I n i t i a l t e s t s produced what appeared to be some s i g n i f i c a n t r e s u l t s . However, these proved to have high autocorrelation i n the re s i d u a l s . Following more rigorous t e s t i n g to remove the autocorrelation, the relationships broke down and the hypotheses had to be rejected. ACKNOWLEDGEMENT The writer wishes to thank Dr. Trevor Heaver and Dr. William Waters for their helpful and encouraging comments during the preparation of this study. Particular thanks go to Mr. Henry Pold for his invaluable help with the computer programs used i n this thesis, and to Mrs. Vivian Cook for her excellent job of reproduction. TABLE OF CONTENTS CHAPTER PAGE I. INTRODUCTION 1 Purpose of the Study 1 Importance of the Study 1 Limitations and Scope of the Study 11 Introduction to the Text 15 II. IDENTIFICATION OF SHIPPING MARKETS 17 Introduction 17 Shipping Markets 17 1. The O i l Tanker Freight Market 18 2. The Dry Cargo Liner Freight Market 23 3. The Dry Cargo Bulk Freight Market 28 Commercial Arrangements i n the Dry Bulk Cargo Freight Market 35 1. Direct Ownership 37 2. Negotiated Fixtures 38 3. Open Market Fixtures 40 Summary 42 III. THE FORCES OF SUPPLY AND DEMAND IN THE SHIPPING MARKETS 44 i i CHAPTER PAGE III. (continued) Introduction 44 Economic Portrayal of the Short Term Supply and Demand for Shipping 46 Ela s t i c i t y of Demand 50 Shifts i n Demand 51 Forces of Demand 55 Forces of Supply 62 Forces of Supply Under the Continuation Rate. 65 Forces of Supply Under the Lay-up Rate 69 Relationship Between Expectations and Supply and Demand i n Shipping Markets 72 Summary 75 IV. EMPIRICAL INVESTIGATION OF CHARTER RATE LEVELS . 77 Introduction 77 The Use of Multiple Regression as a S t a t i s t i -cal Device 78 Variables Included i n Analysis 85 1. Relationship Between Rates 90 Voyage and Time Charter Rates 90 Voyage, Time and Tanker Rates 96 2. Relationship Between Laid Up Cargo Tonnage and Charter Rates 103 1 1 X CHAPTER PAGE IV. (continued) F i t t i n g a Logarithmic Curve 107 3. The Relationship Between Shipbuilding Activity and Charter Rates 113 Charter Rates and Orders for New vessels ... 114 Charter Rates and Vessels Launched 119 Charter Rates and Ship Completions 123 4. Relationship Between World Sea Trade and Charter Rates 127 Summary 131 V. EXAMINATION OF THE CHANGES IN CHARTER RATES WITH THE CHANGE IN SELECTED VARIABLES 134 Introduction 134 Use and Method of F i r s t Differences of Data. 135 Regression Results Using F i r s t Differences . 136 Further Testing of Data 142 VI. COMPARISON OF RESULTS WITH OTHER EMPIRICAL STUDIES ON THE SHIPPING INDUSTRY. . . . . 148 Introduction 148 Correlation Analysis as Performed by Leslie Jones 148 Regression Analysis as Performed by Thomas Bates 151 X V CHAPTER PAGE VI. (continued) Analysis of Tankership Rates as Performed by Z.S. Zannetos 153 Summary 154 VII. SUMMARY 156 Suggestions for Further Research 158 BIBLIOGRAPHY 162 APPENDIX A. Notes on Construction of Freight Indices 166 APPENDIX B. Durbin-Watson Test for Autocorrelation . 170 APPENDIX C. Notes on Sources of Statistics 175 APPENDIX D. Functional Relationships Tested i n Regression Analysis 178 V LIST OP TABLES TABLE PAGE I. Growth of World seaborne Trade .. 19 II. Shipments of Six Major Bulk Commodities 33 III. Movements of Six Major Bulk Commodities 57 IV. Correlation Matrix - 1960 - 1968 Tramp Shipping Variables 89 V. Results: Time Index^. Regressed on Voyage Index 97 VI. Results: Voyage Indext Regressed on Time Index 98 VII. Results: Voyage Index t Regressed on Tanker Index 100 VIII. Results: Time Index t Regressed on Tanker Index 101 IX. Results- Laid Up Cargo t Regressed on Voyage Index 105 X. Results: Laid Up Cargo Regressed on Time Index 106 XI. Results: Log Laid Up Cargo t Regressed on Log Voyage Index I l l XII. Results: Log Laid Up Cargo t Regressed on Log Time Index 112 XIII. Results: New Orders Bulk Carriers Regressed on voyage Index , 117 XIV. Results: New Orders Bulk Carriers t Regressed on Time Index 118 v i TABLE PAGE XV. Results: Voyage Index^. Regressed on Launched Cargo Vessels 121 XVT. Results: Time Index t Regressed on Launched Cargo Vessels • 122 XVII. Results: Time lndex f c Regressed on Completed Cargo Vessels 125 XVIII. Results: Voyage Indexfc Regressed on Completed Cargo Vessels 126 XIX. Results- Voyage Indexfc Regressed on world Sea Trade 129 XX. Results: Time Indext Regressed on World Sea Trade 130 XXI. Results: Voyage Index t Regressed on World Sea Trade 132 XXII. Time Index t Regressed on Voyage Index ... Time Index t Regressed on Voyage Index ... 137 XXIII. Laid Up Cargofc Regressed on Time index Laid Up Cargo t Regressed on Time index ... 138 XXIV. Voyage Index^. Regressed on Completed Cargo Vessels Voyage Index t Regressed on Completed Cargo Vessels 139 XXV. Time Index Regressed on World Sea Trade Time Index^ Regressed on World Sea Trade . 140 XXVI. Correlation Between Index of Freight Rates and Tonnage Launched 150 XXVII. Five and One Per Cent Sigificance Points for 5L d, and dy 172 XXVIII. Five and One Per Cent Significance Points for v i i LIST OF FIGURES FIGURE PAGE I. Short Term Demand and Supply of Tramp Tonnage 48 II. United Kingdom Chamber of Shipping Tramp Ship-ping Indices 93 III. Scatter Diagram of Laid Up Cargo Tonnage and Time Index 1960 - 1967 108 IV. Scatter Diagram of Laid Up Cargo Tonnage and Voyage Index 1960- 1967 109 V. Scatter Diagram of Laid Up Cargo Tonnage and Voyage Index (First Differences).1960 - 1967 .. 143 VI. Scatter Diagram of Laid Up Cargo Tonnage and Time Index (First Differences) 1960 - 1967 .... 144 CHAPTER I INTRODUCTION PURPOSE OF THE STUDY The main purpose of t h i s study i s to explore, analyze, and attempt to i s o l a t e some of the underlying causes for the fluctuations i n tramp shipping rates during the period 1960 -1968. To achieve t h i s goal, i t i s necessary to review the basic structure of the tramp shipping industry and then to test empirically hypotheses about the operation of the charter market. An attempt w i l l be made, using a multiple regression analysis, to determine the impact of supply and demand variables on charter rates. I f s i g n i f i c a n t r e l a t i o n -ships are found, i t may be possible to suggest p o l i c i e s or procedures appropriate for management responsible for vessel chartering. IMPORTANCE OF THE STUDY The most obvious feature about f r e i g h t rates i n the shipping industry i s t h e i r v i o l e n t c y c l i c a l nature. This i s e s p e c i a l l y true i n the tramp shipping market for reasons which w i l l be discussed l a t e r but i t i s also true to a lesser 2 degree i n a l l shipping markets. This fact has l e d to many controversial discussions among authors i n the shipping area as to the causes of these f l u c t u a t i o n s . Often they are at t r i b u t e d to the i n t e r n a t i o n a l nature of shipping and i t s consequent lack of government controls and regulations. Discussions concerning national f l e e t s , subsidies, cargo preference and rate discrimination are ubiquitous i n any l i t e r a t u r e written on the subject. However, to the knowledge of the writer, there has never been a unanimity of opinion except i n a very general way. S p e c i f i c s are avoided, and the subject i s usually dismissed as being of an academic rather than of a p r a c t i c a l nature. This i s not the case however. Apart from the i n s i g -n i f i c a n t amount c a r r i e d by commercial a i r c r a f t , i n t e r n a t i o n a l trade i s p r a c t i c a l l y a l l performed by ships. The sheer si z e of the world f l e e t i s intimidating: over 22,000 ships t o t a l -l i n g 275 m i l l i o n deadweight tons and on order there i s a massive investment of almost 100 m i l l i o n tons. 1 What i s determining, the size of the world f l e e t ? The answer i s trade 1 "Factors Influencing the Size of World F l e e t s " , Shipping World and Shipbuilder, November 1969, p. 1544. 3 and the expectations of trade to f i l l these ships. Dry cargo movements have been increasing at an annual rate of about 7 to 8 percent during the 1960's; o i l movements at a somewhat 2 hxgher rate of about 9 to 10 percent. To meet these rates of growth, the dry cargo f l e e t has been expanding at some 5 to 6 percent per year and the tanker f l e e t at about 8 per-3 cent. The point of a l l of these s t a t i s t i c s i s to impress the fa c t that shipping i s a big business. And, as i n any other business, the chief objective of the shipowner i s to make a reasonable p r o f i t and the chief objective of a shipper i s to arrange for h i s products to reach the market at the lowest possible cost. Therefore, nothing could be more p r a c t i c a l from the point of view of a shipowner or a shipper than a better understanding of the nature of ocean shipping rates and the factors that influence the l e v e l of shipping fr e i g h t s charged. For to operate within some defined objec-t i v e s , users and s e l l e r s of shipping services must be able to plan for present and future a c t i v i t i e s with some degree 2 I b i d . 3 I b i d . 4 of c e r t a i n t y . To do t h i s , some i n d i c a t i o n of the present and future behaviour of f r e i g h t rates would be b e n e f i c i a l . T h e o r e t i c a l l y , f r e i g h t rates i n the shipping markets, l i k e any other p r i c e s , are determined by the forces of supply and demand. Unfortunately, too l i t t l e i s known i n a s t a t i s -t i c a l sense about the exact r e l a t i o n s h i p between the supply and demand of shipping as a whole, or even of shipping of p a r t i c u l a r types of commodities on p a r t i c u l a r routes. But the d r a s t i c fluctuations of f r e i g h t rates suggest that the problem of adapting the supply of tonnage to the demand for shipping space i s one of singular d i f f i c u l t y . A close examination of the shipping industry suggests that t h i s may be attributed to several f a c t o r s . The f i r s t i s that, at l e a s t u n t i l recently, the period between the decision to order a ship and the delivery has been so long that the bloom, as i t were, has worn o f f the boom. A decision not to b u i l d , because of a depression, tends to have l i t t l e e f f e c t because i n any large-scale system there i s always a hidden extra capacity which can be brought into use i n an emergency, as the Suez situations of 1956 and again i n 1967 c l e a r l y demonstrated of shipping. 5 The second i s the great d i f f i c u l t y of estimating the d i r e c t i o n , volume and character of i n t e r n a t i o n a l trade flows consisting of large number of separate commodities moving on a p a r t i c u l a r route. I t i s possible, allowed a l i b e r a l use of assumptions, to predict the range of fluctuations i n i n t e r n a t i o n a l trade as a whole. I t i s also r e l a t i v e l y easy, given such a p r e d i c t i o n as the main assumption, to calculate a change i n the pattern of maritime trade i n a p a r t i c u l a r product, though the r e l i a b i l i t y of such estimates may be subject to question. Though in t e r n a t i o n a l trade, as a whole, i s c a r r i e d i n ships, t h i s i s becoming increasingly i r r e l e v a n t to the problem of balancing the a v a i l a b i l i t y of a p a r t i c u l a r type of ship for the p a r t i c u l a r trade for which they are b u i l t . ^ A t h i r d reason for the d i f f i c u l t y of adapting supply to demand i s that shipping has increasingly become a prestige instrument i n which ships are b u i l t not to carry cargoes but to f l y f l a g s . Finding them expensive as mere flagpoles, the ^ This i s exemplified p a r t i c u l a r l y i n the area of tankers which p e r i o d i c a l l y i n f r i n g e on tramp trades i n the carriage of g r a i n . 6 owners then contrive a v a r i e t y of a r t i f i c i a l arrangements to d i v e r t cargoes from f l e e t s that were b u i l t on the assump-t i o n that these cargoes would form part of the load factors which j u s t i f i e d the o r i g i n a l decision to b u i l d them. In such a s i t u a t i o n , oversupply i s endemic, and the magnificent mechanismi of the market tends to be blamed for a s i t u a t i o n for which i t does not carry a shred of r e s p o n s i b i l i t y . There i s a fourth reason why the snipping industry (and with i t the shipbuilding industry) has been unusually vulnerable to economic pressures. I t might be assumed that the greater the complexity of the environment within which i t had to operate, the greater would be the e f f o r t s these industries would make to assess at l e a s t those elements i n the s i t u a t i o n which are predictable or ascertainable. One might expect, for example, a keen, c r i t i c a l , and continuing analysis of the patterns of trade on p a r t i c u l a r routes, and of the pattern of the shipbuilding industry's order book — a unique expression of the c o l l e c t i v e judgment of the world's shipowners — to be undertaken by every large-scale u n i t i n the industry. From t h i s analysis, shippers and owners might attempt to assess the impact that t h e i r findings may have on 7 the i n t e r a c t i n g forces of supply and demand i n the industry. From the facts gleaned from such an analysis, i t may be possible to speculate on future a c t i v i t i e s i n the shipping markets as a whole or even i n s p e c i f i c areas or on s p e c i f i c routes. Unfortunately, there i s an element of paradox i n analysis i n the l a t t e r case, because i n industries such as shipping, a c t i v i t y fluctuates more or l e s s proportionately to economic a c t i v i t y as a whole. This f a c t alone p r a c t i c a l l y negates any e f f o r t s to make sensible predictions on the basis of the data gathered from any p a r t i c u l a r area of a c t i v i t y . This i s so because of the f a c t that a high proportion of the industry's earnings are obtained i n cross trades where pros-p e r i t y may bear l i t t l e or no r e l a t i o n to the l e v e l of economic a c t i v i t y i n the country of o r i g i n of any s p e c i f i c trade. The Canadian-Chinese gr a i n trade may well slump for reasons which no one could a n t i c i p a t e . Moreover, the d i r e c t import and export trades between Canada and second countries are subject to many influences outside Canada's economy. A drought or a b l i g h t i n some of the South American countries can a f f e c t the import trades as much as an earthquake i n I 8 Chile or a change of government i n Ceylon can a f f e c t the export trades. Though i t i s true that i n a world-wide trading system, many of these influences tend to cancel each other out and may, i n t o t a l , be predictable over a s u f f i c i e n t l y long period, i t i s not possible to predict which sector of shipping i s the next to be affected or what scope there w i l l be for ef f e c t i v e redeployment of the resources involved i n each case. When attempting to predict any future l e v e l of a c t i v i t y i n any market, a forecaster l i k e s "other things" to remain equal. They seldom do. As a consequence, attempts at pre d i c t i o n or forecasting are often stymied by the r e c a l -c i t r a n t sectors of the economy which tend to upset c a l c u l a -t i o n s . I t would be tempting to conclude from t h i s argument that the shipping industry should be excluded from "sensible" attempts to forecast any future a c t i v i t y or indeed to peer ahead i n an attempt to estimate which d i r e c t i o n shipping freights w i l l move or the magnitude of th e i r f l u c t u a t i o n s . However, the r i g h t conclusion should be the exact opposite of t h i s . Shippers and owners a l i k e are s t i l l faced with the 9 problem of making investment decisions i n either c a p i t a l goods, or a budget of resources for future periods to continue with t h e i r own operations. In such a s i t u a t i o n , e f f o r t s should be extended i n an attempt to broaden and improve the q u a l i t y of the "stream of relevant f a c t " . For any industry, and shipping i s no exception, must benefit from the enlarge-ment of understanding and the exchange of information. With these thoughts i n mind, and recognizing the grave dangers of the undertaking, the writer f e e l s that any attempt at analysing f r e i g h t rates i n the shipping market may help i n improving the a v a i l a b i l i t y of information i n t h i s import-ant f i e l d . I f one looks at the tramp shipping market pragmatically, there are c e r t a i n relationships that e x i s t i n t u i t i v e l y by the i n t e r a c t i o n of supply and demand v a r i a b l e s . Among these are the following: (a) The volume of l a i d up tonnage should vary inverse-l y with the l e v e l of f r e i g h t r a t e s . (b) The volume of ships completed and ships ordered should bear a close r e l a t i o n s h i p to the l e v e l of rates; the former of course should correlate 10 more c l o s e l y i n a time lag s i t u a t i o n than would the l a t t e r due to the lead time required to b u i l d vessels. In times of r i s i n g rates, the increase i n vessels ordered should bear a close r e l a t i o n -ship to the increase i n rates, (c) Fluctuations i n the amount of world trade should r e l a t e c l o s e l y to the fluctuations i n f r e i g h t rates. One could continue with t h i s l i s t ad infin i t u m but the few examples given are s u f f i c i e n t to suggest many more r e l a t i o n -ships that e x i s t i n the shipping market. However, any l i t e r ' ature written on t h i s subject seems content to dismiss these relationships by suggesting that they e x i s t and making no e f f o r t to supply more precise information. This study then w i l l be very modest i n i t s i n t e n t i o n . I t i s proposed to t e s t empirically the rela t i o n s h i p s between d i f f e r e n t variables of supply and demand i n the shipping market based on the hypothesis: "Fluctuations i n shipping rates i n the tramp shipping industry can be explained by the use of multiple regression models incorporating d e t a i l s of the supply of and demand for shipping space." 11 From the r e s u l t s of t h i s analysis, i t i s hoped that some conclusions can be drawn as to t h e i r a p p l i c a b i l i t y to decision making i n the economic problems i n the shipping industry. LIMITATIONS AND SCOPE OF THE STUDY A major problem encountered i n writing a paper on ocean shipping rates i s i n deciding which information to include and which to leave out. Many factors require consid-eration; c e r t a i n information i s es s e n t i a l i f the fi n i s h e d product i s to be of any s i g n i f i c a n c e , but beyond that, one has to s t r i k e a del i c a t e balance i n deciding what addit i o n a l d e t a i l s to provide i n order to present a comprehensive picture of ocean shipping and s t i l l keep the paper within manageable proportions. I t i s obvious that tankers, cargo l i n e r s and tramps are i n e x t r i c a b l y interwoven when ocean transport i s consid-ered. One cannot be studied without considering the others, and yet each i s worthy of i n d i v i d u a l examination. In an e f f o r t to be somewhat s p e c i f i c , the decision was made i n planning t h i s paper to r e s t r i c t the area of study to that of tramp shipping, or more s p e c i f i c a l l y to a study of the 12 v o l a t i l i t y of shipping rates i n the open tramp market. 5 Unfortunately, i n t h i s , as well as i n any other area of ocean shipping, there i s a paucity of s t a t i s t i c s a v a i l a b l e . This f a c t contributes greatly to the l i m i t a t i o n s of the study. F i r s t , data on shipping rates for the years under study are available only for the tramp market i n the dry cargo trades, thus the irrevocable choice to l i m i t the study to the open tramp market. Secondly, the data must be procured from secondary rather than from primary sources. The v a l i d i t y of secondary data i s sometimes subject to question; however, the d i s p a r i t y between data i n various publications should not be great enough to influence any of the r e s u l t s obtained. Thir d l y , data are not available for one of the major contrib-utory factors to fl u c t u a t i n g tramp shipping rates: the length of haul. This variable naturally indicates the ton-mile performance of tramp vessels and i s a variable which obvious-l y a f f e c t s both the supply and demand of shipping. As the length of haul increases, not only does demand increase, but so also does output. 5 See Chapter II for an explanation of the "open tramp market." 13 Another source of concern was i n finding a proxy to use to measure the fluctuations i n fr e i g h t r a t e s . U.N.C.T. A.D. defines a f r e i g h t rate as the pri c e paid for the carriage of an item of merchandise such as a ton of i r o n ore or wheat, a crate of china or a piece of machinery.