UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Essays on cash holdings of corporations and mutual funds Simutin, Mikhail Vasilevich 2010

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

Item Metadata

Download

Media
24-ubc_2010_fall_simutin_mikhail.pdf [ 1.97MB ]
Metadata
JSON: 24-1.0071109.json
JSON-LD: 24-1.0071109-ld.json
RDF/XML (Pretty): 24-1.0071109-rdf.xml
RDF/JSON: 24-1.0071109-rdf.json
Turtle: 24-1.0071109-turtle.txt
N-Triples: 24-1.0071109-rdf-ntriples.txt
Original Record: 24-1.0071109-source.json
Full Text
24-1.0071109-fulltext.txt
Citation
24-1.0071109.ris

Full Text

Essays on Cash Holdings of Corporations and Mutual Funds by Mikhail Vasilevich Simutin B.A., University of Washington, 2004 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate Studies (Business Administration) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) July 2010 c© Mikhail Vasilevich Simutin 2010 Abstract In this thesis, I study the relationship between excess cash holdings of corporations and mutual funds and future performance of these entities. In the first chapter, I document a positive relationship between corporate excess cash holdings and future stock returns. The difference in returns of portfolios of high and low excess cash firms amounts to 5% annually or 6% after standard three-factor risk adjustment. Firms with more excess cash have higher market betas and earn lower returns during market downturns. High excess cash companies invest considerably more in the future than do their low cash peers, but do not experience stronger future profitability. On the whole, this evidence is consistent with the notion that excess cash holdings proxy for risky growth options. In the second chapter, I document a positive relationship between excess cash holdings of actively managed equity mutual funds and future fund performance. The difference in returns of portfolios of high and of low excess cash funds amounts to over 2% annually, or approximately 3% after standard risk adjustment. I study whether this difference in performance can be explained by the differences in managerial stock selection skills, market-timing abilities, fund liquidity needs, and operating costs. I show that managers of high excess cash funds make more profitable stock purchasing decisions, while low excess cash fund managers make better sell decisions. Neither high nor low excess cash groups exhibit significant market-timing skills; however, funds with volatile excess cash holdings are successful market timers. The difference in returns between high and low excess cash groups is particularly pronounced during periods of low fund flows, suggesting that high excess cash funds are better able to anticipate fund outflows. Finally, I show that high excess cash funds incur significantly lower operating expenses than do their low excess cash peers. I additionally document new important determinants of mutual fund cash balances, showing that funds with riskier or less liquid shareholdings, as well as those with higher return gap measures hold more cash. The determinants I consider jointly explain three times more cross-sectional variation in cash positions than variables studied in prior literature. ii Table of Contents Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Dedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1 Excess Cash and Stock Returns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Excess Cash and Mutual Fund Performance . . . . . . . . . . . . . . . . . . . . . . 3 1.3 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2 Excess Cash and Stock Returns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.2 Excess Cash Holdings: Estimation and Firm Characteristics . . . . . . . . . . . . . 10 2.2.1 Estimation of Excess Cash . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.2.2 Excess Cash and Firm Characteristics . . . . . . . . . . . . . . . . . . . . . . 11 2.3 Excess Cash Holdings and Stock Returns . . . . . . . . . . . . . . . . . . . . . . . . 12 2.3.1 Future Raw Returns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.3.2 Fama-MacBeth Regressions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.3.3 Risk-Adjusted Returns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.3.3.1 Time Series Characteristics . . . . . . . . . . . . . . . . . . . . . . 15 2.3.3.2 Unconditional Risk Adjustment . . . . . . . . . . . . . . . . . . . . 15 2.3.4 Excess Cash Holdings and Market State . . . . . . . . . . . . . . . . . . . . 16 2.4 Excess Cash, Investment, and Profitability . . . . . . . . . . . . . . . . . . . . . . . 17 2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.6 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 iii 3 Excess Cash and Mutual Fund Performance . . . . . . . . . . . . . . . . . . . . . . 34 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.2 Data and Summary Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 3.2.1 Data and Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 3.2.2 Summary Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 3.3 Determinants of Fund Cash Holdings . . . . . . . . . . . . . . . . . . . . . . . . . . 40 3.4 Excess Cash Holdings and Fund Performance . . . . . . . . . . . . . . . . . . . . . . 42 3.4.1 Excess Cash Estimation Methodology . . . . . . . . . . . . . . . . . . . . . . 42 3.4.2 Characteristics of Excess Cash Portfolios . . . . . . . . . . . . . . . . . . . . 43 3.4.3 Performance Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 3.4.3.1 Market Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 3.4.3.2 Fama-French Three-Factor Model . . . . . . . . . . . . . . . . . . . 44 3.4.3.3 Carhart Four-Factor Model . . . . . . . . . . . . . . . . . . . . . . 44 3.4.3.4 Multifactor Model with Liquidity Factors . . . . . . . . . . . . . . 44 3.4.3.5 Ferson-Schadt Conditional Model . . . . . . . . . . . . . . . . . . . 44 3.4.4 Future Fund Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 3.4.4.1 Raw Cash and Future Performance . . . . . . . . . . . . . . . . . . 45 3.4.4.2 Excess Cash and Future Performance . . . . . . . . . . . . . . . . . 46 3.4.4.3 Excess Cash and Future Performance: Fama-MacBeth Regressions 47 3.4.4.4 Excess Cash and Future Performance: Conditioning on Prior Market Return . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 3.5 A Model of Mutual Fund Cash Holdings . . . . . . . . . . . . . . . . . . . . . . . . 48 3.5.1 Model-Based Excess Cash and Future Fund Performance . . . . . . . . . . . 50 3.6 Sources of Relationship Between Excess Cash and Fund Performance . . . . . . . . 51 3.6.1 Excess Cash and Ability to Control Fund Expenses . . . . . . . . . . . . . . 51 3.6.1.1 Model of Costly Stock Trading . . . . . . . . . . . . . . . . . . . . 51 3.6.1.2 Empirical Evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 3.6.2 Stock Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 3.6.3 Market Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 3.6.4 Liquidity Reasons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 3.6.5 Excess Cash and Closed-End Fund Performance . . . . . . . . . . . . . . . . 59 3.6.5.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 3.6.5.2 Summary Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 3.6.5.3 Determinants of Closed-End Fund Cash Holdings . . . . . . . . . . 60 3.6.5.4 Excess Cash Holdings and Closed-End Fund Performance . . . . . 61 3.6.6 Concealing Portfolio Composition . . . . . . . . . . . . . . . . . . . . . . . . 61 3.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 3.8 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 iv 4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 4.1 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Appendices A Appendix to Chapter 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 A.1 Data Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 A.2 Alternative Definitions of Excess Cash . . . . . . . . . . . . . . . . . . . . . . . . . . 92 A.3 Results Obtained Using Equal-Weighted Returns . . . . . . . . . . . . . . . . . . . . 