^ In addition, they point out that i n the tramp market alone there are several basis on which the concept of "freight rate" can be applied: the rate per ton for the tramp on voyage charter (which, however, i s affected by the r e l a t i o n between the size of the consignment and the deadweight of the vessel) and the rate per deadweight ton for the time chartered v e s s e l . Besides these differences, there are those a r i s i n g from differences r e l a t i n g to the party bearing the cost of loading and discharging and of cargo insurance. I t i s obvious that the time charter rates would never include these costs whereas the tramp voyage charter rate may or may not include them; i t may, for example, exclude a l l such costs or i t may include the costs of loading, but not the costs of discharg-ing; i t may not cover the costs of loading but may cover the costs of trimming, may cover discharging but exclude loading. United Nations, Conference on Trade and Development No. TD/B/C.4/38, December, 1968, p. 16. Hereafter referred to as U.N.C.T.A.D. 14 and so on. As can be seen from the above, the term f r e i g h t rate i s loosely applied to a number of d i f f e r e n t arrangements i n the tramp shipping market. As a consequence, the comparison of actual rates r u l i n g i n d i f f e r e n t markets, assuming that these could be obtained, i s d i f f i c u l t . For the purpose of t h i s study, i n t e r e s t i s not i n f r e i g h t rates per se, but c rather i n the v a r i a t i o n s i n f r e i g h t rates over time. Thus published f r e i g h t indices were chosen to act as a proxy for f r e i g h t rates. Today there i s a plethora of index numbers regularly calculated to depict f r e i g h t rates and t h e i r movements. For the purpose of t h i s study, two of the major indices w i l l be used; those calculated by the United Kingdom Chamber of Shipping and those calculated by the Norwegian Shipping 7 News. I t must be pointed out that index numbers provide a measure of changes i n tramp shipping f r e i g h t s ; i t does not measure d i r e c t l y the l e v e l of earnings, s t i l l l e s s does i t 7 See Appendix A for an explanation of the construction of these i n d i c e s . measure p r o f i t or l o s s . In addition, a c e r t a i n degree of caution i s needed i n i n t e r p r e t i n g the evidence on f r e i g h t movements as presented by f r e i g h t rate i n d i c e s . This i s not the place to enter into a s t a t i s t i c a l discussion of the merits of index numbers, but i t should be borne i n mind that any index number i s an average based on a pre-determined weighting pattern. For any p a r t i c u l a r year the weighting shown by the indices need not correspond at a l l c l o s e l y with the cargo movements recorded. This i s simply the i n -herent l i m i t a t i o n of the s t a t i s t i c a l device. INTRODUCTION TO THE TEXT Following t h i s introductory chapter i n which the importance of the problem has been outlined, Chapter II attempts to provide the reader with a short outline of the shipping industry i n general with s p e c i f i c reference to the tramp shipping industry. More s p e c i f i c a l l y , the Chapter shows that the industry i s not a single market, but rather a combination- of related markets with complex relationships between them. 16 Chapter I I I provides some i n s i g h t i n t o the forces of supply and demand which operate i n these markets and which have d i s t i n c t bearing on the l e v e l of fr e i g h t rates which w i l l p r e v a i l at any given time i n the tramp market. Chapter IV and V present an account of the empirical t e s t i n g of the s t a t i s t i c a l data which has been gathered and tenders an analysis of the r e s u l t s obtained through the t e s t i n g techniques employed. Several related studies i n the f i e l d of shipping rates have been attempted over the years and these are out-l i n e d b r i e f l y i n Chapter VI. The r e s u l t s reported i n these studies are examined i n the l i g h t of the r e s u l t s obtained through the empirical t e s t i n g attempted and reported i n Chapter IV of t h i s t h e s i s . Chapter VII presents a b r i e f summary of the study with some suggestions for further research i n the area. CHAPTER II IDENTIFICATION OF SHIPPING MARKETS INTRODUCTION A prerequisite to any analysis i s a thorough knowledge of the factors relevant to the functioning of the industry under study. In the shipping industry t h i s i s e s s e n t i a l because of the complex nature of i t s operations. As a prelude to the analysis of the s p e c i f i c area of tramp shipping rates, several preliminary parameters must be considered. F i r s t , the various shipping markets must be i d e n t i f i e d . Following t h i s , the forces which are at work i n these markets must be i d e n t i f i e d . Having defined these parameters, the procedure becomes one of examining the relationships between these i d e n t i f i e d market forces which i n turn w i l l hopefully lead to conclusions about the behaviour of tramp shipping and o f f e r support to the o r i g i n a l hypothesis as stated. SHIPPING MARKETS Broadly speaking, ocean shipping breaks down into three separate and f a i r l y d i s t i n c t types of markets: 18 1. the o i l tanker f r e i g h t market 2. the dry cargo l i n e r f r e i g h t market 3. the dry cargo bulk f r e i g h t market The delineation of these markets ends i n name only however, since i t should be emphasized that few, i f any, of the trades i n one market are immune to p o t e n t i a l competition from other markets. Thus an e f f e c t i v e l i n k i s established between the f r e i g h t rates e x i s t i n g i n one market with those i n another. 1. THE OIL TANKER FREIGHT MARKET A cursory examination of Table I indicates that the largest part of t o t a l world seaborne trade since 1960 has been i n the carriage of o i l . To meet t h i s demand, the size of the world tanker f l e e t has increased r a p i d l y , and i t s growth over the l a s t few decades has been l i t t l e short of phenomenal. In 1913 the gross tonnage of the world f l e e t of tankers was about 1.3 m i l l i o n tons. 1 By 1960 t h i s had 1 F.D. McCaffrey, "Winds of Change Sweeping T r a d i t i o n a l Methods", I n d u s t r i a l Canada, July, 1963, p. 141. TABLE I GROWTH OF WORLD SEABORNE TRADE 1960 - 1967 YEAR DRY CARGO mi l l i o n s of metric tons OIL m i l l i o n s of metric tons TOTAL m i l l i o n s of metric tons 1960 540 540 1,080 1961 570 580 1,150 1962 600 650 1,250 1963 640 710 1,350 1964 720 790 1,510 1965 770 870 1,640 1966 800 960 1,760 1967 810' 1' 1,050 ( 1 ) 1,860 (1) Pr o v i s i o n a l (1) Source: O.E.C.D. Maritime Transport 1968 increased to over 40 m i l l i o n tons, and by 1968 stood at 69.1 m i l l i o n tons or about 36 percent of the t o t a l world f l e e t . 3 Moreover, the increase i n the o v e r a l l size of the tanker f l e e t was matched by the increase i n the size of the tankers themselves. In 1947, the average tanker was one 4 between 12,000 and 16,000 deadweight tons. Since 1947, the size of the tankers has developed so r a p i d l y that today vessels with deadweight tons f a r i n excess of that are no longer exceptional. Lord Geddes, President of the United Kingdom Chamber of Shipping, expresses t h i s growth succinctly "Tanker companies are among those who have pointed the way with t h e i r giant c a r r i e r s . Last year, a 312,000 - tonner docked at Bantry Bay and ships of 200,000 tons w i l l soon become commonplace...there i s an obvious saving when ships of t h i s magnitude carry crude o i l on the long voyages around A f r i c a , replacing up to 10 voyages by smaller vessels"^ O.E.C.D. Maritime Transport, 1968, p. 105. 3 I b i d . 4 McCaffrey l o c . c i t . Lord Geddes, C.B.E.,D.L. "Shipowners must make a reasonable p r o f i t " , Shipping World and Shipbuilder, January 1969, p. 119. 21 I t i s the l a t t e r part of t h i s statement that sets the stage f o r the overlapping of t h i s tanker market in t o the f i e l d of tramp shipping. This occurs because of the nature of the ownership and operation of the world tanker f l e e t . B a s i c a l l y , the tanker f l e e t can be divided into 6 three groups: 1. tonnage owned and operated by o i l companies which amounts to approximately 35 to 40 percent i n terms of dwt capacity of the t o t a l tanker f l e e t . 2. vessels owned by private shipping firms, but hir e d to o i l companies for periods that cover a l l or most of the ship's useful l i f e . This tonnage i s approximately 40 to 45 percent of the world's tanker deadweight capacity. 3. tankers owned by private shipping firms and owner-operated i n the spot market as tramps for single voyages or for a few consecutive voyages or time chartered to o i l companies for short periods. 6 U.N.C.T.A.D., o p . c i t . Groups 2 and 3 as l i s t e d above contain p r a c t i c a l l y a l l of the large tankers which are now i n service since these large ships generally do not operate under free market 7 conditions. The obvious advantage i n operating costs i s with the large tanker, quick, highly automated and run by a small crew. As a consequence, i n times of receding demand i n the o i l trade, many smaller tankers - as l i s t e d under 3 above - can f i n d no o i l to carry. However, they can carry grain - not so conveniently perhaps, but at a loss l e s s than they would incur by laying up. Thus, i n these bad times, these tankers enter the grain trade - one of the staple trades of the tramps. Theor e t i c a l l y , t h i s action should r e s u l t i n an oversupply of vessels plying the grain trades, with resultant pressure placed on the l e v e l of tramp f r e i g h t rates i n both the very short run ( i n spot or voyage charter rates) and i n the s l i g h t l y longer run (the time charter market). Infringement of the o i l tanker f r e i g h t market on the dry cargo f r e i g h t market does not end here however. Recent U.N.C.T.A.D. o£. c i t . p. 21 23 technology has evolved a new type of combined ship classed as an "0B0" vessel, the i n i t i a l s standing for o r e / b u l k / o i l . "This type of ship f i r s t appeared i n 1964, when two 54,000 tonners were b u i l t . The p r i n c i p l e here i s to u t i l i z e the entire hold capacity for either s o l i d or l i q u i d cargoes, employing a hermetic method of sealing hatch covers when carrying l i q u i d s . These ships overcome the l i m i t a t i o n s on employment of ordinary tankers or o r e / o i l ships, since they are able to transport f u l l cargoes of crude o i l or any of a wide range of commodities."8 Given the v e r s a t i l i t y of t h i s type of vessel, i t can be seen that i n periods of slow demand i n the o i l trade, these vessels can e a s i l y be converted to compete i n the movement of bulk commodities, t r a d i t i o n a l l y the domain of tramps and bulk-carrying vessels. This overlap might also create some disturbing influences on freig h t s i n the dry cargo bulk market. 2. THE DRY CARGO LINER FREIGHT MARKET The dry cargo trade i s now handled by three types of ships: l i n e r s , tramps and bulk c a r r i e r s . For the moment, consideration w i l l be given only to l i n e r s and tramps. The Ibid p. 23 24 d i s t i n c t i o n between a l i n e r and a tramp may be a physical one, i n the sense that l i n e r s are often more elaborately equipped than tramps; but the most important d i s t i n c t i o n i s commercial and depends on the type of commercial arrangements Q under which the ship i s operated. One way to define dry cargo l i n e r shipping i s to say that i t i s c a r r i e d out by vessels which make regular s a i l i n g s out of d e f i n i t e ports and which play a fixed route. An equally important d i s t i n c t i o n of l i n e r shipping i s the f a c t that the cargoes c a r r i e d are generally of a heterogeneous nature which do not lend themselves to shipments by tramp vessels or by industrially-owned f l e e t s . In e f f e c t , l i n e r s consolidate the t r a f f i c of many shippers and charge fixed and f a i r l y stable rates for t h e i r service. Obviously, the cargo l i n e r s experience competition within the framework of l i n e r operations. However, i n most instances they are able to control t h i s competition. One of the most successful methods they employ i s the e s t a b l i s h -ment of f r e i g h t conferences. A fr e i g h t conference, i n 9 I b i d , p. 2. 25 simple terms, i s an association of shipowners who make an agreement between themselves to regulate the terms and conditions under which they w i l l operate i n competition with each other. This arrangement leads one to the conclusion that the dry cargo l i n e r market operates autonomously and i s not subject to competition from other markets or indeed, by the very nature of the cargo that l i n e r s carry, they themselves would not undertake to in f r i n g e on any other market. This i s not so. Under several conditions, cargo l i n e r s operate i n d i r e c t competition with the tramp market. The general nature of the operation of cargo l i n e r s and the high stowage factor of the general cargo that they carry implies that the ship's carrying capacity i s seldom utilized. However, once the ship i s committed to s a i l i n g , there i s very l i t t l e difference i n s a i l i n g with an empty ship or a f u l l ship. I t i s t h i s c h a r a c t e r i s t i c which has prompted cargo l i n e r s to carry a wide range of tramp-type commodities out of necessity rather than a desire to do so. In many cases, bulk materials are taken on to f i l l up the ship, e s p e c i a l l y the lower holds because these spaces are not 26 suitable for the more de l i c a t e classes of cargo, or, more l i k e l y there i s not enough other cargo to constitute a f u l l load."1"^ Due to the fact that rates on tramp cargoes are more f l e x i b l e than those applying to pure l i n e r cargoes, i n such instances the commodities referred to are c a r r i e d by the l i n e r s at "open rates".'''"'" F i s s e r suggests two other p o s s i b i l i t i e s of the l i n e r trade encroaching upon the domain of tramp shipping: " F i r s t of a l l , many l i n e r companies maintain a park of reserve vessels, which can be placed into service following the demand without being dependent upon unstable rate structures of the time charter market. These reserve f l e e t s consist mainly of high-class tonnage, which during periods of reduced i n t e n s i t y of the l i n e r trade w i l l t r y to f i n d employment on the free market, thus competing with tramp operations for cargo at a moment at which the demand for tramp tonnage i s already slackening on account of the d e c l i n i n g time charter demand....Moreover, l i n e r companies w i l l contract the transport of large scale bulk shipments, d i s t r i b u t e d over a longer period of time, i n order to obtain a long-term security of minimum employment, necessary for the maintenance of a regular service."- 1- 2 10 i b i d , p. 29. 11 When a rate i s "open" i t means that no rate i s s p e c i f i e d i n the rate schedule for the cargo item concerned and the rate i n question i s fixed f r e e l y between the l i n e r operator and the respective shipper. F.M. F i s s e r , Tramp Shipping, Schunemann, Bremen, 1957, pp. 2 9 - 3 0 . A l o g i c a l r e s u l t gleaned from the above i s that i n such trades tramp shipping has to keep i t s rates at rock bottom, i f , considering the slower type of transport, i t . . 13 wants to remain competitive. These few examples suggest that tramp ships are always open to competition from l i n e r companies. Most l i n e r companies, however, maintain that they are not de l i b e r a t e l y i n competi-t i o n with tramp shipping, although i t i s i n e v i t a b l e that t h e i r spheres of a c t i v i t y overlap to some e x t e n t . 1 4 Regard-l e s s of t h e i r intentions, l i n e r s do exert a d e f i n i t e pressure on tramp shipping rates i n such instances. On the other hand, and for quite another reason, the operations of l i n e r and tramp shipping i n t e r a c t with each other. Since l i n e r companies operate on fixed schedules and under c e r t a i n conditions o f f e r incentives to shippers for 1 3 I b i d . 14 H. Gripaxs, Tramp Shipping. London: Thomas Nelson and Sons, 1959, p. 116. 28 exclusive f r e i g h t i n g r i g h t s , there are periods of peak demand that the l i n e r s are unable to cope with. Tramp shipping becomes a necessary complement to l i n e r trades i n these cases which oblige t h i s trade to employ tramps on time charter, i n order to be able to meet the requirements expected to be f u l f i l l e d by t h i s trade i n some t r a f f i c areas.' 3. THE DRY CARGO BULK FREIGHT MARKET The character and nature of the dry cargo bulk f r e i g h t market have changed considerably i n the l a s t few decades. To understand the relationships f u l l y as they e x i s t i n today's market, a b r i e f resume of the changes that have taken place i n t h i s area would be of be n e f i t . T r a d i t i o n a l l y , the ships operating i n t h i s market were a l l tramp ships. In essence the tramp ship was prepared to carry any cargo between any ports at any time. A d e f i n i -t i o n of the vessels used and the nature of t h e i r operations i s s u ccinctly expressed by F i s s e r : 15 In order to control the competition of l i n e s which are not members, the conferences sometimes e s t a b l i s h a system of dual rates under which shippers of cargo who reserve t h e i r shipments exclusively to the ships of the members of the conference receive lower f r e i g h t rates than shippers who use shipping l i n e s which are not members of the conference. 1 6 F i s s e r , op_. c i t . p. 27. 29 "A tramp ship may be defined as a vessel, mostly of medium size and without any special purpose equip-ment, not engaged i n regular trade but following cargo wherever i t i s offered at the most favorable conditions, i . e . , not adhering to a f i x e d route, she w i l l touch only p r o f i t a b l e ports...."I? The ultimate i n precise attempts at d e f i n i t i o n was reached i n c e r t a i n B r i t i s h l e g i s l a t i o n which de t a i l e d c l o s e l y the s p e c i f i c a t i o n s which could and could not be admitted to the class of tramps. I t concluded: "...a tramp voyage i s one i n the course of which a l l the cargo i s c a r r i e d under charter party."18 I t i s t h i s phrase rather than the d e t a i l s which i s r e a l l y s i g n i f i c a n t for our purposes. This w i l l become cle a r when we discuss the d i f f e r e n t commercial arrangements under which the dry bulk cargo i s handled. The service performed by these vessels consisted of transporting bulky goods of low value, such as ore, coal, grain, etc., under conditions where speed was a much l e s s important consideration than cheapness of transport. F i s s e r , o p . c i t . p. 5. 1 8 B r i t i s h Shipping (Assistance) Act, 1935. 30 When the amounts of global t r a f f i c of bulk commodities was i n s i g n i f i c a n t , t h i s type of tramp vessel was e n t i r e l y adequate. However, o v e r a l l increases i n the annual shipment of these bulk commodities, and increased competition i n i n t e r n a t i o n a l markets for the sale of these commodities, prompted some new thinking on the part of shipowners and users of shipping services on the problem of transporting enormous quantities of t h i s r e l a t i v e l y low cost merchandise over great distances. A p a r t i a l solution became one of reducing the costs of sea f r e i g h t s and cargo handling charges. More s p e c i f i c a l l y : "Prior to 1939, and indeed for more than another decade, sea f r e i g h t s did not weigh heavily i n the ca l c u l a t i o n s of dry cargo charterers.... By and large, users did not c r i t i c i s e the types of vessels put before them by shipowners. They grumbled about high freig h t s i n s h o r t - l i v e d booms and enjoyed the low rates p r e v a i l i n g i n the far longer periods of recession which followed. But neither the charterers nor the shipowners made any e f f o r t to devise means by which fre i g h t s could be permanently cheapened, with no s a c r i f i c e of seaworthiness i n ships, and s t i l l leave t h e i r owners a profit."19 C.P.H. Cufley, F.I.C.S. "Some Implications of the Bulk C a r r i e r Era", Shipping World and Shipbuilder, June 16, 1966, p. 471. 31 As a consequence of t h i s new thinking, researchers i n the shipping industry began exp l o i t i n g the economies of increasing ship size and carrying capacity, and i n the early 1950's the age of the bulk c a r r i e r a r r i v e d . " U n t i l t h i s time the norm of tramp ships engaged i n l i f t i n g bulk cargoes was around 9,500 tons deadweight. This was the size of the ' L i b e r t i e s ' and the 'Empires' b u i l t during the Second World War. The f i r s t vessels to be dubbed as bulk c a r r i e r s were h a l f again as big as the ' L i b e r t i e s ' and i n a very few cases venturesome shipowners placed orders for 20,000 tonners."^ 0 Since t h i s early beginning, the increase i n the size of bulk c a r r i e r s has been l i t t l e l e s s than phenomenal. According to "Fairplay", at the middle of 1968, dry bulk c a r r i e r s of over 150,000 tons had been ordered or delivered. Admittedly, i t i s somewhat loose terminology to use ship capacity as the sole c r i t e r i o n for defining a bulk c a r r i e r . The term could be applied with equal j u s t i c e to any vessel capable of being r a p i d l y loaded or discharged when carrying homogeneous bulk cargoes - which would include many of the modern tramps. But for the purposes of t h i s study, a bulk 2 0 I b i d . 21 "World Ships on Order", Supplement to Fairplay  Shipping Journal, August 22, 1968, p. 40. 32 c a r r i e r w i l l be considered as a larger-than-ordinary ship which charterers h i r e i n the expectation of paying lower f r e i g h t s . Most important however, i s i n the f a c t that the advent of the bulk c a r r i e r i n the dry cargo bulk market has had s i g n i f i c a n t impact on the absolute amount of cargo which was t r a d i t i o n a l l y l i f t e d by the conventional tramp ship. An examination of Table II permits an assessment of the impressive growth of bulk c a r r i e r employment i n the main bulk trades. The tremendous upsurge i n the size and carrying capacity as exhibited by the bulk c a r r i e r has led some authors to suggest that the death k n e l l has been scunded for the conven-t i o n a l tramp ship as i t exists today. Bes believes that i t s 21 days are over. He i s supported by the action of the Court Line who, i n 1962, disposed of t h e i r general purpose tramp f l e e t to go i n for bulkers and tankers, although these are sometimes l e t on voyage charter i n the spot market, which 21 j . Bes, Bulk C a r r i e r s . London: Barker and Howard, 1965. 33 TABLE II SHIPMENTS OF SIX MAJOR BULK COMMODITIES YEAR TONNAGE SHIPPED million tons TRANSPORT PERFORMANCE 1,000 million ton-miles 1960 1961 1962 1963 1964 1965 1966 1967 TOTAL 233 244 251 274 315 336 349 361 BULK CARRIERS 65 83 104 131 170 196 218 248 OTHER VESSELS 168 161 147 143 145 140 131 113 TOTAL 767 853 875 980 1,178 1,298 1,396 1,501 BULK OTHER CARRIERS VESSELS 216 299 383 508 667 777 917 1,115 551 554 492 472 511 521 479 386 Source: O.E.C.D. Maritime Transport 1968. 34 22 brings them under our d e f i n i t i o n of a tramp. 23 On the other hand, other authors f e e l that bulk c a r r i e r s have already f a i r l y well caught up with the f a c i l i t i e s of the world's harbours, and t h i s i s undoubtedly the factor that w i l l provide an e f f e c t i v e brake to t h e i r development. I t i s c e r t a i n that the deadweight of bulk c a r r i e r s could be increased at the same pace as that of tankers, but t h i s would prove far i n excess of the o v e r a l l rate for the expansion of ports and f a c i l i t i e s . From t h i s , economic unbalance would quickly be induced by a r e s t r i c t i o n of the f i e l d s of operation. Thus, since dividends are of paramount importance to shipping companies, idealism must be tempered with p r a c t i c -a b i l i t y . I t i s a f a i r assumption that the medium sized tramp vessel w i l l f i t the b i l l for years to come. This assumption i s r e i t e r a t e d by Cufley who i s adamant i n h i s 22 Keith R. Studer, The Evaluation of Long Term Charter  Rates i n the Shipping Industry, U.B.C. Unpublished Term Paper, Commerce 544, 1969. . . . . . . . . "Those Handy Tramps of the Future", Shipping World  and Shipbuilder, November 18, 1965, p. 407. Also see S.J.M. Nolan, Bulk C a r r i e r s , Past, Present and Future, Mimeo, London, 1966. 