95 A.4 ECM Returns Conditional on Book-to-Market, Size, and Leverage . . . . . . . . . . 95 A.5 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 B Appendix to Chapter 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 B.1 Determination of Fund Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 B.2 Simplified Excess Cash Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 B.3 Transition Probabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 B.4 Temporary vs. Permanent Excess Cash . . . . . . . . . . . . . . . . . . . . . . . . . 109 B.5 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 v List of Tables 2.1 Determinants of Cash Holdings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.2 Characteristics of ECM Deciles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.3 ECM Decile Portfolio Returns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 2.4 Fama-MacBeth Regression Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 2.5 Time Series Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2.6 Unconditional Risk Adjustment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.7 ECM Decile Portfolio Returns Conditional on Market State . . . . . . . . . . . . . . 26 2.8 Annual Profitability, Investment Activity, Cash Holdings and Leverage Around Inclusion into High or Low ECM Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 3.1 Summary Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 3.2 Determinants of Fund Cash Holdings . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 3.3 Characteristics of Funds in Different Excess Cash Groups . . . . . . . . . . . . . . . 66 3.4 Fund Raw Cash Holdings and Future Performance . . . . . . . . . . . . . . . . . . . 67 3.5 Fund Excess Cash Holdings and Future Performance . . . . . . . . . . . . . . . . . . 68 3.6 Fama-MacBeth Regression Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 3.7 Fund Excess Cash Holdings and Future Performance Conditional On Positive Market Runup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 3.8 Model-Based Excess Cash Holdings and Future Performance . . . . . . . . . . . . . . 71 3.9 Future Expenses and Liquidity vs. Excess Cash . . . . . . . . . . . . . . . . . . . . . 72 3.10 Fund Excess Cash Holdings and Future Performance Conditional on Fund Size and Liquidity of Holdings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 3.11 Future Performance of Stocks Bought and Sold by Funds in Different Excess Cash Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 3.12 Market Timing of Excess Cash Groups . . . . . . . . . . . . . . . . . . . . . . . . . . 75 3.13 Market Timing Conditional on Volatility of Excess Cash . . . . . . . . . . . . . . . . 76 3.14 Fund Excess Cash Holdings and Future Performance Conditional on Future Fund Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 3.15 Summary Statistics: Closed-End Funds . . . . . . . . . . . . . . . . . . . . . . . . . 79 3.16 Determinants of Closed-End Fund Cash Holdings . . . . . . . . . . . . . . . . . . . . 80 3.17 Excess Cash Holdings and Future Performance of Closed-End Funds . . . . . . . . . 81 A.1 Determinants of Cash Holdings – Modified Regression Specification . . . . . . . . . . 97 vi A.2 ECM Decile Portfolio Returns – Modified Regression Specification . . . . . . . . . . 98 A.3 Fama-MacBeth Regression Results – Modified Regression Specification . . . . . . . . 99 A.4 ECM Decile Portfolio Returns – Simplified Excess Cash Definition . . . . . . . . . . 100 A.5 Fama-MacBeth Regression Results – Simplified Excess Cash Definition . . . . . . . . 101 A.6 ECM Decile Portfolio Equal-Weighted Returns . . . . . . . . . . . . . . . . . . . . . 102 A.7 Unconditional Risk Adjustment: Equal-Weighted Returns . . . . . . . . . . . . . . . 103 A.8 ECM Decile Portfolio Equal-Weighted Returns Conditional on Market State . . . . . 104 A.9 ECM Decile Portfolio Returns Conditional on BM, Size, and Debt: Equal-Weighted Returns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 A.10 ECM Decile Portfolio Returns Conditional on BM, Size, and Debt: Equal-Weighted Returns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 B.1 Fund Excess Cash Holdings and Future Performance: Simplified Excess Cash Definition110 B.2 Excess Cash Groups Transition Probabilities . . . . . . . . . . . . . . . . . . . . . . 111 B.3 Fund Excess Cash Holdings and Future Performance: . . . . . . . . . . . . . . . . . . 112 vii List of Figures 2.1 Time Series of High Minus Low ECM Portfolio Returns . . . . . . . . . . . . . . . . 28 2.2 Excess Cash and Investment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.3 Time Series of High Minus Low ECM Portfolio Returns . . . . . . . . . . . . . . . . 30 3.1 Fund Cash Holdings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 3.2 Cumulative Abnormal Returns of Excess Cash Groups . . . . . . . . . . . . . . . . . 83 3.3 Performance of High–Low Excess Cash Portfolios . . . . . . . . . . . . . . . . . . . . 84 3.4 Effects of Costly Stock Trading on Cumulative Change in Cash . . . . . . . . . . . . 85 viii Acknowledgements I have been very fortunate to be surrounded by many people who have helped, encouraged and motivated me. I am grateful to Adlai Fisher and Murray Carlson (my committee co-chairs) who provided support, thoughtful feedback and inspiration. It is by working with them that I learned the seriousness and joy of research. I will treasure the memories of us working late nights in Adlai’s office, meeting at Max’s deli, and sitting in the hallway typing away on the laptop with the power out. I also benefited immensely from the help and feedback of Glen Donaldson and Ron Giammarino, both of whom I thank for the generosity with which they gave me their time. I also thank Lorenzo Garlappi, Rob Heinkel, Thomas Hellmann, and other faculty members for having their doors always open. I would like to acknowledge the support of my fellow Ph.D. students, to all of whom I am thankful for helping make it such wonderful six years. Special thanks to Oliver Boguth, who went the entire road with me and was always a great person to talk to. Thank you to David Newton and Thomas Ruf for being the two guys to whom I could always turn, and of course for all those lunchtime discussions and arguments. No words could describe how grateful I am to my parents, Vasily and Valentina Simutin, and to my brother Vasily who supported and encouraged me distantly from Russia. And similarly no words could describe my deepest gratitude to my American parents, Harold and Joyce Carr, without whose generosity and unconditional faith in me I certainly would not have achieved what I have. Finally, thank you to my loving fiancée Annabel, who was always there in good days and in bad with all the support I could ever ask for. ix Dedication To my Russian and American parents, my brother Vasily, and Annabel for your love and support. x Chapter 1 Introduction Cross-sectional differences in cash holdings can be striking even among seemingly comparable fi- nancial entities. Corporations competing in the same industry often hold remarkably different cash positions, as do mutual funds pursuing similar investment objectives. For example, Amazon.com ended 2009 holding $3.4 billion in cash and equivalents, accounting for 25% of the firm’s total assets. For Barnes and Noble, the cash-to-assets ratio at the same time was just 1%. Similarly, at the end of 2007 AIM Large Cap Growth Fund carried just 0.33% of total net assets in cash, while the corresponding number for Allianz RCM Large-Cap Growth Fund was 10.30%. Yet, despite such differences, corporate cash holdings have started to receive attention of empirical researchers only recently, and cash positions of mutual funds and their effects on fund performance have largely been overlooked. In this thesis, I study how cash holdings in excess of the level predicted by the characteristics of a financial entity – “excess cash” – relate to its future performance. I explore this relationship under two organizational structures: corporations and mutual funds. I emphasize excess cash because, as a discretionary amount, it has the potential to capture unobservable information that relates to the organization’s future performance. Corporate excess cash may be indicative of the firm’s prospects, and mutual fund cash positions may reflect managerial abilities. In Chapter 2 of the thesis, I show that corporate excess cash holdings relate positively to future stock returns. I provide evidence that this finding is consistent with the notion that corporate excess cash proxies for risky growth options. In Chapter 3, I document new important determinants of cash positions of actively managed mutual funds and demonstrate a positive link between a fund’s excess cash holdings and future performance. I tie this positive relationship to managerial abilities. 