35 24 defense of the general purpose tramp. COMMERCIAL ARRANGEMENTS IN THE DRY BULK CARGO FREIGHT MARKET Accompanying t h i s change i n the physical make-up of the dry bulk cargo f l e e t i s the change i n commercial arrange-ments under which dry bulk cargo now moves. To aid i n the understanding of the analysis which follows i n Chapter IV, i t i s important that the d i s t i n c t i o n between these contract-ual arrangements be understood. Today, the dry cargo bulk f r e i g h t market can be divided i n t o several sections with the following broad groups being i d e n t i f i e d ; 1. Direct Ownership 2. Negotiated f i x t u r e s * 3. Open Market f i x t u r e s U.N.C.T.A.D. presents a comprehensive discussion about the d i s t i n c t i o n between these commercial arrangements: 2 4 C.F.H. Cufley. The Ideal Tramp for the 1970's. London: Barker and Howard, 1966. * "Fixture" i s a standard term used to describe a chartering contract. 36 "The factor determining whether a f i x t u r e i s arranged through the open market or i s negotiated i s l a r g e l y the length of time which i t i s to cover. At one extreme, a f i x t u r e for a single voyage from A to B i s almost i n v a r i a b l y arranged i n the open market. At the other extreme, a charter for twenty years i s always a matter for negotiation between the prospective charterer and the shipowner. Between these two extremes, i t can be said that any f i x t u r e for l e s s than one year i s l i k e l y to be arranged through the market, while anything for a period of three years or more i s almost c e r t a i n l y arranged by negotiation. Fixtures of between one and three years may be arranged i n either way. Most f i x t u r e s , however, tendc.. to be either under one year or over three years, with r e l a t i v e l y l i t t l e i n between. I t i s therefore convenient i n exposition to tre a t open market f i x t u r e s as covering everything under one year and negotiated f i x t u r e s as covering anything over three years. Using t h i s discussion as a c r i t e r i o n for segmenting the dry cargo bulk market contractually, one thing i s apparent. Chapter I alluded to the paucity of s t a t i s t i c s available which i n turn l i m i t e d t h i s i n v e s t i g a t i o n of the tramp ship-ping market to two types of vessel chartering: voyage charters (on a t r i p or spot market basis) and time charters (consisting of f i x t u r e s not exceeding the period of one year). As a consequence, i n t e r e s t i n the bulk cargo market i s with the group defined above as "open market f i x t u r e s " . Nevertheless U.N.C.T.A.D. op_. c i t . p. 31. 37 i t i s important to consider the rela t i o n s h i p s between these 'sub-markets 1 markets within the t o t a l i t y of the dry bulk cargo f r e i g h t market. 1 . DIRECT OWNERSHIP In the dry cargo market, d i r e c t ownership of vessels i s most prominent i n trades such as meat and f r u i t where 2 ships are generally regarded as l i n e r s rather than tramps. However, i n c e r t a i n cases such as the ore and sugar trade, some companies f i n d that the extent of t h e i r trade i n these dry bulk commodities warrants d i r e c t ownership of vessels. I f a l l journeys by these vessels are for outward trade, then t h e o r e t i c a l l y a back haul of commodities for the inward journeys would be desirable as opposed to a return voyage i n b a l l a s t . This arrangement would r e s u l t i n two things. F i r s t i t would decrease the o v e r a l l cost of the transport of the commodities thereby enhancing the competitive p r i c i n g p o s i t i o n of the company's products. On the other hand, i t would decrease the amount of cargo U.N.C.T.A.D. p_£. c i t . p. 3. 38 available for -vessels operating i n the open market. Since any remuneration above the operating costs of the vessel would be advantageous, the rates charged under such arrange-ments could well be below the p r e v a i l i n g spot rates for s i m i l a r contracts by vessels operating i n the open market. The assumption i s that t h i s would create a downward pressure on open market rates under these conditions. Unfortunately, no authority can be found to substantiate t h i s assumption, and i t may be that speculation i n t h i s respect i s unfounded. Regardless, even i f no back hauls are contracted by these ships, they have i n d i r e c t l y reduced the amount of bulk cargo available for tramp l i f t i n g s and t h i s may have bearing on the p r e v a i l i n g rates i n the open market. 2. NEGOTIATED FIXTURES Negotiated f i x t u r e s i n the dry bulk cargo market flourished with the introduction of bulk c a r r i e r s and highly s p e c i a l i z e d vessels.: The advantage of these long term arrange ments to both shipowner and charterer i s t h e i r security and s t a b i l i t y — the shipowner knows that h i s ship w i l l be 39 earning for the period of the charter and the charterer knows that h i s shipping needs w i l l be met. This aspect i s e s p e c i a l l y c r u c i a l i n the case of highly s p e c i a l i z e d vessels since "... the r i s k s of operating i n the short term market are such that shipowners w i l l r a r e l y , i f ever, b u i l d such a vessel unless they have a contract or charter of s u f f i c i e n t length to enable them to amortize the c a p i t a l cost of; the v e s s e l . Such a vessel would then l i k e l y to enter the short term market only for the residue of i t s active l i f e beyond 27 the amortization period." The open market, or short term arrangements do not remain unaffected by these long term charters. The two are intimately related because i t i s obvious that i n many (except the highly specialized) trades, the a l t e r n a t i v e to both charterer and shipowner to a long term arrangement i s a series of short term arrangements. This decision i s c l o s e l y related to the long term casts and expectations of both shipowner and shipper. Therefore, the supply of ships a v a i l a b l e and competing i n the open market may depend to a great extent on the judgment of these people. Either 2 7 Ibid, p. 41. 40 decision w i l l d i r e c t l y a f f e c t the l e v e l of rates that w i l l p r e v a i l at any given time i n the open market. 3. OPEN MARKET FIXTURES The mechanics of the open market are far more complicated than those for negotiated long term arrangements involving only one shipowner and one charterer. The open market embraces the aggregate at any given time of tramp shipowners seeking employment of t h e i r vessels and shippers requiring the services of tramp ships for a l i m i t e d period. We have intimated that there i s great i n t e r a c t i o n between t h i s open market and the other i d e n t i f i e d markets of shipping. This statement can best be exemplified by observing the related operations of the open market. Cufley outlines these opera-28 tions admirably when he l i s t s the functions of tramp ships: 1. To provide shipping space for a l l commodities whose movements cannot be predicted with accuracy. 2. To transport marginal tonnage requirements i n respect of those commodities where the bulk of C.F.H. Cufley. "The Ideal Tramp..." o£. c i t . p. 10 41 the t r a f f i c i s l i f t e d by integrated or h i r e d f l e e t s . 3. To provide cargo l i n e r s with a reserve of space to deal with seasonal: and other f l u c t u a t i o n s . 4. To o f f e r lower rates than conference t a r i f f s for shipload l o t s . 5. To provide a pool of shipping available for emergencies. The importance of t h i s market i n the shipping industry cannot be overemphasized however. Even though i t would appear that the ships i n the open market perform what i s sometimes designated as the "skirmish work" of i n t e r n a t i o n a l 29 trade, the seemingly aimless wanderings of the vessels employed i n the open market are a c t u a l l y planned with great care and foresight. A cargo i s always desirable; but some cargoes are more desirable than others, t h e i r d e s i r a b i l i t y being dependent to a very large degree on one important feature — p r o f i t . For i n t h i s segment of shipping, as i n H.C. C a l v i n and E.G. Stuart. Merchant Shipping  Industry, New York: John Wiley and Sons Inc. 1925, p. 55. 42 any other segment, the chief objective of the shipowner i s to make a reasonable p r o f i t . The amount of p r o f i t to be made depends i n turn on the f r e i g h t rate which can be charged for the transport of the commodities. I t follows that the pr e v a i l i n g f r e i g h t rate i n any trade w i l l depend on the number of ships vying for that trade and the amount of cargo available for transport. In other words, the i n t e r -action of the demand for and the supply of shipping space. SUMMARY In t h i s chapter we have t r i e d to sketch b r i e f l y the underlying market structure of the shipping industry. I t i s apparent from the discussion that the industry i s not concerned with a single market but with several, or many, and t h i s must be kept i n mind when we attempt to explain the r i s e and f a l l i n p r e v a i l i n g market prices (rates). Nevertheless, i t should also be remembered that there are so many l i n k s and connections between these markets, that a clear picture becomes d i f f i c u l t to draw. When, as i s frequent l y the case, routes of competitors coincide only i n p a r t i c u l a r segments and d i f f e r i n others, one may even argue that no single market e x i s t s . 43 Notwithstanding these conceptual d i f f i c u l t i e s i n i s o l a t i n g the markets i n the shipping industry, we have l a i d the groundwork for an i n v e s t i g a t i o n of the forces of supply and demand at work within these markets which we have stated are instrumental i n determining the l e v e l of shipping rates that ex i s t at any given time. I t i s these forces and the relationships between them which we w i l l attempt to i d e n t i f y x i n Chapter I I I . CHAPTER III THE FORCES OF SUPPLY AND DEMAND IN THE SHIPPING MARKETS INTRODUCTION Taking as a point of departure the premise that the l e v e l of rates i n the shipping markets are determined by the i n t e r a c t i o n of supply and demand, the objective of t h i s Chapter i s to i d e n t i f y the forces at work i n the market which are l i k e l y to exert pressure on the demand for ship-ping space and the supply of vessels. S p e c i f i c reference w i l l be made to the tramp or open shipping market, but, as has been argued previously, a l l of the shipping markets are related, and therefore i n many cases, the discussion can be applied equally well to segments of the other markets of the shipping industry. One more important q u a l i f i c a t i o n must be made. Interest i n t h i s study i s with the fluctuations i n shipping rates, and since "... the demand for the supply of vessels to the open market fluctuates i n the short run, charter rates i n the open market also fluctuate within wide l i m i t s . Rates 45 i n the long term market are much more stable and le s s subject to v a r i a t i o n s a r i s i n g from short term movements i n demand conditions."" 1' Consequently, the area of concern i n t h i s chapter w i l l be r e s t r i c t e d to the short run. However, i t must be remembered that the operations of the long term market w i l l have d i s t i n c t bearing on rates i n the short term e.g. the incidence of long term chartering of vessels. Part of the d i f f i c u l t y here involves the time dimension — the delineation between the long and short run. For the sake of analysis we can define the long run as a period of time s u f f i c i e n t for new tonnage to be b u i l t and delivered. In t h i s respect, the short run w i l l be confined to periods of l e s s than one year "because ships have a long period of 2 gestation — from 12 to 18 months...." . This w i l l also bring the discussion within the parameters of voyage and time charters as previously defined. 1 U.N.C.T.A.D. 0 £ . C i t . p. 3. 2 C. O'LOughlin, The Economics of Sea Transport. Oxford: Permagon Press, 1967, p. 86. However, there i s a trend towards shorter construction periods. See Z.S. Zannetos, The Theory of O i l Tankership Rates. Cambridge, Mass.: The M.I.T. Press, 1966, p. 97. 46 At t h i s point i t may be worthwhile to set out some basic assumptions about demand and supply. The demand for world shipping i s represented by the volume of goods to be ca r r i e d and the distances over which they are to be c a r r i e d . We can further surmise that the demand for shipping services i s "derived" from the world demand for commodities and products and t h e i r world supply. The supply of shipping we can state as being represented by the world f l e e t , or more s p e c i f i c a l l y , by the volume of unfixed tonnage i n the world available for f i x i n g . ECONOMIC PORTRAYAL OF THE SHORT TERM SUPPLY AND DEMAND FOR SHIPPING For the sake of exposition, Figure I shows the r e l a t i o n -3 ship of the short term demand and supply of tramp tonnage. An examination of the figure shows that the t o t a l supply of shipping i s i n e l a s t i c down to a l e v e l of low rates. I t i s at t h i s point of i n f l e x i o n (the point at which the supply of 3 This argument, including Figure I i s taken from S.G. Sturmey, On the P r i c i n g of Tramp Ship Freight Service Bergen: In s t i t u t e of Shipping Research, 1965, pp. 5 - 9 . 4 7 shipping changes from e l a s t i c to i n e l a s t i c ) that the "lay-up r a t e " 4 i s encountered. Below t h i s lay-up rate the supply i s more e l a s t i c and i t i s t h i s e l a s t i c i t y which prevents rates from f a l l i n g c a t a s t r o p h i c a l l y . Even i n the deepest of slumps rates cannot f a l l below the lay-up rate of the ship with the lowest voyage costs per ton unless they f a l l to a l e v e l at which no trade takes place. Lay up of ships w i l l be touched on below but i t w i l l s u f f i c e to merely show the shape of the supply curve at t h i s point. The l e v e l of f r e i g h t rates at any one time i s given by the i n t e r s e c t i o n of the short term demand and supply curves i n terms of commodity ton/miles, when demand i s high (for example Dj D|) as shown i n Figure I, the demand curve crosses the supply curve i n i t s i n e l a s t i c range. ^ Although the lay up of ships w i l l be touched on l a t e r , the "lay-up rate" can be defined here as the rate at which i t pays a shipowner to lay up h i s ship rather than continue i t i n service. I t may be represented as the costs which would be incurred by operating the vessel (and hence avoided by not operating) minus the costs of layi n g up. The lay-up rate for each ship i s d i f f e r e n t , depending on the costs of operation of the ship concerned. On old ships the lay-up rate i s generally higher than on new ships, since the older ships are usually l e s s economical to operate i n terms of f u e l costs and repair b i l l s . Hence, i n a period of f a l l i n g rates i t i s the older ships which are the f i r s t to move int o lay up. 48 Commodity Ton/Miles FIGURE I SHORT TERM DEMAND AND SUPPLY OF TRAMP TONNAGE 49 Any change i n demand i s l a r g e l y taken up by a change i n p r i c e , with l i t t l e change i n the amount of tonnage supplied; quite small changes i n demand lead to large changes i n f r e i g h t r a t e s . I f demand i s lower (D-^  D£) so that the demand curve cuts the supply surve where i t i s becoming e l a s t i c , changes i n demand have a smaller e f f e c t on the volume of tonnage supplied. As demand f a l l s (to D| Dj ) the demand curve in t e r s e c t s the supply curve i n the range i n which the e f f e c t on price i s reduced and the e f f e c t s of the amount of tonnage offered by owners i s increased. I f demand continues to shrink, a point would eventually be reached at which rates could f a l l no farther and a l l future f a l l s i n demand a f f e c t only the volume of tonnage kept i n operation. This short resume serves to explain the existence of the short term demand and supply curves as they p r e v a i l on the macro l e v e l . However, i n t e r e s t here l i e s i n why they appear i n the form that they do and what forces determine where the l e v e l of price w i l l be i n equilibrium at t h e i r point of i n t e r s e c t i o n . A good beginning might be to examine the e l a s t i c i t y of demand. 50 ELASTICITY OF DEMAND At any l e v e l of demand, there i s a c e r t a i n e l a s t i c i t y with regards to p r i c e . As the l e v e l of rate v a r i e s , so the demand to carry some commodities w i l l vary, though i n the short run, the extent of the v a r i a t i o n i s l i k e l y to be small. At any one time the o v e r a l l demand for carriage may be f a i r l y i n e l a s t i c with respect to p r i c e . This r e s u l t s from the low price e l a s t i c i t y of the goods being transported which are usually basic raw materials. However, i n the carriage of many bulk commodities t h i s o v e r a l l price i n -e l a s t i c i t y i s becoming i r r e l e v a n t to the problem of balancing supply and demand on c e r t a i n routes. International competi-t i o n for many manufactured products which require basic raw materials for production has become increasingly severe. These raw materials include such items as coal, i r o n and manganese ore, phosphate rock and sulphur. Equally f i e r c e i s the struggle for markets on the part of those who supply the manufacturers with these e s s e n t i a l raw materials. Grade for grade there i s l i t t l e v a r i a t i o n i n the nature of these materials i n whatever part of the world they are produced. 51 And given an approximate p a r i t y of technological progress i n the developed lands, the same applies to the products of manufacturers. In these circumstances, perhaps the only way to get an edge on one's competitors i s to cut the costs of sea f r e i g h t s . Thus the combination of sea fre i g h t s and raw material cost i s becoming increasingly important i n the choice of sources of supply. Therefore, even though the o v e r a l l e l a s t i c i t y of demand for the carriage of these primary products may be very low, the demand can very e a s i l y s h i f t from area to area, which i n the short run or long run w i l l create imbalances between the supply of ships and the demand for shipping space on various routes. SHIFTS IN DEMAND As an aid to understanding why tramp rates fluctuate, l e t us consider further these " s h i f t s " i n demand. In the tramp market they can occur very r a p i d l y , a fa c t which can be attributed to two factors: the highly competitive nature of the tramp market and the lack of control of tramp rates. Emery Troxel suggests that the transport of bulk commodities i n the tramp shipping industry shows some signs 52 5 of pure competition. To substantiate t h i s statement, he postulates the reason for t h i s i s the f a c t that the supply side of the tramp shipping market i s characterised by a large number of firms each consisting of a small number of vessels. And because of the nature of the operation of the vessels themselves of engaging i n transport services wherever net returns w i l l be most advantageous, the ships always stand ready to follow l o c a t i o n a l changes i n transport demand. Consequently, the large number of tramp operators are not usually faced with a few large buyers of t h e i r services, since the general "non s p e c i a l i z a t i o n " of these vessels allows t h e i r owners or operators to choose among the many shippers of the various types of merchandise requiring t h i s kind of transportation. In other words, because the tramp operator can i n most cases transport grain, as well as coal, f e r t i l i s e r , , sugar, or many other such products, he also faces a competitive market on the demand side. 5 Emery Troxel, Economics of Transport. New York: Rinehart &. Company Inc., 1955, p. 420. 53 However, the degree of market competition declines as s p e c i a l i z a t i o n increases, be i t on the supply side or 6 the demand side. Concerning the former, ships are not t o t a l l y homogeneous, i f only because they d i f f e r according to tonnage, speed or cargo handling equipment. This d i f f e r -e n t i a t i o n on the supply side a f f e c t s demand because d i f f e r e n t routes have varying requirements for vessels of p a r t i c u l a r speed, carrying capacity or cargo handling equipment, at l e a s t within c e r t a i n l i m i t s . To t h i s may be added that on the demand side, once an operator has chosen to handle a p a r t i c u l a r type of cargo — a l b e i t the most advantageous to him — frequently the degree of competition i s reduced since the number of shippers of the commodity i n question may be small. Then the f r e i g h t rate charged w i l l , within c e r t a i n l i m i t s , depend on the bargaining power of the p a r t i e s . A further aid to the competitive nature of the tramp market i s the quick and easy communication of information about the current conditions i n a l l markets. This takes place through the medium of shipbrokers, who with t h e i r ^ Claudip Escarpenter, The Economics of i n t e r n a t i o n a l  Ocean Transport. Madison: The University of Wisconson Press, 1965, p. 29. 54 extensive network of contact a l l over the world, bring charterers and owners together; prices which have been negotiated are r a p i d l y c i r c u l a t e d so that the tone of the market can be monitored. A very large part of the world chartering i n the tramp market i s done either i n New York or on the B a l t i c Exchange i n London, which perform a l l the functions of an organized market. The existence of a market and of brokers i s not s u f f i -c i e n t to ensure that the market w i l l behave i n a p e r f e c t l y competitive way. This i s helped by other considerations. Neither e x i t nor entry i n t o the tramp market i s r e s t r i c t e d i n any way. Concerning the former, i t i s the "mobility" of the ships themselves which allows them to move e a s i l y out of one market to seek cargo i n another. The l a t t e r i s unhindered by any insurmountable economic or f i n a n c i a l considerations since the vessel i s a c t u a l l y the firm, under the j u r i s d i c t i o n of the captain for most of the time 7 and supervised only loosely by head o f f i c e . In addition, and because of the in t e r n a t i o n a l nature of tramp shipping, the market i s unhampered by any type of regulation or 7 Zannetos, op. c i t . pp. 180 - 181. 55 a r t i f i c i a l p r i c e c o n t r o l s 8 which would tend to d i s t o r t the competitive environment. The foregoing suggests several reasons why f r e i g h t rates i n the tramp market are able to fluctuate so r a p i d l y . I t remains now to i d e n t i f y and investigate some of the forces of demand and supply that cause these rapid s h i f t s and imbalances to occur. FORCES OF DEMAND The demand for ships depends on the state of world trade, since, as has been mentioned, i t i s derived from the demand for a l l the various commodities which move by sea. The volume of trade has been increasing at a steady pace over the years as was i l l u s t r a t e d i n Table I. From the point of view of transport performance, ton miles i s probably a better measure than merely tons and i n these terms the 8 This has been t r i e d several times i n the past however. Ultimately, the schemes received opposition from the tramp owners who believed that normal market procedure was the only sure and s a t i s f a c t o r y way of adjusting supply and demand. The reader i s directed to H. Gripaios. Tramp Shipping, Chapter 7 for a discussion of the various schemes that have been attempted for f r e i g h t rate s t a b i l i z a t i o n . 