1.1 Excess Cash and Stock Returns Corporate cash holdings have been receiving increasing attention in the recent literature. The principal focus of many studies has been to understand the drivers of cash positions. Opler et al. (1999) were the first to study the determinants of cash holdings, showing that size, book-to-market ratio, past cash flows, and other firm characteristics affect cash balances carried by companies.1 1Bates et al. (2009) confirm the findings of Opler et al. (1999), document an increase in corporate cash holdings since 1980 and explore reasons for this increase. Kim et al. (1998), Almeida et al. (2004), and Riddick and Whited (2009) explore the trade-off between the low and taxable returns that high cash balances produce and the reduced de- pendence on costly external financing they provide. Foley et al. (2007) provide a tax-based explanation for differences in cash holdings. 1 In the first essay of this thesis, I study how cash holdings in excess of the level predicted by firm characteristics (“excess cash”) impact stock returns. I emphasize excess cash because it has the potential to capture unobservable information about firm prospects that is not reflected in the usual proxies such as book-to-market ratio. Information captured by excess cash may relate to a firm’s future raw and abnormal stock returns, risk, investment, and profitability in two distinct ways. First, unusually high excess cash levels may indicate managerial concerns about future operating cash flows and investment opportunities, hinting at a negative link between excess cash holdings and returns, investment, and profitability. Alternatively, firms facing costly external financing may build up cash reserves in anticipation of future investment opportunities, implying that excess cash can relate positively to risk, future investment, and expected returns.2 The empirical evidence I present is supportive of the latter argument. I document a positive relationship between corporate excess cash holdings and future stock returns.3 I find that high excess cash firms outperform their low excess cash peers by 5% annually, or 6% after standard risk adjustment. Consistent with excess cash serving as a proxy for growth opportunities, high excess cash firms have higher market betas and incur significantly higher in- vestment expenditures in the future. I also show that high excess cash stocks firms underperform their low excess cash peers during market downturns. This finding, while somewhat surprising, is consistent with the idea that excess cash holdings correlate with growth opportunities. During market downturns, the value of such investment opportunities falls and the performance of high excess cash firms suffers, while the opposite is true during expansions. This essay builds on the growing literature examining the value of cash holdings. Faulkender and Wang (2006) include lagged cash as a control for explaining changes in firm value, but focus on the contemporaneous relationship between stock returns and changes in firm characteristics. Harford et al. (2003) find that during and immediately following an industry sales decline, firms with larger cash reserves invest more. Pinkowitz and Williamson (2004) study the marginal value of cash, but their focus is on the cross-sectional variation related to the investment opportunity set of the firm. Pinkowitz et al. (2006) and Dittmar and Mahrt-Smith (2007) explore the link between corporate governance and value of cash holdings. Mikkelson and Partch (2003) look at firms that held over 25% of assets in cash in each of the previous five years and find that such firms have better operating performance in the following 5-year period.4 By contrast with other literature, I focus on excess cash holdings, carefully control for other predictors of stock returns, condition on the market state, explore levels of risk, investment and profitability, and conclude that excess cash holdings proxy for risky growth options. 2For recent literature examining the relationship between risk and investment, see Berk et al. (1999), Gomes et al. (2003), Carlson et al. (2004, 2006), Zhang (2005), and Bernardo et al. (2007). 3I define excess cash following Opler et al. (1999) as the residual from cross-sectional regressions of cash-to-assets ratios on variables previously determined to explain cash holdings. I also consider alternative definitions of excess cash. 4Other related papers include Hovakimian (2009), McLean (2009), Morellec and Nikolov (2009), Palazzo (2009), and Simutin (2009), and Yan (2006). 2 1.2 Excess Cash and Mutual Fund Performance Cash holdings of mutual funds have received significant attention in the media, but have largely been ignored in the academic literature.5 The two main exceptions are Chordia (1996) and Yan (2006) who study the link between cash holdings and a number of fund characteristics. In the second essay of this thesis, I document new important determinants of mutual fund cash holdings and study how cash balances in excess of the level needed to conduct normal operations impact fund performance. I again emphasize excess cash because, as a discretionary amount, it has the potential to capture information about otherwise unobservable fund characteristics that affect fund performance. Information captured by excess cash may reflect, among other things, stock-picking skills, market-timing abilities, the investment opportunity set of the manager, and managerial expectations about liquidity needs of the fund. I define excess cash both (i) empirically, as the residual from cross-sectional regressions of cash- to-total net assets ratio on fund characteristics, and (ii) theoretically, as the difference between actual cash position and the target balance predicted by a model of optimal fund cash holdings that I develop. Using either definition, I find that funds with high excess cash holdings outperform those with low excess cash by 2% per year, increasing to 3% after standard risk-adjustment. To understand why high excess cash funds outperform their low excess cash peers, it is helpful to recognize that fund cash holdings are affected by exogenous flows, which include withdrawals, deposits and dividends, and by endogenous managerial decisions about purchases and sales, which in turn affect expenses incurred by the fund. Because I control for the differences in recent fund flows in defining excess cash, it is unlikely that the positive relationship between excess cash and fund performance is due to fund flow shocks. Instead, I conjecture that it is attributable to managerial decision to adjust the fund’s cash holdings. Adjustments to cash positions may reflect (i) managerial proficiency at controlling transaction costs of the fund, (ii) the manager’s stock-picking abilities and investment opportunities, (iii) managerial market-timing skills, and (iv) the manager’s aptitude at anticipating future fund flows. I develop these four hypotheses in detail and find empirical evidence supporting each conjecture. More specifically, I find that managers of high excess cash funds (i) incur lower expenses in the future, (ii) have superior ability to identify which stocks to purchase, (iii) are marginally more skilled than low excess cash fund managers at timing the market, and (iv) are better able to anticipate future fund flows. Central to my analysis is the definition of excess cash. To calculate excess cash, I thoroughly explore the determinants of cash holdings of mutual funds. In particular, I show that funds holding riskier, less liquid, or low dividend-paying stocks, as well as those run by managers with lower return gap measure of Kacperczyk et al. (2008) carry more cash. Compared to the determinants of fund cash holdings studied in the prior literature, the characteristics I consider explain three times 5For recent examples, see “Fund’s Extra Cash Holds Opportunities”, Wall Street Journal, April 8, 2009, page C13; “More Stocks Funds Declare Cash King”, Wall Street Journal, April 9, 2009, page C9; “Cash Regains Its Asset Status”, Barron’s, August 17, 2009, page 24; “Harvard, Yale Are Big Losers in ‘The Game’ of Investing”, Wall Street Journal, September 11, 2009, page A1. 3 more cross-sectional variation in cash positions. In the vast literature exploring the factors affecting mutual fund performance, surprisingly little research has been devoted to studying the role played by fund cash holdings. Chordia (1996) develops a model of mutual fund fee structures and empirically links fund cash holdings to load fees and uncertainty about redemptions. Yan (2006) identifies additional determinants of fund cash positions and focuses on studying the relationship between aggregate cash holdings and market returns. He finds no link between raw cash balances and future fund performance. By contrast, I focus on excess cash holdings and document a positive relationship between excess cash and fund performance. I study the sources of this relationship, linking it to managerial stock-picking and market-timing skills, liquidity needs and operating costs of the funds. I additionally document a number of new important determinants of fund cash holdings, relating cash balances to, among other things, risk and liquidity of fund shareholdings and to fund return gap. In related work, Dellva and Olson (1998) study a 1987-1992 sample and obtain a significantly positive coefficient when regressing fund returns on – among other fund characteristics – cash holdings. Their focus, however, is on the effects of fund expenses, rather than cash holdings, on performance. More recently, Baker et al. (2009) find a positive link between cash holdings of institutional funds (i.e., funds investing on behalf of endowments and other institutions) and future returns. Methodologically, this essay builds on the work of Simutin (2010) who documents a positive relationship between corporate excess cash holdings and future stock returns, and on the studies analyzing the link between excess CEO compensation and firm performance (e.g., Brick et al., 2006). This essay also contributes to the literature studying market-timing skills (e.g., Treynor and Mazuy, 1966; Henriksson and Merton, 1981; Chang and Lewellen, 1984; Henriksson, 1984; Cumby and Glen, 1990; Becker et al., 1999; Jiang et al., 2007) and stock-picking abilities of the managers (e.g., Chen et al., 2000; Kacperczyk et al., 2005; Cremers and Petajisto, 2009), as well as to the literature exploring the importance of mutual fund liquidity and fund flows (e.g., Sirri and Tufano, 1998; Edelen, 1999). The rest of the thesis proceeds as follows. In Chapter 2, I explore the relationship between corporate excess cash holdings and stock returns. In Chapter 3, I study the determinants of cash holdings of actively managed mutual funds, document a positive link between excess cash and future fund performance, and attribute this relationship to managerial abilities. Chapter 4 provides concluding remarks. 4 Bibliography Almeida, H., Campello, M., and Weisbach, M. S. (2004). The cash flow sensitivity of cash. Journal of Finance, 59(4): 1777–1804. Baker, H. K., Haslem, J. A., and Smith, D. M. (2009). Performance and characteristics of actively managed institutional equity mutual funds. Journal of Investing, 18(1): 27–44. Bates, T. W., Kahle, K. M., and Stulz, R. M. (2009). Why do us firms hold so much more cash than they used to? Journal of Finance, 64(5): 1985–2021. Becker, C., Ferson, W., Myers, D. H., and Schill, M. J. (1999). Conditional market timing with benchmark investors. Journal of Financial Economics, 52(1): 119–148. Berk, J. B., Green, R. C., and Naik, V. (1999). Optimal investment, growth options, and security returns. Journal of Finance, 54(5): 1553–1607. Bernardo, A. E., Chowdhry, B., and Goyal, A. (2007). Growth options, beta, and the cost of capital. Financial Management, 36(2): 5–17. Brick, I. E., Palmon, O., and Wald, J. K. (2006). Ceo compensation, director compensation, and firm performance: Evidence of cronyism? Journal of Corporate Finance, 12(3): 403–423. Carlson, M., Fisher, A., and Giammarino, R. (2004). Corporate investment and asset price dy- namics: Implications for the cross-section of returns. Journal of Finance, 59(6): 2577–2603. Carlson, M., Fisher, A., and Giammarino, R. (2006). Corporate investment and asset price dynam- ics: Implications for seo event studies and long-run performance. Journal of Finance, 61(3): 1009– 1034. Chang, E. C. and Lewellen, W. G. (1984). Market timing and mutual fund investment performance. Journal of Business, 57(1): 57–72. Chen, H. L., Jegadeesh, N., and Wermers, R. (2000). The value of active mutual fund management: An examination of the stockholdings and trades of fund managers. Journal of Financial and Quantitative Analysis, 35(3): 343–368. Chordia, T. (1996). The structure of mutual fund charges. Journal of Financial Economics, 41(1): 3–39. 5 Cremers, K. J. M. and Petajisto, A. (2009). How active is your fund manager? A new measure that predicts performance. Review of Financial Studies, 22(9): 3329–3365. Cumby, R. E. and Glen, J. D. (1990). Evaluating the performance of international mutual funds. Journal of Finance, 45(2): 497–521. Dellva, W. L. and Olson, G. T. (1998). The relationship between mutual fund fees and expenses and their effects on performance. Financial Review, 33(1): 85–104. Dittmar, A. and Mahrt-Smith, J. (2007). Corporate governance and the value of cash holdings. Journal of Financial Economics, 83(3): 599–634. Edelen, R. M. (1999). Investor flows and the assessed performance of open-end mutual funds. Journal of Financial Economics, 53(3): 439–466. Faulkender, M. and Wang, R. (2006). Corporate financial policy and the value of cash. Journal of Finance, 61(4): 1957–1990. Foley, C. F., Hartzell, J. C., Titman, S., and Twite, G. (2007). Why do firms hold so much cash? A tax-based explanation. Journal of Financial Economics, 86(3): 579–607. Gomes, J., Kogan, L., and Zhang, L. (2003). Equilibrium cross section of returns. Journal of Political Economy, 111(4): 693–732. Harford, J., Mikkelson, W., and Partch, M. (2003). The effect of cash reserves on corporate investment and performance in industry downturns. University of Washington Working Paper. Henriksson, R. D. (1984). Market timing and mutual fund performance - an empirical-investigation. Journal of Business, 57(1): 73–96. Henriksson, R. D. and Merton, R. C. (1981). On market timing and investment performance ii: Statistical procedures for evaluating forecasting skills. Journal of Business, 54(4): 513–533. Hovakimian, G. (2009). Determinants of investment cash flow sensitivity. Financial Management, 38(1): 161–183. Jiang, G. J., Yao, T., and Yu, T. (2007). Do mutual funds time the market? Evidence from portfolio holdings. Journal of Financial Economics, 86(3): 724–758. Kacperczyk, M., Sialm, C., and Zheng, L. (2005). On the industry concentration of actively managed equity mutual funds. Journal of Finance, 60(4): 1983–2011. Kacperczyk, M., Sialm, C., and Zheng, L. (2008). Unobserved actions of mutual funds. Review of Financial Studies, 21(6): 2379–2416. Kim, C. S., Mauer, D. C., and Sherman, A. E. (1998). The determinants of corporate liquidity: Theory and evidence. Journal of Financial and Quantitative Analysis, 33(3): 335–359. 