56 r i s e i n trade was also shown. The s t a t i s t i c s for the trans-port performance of bulk commodities was depicted i n Table I I , but these are assembled again i n Table I I I with the addition of the average length of haul. I t should be mentioned i n t h i s regard that the two components of demand — volume and distance — need not necessarily move together. This was c l e a r l y i l l u s t r a t e d a f t e r the closure of the Suez Canal i n 1967, which l e d to the lengthening of many trade routes and so to an increase i n ton/miles demanded without any necessary increase i n the volume of goods to be moved. An increased l e v e l of the volume of trade may not be s u f f i c i e n t impetus for a general r i s e i n the l e v e l of rates. On the contrary, i f the demand for trade arises on the routes which ships have previously t r a v e l l e d i n b a l l a s t , then these ships w i l l merely absorb the cargo and the e f f e c t s on the fr e i g h t rates i s l i k e l y to be depressive i n the reverse d i r e c t i o n . This f a c t i s exemplified i n Lewis' explanation of the behaviour of inward and outward f r e i g h t rates. He states: " . . . i f outward cargoes are smaller, some boats w i l l be going out i n b a l l a s t , and the outward rate w i l l be at a minimum. I f , then, the demand for ships inward increases, the inward rate may r i s e , but the outward 57 TABLE I I I MOVEMENTS OF SIX MAJOR BULK COMMODITIES YEAR TONNAGE SHIPPED TRANSPORT PERFORMANCE AVERAGE TRANS-PORT DISTANCE mi l l i o n s of tons 1,000 m i l l i o n Nautical ton miles miles 1960 1961 1962 1963 1964 1965 1966 1967 233 244 251 274 315 336 349 361 767 853 875 980 1,178 1,298 1,396 1,501 3,290 3,500 3,500 3,600 3,700 3,900 4,000 4,200 Source: Maritime Transport. 1968 p. 28. 58 rate cannot f a l l because i t i s already at a minimum. " 9 Lewis goes on to explain that rates on outward journeys may indeed r i s e i f shipping i s i n a boom period and queuing delays caused by increased inward cargoes would delay out-ward cargoes to the point where waiting was more expensive than seeking cargo elsewhere. His general conclusions were: "There i s a tendency for inward and outward rates to move inversely on any route when the general conditions elsewhere are stable; and there i s a tendency for them to move together i n the same d i r e c t i o n when market conditions are generally changing. The foregoing serves to complicate the problem of i s o l a t i n g the e f f e c t s of increased trade movement on the l e v e l of f r e i g h t rates. And, dealing as we are with f r e i g h t rate averages as computed i n the f r e i g h t rate indices, the problem assumes greater proportions. W. Arthur Lewis, Overhead Costs. London: George A l l e n & unwin Ltd., 1949, p. 96. Ibid, p. 97 59 Any consideration of the short run demand for tonnage space must take in t o account the d i s t i n c t seasonal movements that i t displays. The seasonal movement i n dry cargo bulk f r e i g h t s i s caused mainly for the shipment seasons for gra i n and to a smaller extent, sugar. 1 1 In addition, the seasonal movement i s further aggravated by demands for fu e l i n the northern winters and can be seen with the "comparative strength ening of fr e i g h t s i n Autumn as anxious shippers attempt to complete shipments to and from areas l i k e l y to be closed by 12 i c e . . . . " However, t h i s does not constitute an insurmount-able problem since there are several methods for removing seasonality from data p r i o r to empirical t e s t i n g . In addition, there are q u a l i t a t i v e aspects of demand which are impossible to quantify and which are not amenable to proposition or empirical v a l i d a t i o n such as we hope to perform. However, many of these aspects are germane to the problem of tramp shipping rate flu c t u a t i o n s , and as such, cognizance should be taken of them. 11 . . Grxpaios, op_. c i t . p. 16. 1 2 I b i d , p. 17. I t has been said that tramp f r e i g h t rates i n the short term are governed by changes i n demand a r i s i n g from 13 natural disasters, wars, crop f a i l u r e s and so on. The truth of t h i s statement i s d i f f i c u l t to dispute. A crop f a i l u r e or a bumper harvest can reverse the flow of commod-i t i e s or remove t r a f f i c to another corner of the world. In the area of tramp shipping, the impact of these events i s personified since: "No les s than 66% of annual tramp movements by weight i s made up of a g r i c u l t u r a l produce whose production i s d i r e c t l y covered by c l i m a t i c factors; a further 18%, consisting of s o l i d f u e l and f e r t i l i z e r , i s to some extent dependent on the weather. "-1-4 The wars and natural disasters are also i n c l i n e d to cause v i o l e n t upheavals i n the pattern of trade. The e f f e c t that these upheavals have on the l e v e l of f r e i g h t rates i n the tramp industry i s i n c r e d i b l e . Such an event occurred i n 1967 with the closure of the Suez Canal, and the eff e c t s of t h i s closure on tramp shipping i s succinctly reviewed by "Fairplay" i n t h e i r annual f r e i g h t market review, and i l l u s t r a t e s p r e c i s e l y how an event such as t h i s can a f f e c t 1 3 U.N.C.T.A.D. 0 £ . c i t . p. 39. 14 . Studer. op_. c i t . p. 10. 61 shipping rates: "The scene was set for a mightly f r e i g h t - r a t e war i n the gr a i n trades between these giant dry-cargo trans-ports and the s u r f e i t of tankers made redundant by the seasonal recession i n the demand for o i l . I t seemed inevi t a b l e that dry-cargo charter rates would plumet to new low depths of u n p r o f i t a b i l i t y and, ultimately, force many ships into lay-up berths. However, the gods of war smiled kind l y upon tramp shipowners - as they have done so often i n the past - and the erstwhile scene of depression was transformed within 24 hours to one of joyous optimism. Instead of a b a t t l e between the big dry-cargo bulk c a r r i e r s and tankers, they joined forces to combat the threat l e v e l l e d at the Western countries' o i l suppliesi During the two months of June and July, somewhere i n the region of 70 bulk c a r r i e r s , ranging i n sizes from 20,000 tons to 80,000 tons deadweight, and aggregating approximately 2,750,000 tons dead-weight, were hire d for varying periods on time charter up to a year. Most of these were taken as replacements for o r e / o i l c a r r i e r s and tankers employed i n the dry cargo trades so as to release them for a switch into the more p r o f i t a b l e o i l - t r a n s p o r t b u s i n e s s . " 1 5 I t i s apparent that the e f f e c t s of such an event would render f r e i g h t fluctuations inconsistent with what otherwise would be the e f f e c t of the normal laws of supply and demand. -With the withdrawal of the o i l - c a r r y i n g capacity from the dry cargo trade, i t would be a while before the supply and demand for shipping i n t h i s market could be 1 : 5 "Annual Freight Market Review", Fair p l a y Shipping  Journal, January 11, 1968, p. 94. 62 reconciled and rates would undoubtedly r i s e . Mention of a l l the minor upheavals would e n t a i l a l i s t which i s i n f i n i t e l y long. Currency r e s t r i c t i o n s , r e c i p r o c a l trade agreements and the vagaries of government p o l i c y are a l l r e f l e c t e d i n some manner i n the demand for shipping space. However, i n these circumstances, as well as i n a l l of the other aforementioned s i t u a t i o n s , the l e v e l of rates w i l l be determined by forces i n t r i n s i c to the supply of ships i n t e r a c t i n g with these fluctuations i n demand. I t i s t h i s side of the coin which must be investigated now. FORCES OF SUPPLY The broad d e f i n i t i o n of the supply of shipping as being represented by the world f l e e t was refined to be the amount of unfixed tonnage i n the world available for f i x i n g . In the short terra supply of shipping space t h i s i s a r e s t r i c t ing f a c t o r . " . . . i t w i l l be found that the amount of tonnage able to enter any p a r t i c u l a r trade i n response to a sudden increase i n demand w i l l only be a small proportion of the t o t a l world tonnage; some ships w i l l be on long term charters to s p e c i f i c trades; l i n e r companies, even i f they have near-empty ships, are t i e d to s p e c i f i c routes and schedules; many 63 vessels are s p e c i a l i z e d to a p a r t i c u l a r cargo, and others are l i m i t e d to p a r t i c u l a r seas and coastal a r e a s . " 1 6 In the short run therefore, i t can be surmised that the supply of tonnage i s f a i r l y r i g i d . However, even i n the short run, the tonnage available for f i x i n g i s not an absolute factor i n governing the supply of shipping space. There are v a r i a t i o n s which can be introduced which could increase t h i s supply of shipping space a v a i l a b l e . A f u l l e r or more e f f i c i e n t use of cargo space i s one way. A further p o s s i b i l i t y i s to increase the speed of the vessel but t h i s i s severely r e s t r i c t e d both by technical and finan-c i a l considerations. "Fuel consumption increases dispro-portionately once the 'economical speed' i s passed, and i n any case, the advantage of greater speed would probably be n u l l i f i e d by an increase i n time spent i n port. With the physical l i m i t a t i o n s placed on cargo handling by dock accom-odation and equipment, rapid turn round may not be possible. Even though i t would pay to incur overtime for loading and discharging at a time when fr e i g h t s were high, the saving i s further o f f s e t "by the general port congestion which 16 O'loughlm, op_. c i t . p. 70. 1 7 Jones, op_. c i t . p. 33. 64 18 normally accompanies active trade conditions." Any l i k e -lihood of such adjustments to supply would only occur, i f at a l l , when fre i g h t s were s u f f i c i e n t l y high to warrant such action. Therefore, the e l a s t i c i t y that can be introduced in t o the e x i s t i n g supply (even at high freights) i s confined within narrow l i m i t s , and i n these circumstances, the supply of tonnage i s very i n e l a s t i c . The amount of tonnage that w i l l be operating at any given time w i l l be dependent on the l e v e l of rates. Professor Sturmey suggests that the tramp operator goes into the market with three rate l e v e l s i n mind: (1) the continuation rate, which year-in year-out w i l l provide a l e v e l of p r o f i t s u f f i c i e n t to keep the owner happily i n business; (2) the transfer rate which w i l l enable him to more or les s break even but leave him wondering whether he would not be wiser to use h i s c a p i t a l and s k i l l elsewhere. "Logically, t h i s rate should be no l e s s than the t o t a l costs; i n practice the shipping company may have investments which w i l l produce a s u f f i c i e n t 18 Gripaios, op_. c i t . p. 64. 65 return that the whole outcome of the business i s p r o f i t a b l e , while management may regard shipowning as a way of l i f e or a national duty of such importance that f a i l u r e to cover t o t a l costs over a period of years w i l l not lead to the transfer of c a p i t a l from the industry. (3) the lay up rate, which i s equal to the l e v e l of voyage costs minus the costs of lay up, and which i s the rock-bottom l e v e l that the owner w i l l accept; i n general t h i s 20 " i s about 5% below the l e v e l of voyage costs." Let us consider farther the forces of supply as they would operate under conditions of the "continuation rate" and the "lay up r a t e " since i t i s within these two situations that rate fluctuations are most l i k e l y to occur. FORCES OF SUPPLY UNDER THE CONTINUATION RATE Disregarding for the moment the extreme i n e l a s t i c i t y of supply of shipping at very high f r e i g h t s , i t i s found that i n the operations of a normal market (where conditions are such that very high or low p r o f i t s are the exception rather than the rule) at a l l l e v e l s of rates above the l e v e l 19 Sturmey, On the P r i c i n g .... op_. c i t . , p. 5. I b i d . 66 where the most expensive to run vessels move i n t o lay-ups, the t o t a l supply of shipping i s also i n e l a s t i c . However, there i s more f l e x i b i l i t y within t h i s range. This happens when, as has been mentioned i n Chapter I I , l i n e r s w i l l move into tramp trades and o i l - c a r r y i n g vessels can be cleaned out for the carriage of grain. The r e s u l t s of t h i s f l e x i b i l i t y are such that at any given time there i s a c e r t a i n volume of f l o a t i n g tonnage available to meet the demand. The presence of t h i s reserve supply, and the mobility of the vessels themselves of being able to reach most markets i n a short time, even i f they are i n distant waters when a r i s e i n demand occurs, " i s the p r i n c i p a l cause of the sensitive nature of the f r e i g h t market inasmuch that i n the short run an increase or decrease of the available tonnage, i n any one market, brings about 21 minimum or maximum rates." During the normal operation of a f a i r l y p r o f i t a b l e market, there are other factors which would undoubtedly upset any balance between supply and demand. One of these 2 1 Jones, o£. c i t . p. 32. 67 i s the i n s t i t u t i o n a l structure of the industry i t s e l f . The presence of large shippers such as o i l companies and ore importers, who do not have s u f f i c i e n t tonnage for t h e i r own needs, implies that a change i n ownership p o l i c y of t h e i r behalf can cause new orders to be placed even i f there should be an aggregate oversupply i n the industry. On the other hand, a decision to negotiate more long term arrangements could mean a withdrawal from the stock of tonnage i n the short term market and a case of undersupply would r e s u l t . The discrete nature of ship ordering should be another factor which would s i g n i f i c a n t l y increase the uncertainty of the supply of tonnage i n the short run. In the case of unforseen increases i n demand "rates i n general may reach a l e v e l which w i l l stimulate shipowners to order new tonnage. Further to t h i s , i f a large number of owners i n d i f f e r e n t parts of the world order tonnage independently of each other, as i t a l l comes into service there i s l i k e l y to be a rather sudden oversupply of tonnage and a consequent f a l l O'Loughlin, op_. c i t . p. 88. 68 i n f r e i g h t rates. I t has been suggested that t h i s i s not the case and the more l i k e l y r e s u l t i s that i f f r e i g h t rates decline during the construction period, there may be no pressure for delivery and construction may be slowed down or suspended. In t h i s s i t u a t i o n , the e f f e c t may be widely 23 dispersed over the construction stage. Our empirical work w i l l attempt to t e s t these assumptions to see i f either holds true for the period under study. The foregoing has suggested conditions which ari s e to a l t e r the supply of tonnage during a time of normal p r o f i t a b l e trading for tramp shipping. I t has been concluded that there are c e r t a i n pressures exerted on rates that w i l l increase or decrease them coincident with the amount of tonnage a v a i l a b l e . However, the impact of these conditions i s magnified i f they occur i n a period of depressed rates. At t h i s time the problem i s made more complex by the introduc t i o n of laid-up ships. Jones, o£. c i t . p. 38. 69 FORCES OF SUPPLY UNDER THE LAY-UP RATE "Laying up s i g n i f i e s i n shipping the removal of tonnage from employment, implying, however, also the i n t e n t i o n and hope of i t s r e a c t i v a t i o n i n a changed market s i t u a t i o n . Laid up and reactivated tonnage w i l l d i r e c t l y and e f f e c t i v e l y influence the supply s i t u a t i o n of tramp tonnage." 2 4 The laying up of vessels can be i n s t i g a t e d by two things: a subsidence i n demand for shipping space or an oversupply of tonnage a v a i l a b l e . In either case freig h t s may become so depressed that a large volume of tonnage, which i n normal conditions would be maintained, i s l a i d up. The l e v e l of rates at which owners w i l l lay up t h e i r vessels w i l l d i f f e r considerably, dependent of course on the type and age of the ship and the l e v e l of national wage costs. Frequently, older ships, whose operations are unprofitable i n high-wage countries, are sold to owners i n low-wage countries. The low-wage costs i n these countries enable the owners to sustain the high repair costs on old 25 ships. I t i s apparent that the older ships retained by 24 F i s s e r . op_. c i t . p. 191. 25 . . . . S.G. Sturmey, B r i t i s h Shipping and World Competition. London: The Athlone Press, 1962, p. 252. 70 high-wage countries w i l l be the f i r s t to move into lay up. Between t h i s highest lay-up rates and the t h e o r e t i c a l low point at which no tramp shipping would operate, the supply i s e l a s t i c . The range of e l a s t i c i t y depends on the d i f f e r e n t l e v e l s of rates at which vessels w i l l move i n and out of lay up. Professor Sturmey argues that the supply of shipping i s i n e l a s t i c i n the short term even at low rate l e v e l s . Because each owner w i l l accept any rate above the lay-up costs rather than lay h i s ship up, changes i n demand are re f l e c t e d i n i t i a l l y not i n changes i n capacity offered but i n changes i n rates. In t h i s way, he suggests, the s i t u a t i o n emerges that tramp shipping can be depressed but operating to capacity. Jones' view disputes t h i s point. He maintains that under conditions of depressed rates and lay ups, even "a s l i g h t l y higher rate would be s u f f i c i e n t to bring on the market f a i r l y rapidly a large amount of serviceable tonnage." An a p r i o r i presumption i s that the former argument i s the most v a l i d . For as Professor Sturmey points out: Sturmey, On the P r i c i n g op. c i t . p. 14. 27 Jones, p_£. c i t . p. 33. 71 "Once rates enter t h i s more e l a s t i c range of the supply curve, the t o t a l volume of tonnage on o f f e r ceases to be important. Demand, and the points at which d i f f e r e n t owners w i l l lay up t h e i r ships, determine the course of f r e i g h t r a t e s . " 2 8 Under an assumption such as t h i s , the implication i s that the l e v e l of rates would be a function of both volume of trade and the l e v e l of laid-up tonnage rather than a straight inverse r e l a t i o n s h i p between f r e i g h t rates and l a i d -up tonnage as envisaged by Jones. Both of these assumptions may be amenable to empirical t e s t i n g as we s h a l l see l a t e r . Complicated as t h i s question of lay up i s , one fact i s c l e a r . The decision of whether or not to lay up a ship i s not a minor one. The actual cost of putting a vessel i n lay up and r e a c t i v a t i n g her may be as much as ^6,000. 2^ A natural conclusion therefore i s that the decision for lay up would be strongly influenced by the expectations of ship-owners as to the future market conditions based on t h e i r evaluation of how long the market depression w i l l l a s t . A mere r e f l e c t i o n on the word 'espectations' suggests that t h i s may be a potent force working on both the supply 28 Sturmey, On the P r i c i n g . . . . , op. c i t . p. 15. 2 ^ O'lLoughlin, op_. c i t . p. 147. 72 and the demand side of the shipping market. I t i s a force that should be investigated to gather a clear picture of the relationships that l i e at the heart of the problem of fr e i g h t rate f l u c t u a t i o n s . RELATIONSHIP BETWEEN EXPECTATIONS AND SUPPLY AND DEMAND IN SHIPPING MARKETS We have mentioned above that the expectations of shipowners may permeate the decision of new vessel ordering and the decision to lay up tonnage during a period of depressed rates. In addition, i t might be surmised that expectations of sudden increases i n the supply of vessels coming on the market would influence the l e v e l of f r e i g h t r ates, since f r e i g h t s have generally declined seriously before the supply of new tonnage from the yards becomes • i 30 av a i l a b l e . The degres of optimism or pessimism of both ship-owners and shippers i s r e f l e c t e d i n the manner i n which vessels are chartered. 3 0 Jones, op_. c i t . p. 32. 73 "What happens i s that i n a period of gently r i s i n g f r e i g h t s , c e r t a i n charterers may begin to grow _ worried about t h e i r a b i l i t y to secure suitable ships for t h e i r cargoes a few months or even a year or more ahead. The f i r s t signs of such a trend are an increase for charterings for "consecutive" voyages, probably at a scale of r i s i n g f r e i g h t s for each successive load, plus an increase i n the number of short-term charters... Once started, such a movement tends to snowball, and more charterers enter the market f e a r f u l that the actions of others i n protect-ing forward loading positions w i l l prejudice t h e i r own chances of getting space to meet t h e i r requirements. Shipowners, on the other hand, would be reluctant to charter out t h e i r vessels for long periods of time i n a r i s i n g market i f expectations were that rates would go higher. In a period of f a l l i n g rates, the reverse s i t u a t i o n i s true. Shippers wish to charter s t r i c t l y on a voyage basis i n a n t i c i p a t i o n of rates decreasing to a lower l e v e l while shipowners would prefer to l e t charters for longer periods i n the hopes of gaining extra p r o f i t s before a slump sets i n . The foregoing suggests that expectations have s i g n i f -icant impact on the supply of tonnage available i n the short term market. However, demand i s also affected. In the case of a very weak market where rates have been at a low U.N.C.T.A.D. op_. c i t . p. 40. 74 l e v e l for some period of time, chartering i s , almost 32 i n v a r i a b i l i t y , on a hand-to-mouth basis. Charterers, i n the main, do not take ships i n forward positions, as they f e e l that rates may go lower s t i l l , and owners are reluctant to f i x ahead at current l e v e l s i n case matters improve. As a r e s u l t , stocks of commodities are sometimes allowed to become depleted to a point at which regular shipments become es s e n t i a l and i n consequence, the e f f e c t upon f r e i g h t rates of even a s l i g h t shortage of tonnage i n a p a r t i c u l a r area may be considerably magnified. In short, what i s asserted here i s that the market sentiment or tones, i e . the degree of optimism or pessimism exhibited by shipowners or shippers, cannot be discounted when one i s t r y i n g to i s o l a t e some of the causes of f r e i g h t rate v a r i a t i o n s . The sentiment of the market stems from psychological factors and represents an expression of the transient feelings of the people a c t u a l l y operating i n the 33 open market. 