6 McLean, R. D. (2009). Share issuance and cash savings. University of Alberta Working Paper. Mikkelson, W. H. and Partch, M. M. (2003). Do persistent large cash reserves hinder performance? Journal of Financial and Quantitative Analysis, 38(2): 275–294. Morellec, E. and Nikolov, B. (2009). Cash holdings and competition. Ecole Polytechnique Federale de Lausanne Working Paper. Opler, T., Pinkowitz, L., Stulz, R., and Williamson, R. (1999). The determinants and implications of corporate cash holdings. Journal of Financial Economics, 52(1): 3–46. Palazzo, D. (2009). Firms cash holdings and the cross-section of equity returns. Boston University Working Paper. Pinkowitz, L., Stulz, R., and Williamson, R. (2006). Does the contribution of corporate cash holdings and dividends to firm value depend on governance? A cross-country analysis. Journal of Finance, 61(6): 2725–2751. Pinkowitz, L. and Williamson, R. (2004). What is a dollar worth? The market value of cash holdings. Georgetown University Working Paper. Riddick, L. A. and Whited, T. M. (2009). The corporate propensity to save. Journal of Finance, 64(4): 1729–1766. Simutin, M. (2009). Excess cash and mutual fund performance. University of British Columbia Working Paper. Simutin, M. (2010). Excess cash and stock returns. Financial Management, Forthcoming. Sirri, E. R. and Tufano, P. (1998). Costly search and mutual fund flows. Journal of Finance, 53(5): 1589–1622. Treynor, J. L. and Mazuy, K. K. (1966). Can mutual funds outguess market. Harvard Business Review, 44(4): 131–136. Yan, X. M. (2006). The determinants and implications of mutual fund cash holdings: Theory and evidence. Financial Management, 35(2): 67–91. Zhang, L. (2005). The value premium. Journal of Finance, 60(1): 67–103. 7 Chapter 2 Excess Cash and Stock Returns 2.1 Introduction Corporate cash holdings can differ dramatically even for seemingly comparable companies.6 For example, BlackBerry manufacturer Research in Motion ended fiscal 2008 with over $1 billion in cash and equivalents accounting for 21% of the firm’s total assets. By contrast, Nokia’s cash-to-assets ratio in the same year reached only 4%. In recent research, authors have attempted to explain the determinants of cash holdings indicating that size, book-to-market ratio, past cash flows, and other firm characteristics affect cash balances carried by companies.7 In this paper, I study how cash holdings in excess of the level predicted by firm characteristics (“excess cash”) impact stock returns. I emphasize excess cash because it has the potential to capture information about firm prospects that is not reflected in the usual proxies such as book-to- market ratio. Information captured by excess cash may relate to a firm’s future raw and abnormal stock returns, risk, investment, and profitability in two distinct ways. First, unusually high excess cash levels may indicate managerial concerns regarding future operating cash flows and investment opportunities, hinting at a negative link between excess cash holdings and returns, investment, and profitability. Alternatively, firms facing costly external financing may build up cash reserves in anticipation of future investment opportunities, implying that excess cash can relate positively to risk, future investment, and expected returns.8 The empirical evidence I present is, on the whole, supportive of the latter argument. I document a positive relationship between corporate excess cash holdings and future stock returns. I define excess cash following Opler et al. (1999) as the residual from cross-sectional regressions of cash-to-assets ratios on variables previously determined to explain cash holdings. This measure of excess cash retains its stock return forecasting ability even after controlling for a variety of firm characteristics known to relate to future returns including book-to-market ratio, asset growth, accruals, and others. Consistent with excess cash serving as a proxy for growth 6A version of this chapter has been accepted for publication. Simutin, M. (2010) Excess Cash and Stock Returns. Financial Management. 7Opler et al. (1999) were the first to study the determinants of cash holdings. Kim et al. (1998), Almeida et al. (2004), and Riddick and Whited (2009) explore the trade off between the low and taxable returns that high cash balances produce and the reduced dependence on costly external financing they provide. Foley et al. (2007) propose a tax-based explanation for differences in cash holdings. John (1993) suggests that firms that are subject to higher financial distress costs may wish to hold more cash. Bates et al. (2009) document an increase in corporate cash holdings since 1980 and explore reasons for this increase. 8Examples of recent literature examining the relationship between risk and investment include Berk et al. (1999), Gomes et al. (2003), Carlson et al. (2004, 2006), Zhang (2005), and Bernardo et al. (2007). 8 opportunities, high excess cash firms have higher market betas and report significantly higher investment expenditures in the future. The difference in the investment-to-assets ratios of the top and bottom excess cash groups reaches nearly 5% in just the first year following portfolio assignment. Interestingly, while this difference slowly attenuates, high excess cash firms invest more than their low cash peers in each of the following ten years. However, over the same ten-year period, firms with high measures of excess cash report profitability figures that are no larger than those of low excess cash companies. If high excess cash does, in fact, proxy for growth options as larger betas and greater investment expenditures of such firms suggest, higher returns earned by the firms with larger cash resources may be viewed as compensation for additional risk. However, I find that controlling for loadings on common risk factors does not eliminate the relationship between excess cash and stock returns. For example, the Fama and French (1993) three-factor alpha of the strategy that is long high excess cash decile and short the group with low values is 0.52% per month. Including factors that control for differences in momentum, asset growth, accruals, and leverage does not eliminate the statistical and economic significance of profits from this strategy. I explore whether firms with higher excess cash earn greater returns in all market states. It is natural to expect that in times of economic downturn, companies with greater excess cash might exhibit better stock performance than those with limited cash holdings. During such times, acquiring external capital may be more costly, meeting financial obligations may be more difficult, and having an extra cash cushion may prove particularly valuable. Curiously, I find that this is not the case. While the average spread between value-weighted returns of firms in high and low excess cash deciles amounts to 0.40% per month, in times of market slowdowns, high excess cash stocks underperform their peers with low excess cash by 0.31%. Conversely, during expansions, the difference in returns of the two groups is positive, exceeding 1% monthly. This finding, while surprising, is consistent with the idea that excess cash holdings correlate with growth opportunities. During market downturns, the value of such investment opportunities falls and the performance of high excess cash firms suffers, while the opposite is true during expansions. This study most closely relates to the recent literature that examines the value of cash hold- ings. Faulkender and Wang (2006) include lagged cash as a control for explaining changes in firm value, but focus on the contemporaneous relationship between stock returns and changes in firm characteristics. Harford et al. (2003) find that during and immediately following an industry sales decline, firms with larger cash reserves invest more. Pinkowitz and Williamson (2004) study the marginal value of cash, but their focus is on the cross-sectional variation related to the investment opportunity set of the firm.9 In independent contemporaneous work, Palazzo (2009) finds no unconditional relationship be- tween raw cash levels and future stock returns, but observes a positive link when conditioning 9Other related papers include Mikkelson and Partch (2003), Pinkowitz et al. (2006), Dittmar and Mahrt-Smith (2007), Hovakimian (2009), and McLean (2009). Yan (2006) and Simutin (2009) study the determinants and impli- cations of mutual fund cash holdings. Morellec and Nikolov (2009) relate cash holdings to the intensity of product market competition. 