3 2 "The Year i n the Freight Markets", The Syren and  Shipping, January 2, 1963, p. 203 3 3 U.N.C.T.A.D. pp. 39 - 40. 75 SUMMARY Chapter I I I has attempted to i d e n t i f y some of the forces which are at work i n the shipping markets. I t was suggested that the forces and the relationships between them were very complex. Moreover, i t i s apparent that the combinations of any, or a l l of them, w i l l r e s u l t i n some disruption i n the equilibrium of the supply of and the demand for shipping. Ultimately, these imbalances w i l l be r e f l e c t e d i n the f l u c t u a t i n g l e v e l of f r e i g h t rates. The discussion i n Chapter I I I has suggested also that several r e l a t i o n s h i p s e x i s t i n the tramp shipping market which should be amenable to empirical v a l i d a t i o n . Placed i n the form of hypotheses, these relationships include the following: 1. There exists an inverse r e l a t i o n s h i p between charter rates and l a i d up tonnage. 2. Periods of high f r e i g h t s w i l l stimulate ship-owners to order new tonnage. 3. The incidence of new vessels entering the ship-ping trade w i l l have a depressing e f f e c t on charter rates. 76 4. The l e v e l of charter rates w i l l be a function of the volume of seaborne trade. In the following chapter, these and several r e l a t e d hypotheses (although not necessarily i n that order), w i l l be subjected to empirical t e s t s . I t i s hoped the r e s u l t s of these tests w i l l e s t a b l i s h whether these commonly-held assumptions can be accepted or rejected on the basis of the data i n hand. I t i s r e a l i z e d that any empirical analysis w i l l be started with several ponderous d i f f i c u l t i e s . The abundance of exogenous variables influencing the operations of the shipping markets i s one. Qualitative variables i n t r i n s i c to the industry i t s e l f complicate any attempts to determine the impact of supply and demand variables (which can be quantified) on rates i n the voyage and time charter markets. But, as was mentioned i n the introduction to t h i s thesis any attempt made to analyse f r e i g h t rates i n the shipping markets may help i n improving the a v a i l a b i l i t y of information i n t h i s important f i e l d . I t i s with t h i s thought i n mind that the analysis i n Chapter IV begins. CHAPTER IV EMPIRICAL INVESTIGATION OF CHARTER RATE LEVELS INTRODUCTION The previous chapters have provided a b r i e f outline of the shipping industry i n general and the tramp shipping industry i n p a r t i c u l a r . Several forces of supply and demand i n the shipping markets have been i d e n t i f i e d . The purpose of t h i s chapter i s to present an analysis of the s t a t i s t i c a l tests which were performed on the quantified data of supply and demand variables and shipping rates i n the tramp shipping industry which have been assembled. I t must be remembered that any s t a t i s t i c a l analysis i s "ex post" and as such i t often leads the unwary i n t o u n r e a l i s t i c assumptions and often to make inappropriate comparisons; " l i e s , damned l i e s , and s t a t i s t i c s " too often has a r i n g of truth about i t . Therefore, caution must be used when i n t e r p r e t i n g the r e s u l t s of any s t a t i s t i c a l tests performed, p a r t i c u l a r l y i f these r e s u l t s are extrapolated to p r e d i c t future events. With reference to the area under study i n t h i s paper, predictions such as future rate l e v e l s 78 would be extremely dangerous considering the background of uncertainty against which the shipping industry operates. However, i f , through s t a t i s t i c a l testing, i t can be deter-mined what impact supply and demand variables have had on voyage and time charter rates i n the tramp shipping industry i n the past, i t may be possible to determine, under c e t e r i s  paribus conditions, what i s the " l i k e l y " behaviour of voyage and time charter rates i n future. I t i s hoped to be accomplished by t e s t i n g the assumptions of the o r i g i n a l hypothesis. I t would be pertinent at t h i s point to r e i t e r a t e the o r i g i n a l hypothesis as stated i n the introduction to t h i s paper: "Fluctuations i n shipping rates i n the tramp shipping industry can be explained by the use of multiple regression models incorporating d e t a i l s of the supply of and demand for shipping space."1 THE USE OF MULTIPLE REGRESSION AS A STATISTICAL DEVICE The use of multiple regression as a t o o l for the 1 See p.10 f f . 79 estimation of parameters i n economic rel a t i o n s h i p s i s widespread. The technical d e t a i l s of the method of estimation are purposely omitted from t h i s paper. The procedure i n the case of a r e l a t i o n s h i p between two variables i s discussed i n most s t a t i s t i c s texts. In the case of more than two variables, the l o g i c a l method for deriving the r e s u l t s i s the same as the simple case invo l v i n g only two variables, but the mathematical tools required f o r understanding i t are more sophisticated. This case i s discussed i n both advanced s t a t i s t i c s and econo-metric t e x t s . I t w i l l be assumed therefore, that the reader i s f a m i l i a r with the s t a t i s t i c a l technique of least-squares regression. However, i t may be b e n e f i c i a l to retrace with some care, the arguments which lead up to the l i n e a r regression hypothesis as i t i s usually stated at the outset of an estimation problem. A b r i e f r e c a l l of the elements of least-squares regression and a review of some of i t s properties which s u b s t a n t i a l l y a f f e c t the formulation of the economic models to which i t i s applied, may help c l a r i f y the problems which accompany the analysis performed below. The purpose of multiple regression i s to estimate the parameters of a relationship which has the form: Y - c f o + ^ X , + * 2 X 2 + + * K X k + u (4:1) This form assumes that the relationship between the variables i s linear with Y being the dependent variable -the one whose fluctuations are explained by the relation-ship; X = (X / ,X A #....X^) i s the set of independent variables -ones which i t i s hypothesized are relevant explan-atory influences of variations in the values of Y: A = (o( o, oCj,....* ) i s the set of parameters - in this case constants, whose true values are unknown -which relate X and Y; u = i s a random error (disturbance term) whose value i s unknown. It i s usually assumed that i t has a mean of zero. The presence of this term in empirical economic relations i s justi f i e d on the following grounds: 1. The set X can rarely include a l l possible 81 factors which influence Y. Consequently, u contains the influences which are omitted either because: a) their individual importance in explaining the behaviour of Y i s very small compared with that of the variables included in X; or b) i t may be either impractical or impossible to quantify these omitted variables. The sample observations on Y and X may be subject to errors of measurement - almost certainly the case with macroeconomic data at least - so that even i f there i s an exact relationship (one without an error term) between variables, the relationship between the measured values w i l l require an error term. St a t i s t i c a l analysis invariably must generalize on various facets of human behaviour. A relationship such as (4:1) implies that two identical changes in the values of X w i l l pro-duce identical changes in Y. However, there i s always some element of randomness in the human 82 element of response to the same stimulus, and this i s accounted for by the presence of u in (4:1). This i s in contrast to the non-randomness expressed in the remainder of the relationship. The inclusion of this error term i s very important to this study as w i l l be seen later. The reason for this i s that the assumption of independence (a mean value of zero) in the error term i s often inappropriate to economic analysis, particularly so when the observations consist of time series such as used in this study. In this case the observed X*s and Y's in their natural order denote success-ive periods over the years under study. In a situation such as this, i t i s often found that adjoining disturbance terms (which refer to consecutive periods) are no longer indepen-dent but positively related. This s t a t i s t i c a l phenomenon i s referred to as ser i a l correlation or autocorrelation. 2 The presence of autocorrelation in the error terms does not negate the use of multiple regression analysis 2 William Addison Neiswanger, Elementary S t a t i s t i c a l  Methods New York: The MacMillan Company 1965, p.649 83 however. It has been shown by Aitken 3 that the method of least squares s t i l l yields the best linear unbiased estimates of the regression coefficients provided the lack of indepen-dence i n the error series i s taken into account. However, i f account i s not taken of this error term in the analysis of a regression equation, any results would lose their significance. If, in fact, the errors were s e r i a l l y corre-lated and the straightforward least-squares formulas are applied directly to the observations Y X , there are three main consequences.4 F i r s t , unbiased estimates of of w i l l be obtained, but the sampling variances of these estimates may be unduly large compared with those achievable by a slightly different method of estimation. Second, i f the usual least-squares formulas for the sampling variances of the regression co-efficients are applied, a serious underestimate of these variances w i l l result. In any case, these formulas are no longer valid, nor are the precise forms of the " t " and "F" 3 A.C. Aitken, "On Least Squares and Linear Com-binations of Observations". Proceedings of Royal Society, Edinburgh, Volume 55, 1934/5 pp. 42-48. Cited in American S t a t i s t i c a l Association Journal, Vol.44, 1949, pp. 34-35. 4 J. Johnston, Econometric Methods. New York: Mcgraw-H i l l Book Co. Inc. 1963, p.179. 84 tests of significance used in a linear model. Third, any predictions based on the formulas would be inefficient, that i s , predictions would have needlessly large sampling variances. A test for detecting significant deviations from a se r i a l independence assumption has been designed by Ourbin and Watson,5 and extensive use w i l l be made of this test in the analysis below. One more brief caveat should be introduced before leaving the discussion of regression analysis as performed on the time series used i n this study. The primary purpose of this type of analysis i s to discover and measure any irregularities which characterize the movement of the data through time. In no way does time series analysis purport to explain the various variables that cause the observed behaviour, nor i s the passage of time i t s e l f a cause of the behaviour. It i s merely a way of measuring the net effect of a number of intricately related variables that cause the 5 J. Durbin and D.S. Watson, "Testing for Serial Correlation i n Least Squares Regression." Biometrika, Vo. 37, Cambridge: University Press, 1950, pp. 409 - 431. See Appendix B 85 observed behaviour, i n relation to the passage of time. Moreover, a high value of R^  (the coefficient of determin-ation) does not necessarily signify a "causal" relationship. We should speak of determination of one thing by another; however, only when a causal relationship can be logically defended; otherwise, an expression such as "associated with" or "accounted for" i s probably much better. VARIABLES INCLUDED IN ANALYSIS Chapter I mentioned the limitations which were im-posed on the choice of variables to be used in the empirical analysis. However, during the course of data collecting, sufficient monthly observations from several publications were available for fifteen variables. These are noted below: a. Voyage Index of Tramp Shipping Rates (VOYIND) b. Time Index of Tramp Shipping Rates (TIMIND) c. Tanker Freight Rate Index (TNKIND) d. New Order of Cargo Vessels (NEWCAR) See Appendix C for explanation of and sources of the variables used. 86 e. New Orders of Tankerships (NEWTNK) f. New Orders of Bulk Carriers (NEWBLK) g. New Cargo Vessels Launched (LANCAR) h. New Tankerships Launched (LANTNK) i . New Bulk Carriers Launched (LANBLK) j . Trials and Completions of Cargo Vessels (T&CCAR) k. Trials and Completions of Tankerships (T&CTNK) 1. Trials and Completions of Bulk Carriers (T&CBLK) m. World Sea Trade (WORSEA) n. Laid Up Cargo Vessels (LUCAR) o. Laid Up Tankerships (LUTNK) It i s apparent that the general hypothesis as out-lined i s only capable of being proven true i f the variables chosen to be the independent variables do indeed influence the behaviour of rates (i.e., the dependent variable as implied in the hypothesis). Every investigator presumes that he has selected the proper variables, but i t i s con-ceivable, for example, that i t i s the level of rates that influences the independent variables and not the reverse. In an effort to establish a framework within which to conduct the s t a t i s t i c a l tests on the data, i t was 87 decided to examine therelationships between the variables i n four distinct categories: 1. between the various categories of rates i.e., voyage, time and tanker rates. 2. the relationship between la i d up tonnage and charter rates. 3. the relationship between charter rates and the various stages of activity i n the shipyards i.e., ship ordering, ship launching and ship completions. 4. the relationship between the demand for shipping space as indicated by world sea trade, and charter rates. In each category, only selected combinations of variables were chosen to exhibit i n the text of this thesis. This i s by no means the extent of testing that was performed, and the reader i s directed to Appendix D for a more complete l i s t of the tests performed in each category or combination of categories as li s t e d above. Early investigation of the relationship between each pair of variables consisted of a series of simple plotted 88 graphs of the monthly data for comparison against each other. Following this, a series of scatter diagrams were made by plotting one variable against the other. This i s not a particularly sophisticated method of discovering relationships between variables, but i t does have advantages insofar as i t might reveal i f there are any f a i r l y obvious coincident variations in the pairs of variables. The graphs and scatter diagrams are not exhibited here primarily because of space limiiations and the wish to avoid confusion. The very short-run fluctuations in monthly data make visual comparison quite d i f f i c u l t and i t was therefore decided to construct a correlation matrix of a l l fifteen variables to see i f any, or a l l of the paired series had some degree of closeness. Table IV i s a correlation matrix incorporating the majority of these variables. TABLE IV VQYIND TIMIND TNKIND LUCAR LANCAR T&CCAR NEWBLK LUTNK WORSEA VOYIND 1.000 TIMIND 0.8960 1.000 TNKIND 0*3639 0.2581 1.000 LUCAR -0.4918 -0.6635 -0.0061 1.000 LANCAR -0.6116 -0.6624 -0.3662 0.5656 1.000 T&CCAR -0.5998 -0.6666 -0.3046 0.5379 0.7301 1.000 NEWBLK 0.4914 0.5132 0.1121 -0.4794 -0.4078 -0.3407 1.000 LUTNK -0.5888 -0.5579 -0.3520 0.5386 0.7125 0.7703 -0.2855 1.000 WORSEA 0.4639 0.6178 0.1718 -0.6100 -0.6651 -0.7340 0.2082 -0.8116 1.000 CORRELATION MATRIX - 1960 - 1968 TRAMP SHIPPING VARIABLES co 10 90 1. RELATIONSHIPS BETWEEN RATES  VOYAGE AND TIME CHARTER RATES A good starting point of an analysis of the impact of supply and demand variables on charter rates would be to investigate the relationship between the two rates of con-cern in this study - voyage and time rates. Of course i t should be realized that the time index and voyage index are prices i.e., "market equilibrium points", and i t i s assumed that supply and demand relationships affect both rates although to a different degree. Therefore, a com-parison of the two indices i s , in effect, an examination of the behaviour of the relative timing of these equilibria. It i s worth investigating as i t could be established i f common arguments about the behaviour of the two rates are valid for the period which i s under study in this thesis. There i s a close connection between voyage and time rates, especially in the short run. W.R. Lewis explains this interrelation: "Time charter rates are linked with voyage charter rates both on the supply side and on the dem-and side. On the supply side, the shipowner has the alternative of seeking a time or voyage charter. On the demand side, some merchants are indifferent as between time and voyage charters and even i f this were not so, speculators would take ships on time charters and relet them on voyage charter i f the former rates were out of step. The two rates must therefore move together." 7 This speculation does exist however, and i s associated with the expectations of shipper and shipowners mentioned in Chapter III. If i t were not for the po s s i b i l i t y of this speculation, there would be absolutely no difference between the trend of charters for a voyage and those for several voyages i.e., time charters. But since uncertainty and speculation - forecasting market developments and acting accordingly - always does exist, there may be differences in rate trends between the two markets. It i s further postulated by Branch 8 that since time chartering extends into the future farther than a mere voyage, rates in this market tend to reflect longer run expectations and do not give the same weight to current market conditions. When i t i s expected that the market w i l l 7 W.R. Lewis, op.cit. p.91 Q Alan E. Branch. The Elements of Shipping. London: Chapman & Hall, 1964, pp 80 - 81. 92 strengthen, time charter rates w i l l rise more rapidly than those for voyages and w i l l tend to be above the current voyage rates. On the other hand, when the market i s weaken-ing, owners are anxious to f i x for as long a period as possible so as to hedge against further deterioration; charterers hold off in anticipation, and the long term rates tend to f a l l faster than voyage rates and w i l l tend to be below the current voyage rates. From the discussion above, a general hypothesis can be formulated: "Generally speaking, voyage and time charter rates move in the same direction, but because time charter rates depend on longer range market expectations, they tend to fluctuate more widely then voyage rates." Although simple graphs are not an ideal method of research to support or reject an hypothesis, the relation-ship of interest between the two rates i s not a determin-i s t i c one and therefore Figure II w i l l suffice to i l l u s t r a t e the point. A visual examination of Figure II implies that the behaviour of voyage and time charter rates i s indeed as surmised above at least for the period 1960 - 1965. From INDEX us — Voyage charter index — Time charter index FIGURE II YEAR UNITED KINGDOM CHAMBER OF SHIPPING . TRAMPSHIPPING INDICES ... 1960 - 1968 (1960 = 100) 94 this point on there seems to be a slight change i n the pattern and time charter rates appear to be consistently higher than voyage charter rates. This i s an interesting phenomenon, but one for which this investigator has no ready explanation. This visual observation of a seeming discrepancy from an accepted argument over previous years, prompted investigation about other commonly held opinions concerning time and charter rates. W.R. Lewis states, "... the time charter rate should depend not so much on the current voyage rate as upon the expected trend of voyage rates in future months." Alan Branch supports this assumption when he says, "... the current time charter rate tends to reflect the expected trend of voyage rates in the future."^-® How-ever, there i s also an opposing school of thought that suggests what happens i n the voyage market "today" influences what w i l l happen i n the time market "tomorrow". The discussion here suggests that there are different hypotheses which could be formulated, each probably giving different results. However, i f at f i r s t i t i s assumed that 9 W.R. Lewis, loc. ext.  1 0 Alan E. Branch, op.cit. pp. 80 - 81 95 the former argument i s true, then an hypothesis could be constructed as a basis for testing this relationship to see i f common assumptions are borne out by fact for the period under study in this thesis. 1. "During the years 1960 - 1968, the closest relationship between voyage and time charter rates w i l l occur when the level of the time charter index precedes that of the voyage charter index, thus implying that the voyage index i s a function of the time index for some lagged period." Alternatively, i f i t i s assumed that the second argument i s more valid then a competing hypothesis i s just the reverse: 2. "During the years 1960 to 1968, the closest relationship between voyage and time charter rates w i l l occur when the level of the voyage charter index precedes that of the time charter index, thus implying that the time index i s a function of the voyage index for some lagged period." The results of the regressions run on the variables of the two functional relationships as outlined in the hypotheses are presented in Table V 1 1 and Table VI. On the basis of 1 1 A brief example using the subscripts in Table V and subsequent tables in this chapter may aid i n interpret-ing the results: 1. TIMINDt » f (V0YIND)t merely means that this month's (t) time index i s a function of this month's (t) voyage index. 2. TIMINDt - f (V0YIND)t_i means that this month's (t) time index i s a function of last month's (t-1) voyage index. See bottom of page 96 for Note 96 Table V, hypothesis No.l i s rejected. On the other hand, hypothesis No.2 cannot be rejected because the data would seem to indicate the movement of the indices are as postu-lated. Although the movement i s not particularly strong, which would raise doubts as to i t s validity, i t would appear to indicate that the voyage index i s a good indicator of the behaviour of the time index in future periods rather than the reverse as argued by Lewis and Branch. To these tentative conclusions must be added a word of caution. The results in both tables indicate a high autocorrelation in the residuals as evidenced by the low 12 Durbin-Watson Stat i s t i c , and thus definite conclusions must be held in abeyance u n t i l such time as this problem of serial correlation can be investigated i n a later chapter. VOYAGE. TIME AND TANKER RATES The discussion i n Chapter II implied that the tramp shipping market i s always vulnerable to competition from Note: (t-1) to (t-6) indicate lags of from one to six months. Sx.y a Standard Error of Estimate D.W.S. = Durbin Watson Stat i s t i c 12 See Appendix B for significant levels for the Durbin Watson Test. TABLE V INDEPENDENT VARIABLE RESULTS: TIME INDEX REGRESSED ON VOYAGE INDEX REGRESSION CONSTANT COEFFICIENT t-VALUE R" F RATIO Sx.y D.W.S, V O Y I N D t V O Y I N D t _ 1 V O Y I N D t _ 2 V O Y I N D t 't-3 VOYIND t-4 VOYIND t-5 VOYINDt_e -23.12 -24.06 -18.07 -12.46 - 7.95 - 1.17 8.23 1.26 1.27 1.21 1.16 1.12 1.06 0.97 18.60* 19.46* 15.87* 13.42* 12.01* 10.35* 3.63* 0.80* 346.12 8.24 0.86 0.81* 378.92 7.94 1.23 0.75* 252.04 9.31 1.19 0.68* 180.19 10.51 0.97 0.64* 144.41 11.31 0.67 0.57* 107.19 12.42 0.54 0.48* 74.53 13.68 0.42 * Significant at .01 level <x> TABLE VI INDEPENDENT VARIABLE RESULTS: VOYAGE INDEX REGRESSED ON TIME INDEX REGRESSION CONSTANT COEFFICIENT t-VALUE F. RATIO Sx.y D.W.S, TIMINDt TIMINDt_1 TIMINDt_2 TIMIND* 't-3 TIMIND. t-4 TIMIND. t-5 TIMIND. t-6 35.99 42.56 50.94 59.88 68.15 74.19 78.19 0.63 0.57 0.50 0.42 0.35 0.30 0.26 18.60* 0.80* 346.1 12.77* 0.66* 160.0 9.20* 0.50* 82.69 6.86* 0.36* 45.04 5.27* 0.24* 26.15 4.34* 0.18* 17.48 3.72* 0.14* 13.21 5.85 0.88 7.81 0.77 9.48 0.51 10.85 0.37 11.82 0.32 12.34 0.26 12.56 0.20 •Significant at .01 level 00 99 tankers which, at times of receding demand in the carriage of o i l , are able to convert to grain-carrying f a c i l i t i e s . If this i s so, then i t seems conceivable that when tanker rates f a l l , the incidence of these ships entering the tramp market w i l l have an influence on tramp rates. In other words, when tanker rates f a l l , tramp rates should be forced down because of an oversupply of vessels vying for the bulk trade. From this assumption, an hypothesis could be formulated con-cerning the impact which changes in tanker rates w i l l have on voyage and time charter rates. "A f a l l in tanker rates w i l l be followed shortly thereafter by a f a l l in voyage and time charter rates with the reverse situation when tanker rates r i s e . " The relationships to be investigated consist of: Voyage Index = f (Tanker Index) Time Index = f (Tanker Index) The results of the series of regressions run on these func-tional relationships are depicted in Tables VII and VIII. The association i s not strong; however, the majority of them are considered to be s t a t i s t i c a l l y significant at the 95% confidence level. Two things can be inferred from the table. TABLE VII RESULTS: VOYAGE INDEX t REGRESSED ON TANKER INDEX INDEPENDENT VARIABLE CONSTANT REGRESSION COEFFICIENT t-VALUE R2 F. RATIO Sx.v D.W.S. TNKINDt. 91.12 0.27 3.60* 0.13 12.97* 12.28 0.21 TNKINDt_1 93.04 0.24 3.10* 0.10 9.66* 12.56 0.28 TNKINDt_2 93.68 0.23 2.94* 0.09 8.65* 12.69 0.24 TNKINDt_3 94.95 0.21 2.65* 0.07 7.06* 12.86 0.22 TNKIND t - 4 97.84 0.16 2.03# 0.04 4.13# 13.15 0.22 TNKINDt_5 99.75 0.13 1.64 0.03 2.71 13.32 0.21 TNKINDt_6 100.38 0.13 1.55 0.02 2.40 13.36 0.22 * Significant at .01 level # significant at .05 level INDEPENDENT VARIABLE CONSTANT TABLE VIII RESULTS: TIME INDEXt REGRESSED ON TANKER INDEX REGRESSION COEFFICIENT t-VALUE R2 F RATIO Sx.y D.W.S, TNKIND. TNKIND t-1 TNKIND t-2 TNKIND t-3 TNKIND. t-4 TNKIND. t-5 TNKIND t-6 96.30 94.95 96.10 96.14 100.29 102.41 102.77 0.27 0.30 0.28 0.28 0.21 0.18 0.18 2.46* 0.06 6.06# 17.94 0.20 2.67* 0.07 7.15* 17.89 0.19 2.49* 0.06 6.23# 18.04 0.21 2.49* 0.07 6.20# 18.13 0.17 1.86* 0.04 3.48 18.51 0.20 1.56 0.02 2.45 18.71 0.18 1.52 0.02 2.33 18.80 0.19 * Significant at .01 level # Significant at .05 level 102 Fi r s t , a one month lag of the tanker index on the time index shows a slightly stronger relationship (although probably not significantly different from a 0, 2, or 3 month lag) and might offer some support to the hypothesis with respect to time charter, rates. Secondly, the stronger relationship shown by the tanker and voyage index on a no-lag basis, although appearing to reject the hypothesis with respect to voyage rates, may only be indicative that rate adjustments to competitive pressures i n the voyage market are instantaneous, or by the fact that voyage rates may become depressed by the expectation of a transfer of tanker vessels -into the voyage market. The monthly index, based as i t i s on averages, would not reflect this association i f i t i s a short time-lag situation. Very l i t t l e could be said of these results even though they are significant and logically valid. However, i t i s important to note that they do show some evidence of short (0 to 3 month) lags. It could be postulated that a relationship does exist as hypothesized, but very l i t t l e of the variation of voyage or time rates can be explained by the variation in tanker rates. In addition, the evidence 103 of autocorrelation pua/ents us from drawing any pointed conclusions. 2. RELATIONSHIP BETWEEN LAID UP CARGO TONNAGE  AND CHARTER RATES Chapter III established the fact that the movement of ships into lay up follows a period of low rates where some ships find that revenues do not cover the variable cost of operation. Logically, i t would be expected that a high level of rates would be associated with a low level of l a i d up tonnage and vice versa. It can therefore be hypothesised: 1. "The level of l a i d up cargo tonnage w i l l vary inversely with the level of time and voyage rates." It i s also known from previous discussion that ship-owners w i l l accept any rate above the lay-up cost rather than lay their ships up. Consequently, i t would be expected that the incidence of l a i d up tonnage would lag somewhat behind the level of charter rates. This assumption in the form of a second hypothesis would be: 2. "The level of l a i d up cargo tonnage w i l l lag the level of rates by some time period." If consideration i s given to the fact that owners are re-104 luctant to lay up vessels because of the costs involved (particularly i f conditions in the market are expected to improve), i t might also be assumed that on a voyage basis, owners would continue to operate ships until the time that lay-ups became absolutely necessary because of economic considerations. On the other hand, the length of commitment involved i n a time charter should dictate an earlier lay up of these vessels. A third hypothesis can therefore be formulated: 3. "The lagged relationship between la i d up cargo tonnage and the time index w i l l be less than the equivalent relationship with regards to the voyage index." The results of the regressions run on the functional relation ships: Laid Up Cargo Tonnage =» f (Voyage Index) Laid Up Cargo Tonnage » f (Time Index) are l i s t e d in Tables IX and X. Several things can be infer-red from these results. F i r s t , the negative signs of a l l of the regression coefficients support the assumption of an inverse relationship and thus hypothesis 1^  i s accepted for both the time and the voyage index. Secondly, also for both TABLE IX INDEPENDENT VARIABLE CONSTANT RESULTS: LAID UP CARGO, REGRESSED ON VOYAGE INDEX REGRESSION COEFFICIENT t-VALUE R* F RATIO Sx.y D.W.S. VOYINDt VOYIND t-1 VOYIND t-2 VOYIND t-3 VOYIND t-4 VOYIND t-5 VOYIND t-6 4676.39 5058.09 5443.51 5780.41 5896.24 5949.31 5923.66 -33.26 -37.01 -40.74 -43.97 -45.20 -45.80 -45.86 -5.20* 0.24* 27.11 -6.26* 0.31* 39.25 -7.49* 0.40* 56.16 -8.77* 0.48* 77.01 -9.48* 0.52* 90.053 -9.94* 0.55* 98.82 -11.70* 0.63* 137.03 776.55 0.23 718.03 0.29 660.22 0.35 607.44 0.40 577.24 0.47 558.02 0.46 474.33 0.40 •Significant at .01 level TABLE X INDEPENDENT VARIABLE CONSTANT RESULTS: LAID UP CARGO, REGRESSED ON: TIME INDEX REGRESSION COEFFICIENT t-VALUE R2 F RATIO Sx.- D.W.S, TIMIND, TIMIND t-1 TIMIND t-2 TIMIND t-3 TIMIND t-4 TIMIND t-5 TIMIND t-6 4689.69 4856.76 4940.02 4945.58 4927.10 4783.32 4561.39 -31.86 -33.56 -34.48 -34.69 -34.69 -33.55 -31.90 -8.17* 0.44* 66.85 -9.52* 0.51* 90.72 -10.57* 0.57* 111.84 -11.09* 0.60* 123.19 -11.42* 0.61* 130.60 -10.74* 0.59* 115.44 -11.13* 0.61* 124.01 667.26 0.37 603.06 0.40 557.97 0.49 534.72 0.59 518.98 0.55 533.76 0.52 489.31 0.33 •Significant at .01 level the time and the voyage index, hypothesis 2_ i s accepted although i t w i l l be noticed that the l a i d up/voyage index relation appears to explain better as the lag i s increased while the l a i d up/time index relation appears more steady. Hypothesis 3. received no support from the results and thus i t must be rejected. As a further i l l u s t r a t i o n , the data for l a i d up cargo tonnage and the voyage and time indices were plotted on scatter diagrams. Figures III and IV reveal a relation-ship that appear non-linear. An obvious line of best f i t seemed to be concave rather than straight. Consequently, i t was decided to use a logarithmic transformation of the variable observations in an effort to improve the f i t . FITTING A LOGARITHMIC CURVE In the same way that i t i s possible to represent relations mathematically by a straight line, i t i s possible to represent them by curves of various types. There are three simple types of logarithmic curves, and i f we use Y to represent the logarithms of the Y values, and X to re-present the logarithms of the X values, the original LAID UP TONNAGE ( 000's TONS ) 9 <3> © © I Loo I V © 0 O 1oo L 9 0 0 <? Z o o J | i | J~ (fioe> o 7f AS- 9S~ "5" /as- f3S~ TIME INDEX FIGURE III SCATTER DIAGRAM OF LAID UP CARGO TONNAGE AND TIME INDEX 1960 - 1967 00 LAID UP TONNAGE ( 000'S TONS ) iloa 1 © Ztoo (3 0 ffoo Mao . Coo -Zoo o o o o © o 0 « O 0 0 © 7S SS pj- /OS i'S~ /zs~ '3S '+S VOYAGE INDEX FIGURE IV H SCATTER DIAGRAM OF LAID UP CARGO TONNAGE °, AND VOYAGE INDEX 1960 - 1967 110 equations and their transformations w i l l be as follows: (a) log Y « a + bX, to Y » a + bX (b) log Y * a + b log X, to Y = a + bX (c) Y - a + b log X, to Y » a + bX In each case i t i s evident that the new equation i s identical i n form with the simple straight-line equation: Y - a + bX and the same method of regression may therefore be used for finding a line of best f i t from the data. Of the three equations (a, b, and c) l i s t e d above, (b) seemed to preside the most satisfactory f i t and a series of regressions identical to those performed above were run on the transformed data. The results are exhibited in Tables XI and XII. It i s apparent from the tables that the assumption of a non-linear relationship was worthy of merit and the f i t of the data was improved somewhat. How-ever, there i s no added support or rejection given to any of the three hypotheses from this transformed data. In spite of the fact that the results in Tables VII to XII are a l l significant at the 99% level of confidence, no definite conclusions can be drawn because of the presence TABLE XI RESULTS: LOG LAID UP CARGOt REGRESSED ON LOG VOYAGE INDEX INDEPENDENT VARIABLE CONSTANT REGRESSION COEFFICIENT t-VALUE R' F RATIO Sx.y D.W.S, LOG VOYINDt 23.16 -3.52 -6.19* 0.31* 38.35 0.64 0.14 LOG VOYINDt_ •1 25.10 -3.94 -7.53* 0.40* 56.83 0.59 0.17 LOG VOYINDt_ 2 26.89 -4.32 -9.13* 0.50* 83.52 0.53 0.21 LOG VOYINDfc_ •3 28.19 04.60 -10.70* 0.58* 114.60 0.48 0.30 LOG VOYINDt_ •4 28.94 -4.77 -12.02* 0.64* 144.63 0.45 0.36 LOG VOYIND^ t-•5 29.38 -4.86 -13.14* 0.68* 172.67 0.42 0.47 LOG VOYIND4. .a. 29.68 04.93 -15.19* 0.74* 230.81 0.36 0.50 •Significant at .01 level INDEPENDENT TABLE XII RESULTS: LOG LAID UP CARGOt REGRESSED ON LOG TIME INDEX REGRESSION VARIABLE CONSTANT COEFFICIENT t-VALUE R2 F RATIO Sx.v D.W.S. LOG TIMINDt 23.21 -3.50 -10.55* 0.56* 111.34 0.51 0.29 LOG TIMIND. •1 24.14 -3.70 -12.65* 0.65* 160.18 0.45 0.33 LOG TIMINDt_ •2 24.72 -3.83 -14.77* 0.72* 218.26 0.40 0.43 LOG TIMINDt_ •3 24.88 ^3.87 -16.05* 0.75* 257.71 0.37 0.61 LOG TIMINDt_ •4 24.90 -3.87 -16.97* 0.78* 288.01 0.35 0.60 LOG TIMINDt_ •5 24.55 -3.80 -16.36* 0.77* 267.92 0.35 0.69 LOG TIMIND. a 24.11 -3.17 -16.60* 0.77* 275.67 0.34 0.53 •Significant at .01 level 113 of the low Durbin Watson Statistic indicating auto-correlation. Before attempting to eliminate this problem, the relationship between shipbuilding and charter rates i s examined. 3. THE RELATIONSHIP BETWEEN SHIPBUILDING  ACTIVITY AND CHARTER RATES An important area to investigate i s monthly activity in the shipbuilding industry to discover i f this activity has any impact on charter rates, or, alternatively, i f i t i s more l i k e l y that the rates existing in the market influence the building of vessels. It i s apparent that the amount of tonnage bu i l t in any one month i s very small com-pared to the total existing tonnage of the word fleet but i t does constitute additions to the supply of shipping space available. If shipowner's expectations are reflected in the behaviour of rates as has been suggested, we might expect different reactions during the shipbuilding cycle. This cycle can be divided into three distinct phases: (1) vessel ordering, (2) vessel launching, and (3) completion of the vessel. These phases are investigated i n this order below. CHARTER RATES AND ORDERS FOR NEW VESSELS 114 It has been suggested by Jones1"5 and O'Loughlin that tramp freight rates are subject to wide fluctuations which are accentuated by the eagerness with which ship-owners place orders for new tonnage during every period of high freights. It can therefore be hypothesised that: 1. "There w i l l be a positive relationship between the level of new orders and the level of charter rates." In addition, i t can be assumed that shipowners would only be encouraged to order new vessels i f they expect the high level of freight rates to continue. Therefore, i t i s to be expected that there w i l l be a slight time lag between the level of freights and the level of new orders; con-sequently, a second hypothesis would be; 2. "The level of charter rates w i l l precede the ordering of new vessels by some time period." However, recognising the fact that tramp rates are Jones, op. c i t . p. 34 O'Loughlin, op.cit. p.88 Characterized by c y c l i c a l movements of "boom" and "slumps" as revealed in Figure II, i t might be expected that because of the lead time between the ordering and delivery of a vessel, the placing of orders to take advantage of "boom" periods in tramp shipping would be too late. By the time the major portion of these vessels were delivered, the market may have levelled off and they would not secure the advantages of the high freights existing at the time of ordering. Thorburn 1^ has suggested that building costs at shipyards vary very much with the trade situation as do, to a greater extent, the shipyard price of new vessels. He postulates that reductions by 25 to 50 per cent from the boom years to slump years may occur. Therefore, he con-cludes that this factor probably contributes towards induc-ing some shipowners to order new vessels in times of bad trade. If i t can be assumed that this could be a major con-sideration by prospective owners, then we are faced with a completely opposite hypothesis to No. 1, which seems to be valid from a conceptual standpoint. This hypothesis would 15 Thomas Thorburn, Supply and Demand of Water Transport. Stockholm; The Business Research Institute. lybU, p. 92. 116 be cast as the following: 3. "There w i l l be a negative relationship between the level of new orders for vessels and charter rates." Table XIII and Table XIV present the results of regression tests run on the functional relationships: New Orders (Bulk Carriers) = f (Voyage Index) New Orders (Bulk Carriers) = f (Time Index) On the basis of the s t a t i s t i c a l results, support i s given to hypothesis No.l since the data appears to sub-stantiate a positive relationship between the two variables. For the same reason, hypothesis No.3 must be rejected. With reference to hypothesis No.2, the data would suggest that there i s a slight lag between the level of charter rates and new vessel ordering and thus the hypothesis cannot be rejected. Although the magnitude of this variation between a lag and no-lag situation i s small, according to the data i t cannot be ignored. This i s evident when Table XIV i s considered where the pattern of lag i s quite noticable by the obvious decline in the f i t of the data following a one month lag of the independent variable. In addition to the fact that the results for both the time and voyage index TABLE XIII INDEPENDENT VARIABLE RESULTS: NEW ORDERS BULK CARRIERS REGRESSED ON VOYAGE INDEX CONSTANT REGRESSION COEFFICIENT t-VALUE F RATIO Sx . i D.W.S. VOYIND, VOYIND t-1 VOYIND t-2 VOYIND t-3 VOYIND t-4 VOYIND. t-5 VOYIND t-6 -617.93 -777.32 -651.58 ^691.85 -648.54 -546.63 -497.63 8.69 10.17 9.02 9.39 9.02 8.07 7.63 5.20* 6.44* 5.41* 5.67* 5.35* 4.60* 4.27* 0.24* 27.05 203.29 1.94 0.33* 41.48 192.07 1.90 0.26* 29.32 202.4 2.06 0.28* 32.24 200.63 1.92 0.26* 28.71 204.00 1.89 0.20* 21.17 212.39 1.82 0.18* 18.30 215.99 1.70 •Significant at .01 level TABLE XIV INDEPENDENT VARIABLE RESULTS: NEW ORDERS BULK CARRIERS, REGRESSED ON TIME INDEX CONSTANT REGRESSION COEFFICIENT t-VALUE R* F RATIO Sx.y D.W.S, TIMIND. TIMIND t-1 TIMIND t-2 TIMIND t-3 TIMIND t-4 TIMIND t-5 TIMIND t-6 -408.41 -450.99 -380.32 -346.73 -268.53 •^231.11 -108.27 6.44 6.83 6.22 5.94 5.27 4.93 3.86 5.51* 0.24* 26.65 203.7 1.91 5.93* 0.29* 35.22 197.05 1.91 5.21* 0.24* 27.19 204.40 1.91 4.87* 0.22* 23.74 208.54 1.79 4.18* 0.17* 17.51 215.27 1.73 3.84* 0.15* 14.78 219.43 1.61 2.89* 0.09* 8.40 227.90 1.57 •Significant at .01 level 119 regressed on new orders are a l l s t a t i s t i c a l l y significant at the 99 per cent confidence level, i t w i l l be noticed that there i s an absence of autocorrelation as evidenced by the Durbin Watson Stat i s t i c which i s well within the significant limits as outlined in Appendix II. From this discussion, i t can be said with confidence that the level of charter rates has a definite influence on the ordering of new bulk carriers. However, judgement i s reserved i n stating the amount of variation i n new orders that can be explained by the variations i n rates particular-ly i n a lag situation because the validity of the s t a t i s t i c a l results i s marginal. CHARTER RATES AND VESSELS LAUNCHED The launching of vessels precedes the actual delivery of vessels to shipowners by a periodof several weeks or even 16 months depending on the type and size of vessel. In other words, this stage of shipbuilding i s somewhat closer to the time when these vessels w i l l add to the existing supply than The source from which the data on launchings was extracted revealed a period of from six weeks to five months between launching and completing of vessels. 120 i s the case with vessel ordering. Therefore, i t i s conceiv-able that the launching of vessels could have an effect on freight rates, i f rates are in any way influenced by these impending additions to supply. An hypothesis following this line of reasoning could be stated: 1. "The level of charter rates w i l l be inversely related to the launching of new vessels, with a strengthening of this relationship shown when the launching of vessels precedes the rates." However, i t i s also well known that shipyards have the capacity to vary the scheduling of their launchings by either slowing construction or suspending i t entirely i f there i s no pressure for delivery. This pressure would certainly be absent i f a shipowner f e l t he would be taking delivery of a vessel during a slump period in rates. This dichotomy intro-duces the opposing hypothesis of an opposite functional relationship i.e., launching as a function of rates rather than the reverse. The hypothesis would be: 2. "The launching of cargo vessels w i l l be positively related to the level of charter rates." Tables XV and XVI present the results of the regres-sions run on the relationships accompanying hypothesis 1. TABLE XV RESULTS: VOYAGE INDEX^ . REGRESSED ON LAUNCHED CARGO VESSELS INDEPENDENT REGRESSION VARIABLE , CONSTANT COEFFICIENT t-VALUE R 2 F RATIO Sx.y D.W.S. LANCAR 122.33 -0.08 -7.12* 0.37* 50.78 10.43 0.66 t -LANCAR^ 122.46 -0.08 -7.11 0.37* 50.55 10.47 0.62 LANCARt_2 122.71 -0.08 -7.07* 0.37* 49.98 10.53 0.61 LANCAR 121.74 -0.07 -6.27* 0.32* 39.32 11.01 0.61 LANCAR,. . 122.11 -0.07 -6.38* 0.33* 40.76 10.99 0.47 t-4 LANCARt_5 121.13 -0.07 -5.67* 0.28* 32.17 11.44 0.50 LANCAR 120.54 -0.06 -5.23* 0.25* 27.36 11.69 0.39 t-6 •Significant at .01 level TABLE XVI RESULTS: TIME INDEXfc REGRESSED ON LAUNCHED CARGO VESSELS INDEPENDENT VARIABLE CONSTANT REGRESSION COEFFICIENT t-VALUE R2 F RATIO Sx.y D.W.S. LANCARt 135.05 -0.122 -8.15* 0.43* 66.44 13.91 0.68 LANCARt _ 134.63 -0.119 -7.70* 0.41* 59.31 14.27 0.74 t-1 LANCARt_2 135.80 -0.123 -8.13* 0.44* 66.23 13.94 0.64 LANCARt_3 135.63 -0.122 -7.91* 0.43* 62.67 14.15 0.68 LANCARt_4 135.39 -0.120 -7.56* 0.41* 57.21 14.47 0.66 LANCAR,. - 135.04 -0.117 -7.27* 0.39* 52.85 14.74 0.65 t—D LANCARt_6 135.21 -0.117 -7.25* 0.39* 52.65 14.78 0.56 •Significant at .01 level 123 It can be inferred from the results that the hypothesis i s supported with regards to an inverse relationship between the two variables. The data, however, offer; no support to the assumption of a time-lag situation unless one were to construe the very slight movement in this respect exhibit-ed by a two-month lag of the launchings on the time index as being s t a t i s t i c a l l y valid, which i s doubtful. Hypothesis 2 i s rejected immediately without further testing because of the inverse relationship shown in the previous tests which would be identical i f tests were performed with the depend-ent and independent variables reversed. Conclusions drawn on the basis of these results would again be indecisive because of the presence of autocorrelation. CHARTER RATES AND SHIP COMPLETIONS Ship completions constitute actual additions to the supply of shipping. At this stage the vessels are now available for actual service in the transporting of com-modities. An argument frequently advanced i s that a f a l l in freights i s due to an increase in total tonnage, or over-17 production. In spite of the fact that, as has been 1 7 Jones, op. c i t . p.37 124 mentioned, the absolute monthly increases are very small compared to the total size of the existing fleet at the time,^8 this does not necessarily mean this i s not a valid argument, but i t i s assumed that the impact would be quite small. It would also be expected, assuming an inverse relationship exists, that there would be a lag of some time period between the level of ship completions and that of rates. From this argument i t can be hypothesised: "The level of charter rates w i l l be inversely related to the completion of cargo tonnage with a strengthening of this relationship revealed as the completion of vessels precedes the level of rates." The regression results of the voyage and time index as a function of the completion of cargo vessels are exhibited in Tables XVII and XVIII. The results offer support to the hypothesis regarding the negative relationship between the two indices and completions, but to accept i t on the basis of a lag relationship as stated i s questionable. It For 1967, according to our data, the highest monthly total for the completion of a l l dry cargo vessels was approximately one-quarter of one per cent of the total merchant fleet (over 1000 G.R.T.) on June 30 of that year. (Shipping S t a t i s t i c s . Bremen: Institute of Shipping Economics 1968, p. II-4). TABLE XVII REGRESSED ON COMPLETED CARGO VESSELS INDEPENDENT REGRESSION VARIABLE CONSTANT COEFFICIENT t-VALUE R F RATIO Sx.y D.W.S. T & CCARt 133.06 -0.1258 -8.24* 0.44* 6 7 . 9 9 1 3 . 