9 on size and book-to-market. His primary empirical results are consistent with the findings that I document. He additionally focuses on the ability of a cash factor to serve as a risk proxy and proposes a model with costly equity financing in which firms whose cash flows are correlated with an aggregate shock hedge a cash shortfall by increasing their savings. By contrast, I focus on excess cash holdings, carefully control for other predictors of stock returns, condition on the market state, and explore levels of risk, investment, and profitability. Prior literature documents a negative relationship between investment and future stock returns (Titman et al., 2004).10 Therefore, it may seem somewhat puzzling that this paper finds that excess cash firms have both greater future returns and higher future investment. However, I find no relationship between excess cash and lagged or contemporaneous investment, but document that high excess cash firms invest more only in the future. Indeed, the reason why the positive correlation between excess cash and future stock returns has not been discussed in the prior literature may, in part, relate to the commonly used approach of skipping up to 18 months between fiscal year end and the inclusion of a stock into a portfolio. This method confounds two effects: 1) higher returns prior to the exercising of growth options and 2) lower returns following their exercise. This paper focuses on the former effect and confirms that firms with excess cash are temporarily riskier and earn higher returns as they prepare to exercise their growth options. In the future, these options are gradually exercised, as evidenced by significantly higher investment-to-assets ratios of the high excess cash firms. The rest of the paper proceeds as follows. Section 2.2 describes the data and discusses the characteristics of firms with different levels of excess cash. The empirical relationship between excess cash and future returns is examined in Section 2.3. Section 2.4 studies the relationship between excess cash holdings and future investment and profitability. Section 2.5 provides my concluding remarks. The appendix contains data definitions, results obtained using alternative methods of estimating excess cash, and additional robustness checks. 2.2 Excess Cash Holdings: Estimation and Firm Characteristics Cross-sectional cash holdings can vary substantially depending on the nature of a firm’s business and recent activities of the firm. To account for such differences, I focus on a measure of excess cash; that is, holdings above what one would expect for companies in a similar line of business with similar characteristics. In this Section, I discuss the data and methodology used in constructing excess cash measures (ECM) and study the characteristics of firms with different levels of ECM. 2.2.1 Estimation of Excess Cash Opler et al. (1999) thoroughly explore the determinants of cash holdings, and I use their findings as a guide for determining excess cash. More specifically, to obtain an excess cash measure for stock 10Anderson and Garcia-Feijóo (2006) document a negative relationship between investment growth and subsequent stock returns. Chen et al. (2007) find that investment by a firm negatively affects stock prices of its competitors. 10 i in month t, I use all stocks that have fiscal year ends between t − 11 and t. In each month t, I run a cross-sectional regression: Ciτ = γ0t + γ1tMBiτ + γ2tSizeiτ + γ3tCPXiτ + γ4tWCiτ + γ5tLTDiτ + γ6tRDiτ + γ7tCFiτ + γ8tσINDiτ + it, where variable definitions follow those in Opler et al. (1999): C is the log of ratio of cash to total assets less cash; market-to-book ratio (MB) is measured as the book value of assets, less the book value of equity, plus market value of equity, divided by assets; Size is the log of real (adjusted by CPI) assets; CPX is the ratio of capital expenditures to assets; WC is the ratio of net working capital calculated without cash to assets; LTD is the ratio of long-term debt to assets; RD is the ratio of research and development expense (R&D) to sales; CF is the ratio of cash flow to total assets; and σIND, industry sigma, is the mean of standard deviations of CF over ten years for firms in the same two-digit SIC industry. I also include industry dummies based on Kenneth French’s 17 industry definitions and a dividend dummy.11 τ refers to the fiscal year that ended between t− 11 and t, and all variables with the τ subscript use the most recent data available for firm i. ECM as of the end of month t is defined as the residual it from this regression. This study focuses on the U.S. corporations in the 1960-2006 period with valid CRSP and Compustat data and excludes all financial firms (SICs = 6XXX) and utilities (SICs = 49XX).12 Table 2.1 presents the results of regressions used to estimate excess cash measures. Similarly to Opler et al. (1999) who focus on the 1971-1994 period and Bates et al. (2009), who study the 1980-2006 sample, I find that cash holdings increase with ratios of market equity to book equity, R&D to sales, cash flow to assets, as well as industry sigma, and decline with size, ratio of working capital to assets, and leverage. While Opler et al. (1999) and Bates et al. (2009) observe that the effect of the ratio of capital expenditures to assets on cash holdings is sensitive to regression specifications, I document that in my extended sample it relates negatively to cash holdings. I also find that dividend paying firms maintain cash-to-assets ratios that are no different from those of non-dividend paying companies. 2.2.2 Excess Cash and Firm Characteristics To study the relationship between characteristics of firms and their excess cash levels, at the end of each calendar year τ , I assign companies into excess cash deciles and obtain the most recent values of the characteristics of interest for each firm. All accounting measures such as cash, book equity, or debt of a given firm refer to the most recent year τ observation for that company. 11Bates et al. (2009) use a similar specification to explain corporate cash holdings. The findings of this paper are robust to alternative reasonable definitions of excess cash, which I explore in Appendix B. Industry definitions are from Kenneth French’s data library, http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html. 12None of the results are affected by retaining these firms; however, they could be misleading because financials (utilities) tend to hold a large (small) fraction of their assets in cash and equivalents. 