8 4 0.66 T & CCAR 133.55 -0.1275 -8.38* 0.45* 70.36 13.75 0.69 U« *™ JL. T & CCARt_2 133.37 -0.1252 -8.09* 0.44* 65.56 13.98 0.63 T & CCARt_3 132.64 -0.1194 -7.36* 0.39* 54.23 14.59 0.60 T & CCAR A 131.67 -0.1131 ^6.71* 0.35* 45.12 15.85 0.63 t-4 T & CCAR. _ 132.44 -0.1169 -7.03* 0.38* 49.43 14.93 0.57 t — D T & CCAR. - 133.14 -0.1199 -7.29* 0.40* 53.27 14.74 0.59 •Significant at .01 level TABLE XVIII INDEPENDENT VARIABLE RESULTS: VOYAGE INDEX* REGRESSED ON COMPLETED CARGO VESSELS CONSTANT REGRESSION COEFFICIENT t-VALUE R F RATIO Sx.y D.W.S, T & CCAR T & CCAR t-1 T & CCAR t-2 T & CCAR t-3 T & CCAR t-4 T & CCAR t-5 T & CCAR t-6 120.69 120.93 120.68 120.20 119.16 119.00 119.55 -0.080 -0.081 -0.079 -0.075 -0.069 -0.067 ^0.069 -6.91* 0.35* 47.65 10.55 0.60 -6.98* 0.36* 48.73 10.55 0.57 -6.67* 0.34* 44.50 10.76 0.54 -6.11* 0.31* 37.45 11.10 0.47 -5.35* 0.26* 28.70 11.58 0.48 -5.15* 0.24* 26.54 11.74 0.46 -5.38* 0.26* 28.96 11.60 0.44 •Significant at .01 level 127 cannot be rejected because the data does indicate the relationship appears to move in the direction as hypo-thesised, particularly i n the results encompassing the voyage index. The s t a t i s t i c a l v a l i d i t y i s in doubt however since the problem of autocorrelation i s in evidence. 4. RELATIONSHIP BETWEEN WORLD SEA  TRADE AND CHARTER RATES Chapter II referred to the demand for shipping as being represented by the volume of goods to be carried and the distances over which they are carried. As such, expectations would be that the volume of seaborne trade would be closely related to charter rates. Granted, there i s a b i t of distortion introduced into this relationship, because, as has been previously mentioned, over half of the world seaborne trade i s accounted for by the carriage of o i l . Consequently, during an investigation of charter rates as a function of world sea trade, cognizance must be taken of the fact that data used in this thesis include this com-ponent. However, i t i s assumed here that international trade i n o i l i s not generally characterized by the seasonal fluctuations which are found in dry cargo trade, and changes 128 in demand for tanker capacity i s generated not by increases in the volume of o i l to be transported, but rather by de-mands for longer hauls and, in turn, to the supply of tanker ships. Thus the relationship should be relatively the same with the volume of o i l trade in or out of the data. An hypothesis can therefore be cast: 1. "The level of charter rates w i l l bear a positive relationship to the volume of world seaborne trade." Also, given what i s known about the supply curve for tramp shipping in the short run, i.e., i t i s relatively inelastic, and i f competition i s as ubiquitous in the tramp market as has been maintained, then another assumption would be that any change in the demand for shipping space due to an increased volume of trade would be reflected almost instantaneously i n charter rates. Following this line of reasoning, another hypo-thesis can be indited: 2. "The closest relationships between charter rates and the volume of seaborne trade w i l l be found when using concurrent observations of the data." The results of Tables XIX and XX bear out the assumption of a positive relationship existing between TABLE XIX INDEPENDENT VARIABLE CONSTANT RESULTS: VOYAGE INDEX REGRESSED ON WORLD SEA TRADE REGRESSION COEFFICIENT t-VALUE F RATIO Sx.y D.W.S, WORSEA. WORSEA. t-1 WORSEA t-2 WORSEA t-3 WORSEA t-4 WORSEA t-5 WORSEA t-6 77.06 78.31 79.41 79.54 78.81 75.69 75.68 0.2556 0.2461 0.2376 0.2376 0.2446 0.2729 0.2757 4.82* 0.21* 23.30 11.68 0.23 4.53* 0.19* 20.58 11.88 0.28 4.30* 0.18* 18.51 12.05 0.28 4.25* 0.18* 18.12 12.13 0.26 4.31* 0.18* 18.58 12.16 0.29 4.75* 0.22* 22.58 11.96 0.23 4.63* 0.21* 21.43 12.03 0.26 •Significant at .01 level TABLE XX INDEPENDENT VARIABLE. . RESULTS: TIME INDEX. REGRESSED ON WORLD SEA TRADE REGRESSION CONSTANT COEFFICIENT t-VALUE R^  F RATIO Sx.y D.W.S, WORSEA WORSEA t-1 WORSEA t-2 WORSEA t-3 WORSEA t-4 WORSEA. t-5 WORSEA t-6 55.17 57.22 60.46 59.14 58.38 55.06 53.45 0.4796 0.4648 0.4399 0.4525 0.4611 0.4922 0.5096 7.24* 0.38* 52.47 14.60 0.33 6.80* 0.35* 46.33 14.96 0.32 6.21* 0.31* 38.67 15.44 0.37 6.40* 0.33* 41.01 15.35 0.29 6.41* 0.33* 41.17 15.39 0.35 6.77* 0.36* 45.87 15.14 0.31 6.79* 0.36* 46.10 15.16 0.32 •Significant at .01 level 131 the two variables and hypothesis 1 i s accepted in this respect (with reservations, however, because of the evidence of autocorrelation). Hypothesis 2, i s not rejected com-pletely only because the data appears to show very l i t t l e variation over the period but i t would have to be said that the results are inconclusive. In addition, autocorrelation i s again ever present in the residuals. It was also recognized that the demand for the carriage of seaborne trade i s reputed to dispify distinct seasonal movements. To see i f the " f i t " of the data could be improved somewhat, the world sea trade data were season-a l l y adjusted and the same series of regressions were run as a further test of the o r i g i n i hypothesis. For the sake of exposition, Table XXI presents the results of the voyage index regressed on seasonally adjusted world sea trade. As can be seen, the results offer no further support or rejection of the hypotheses than has already been estab-lished. SUMMARY This chapter has examined, by means of graphs, TABLE XXI INDEPENDENT VARIABLE** RESULTS: VOYAGE INDEX REGRESSED ON WORLD SEA TRADE CONSTANT REGRESSION COEFFICIENT t-VALUE R' F RATIO Sx.y D.W.S, WORSEA, WORSEA t-1 WORSEA t-2 WORSEA t-3 WORSEA t-4 WORSEA t-5 WORSEA. t-6 77.36 78.90 79.31 79.49 79.26 76.62 75.35 0.2529 0.2408 0.2384 0.2379 0.2409 0.2647 0.2784 4.79* 0.21* 23.02 11.69 0.24 4.45* 0.19* 19.86 11.92 0.29 4.33* 0.18* 18.83 12.04 0.28 4.28* 0.18* 18.36 12.11 0.26 4.26* 0.18* 18.22 12.18 0.29 4.63* 0.21* 21.49 12.03 0.25 4.68* 0.21* 21.91 12.00 0.27 •Significant at .01 level ** Seasonally adjusted 133 figures and s t a t i s t i c a l tests, several relationships which were hypothesised to exist between charter rates and various variables of supply and demand in the tramp shipping indust-ry. The data appeared to support or reject several of the hypotheses put forth; however, i t was impossible to draw any definite conclusions concerning any relationships because in a large majority of the tests performed, i t was obvious that the problem of autocorrelation i n the residuals was present. From the discussion presented at the begin-ning of this chapter, i t was apparent the results could not be accepted at face value and that some further manipulation of the dataware necessary i f there were to be any meaning-ful results obtained. It i s this problem of autocorre-lation and the attempt to eliminate i t from the data which is the area of concern in the following chapter. CHAPTER V EXAMINATION OF THE CHANGE IN CHARTER RATES WITH THE CHANGE IN SELECTED VARIABLES INTRODUCTION The problem of autocorrelation in time series analysis proved to be very prevalent i n the s t a t i s t i c a l testing performed and reported i n Chapter IV. However, i t has been mentioned that the method of least squares s t i l l produces the best linear unbiased estimates of the regression coefficients provided account i s taken of the lack of independence in the error series (i.e., residuals). There are several methods of dealing with this phenomenon of autocorrelation, one of which i s to examine the month-to-month changes in the data rather than the absolute monthly values. 1 This method - albeit a crude one - i s 1The common sense meaning of a regression between f i r s t differences i s as follows: If (say) high original values of price are associated with low original values of consumption and vice versa, i t follows logically that a change from a high price to a low price w i l l be associated with a change from a low consumption to a high consumption figure. In s t r i c t l y random samples, linear regression co-efficients estimated from f i r s t differences w i l l closely approximate those obtained from the original values. (See: 135 commonly referred to as " f i r s t differences". USE AND METHOD OF FIRST DIFFERENCES OF DATA The justification for using a f i r s t difference transformation of data i s expressed by Fisser: "The primary device for the estimation of short run reactions i s the use of f i r s t differences of the data. Aside from the fact that such use often (but by no means always) has the conven-ience of reducing or eliminating autocorrelation in the residuals, or reducing multicollinearity, i t seems analytically the correct form for the estimation of short-run functions." 2 In this thesis, the primary concern i s with short-run reactions, and thus the method seemed the proper one to use. Transforming the observation on the variables to f i r s t differences i s accomplished by the following: Y ± * Y| - Y• - Y ^ x X j i * X$i m - X ^ U - D t for j = 1,2 ,K. (See: Mordecai Ezekiel and Karl A. Fox, Methods of Correlation and Regression Analysis. New York: John Wiley and Sons, Inc., 1959. p. 342) F.M. Fisher. A Pri o r i Information and Time Series  Analysis, r Amsterdam: North-Holland" Publishing Co. 1962, pp. 22-23. 136 Without delving into the intricacies of mathematical derivation, i t w i l l suffice to say here that under this method, the autocorrelation of the disturbance term of the f i r s t difference equation tends to zero and i t i s thus held that the transformation to f i r s t differences may therefore serve to deal with the worst cases of auto-3 correlated disturbances. REGRESSION RESULTS USING FIRST DIFFERENCES The foregoing method was applied to a l l data and the transformed variables were then subjected to a regres-sion analysis, using the same format of tests as used i n the series of regressions run previously, to discover i f : 1. the method used would be successful in removing the autocorrelation i n the residuals, and 2. the results of the tests would support or reject the hypothesis as set out in Chapter IV. To avoid confusion and repetition of figures. Tables XXII - XXV present partial comparative results between one 3 J.S. Cramer. Empirical Econometrics, Amsterdam: North Holland Publishing Company, 1969, p.89. TABLE XXII A TIME INDEX. REGRESSED ON ^ VOYAGE INDEX TIME INDEXt REGRESSED ON VOYAGE INDEX (RE: TABLE V) INDEPENDENT REGRESSION INDEPENDENT REGRESSION VARIABLE AVOYIND, AVOYIND t-1 A VOYIND t-2 AVOYIND t-3 AVOYIND t-4 AVOYIND t-5 AVOYIND t-6 COEFFICIENT t-VALUE R D.W.S. VARIABLE COEFFICIENT t-VALUE R D.W.S. 0.5825 0.3616 0.0081 -0.1099 0.1391 0.1271 0.1303 4.60 0.20 2.66 VOYIND 2.63 0.07 2.45 VOYIND t-1 0.05 0.00 2.17 VOYIND t-2 -0.74 0.006 2.15 VOYIND t-3 0.93 0.01 2.17 VOYIND t-4 0.84 0JD09 2.21 VOYIND t-5 0.66 OJ0O9 2.18 VOYIND t-6 1.26 1.27 1.21 1.16 1.12 1.06 0.97 18.60 0.80 0.86 19.46 0.81 1.23 15.87 0.75 1.19 13.42 0.68 0.97 12.01 0.64 0.67 10.35 0.57 0.54 8.63 0.48 0.42 TABLE XXIII RESULTS: A LA I D UP CARGO, RESULTS: LAID UP CARGO REGRESSED ON -A TIME INDEX INDEPENDENT REGRESSION VARIABLE REGRESSED ON TIME INDEX (RE: TABLE X). INDEPENDENT REGRESSION A TIMIND+ A TIMIND t-1 A TIMIND t-2 A TIMIND t-3 A TIMIND t-4 A TIMIND t-5 A TIMIND TIMIND t-6 COEFFICIENT t-VALUE R D.W.S. VARIABLE 1.34 0.27 0.0009 2.38 -6.03 -1.25 0.0187 2.45 -4.93 -1.02 0.0126 2.48 -1.66 -0.34 0.0014 2.46 -7.49 -1.54 0.0291 2.16 0.32 0.07 0.0001 1.05 -3.19 -1.13 0.0162 1.16 COEFFICIENT t-VALUE R D.W.S, -31.86 TIMIND+ , -33.56 t—1 TIMINDt_2 -34.48 TIMIND t_ -34.69 TIMINDfc-4 -34.69 TIMINDt_5 -33.55 TIMIND^ r -31.90 t-6 -8.17 0.44 0.37 -9.52 0.51 0.40 —10.57 0.57 0.49 -11.09 0.60 0.59 -11.42 0.61 0.55 -10.74 0.59 0.52 -11.13 0.61 0.33 CJ. C O . TABLE XXIV RESULTS: A VOYAGE INDEX^ REGRESSED ON COMPLETED CARGO VESSELS RESULTS: VOYAGE INDEX INDEPENDENT REGRESSION VARIABLE T&CCARt A T&CCARt_1 A. T&CCAR n t-2 A T&CCARt_3 A T&CCARt_4 A T&CCAR, c t-5 AT&CCAR^ . t-6 COEFFICIENT t-VALUE R D.W.S. -0.0007 -0.0089 -0.0040 -0.0045 0.0101 0.0105 0.0008 -0.0743 0.0001 1.99 -0.0042 0.0028 2.09 -0.4604 0.0026 2.09 -0.4973 0.0030 2.07 1.1163 0.0153 2.11 1.1729 0.0171 2.16 0.0829 0.0001 2.11 REGRESSED ON COMPLETED CARGO VESSELS (RE: TABLE XVIII) INDEPENDENT REGRESSION VARIABLE COEFFICIENT t-VALUE R D.W.S. T&CCARt T&CCARt^ T&CCARt_.2 T&CCAR. , t-3 T&CCARt-4 T&CCARt_5 T&CCARt_6 -0.080 -0.081 -0.079 -0.075 -0.069 -0.067 -0.069 -6.91 0.35 0.60 -6.98 0.36 0.57 -6.67 0.34 0.54 -6.11 0.31 0.47 -5.35 0.26 0.48 -5.15 0.24 0.46 -5.38 0.26 0.44 H Co VO TABLE XXV RESULTS: A TIME INDEX, RESULTS: TIME INDEX REGRESSED ON A WORLD SEA TRADE REGRESSED ON WORLD SEA TRADE (RE: TABLE XX) INDEPENDENT REGRESSION INDEPENDENT REGRESSION VARIABLE COEFFICIENT t-VALUE R D.W.S. VARIABLE COEFFICIENT t-VALUE R D.W.S, A WORSEA. A WORSEA t-1 A WORSEA t-2 A WORSEA. t-3 A WORSEAt_4 A WORSEAt_5 A WORSEA _ t-6 0.0622 0.7484 0.0066 2.18 -0.0164 -0.1963 0.0005 2.16 -0.0065 -0.0767 0.0001 2.17 -0.0095 -0.1121 0.0002 2.17 0.0208 0.2429 0.0007 2.17 -0.0452 -0.5261 0.0035 2.12 0.0602 0.7138 0.0065 2.24 WORSEAt 0.4796 WORSEAt_1 0.4648 WORSEAt__2 0.4399 WORSEA^  3 0.4525 WORSEA . 0.4611 t-4 WORSEA, r 0.4922 t—5 WORSEA,. . 0.5096 t-6 7.24 0.38 0.33 6.80 0.35 0.32 6.21 0.31 0.37 6.40 0.33 0.29 6.41 0.33 0.35 6.77 0.36 0.31 6.79 0.36 0.32 H O 141 selected set of regressions out of each category outlined and examined i n the previous chapter. Several points can be commented upon from these results, which are typical of those obtained from a l l the regressions. F i r s t , referring to the primary reason for using f i r s t differences, in a l l cases the positive autocorrelation i n the residuals has been eliminated. However, notice should be given to Table XXII concerning two relationships: (1) Time Indext = f (Voyage Index) and (2) Time Index t = f (Voyage Index) t_-j_« The results reveal that the former i s negatively autocorrelated while the latter i s i n the area of uncertainty as outlined by the significant limits of the Durbin and Watson Test. This of course indicates that no conclusions could be based on these results, which are the only ones that approach being statistically signif-icant i n the four tables. Secondly, i t i s d i f f i c u l t to explain the reason for the changing of signs i n the regression coefficients. However, this may be due i n total to the apparent poor f i t of the data using first-difference transformations. 142 Finally, on the basis of the lack of s t a t i s t i c a l significance in practically a l l of the results, i t would be impossible to draw any conclusions from the results to support or reject any of the hypotheses put forth i n the previous chapter. It would also appear to negate any conclusions concerning various relationships that had been drawn on the basis of results obtained from absolute levels of the data. The poor results of the foregoing tests with f i r s t differences of the data, prompted a further investigation. At this point the transformed data ( f i r s t differences) were plotted to see i f anything other than a linear relationship was revealed. Figures V and VI show that this was not the case. FURTHER TESTING OF DATA Several manipulations of the data were considered i n some detail : 1. Monthly data i s often subject to seasonal variation, and this was removed in cases where i t was thought, on general grounds, that i t might exist. The program used was incorporated into the computer regression package. In effect, the program f i t s a moving average LAID UP TONNAGE ( 000'S TONS ) a. /»»"oo £os I OOOO IS a G> . o <6 0 G> - i . -2 /6 VOYAGE INDEX FIGURE V SCATTER DIAGRAM OF LAID UP CARGO TONNAGE AND VOYAGE INDEX (FIRST DIFFERENCES) 1960 - 1967 LAID UP TONNAGE ( 000*s TONS ) <9 0 © - 9 - S"oo o •CD 0 - o . / o o £ o ^ -/r 4^  TIME INDEX FIGURE VI SCATTER DIAGRAM OF LAID UP CARGO TONNAGE AND TIME INDEX (FIRST DIFFERENCES) 1960 - 1967 145 to the series, works out the ratio of the actual to the average and constructs a factor for each month which i s divided into the actual to give the deseason-alised figure. 2. Logarithmic transformations were used on some of the series for the usual reasons (e.g., that price changes are relative, or that the combined efforts of different variables are multiplicative. 3. Regressions were completed on quarterly averages of the data to eliminate some of the extremely short-run fluctuations. 4. Some of the data were tested with absolute levels of some variables with f i r s t differences on the others. It was thought that some of the variables may have very l i t t l e change from period to period, and that i t may show a better relationship i f some of the absolute levels were regressed with changes. 5. The number of observations was shortened to eliminate the extreme fluctuations encountered i n rates during the Suez c r i s i s of 1967. The data was thus reduced to 87 observations from the original 108. 6. The linear trend was also removed from some of the 146 data i f i t seemed that the series would contain i t . However, working with f i r s t differences on the data does primarily the same job of removing this trend; consequently this method was used to a greater extent. In each of the aforementioned cases, exhaustive testing was completed using many different combinations of variables, incorporating both leads and lags of the dependent and independent variables. On the whole, approximately 3000 such models were tested. Appendix D contains a l i s t of some of the models that were tested and the results obtained. Several of those tested produced reasonably significant results upon using absolute levels of the data. Unfortunately the same recurring problem of high autocorrelation i n the residuals was encountered i n every model tested. Moreover, when regressions were run on the same models using f i r s t differences of the data, the effects of autocorrelation in the residuals were eliminated (with the exception of those which became negatively autocorrelated) but at the expense of any s t a t i s t i c a l significance at any acceptable level of confidence. 147 Before presenting a chapter on a summary and conclusions which have been drawn from the empirical investigation performed here, a brief chapter i s presented to reflect on some of the studies which have attempted to do similar work i n the area of shipping and to discover what conclusions have been drawn from these works. This follows immediately i n Chapter VI. CHAPTER VI COMPARISON OF RESULTS WITH OTHER EMPIRICAL STUDIES ON THE SHIPPING INDUSTRY INTRODUCTION In the light of the results of empirical testing reported i n Chapter IV, i t seemed pertinent to reflect on the results of some of the other work that had been performed i n the area. It i s not the intention here to dwell on the defects of any study inparticular as most of them would be acknowledged by the authors themselves. Indeed, the studies quoted may be entirely without defects at a l l . But i t would be desirable to compare some of the results obtained by other authors to those obtained here, and possibly suggest some reasons for any disparity between them. CORRELATION ANALYSIS AS PERFORMED BY LESLIE JONES In his book Shipbuilding i n Britain, Leslie Jones devotes a section 1 to a brief correlation analysis between 1 Jones, o£. c i t . pp. 37 - 40. 149 freight rate indices and indices of shipbuilding activity (i.e. index of tonnage commenced, index of tonnage under construction and index of tonnage launched). In one particu-lar table (reproduced i n Table XXVI), the author compares (by correlation coefficients) the index of freight rates and tonnage launched over the period 1920 - 1939 and concludes from the results that because of sharp changes i n freight rates: "... there i s an almost immediate effect on tonnage launched...simply because there i s no pressure for delivery. A f a l l i n freights has an immediate effect on launchings as well as a deferred effect due to the fa l l i n g off i n tonnage commenced."2 The author also concludes that the results of the table indicate that tonnage launched has no effect on freight rates because of the lack of any high correlation for some lag that shows tonnage launched preceding freights. Although Jones' data i s an index of ship launchings rather than absolute levels as used i n this thesis, and, i n addition, i s based on quarterly data rather than monthly as used here, i t may not be an ideal methodology to compare 2 Jones, op_. c i t . p. 38. 150 TABLE XXVI Correlation between Index of Freight Rates and Tonnage Launched 1920 - 1939 Lags i n Quarters 0 1 2 3 4 5 6 Freights preceding Launchings 0.51 0.53 0.52 0.51 0.36 0.32 0.11 Launchings preceding Freights 0.51 0.48 0.47 0.40 0.13 0.08 0.05 Source: Jones, o£. c i t . p. 39. the two results. However, any positive or negative relation-ship would be revealed using comparable techniques regard-less of this difference. With reference to Tables XV and XVI on pages 123 and 124 of this thesis, quite a strong negative relationship between launched cargo ships and charter rates i s observed, rather than a positive relationship as revealed by Jones' results. This might be due to significant differences between the two time periods. Jones was observing data 151 between the years 1919 and 1939, and i t i s not inconceivable to suggest that the effects of increased technology and consequent reduction i n the time required to build a vessel, could very well have altered the effects of this variable. This writer does feel however, that correlation analysis alone i s a very unsophisticated method of research for drawing any conclusions about the effects of one variable on the other. This statement does not purport to oast any dispersions on the validity of Jones' work per se, but, as has been observed i n the testing performed here, i t does point out the inherent limitation of correlation analysis of time series. For, as shown i n Chapter IV, i f the relation-ships as shown by correlation analysis of time series are not put to further testing, and attention paid to autocor-relation i n the residuals, any conclusions drawn on i n i t i a l correlation tests could be very misleading. REGRESSION ANALYSIS AS PERFORMED BY THOMAS BATES Another study which enforced interest i n the nature of tramp shipping rates was a regression analysis of ocean 3 tramp rates as performed by Thomas Bates. The purpose of 3 Thomas H. Bates. A Linear Regression Analysis of  Ocean Tramp Rates. Transportation Research. Vol. 3 No. 3. Oxford: Pergamon Press, September, 1969, pp. 377-391. 152 Bates' study was* through a multivariate regression analysis to better understand the nature of ocean tramp shipping rates and to discover how these rates are influenced by various important factors such as distance of haul, volume of ship-ment, trade route, season, year (i.e., the forces of demand that existed i n a given year) and the terms upon which a rate i s quoted. The analysis done by Bates i s not s t r i c t l y analagous to the one done i n this paper due to the fact that Bates concentrated on one commodity and confined his observations to two separate years - 1959 and 1963. His work was on a method of cross-sectional data rather than time series as used i n this study. Consequently, he couldn't have encounter-ed problems of autocorrelation i n the error term, since these arise with time series only. Moreover, his analysis made no reference to the s t a t i s t i c a l phenomenon so i t can be assumed he did not. However, his conclusions did enforce one of the major assumptions mentioned i n the introduction to this thesis - that exogenous forces (p o l i t i c a l and economic) were one of the prime contributors to tramp freight rate fluctuations. 153 ANALYSIS OF TANKERSHIP RATES AS PERFORMED BY Z.S.ZANNETOS We have mentioned that the tanker market i s ine x t r i -cably associated with the tramp shipping market and, as such, i t was f e l t that the book by Zannetos, The Theory of O i l  Tankership Rates, would include some empirical tests whose results could be compared with those performed on the data for tramp shipping rates used i n this study. Zannetos went so far as to construct, by means of correlation and multiple regression analysis, a model for the formulation of long term tanker rates i n the short run. He considered his results to be very satisfactory and concluded: "It i s shown that under periods of low rates the most significant factors affecting the level of long term rates under depressed market conditions are the size of the vessel (negative impact), the changes i n idle capacity (negative), and the level of outstanding orders for new vessels (positive). Under periods of high rates, the most important factors shown by the regression model to affect long term rates are the outstanding orders for new vessels (positive relation), the size of the chartered vessel (negative), the level of spot rates (positive), the duration of the charter (negative), and the load time from contract to delivery (negative). The quantitative significance of these factors as well as the ranges and extent of their applicability were intensively analyzed." 4 4 Z.S. Zannetos, The Theory of O i l Tankership Rates, Cambridge: The M.I.T. Press, 1966, p. 244. 154 Considering the fact that Zannetos was working with a time series using 1,048 observations over the period 1950 5 - 1957, i t would be interesting to discover i f he had considered the problems of autocorrelation as discovered i n our empirical testing i n this paper. If he had not, the inclination would be to look at the significance of his results and his predictive models with some apprehension. SUMMARY As we mentioned at the beginning of this chapter, i t i s not the intention here to disparage the results of the aforementioned studies. The primary purpose for mentioning them was simply to point out, and possibly justify somewhat, the reasons that such poor results were obtained by the analysis i n this paper as compared to the results obtained by other authors i n the area. In the empirical work performed on the data i n this study, the problem of autocorrelation was a highly significant factor, and the validi t y of the regression results was eliminated by using a method to cope 5 Ibid, p. 219. with this problem. It would be f a i r to say that results obtained without consideration of the autocorrelation i n the error terms would, i n the opinion of this author, be highly suspect. CHAPTER VII SUMMARY The attempt to investigate the relationships between supply and demand variables and freight rates i n the shipping industry i s not an original undertaking. This investigator did feel however, that the area of tramp shipping per se had not received any great consideration "empirically. The results of such an investigation, i f positive, would hope-fu l l y be of benefit to management responsible for the charter-ing of tramp vessels, insofar as they may be able to better understand forces which affect freight rates and possibly assess the future behaviour of voyage and time charter rates. On the other hand, i t i s an interesting empirical question i n i t s e l f to see i f the behaviour of tramp rates can be explained -and economic hypothesis about shipping can be verified or rejected. By means of rigorous empirical testing of various data i n t r i n s i c to tramp shipping, the author attempted to accomplish just this. On the basis of i n i t i a l tests on the absolute levels of the data, i t was f e l t that some progress was made. However, f i r s t tests revealed a high incidence 157 of positive autocorrelation i n the residuals which i s common to most time series data and i t was found necessary to use a transformation of the data to f i r s t differences i n an effort to establish s t a t i s t i c a l l y valid results i n a l l respects. It was very surprising to the author to find the very great disparity between results obtained through the use of absolute levels of the data and those obtained through the use of f i r s t differences. It was anticipated that results would be somewhat poorer, for as Fisser states: " . . . f i r s t difference regressions often yield correla-tion and significance apparently a good deal lower than the corresponding figures reported when other methods of trend removal are used (or when no trend adjustment i s made)j however, this i s largely the reward of honesty." But i t was not expected that the results would lose their s t a t i s t i c a l significance entirely. The economic hypotheses tendered i n Chapter IV were based on what seem to be valid arguments and i t i s most revealing to find that they could not be supported s t a t i s t i c a l l y by our data. 1 F.M. Fisher, op_. c i t . pp. 25-26. 158 Assuming that a l l of the conditions of regression analysis have been met i n the method of estimation used i n this study, the original hypothesis as set out at the begin-ning of this thesis must be rejected because of the lack of significant results. SUGGESTIONS FOR FURTHER RESEARCH The author feels that there must be some explanation for the complete lack of s t a t i s t i c a l significance encountered i n the empirical testing of our data and the following are put forth for consideration and possible suggestions for further research i n the area. F i r s t , and the most obvious observation that could be made i s that improper variables of supply and demand were used i n the tests performed. This does not necessarily mean that the variables used are i n no way related to the fluctua-tions i n freight rates but rather that they may be completely interdependent on some other motivating force for which no quantitative data were available. Among these may be the following: 159 (a) the cost of shipbuilding (b) the rate of technological obsolescence and the age distribution of the existing vessels. Technology and age determine remaining economic l i f e , consequently, they reflect retirements and replacements of vessels. (c) monthly categorization of size groups i n the fleet that i s available for l i f t i n g . (d) a more refined delineation of world trade with regards to types of commodities carried and the time periods during which they are carried. Included here also would be the volume and length of haul of these commodities. (e) the pattern of ownership within the industry as coloured by the existing institutional consid-erations . Category (e) above creates some insurmountable d i f f i c u l t i e s i n data collecting. The increase i n long-term chartering and private ownership of vessels has been discussed previous-l y , but the segregation of these types of vessels from the available published figures of vessels ordered, launched 160 or completed i s an impossible task. This problem causes a distortion i n the data which i n turn i s reflected i n the significance of the regression equations insofar as these vessels should have very l i t t l e direct influence on the short term fluctuations of tramp shipping rates. Indirectly, they do affect the amount of l i f t i n g s available for tramp ships but their short-run effects would be minimal. Secondly, and recalling the discussion i n Chapter III concerning the nature of competition i n the tramp shipping industry, i t i s possible, allowing a l i b e r a l use of assump-tions, that the supply of and demand for shipping space i n response to changes i n freight rates, or even vice versa, possibly occur faster than we imagine and the implication of "perfect competition" i n the tramp shipping industry i s more fact than f i c t i o n . Under such a supposition, data would have to be gathered on a weekly, or even a daily basis to produce any meaningful results. This assumption might receive further substantiation from the fact that tests run on quarterly averages of the data (both absolute levels and f i r s t differences) provided results that were no more signif-icant (and i n cases less significant) than those provided 161 by the monthly data. A third point for consideration i s concerned with the use of f i r s t differences i n the regression tests. Referring again to Figures V and VI the practically zero slope of the regression line indicates that monthly changes or d i f f -erences i n the dependent and independent variables are complete-ly independent events. The relationship shown when absolute levels are used-merely reflects the fact that the dependent and independent variables move together over time; but as has been previously deduced, this i s merely the dependency of one month's observations on the previous month's observa-tions and not a deterministic relationship. Speculation could become endless however, and there would s t i l l be many questions l e f t unanswered about tramp shipping freights. This investigator feels that a small step has been made i n this thesis towards starting to answer some of these questions. At least, i f nothing else, this study has shown that "good" s t a t i s t i c a l results are not necessarily "valid" s t a t i s t i c a l results. 162 BIBLIOGRAPHICAL ENTRIES A. BOOKS Alexandersson, Gunnar and Goran Norstrom. World Ship- ping (An Economic Geography of Ports & Seaborne  Trade). New York: John Wiley & Sons, 1963. Bes, J . Bulk Carriers. London: Barker & Howard, 1965. Branch, Alan E. The Elements of Shipping. London: Chapman and Hall Ltd., 1964. Burley, Kevin. British shipping & Australia. Cambridge: Cambridge University Press, 1968. Calvin, H.C. and E.G. Stuart. Merchant Shipping Industry. New York: John Wiley & Sons Inc., 1925. Cramer, J.S. Empirical Econometrics. Amsterdam: North-Holland Publishing Company, 1969. Cufley, C.H.F. Ocean Freights & Chartering. London: Barker & Howard, 1962. Cufley, C.F.H. The Ideal Tramp for the 1970's. London: Barker & Howard, 1966. Escarpenter, Claudio. The Economics of International Ocean Transport. Madison: The University of Wisconson Press, 1965. Ezekiel, Mordecai and Karl A. Fox. Methods of Correla- tion and Regression Analysis. New York: John Wiley & Sons Inc., 1959. Fayle, C. Ernest. A Short History of the World's Shipping Industry. London: George Allen & Unwin Ltd., 1934. 163 Fisher, P.M. A P r i o r i Information and Time Series  Analysis. North-Holland Publishing Co." Amsterdam^ I962. . - - .. - ... - V " Fisser, F.M. Tramp Shipping. Bremen: Weltschiffahrt Archiv, 1957. Goss, R.O. Studies i n Maritime Economics. Cambridge: The University Press, 1956. Gripaios, Hector. Tramp Shipping. London: Thomas Nelson & Sons Ltd., 1959. Jones, Leslie M.A. shipbuilding i n Britain (mainly between the two world wars). Cardiff: University of Wales Press, 1957. Kane, Edward J . Economic sta t i s t i c s & Econometrics. New York: Harper & Row, Publishers, 1968. Lewis, W. Arthur. Overhead Costs. London: George Allen & Unwin Ltd., 1949. Metcalfe, J.V. The Principles of Ocean Transportation. New York: Simmons Boardman, 1959. O'Loughlin, Carleen. The Economics of Sea Transport. Oxford: Permagon Press, 1967. Sturmey, S.G. B r i t i s h Shipping and World Competition. London: The Athlone Press, 1962. Sturmey, S.G. on the Pricing of Tramp Ship Freight  Service. Bergen: Institute of Shipping Research, 1965. Thorburn, Thomas. Supply & Demand of Water Transport. Stockholm: FFI, 1960. Troxel, Emery. Economics of Transport. New York: Rinehart & Company Inc., 1955. 164 Zannetos, Zennon S. The Theory of O i l Tankership Rates. Cambridge Mass: The M.I.T. Press, 1966. B. PERIODICALS "Annual Freight Market Review," Fairplay Shipping Journal, (January 11, 1968). Bates, Thomas S. "A Linear Regression Analysis of Ocean Tramp Rates," Transportation Research, Vol. 3, No. 3, (September, 1969), Oxford: Permagon Press. Cochrane, D. and G.H. Orcutt. "Application of Least Squares Regression to Relationships Containing Autocorrelated Error Terms," Washington, D.C.: Journal of the American St a t i s t i c a l Association, (March, 1949). Cufley, C.F.H., F.I.C.S. "Some Implications of the Bulk Carrier Era," The Shipping World and Ship-builder , (June 16, 1966). Durbin, J . and G.S. Watson. "Testing for Serial Correlation i n Least Squares Regression II," Biometrika, (June 1951). "Factors Influencing the Size of World Fleets," The Shipping World & Shipbuilder, (November, 1969). Lord Geddes C.B.E., D.L. "Shipowners must make a Reasonable Profit; 1 The Shipping World and Ship- builder , (January, 1969). McCaffrey, F.D. "Winds of Change Sweeping Traditional Methods," Industrial Canada, (July, 1963). Studer, K.R. "The Evaluation of Long Term Charter Rates i n the Shipping Industry" U.B.C. Term Paper, Commerce 544, 1969. 165 The Shipping World and Shipbuilder, Various Year. "The Year i n the Freight Market',' The Syren & Shipping, (January 2, 1963). "Those Handy Tramps of the Future," The Shipping World  and Shipbuilder, (November 18, 1965). "World Ships on Order" Supplement To Fairplay Shipping  Journal, (August 22, 1968). C. REPORTS, MONOGRAPHS AND THESES British Shipping Assistance Act 1935. Nolan, D.J.M. "Bulk Carriers - Past, Present and Future", Mimeo London 1966. Organization for Economic Co-operation & Development, Maritime Transport, Paris: 1968. United Nations Conference on Trade and Development, No. TD/B/C.4/38 1968. United Nations Monthly Bulletin of Statistics, United Nations New York, 1960 - 1968. U.N. Secretariat - Year Book of International Trade Statistics, United Nations, New York, 1968. APPENDIX A 166 NOTES ON CONSTRUCTION OF FREIGHT INDICES UNITED KINGDOM CHAMBER OF SHIPPING_TRAMP  FREIGHT INDEX (VOYAGE CHARTER)1 Publication of this monthly index started i n 1921, and after an interruption during the Second World War, was resumed i n 1948. The present index i s the result of a revision i n 1961 i n which 1960 was adapted as the base year. The information supplied by Br i t i s h tramp owners showed that on more than 93 percent of the total freight on voyage was earned i n the carriage of coal, grain, sugar, ore, f e r t i l i s e r s , timber and sulphur, and these are the commodities used i n the calcula-tions . The c r i t e r i a adapted i n determining whether or not a particular trade should appear i n the calculations were: (1) that the freights earned by the United Kingdom tramps i n that trade i n 1960 was large enough to make the trade sufficiently important for inclusion and (2) that the reported fixture of 1 Chamber of shipping of The United Kingdom, Tramp  Freight Index, Revised, Base (1960 = 100) London: May 1961, P. 11. 167 ships i n that trade appear with reasonable regularity. For each of the commodities, the main movements of t r a f f i c are ascertained. In certain cases i n which the load-ing or discharging area covers a large range of ports, a particular route i s chosen by the Chamber for the purpose of measuring changes i n freight rates. The Chamber of Shipping index u t i l i z e s 27 routes. The method of calculation of the index i s as follows: 1. For each trade route, the arithmetic mean of the freight rates for fixtures reported each month i s calculated. The arithmetic mean for the year 1960 of these monthly averages forms the basis of comparison of freight rates i n that trade route. 2. To each commodity i s attached an index number which i s the weighted arithmetic mean of the price relatives for the trade routes i n which fixtures were reported i n the month. 3. The index number of tramp shipping freights i s the weighted arithmetic mean of the commodities for the month. 168 CHAMBER OF SHIPPING OF THE UNITED KINGDOM ~" ' " TIME CHARTER INDEX** "~ This index i s meant to supplement the voyage charter freight index; the f i r s t assessment was calculated by the Chamber of Shipping of the United Kingdom i n 1953 and revised i n 1961 using a base year as 1960. Detailed calculation of the index i s as follows: 1. Fixtures for steamers, for vessels of less than 9,000 deadweight tons, for periods of longer than about nine months, or where the ports of delivery and re-delivery are widely separated are not used i n the calculations. 2. The arithmetic mean of the rates for fixtures for motor ships reported i n each month i s calculated, and the arithmetic mean for the year 1960 of these monthly averages forms the basis of comparison. NORWEGIAN SHIPPING NEWS -- TANKER CARGO, . . .. TRIP CHARTER-3 This index i s based on the price relatives derived by 2 Ibid* 3 U.N. Monthly Bulletin of Statistics - various years. 169 comparing current fixtures, not with those reported i n a base period as i n the case of dry cargo, but with a precisely defined standard as set down by the International Tanker Nominal Freight scale Association Limited (Intascale). Fixtures i n the tanker trade are reported as Intascale minus or plus as the market may be at the time of fixing, and the index i s f i n a l l y calculated by taking the arithmetic averages of these pride relatives. 1 APPENDIX B 170 DURBIN-WATSON TEST FOR AUTOCORRELATION The Durbin Watson d-statistic i s used to test for autocorrelation i n least squares regression, usually when the "variables" involved i n the regression are actually time series. If the residual derived from mathematically f i t t e d regression i s Z, then the coefficient of autocorrelation i s calculated as follows: n I r a = i=l z i i-1 n £ -2 i=l Z i Where n i s the sample size and Z again a residual, the formula for calculating the Durbin Watson d-Statistic i s : n d = i-2 ( 2 i - Zi-1) n i=l i where Z. i s the i th residual. That i s , Z. i s the i th x x observed value of Y minus the i th predicted value of Y. 171 It i s not possible, according to Durbin and Watson, to obtain c r i t i c a l values of d exactly. However, upper and lower bounds for the c r i t i c a l values can be obtained. They are denoted by dy and d^ respectively, and depend on the significant level, the number of observations, and the number of independent variables. ONE SIDED TEST FOR POSITIVE AUTOCORRELATION If the calculated value of d i s less than d^ («< ), d i s significant at the o< % level, and i t i s concluded there exists significant evidence of positive autocorrelation. If the calculated value of d i s between d L (o< ), and djj ("<), the test i s inconclusive. If the calculated value of d i s greater than dy (=<), d i s not significant at the % level, and i t i s concluded there i s no significant evidence of positive autocorrelation. The c r i t i c a l points used i n this study are l i s t e d i n Table XXVII. 172 TABLE XXVII Five and One Per Cent Significance Points for d L and d^ 1 1 Independent Variable N Value of d T L Value P = iOl P = .05 P = .01 P = .05 80 1.47 1.61 1.52 1.66 85 1.48 1.62 1.53 1.67 90 1.50 1.63 1.54 1.68 95 1.51 1.64 1.55 1.69 100 ,1.52 1.65 1.56 1.69 N = Number of observations. ONE SIDED TEST FOR NEGATIVE AUTOCORRELATION The value of d i s significant at the *K % level i f i t i s greater than 4 - d ( o< ), insignificant i f i t i s J. Durbin and G.S. Watson, "Testing for Serial Correlation i n Least Squares Regression, Part II", Biometrika, 1951, pp. 173 - 175. 173 less than 4 - dg (<K), and the test i s inconclusive i f d i s between 4 - dy (e<) and 4 - d^ (o<). The c r i t i c a l points used i n this study are l i s t e d i n Table XXVIII. TABLE XXVIII Five and One Per Cent 2 Significance Points for 4-d^ and 4-dy 1 Independent Variable N Value of 4-dL Value of 4-dy P = .01 P = .05 P = .01 P = .05 80 2.53 2.39 2.48 2.34 85 2.52 2.38 2.47 2.33 90 2.50 2.37 2.46 2.32 95 2.49 2.36 2.45 2.31 100 2.48 2.35 2.44 2.31 N = Number of observations. TWO SIDED TEST FOR AUTOCORRELATION The value of d i s significant at the <^  % level i f i t 2 Calculated from Table XXVIII. 174 i s greater than 4 - d y ( << ) or less than cL^  ( «< ). The value of d i s insignificant at the o< % level i f i t i s between % ( ^  ) and 4 - d^ ( «K ). The test i s inconclusive otherwise. APPENDIX C 175 NOTES ON SOURCES OF STATISTICS 1. SHIPS; NEW ORDERS, LAUNCHINGS, AND  TRIALS AND COMPLETIONS These st a t i s t i c s were drawn on a monthly basis from the commercial journal: The Shipping World and Shipbuilder. Each category i s further broken into sub-categories and monthly totals of deadweight tonnage: a. Cargo. b. Tanker. c. Bulk Carriers and Specialized Vessels. Vessels of less than 1000 D.W. Tons were excluded. The reason for this exclusion was the fact that this thesis i s intended to represent primarily vessels which trade internationally and, although vessels of less than 1000 D.W.T. do trade internationally, for most practical purposes they are of a regional rather than international interest. In addition, a l l military, fishing and passenger vessels were also excluded based on the premise that these vessels had no bearing on the shipping of ocean freights (particularly with respect to tramp trades). 176 NEW ORDERS These figures refer to new contracts l e t to ship-building firms by various owners. There i s no indication whether building was commenced at the time of ordering. LAUNCHES These figures refer to the launching of the hu l l of the vessel but minus the superstructure. COMPLETIONS These figures constitute actual tonnage delivered to owners — t r i a l runs having been completed. 2. WORLD TRADE These st a t i s t i c s are based on monthly totals (in metric tons) of international seaborne trade as compiled i n the United Nations Monthly Bulletin of St a t i s t i c s . The monthly figures were adjusted slightly on a monthly percentage to overcome the disparity between monthly totals and aggregate yearly totals as l i s t e d i n the United Nations S t a t i s t i c a l  Year Book. 177 3. LAID UP TONNAGE (CARGO AND TANKER) These figures (in deadweight tons) refer to the total tonnage l a i d up monthly through lack of employment. Totals drawn from monthly issues of Fairplay International Shipping  Journal. 4. FREIGHT INDICES — VOYAGE. TIME AND TANKER See Appendix A /APPENDIX' D 178 FUNCTIONAL RELATIONSHIPS TESTED IN  REGRESSION ANALYSIS The following are some of the relationships tested but not reported i n the text of t h i s thesis: LAID UP CARGO AS A FUNCTION 0F» 1. World Sea Trade 2. Indices and World Sea Trade 3. La i d Up Tanker 4. Ship ( Launchings, Completions ) RATE INDICES AS A FUNCTION OFt 1. Ship Completions ( Cargo, Bulk C a r r i e r , Tanker ) 2. Ship Launchings ( Cargo, Bulk C a r r i e r , Tanker ) 3. World Sea Trade; L a i d Up Cargo 4. World Sea Trade; Laid Up Cargo? Ship ( Completions, Launchings, New Orders ) 5. Laid Up Cargo; Ship Completions 6. New Orders 7. Index; L a i d Up Cargo; World Sea Trade; Ship ( Completions, Launchings, New Orders ) 1 7 9 NEW ORDERS ( CARGO. TANKER. BULK CARRIER )  AS A FUNCTION QF> 1. Rate Indices 2. World Sea Trade Most of the tests were performed with a series of leads, lags, trend removal, seasonally adjusted ( some) and tests were on both original values and f i r s t differences. 

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.831.1-0102070/manifest

Comment

Related Items