11 Table 2.2 presents averages of the selected characteristics of each ECM decile. As one would expect, firms with higher ECM hold a significantly higher fraction of assets in cash, while companies in the highest ECM decile hold, on average, 42% of assets in cash. The comparable figure for firms in the lowest group is just 1.7%. Cash is one of the safest assets, and it is commonly considered to be less risky than assets in place. Thus, it is natural to expect firms with higher ECM to have lower risk. Surprisingly, I find the opposite is true. Table 2.2 demonstrates that firms’ risk, as proxied for by market beta, increases with excess cash. The relationship is surprisingly monotonic. Firms in the lowest ECM decile have an average beta of just 0.86, while those in the top group have risk measures that are nearly 20% higher at 1.02.13 The difference between average loadings of high and low excess cash groups, at 0.16, is highly significant (t-statistic of 5.37). This positive relationship between excess cash and betas can be justified if excess cash proxies for the presence of risky growth options. In the following Section, I will provide further evidence supporting this explanation. The next four columns of Table 2.2 examine the relationship between excess cash and book-to- market ratio, firm size, profitability, and cash flow. While each of these characteristics is lower for the decile of high ECM firms, there is no monotonic relationship between excess cash and either of the variables.14 On average, firms in both high and low ECM deciles are smaller, have lower book-to-market ratios, and generate lower return on assets and lower cash flows relative to firms in the middle groups. The last three columns of Table 2.2 illustrate the generally monotonic relationship between ECM and measures of debt, accruals, and asset growth. Leverage reaches 0.22 for low excess cash firms and gradually declines to 0.15 for companies with high excess cash. This negative correlation between excess cash and leverage is consistent with the idea that firms with limited access to debt financing may accumulate higher levels of cash to ensure they have enough resources to meet financial obligations. Table 2.2 further reports that firms experiencing low accruals or high asset growth in the past tend to have higher ECMs. The monotonic relationships of excess cash with leverage, accruals, and asset growth are interesting, and I will take particular care in ensuring that the findings of this paper are not driven by any of these three characteristics. 2.3 Excess Cash Holdings and Stock Returns What relationship should exist between excess cash holdings and the performance of a company’s stock? High excess cash levels may be indicative of managerial concerns regarding future operating cash flows and investment opportunities, hinting at a negative link between cash holdings and 13I calculate market beta as the sum of slope coefficients (Dimson, 1979) from regressions of daily excess stock returns in year τ on market excess return, its lead, and its lag. The fact that average betas are lower than unity is attributable to the fact that they are calculated as equally-weighted averages over all stocks with valid ECMs. This restricts the sample to firms with valid Compustat data and eliminates smaller stocks that tend to have higher betas. 14Table 2.1 follows Opler et al. (1999) definition of market-to-book ratio (MB) in estimating excess cash, while Table 2.2 and all subsequent tables use the more conventional book-to-market ratio (BM), whose calculation is detailed in the Appendix. Excess cash is orthogonal to MB by definition and relates only weakly to BM, as Table 2.2 indicates. 12 returns. Conversely, in the presence of costly external financing, firms may accumulate cash in anticipation of future investment opportunities implying that cash can relate positively to risk and expected returns. The positive link between excess cash and market beta documented in the previous section is in line with the latter argument. In this section, I present additional evidence supporting this explanation by documenting a positive association between excess cash and future stock returns. 2.3.1 Future Raw Returns I begin the empirical investigation by examining the performance of ten portfolios formed on the basis of excess cash level. In particular, at the end of every month t, I use all common stocks with fiscal years ending between t − 15 and t − 4 to assign stocks into quintiles based on their market betas calculated using daily data from t− 15 to t− 4. This sorting is done to filter out the differences in betas documented in Table 2.2. Within each beta group, I then assign stocks into deciles d on the basis of their excess cash measures computed as of month t − 4.15 Grouping all firms that fall into a given decile d results in ten ECM portfolios with approximately equal market exposure. The position taken in each company at the beginning of month t + 1 is equal to either $1 (when computing equally-weighted returns) or to the market capitalization of the firm as of the end of month t (when computing value-weighted returns). I hold the position without rebalancing for 12 months starting in month t+ 1. Table 2.3 reports average returns and the corresponding t-statistics for each of the ECM deciles and for the differences between high and low ECM portfolios. The same message emerges both from the full sample (1960-2006) and the subsample (1960-1982 and 1983-2006) results. Stocks with higher ECMs earn greater returns in the future. In the full sample, the spread in returns of high and low ECM deciles amounts to 0.40% per month, which is both statistically significant (t-statistic of 4.19) and economically important. These results are similar during the subperiods, with the average return difference reaching 0.33% during 1960-1982 and 0.47% during 1983-2006. Figure 2.1 plots the time series of monthly and cumulative log returns of the high minus low ECM portfolio. Monthly returns fluctuate in the range of 5% between 1960 and the late 1990s, but the portfolio experiences increased volatility and a substantial run-up followed by a decline around the time of the dot-com bubble. The two most extreme returns occur in two consecutive months around the peak of the bubble (25.66% in February 2000 and -12.58% in March 2000).16 15A commonly used approach in the literature is to assign stocks to groups based on data from fiscal year τ and hold the resulting portfolios from July of year τ to June of τ + 1. This lag of up to 18 months is excessive to capture a short-lived effect like the one documented in this paper for ECM and future returns. For this reason, I assume that accounting data is publicly available four months after the fiscal year end. This approach is not uncommon. Indeed, Haugen and Baker (1996) assume just a three-month lag. In unreported results, I use all post-1993 data available from the SEC via EDGAR to determine that just 1% of the companies in my sample file their 10-K reports later than four months following fiscal year end. Excluding those firms does not affect the results in the 1993-2006 sub-period. 16In untabulated results, I find that excluding the dot-com bubble period from the sample does not alter the results of the paper. 13 2.3.2 Fama-MacBeth Regressions The positive relationship between excess cash holdings and future stock performance is intriguing, but as Table 2.2 indicates, cash holdings are correlated with a number of firm characteristics known to relate to future returns. To ensure that excess cash measures do not simply proxy for such characteristics, I use Fama-MacBeth (1973) regressions to control for a number of variables previously linked to future stock returns. Table 2.4 presents average slope coefficients and the corresponding t-statistics from these monthly cross-sectional regressions of monthly returns on lagged ECM and other firm characteristics. Regression (1) confirms the results of Table 2.3 by demonstrating that excess cash holdings are a significant predictor of future stock returns. Regression (2) indicates that controlling for market risk, size, and book-to-market does not diminish the ability of ECM to forecast stock returns. As in Fama and French (1992), beta is unrelated to, while firm size and book-to-market ratio are strongly related to, future stock returns. Table 2.2 documents a generally monotonic relationship between excess cash and both asset growth and accruals, but Specifications (3) and (4) indicate that ECM remains a statistically significant predictor of returns after accounting for these variables. Similarly, Specifications (5)-(9) determine that controlling for investment, cash flow, leverage, momentum, and stock issuance does not eliminate the statistical significance of ECM. While investment, past returns, and share issuance are significant predictors of stock returns, ECM retains its forecasting power in their presence.17 Specification (10) reports that combining multiple predictor variables does not affect the ability of excess cash holdings to forecast returns. The average slope coefficient on ECM, at 0.074, is only slightly lower than that of Regression (1) with no additional controls, and is statistically significant (t-statistic of 3.69).18 Therefore, excess cash measure is not simply proxying for other variables previously documented to relate to future stock returns, but rather is a predictor different from those discussed earlier in the literature. 2.3.3 Risk-Adjusted Returns I now examine whether higher returns earned by the firms with larger excess cash resources may be viewed as compensation for additional risk. I consider a strategy that each month buys the stocks in the top ECM decile, shorts those in the low ECM group, and holds the resulting position for 12 months. I conduct a series of unconditional regressions to find that neither market, nor three- and four-factor models, nor models that include asset growth, accruals, and leverage factors can explain the excess return earned by this portfolio. 17The negative relationship between asset growth and future returns is consistent with the findings of Cooper et al. (2008). Sloan (1996) studies the link between accruals and future returns. Titman et al. (2004), Jegadeesh and Titman (1993), and Daniel and Titman (2006), among others, investigate the relationship between future returns and investment, momentum, and share issuance, respectively. 18In untabulated results, I find that raw cash does relate positively to future stock returns, although this result is weak in several Fama-MacBeth (1973) regression specifications. 14 2.3.3.1 Time Series Characteristics Table 2.5 details the first four moments and other time series characteristics of returns of the high minus low ECM portfolio and several factors. I obtain the commonly used four factors (market, value, size, and momentum) from Kenneth French’s data library, and construct asset growth, accruals, and debt factors following the same procedure used to obtain ECM returns.19 Confirming the results of Table 2.3, the difference in returns between the portfolios of high and low excess cash firms amounts to 0.40% per month, a magnitude comparable to the average return of the value factor. However, due to lower volatility of ECM portfolio returns, the strategy’s Sharpe ratio (0.18) is slightly above that of the value factor. ECM returns are considerably right skewed, with skewness (2.39) exceeding that of any other time series considered. Excess cash portfolio returns are also leptokurtic, with kurtosis comparable to that of size, momentum, and asset growth factors. For completeness, Table 2.5 presents the correlation matrix of returns of the ECM strategy and the factors. Excess cash portfolio exhibits a positive correlation with market (correlation coefficient of 0.23), size (0.34), and momentum (0.14) returns, and is negatively correlated with value (-0.47), asset growth (-0.24), accruals (-0.17), and debt (-0.54) factors. Given these high correlations, it is particularly important to consider risk adjustment that controls for these factors, which is what I explore next. 2.3.3.2 Unconditional Risk Adjustment Table 2.6 presents the results of the unconditional regressions of high minus low ECM portfolio returns on a number of factors. Specification (1) confirms that the high ECM decile outperforms the low ECM group by 0.40% per month (t-statistic of 4.19). Market model Regression (2) reports that market excess return alone is insufficient to explain the profits of the investment strategy (alpha of 0.34%). Both the Fama-French (1993) three-factor and the Carhart (1997) four-factor models (Regressions (3) and (4), respectively) only augment the returns of the strategy when compared to the case of no risk adjustment in Specification (1). In particular, the three-factor alpha amounts to 0.52% monthly (t-statistic of 6.10), while the four-factor alpha stands at 0.47% (t-statistic of 5.36). The loading on the value factor is strongly negative, while the size and momentum betas are significantly positive. Regressions (5)-(8) of Table 2.6 consider risk adjustment with asset growth, accruals, and debt factors. Specifications (5) and (6) indicate that the loadings on asset growth and, to a lesser degree, accruals factors are strongly negative. The R2 values, however, are low, and the profitability of the ECM portfolio remains both statistically and economically meaningful. Interestingly, inclusion of the leverage factor alone in Regression (7) produces a higher adjusted R2 (29.21%) than does the four-factor model. Yet, despite the high R2 and a large loading on the debt factor, the returns of the high minus low ECM portfolio retain their significance (alpha of 0.46% with t-statistic of 5.74). 19More precisely, I assign stocks into deciles on the basis of lagged asset growth, accruals, or debt. Each month I take a long position in the top decile while shorting the bottom group and hold the resulting portfolio for 12 months. The returns from such high minus low portfolios define the three factors. 15 Combining the three factors in Specification (8) renders the accruals factor insignificant, but the high minus low ECM portfolio remains profitable (alpha of 0.33 Specification (9) considers both the commonly used four factors and the three factors I con- structed to attempt to explain the returns of the ECM strategy. Each factor except momentum is statistically significant, and the adjusted R2 of this specification is higher than that of any other regression considered, but the alpha remains both statistically and economically significant. Thus, none of the commonly used four factors or the asset growth, accruals, and leverage factors can explain the profits from the investment strategy that buys the stocks in the top ECM decile and shorts those in the bottom group. It is tempting to infer a causal relationship between excess cash and future returns, but the profitability of the high minus low ECM portfolio should be interpreted with caution. I consider a number of commonly used models to explain the returns of the high minus low ECM portfolio, and while none of them are able to explain the profitability of the strategy, it may be more prudent to conclude that higher excess cash holdings correlate with, rather than cause, higher future returns. 2.3.4 Excess Cash Holdings and Market State It seems reasonable to expect that cash is particularly valuable during times of economic slowdown. To check this conjecture, I study the relationship between excess cash and future returns conditional on the state of the market. I use market return as a proxy for whether general economic conditions are strong or poor, and assign each month from January 1960-December 2006 into five groups based on the magnitude of market return in that month. Table 2.7 explores the relationship between excess cash holdings and stock returns conditional on the market state. During the times with the lowest market returns, it is the stocks of firms with the highest excess cash that perform the worst (-6.27% per month for high excess cash stocks vs. -5.96% for the low cash group). This is somewhat surprising as it may be intuitive to expect cash to be particularly beneficial during economic downturns. During such times, access to credit may be tight, cash flows may be low, and holding excess cash may prove especially valuable. However, this finding is consistent with the idea that firms build up cash reserves in anticipation of investment opportunities. In down markets, the value of such growth options is likely to fall, resulting in lower stock returns for firms with high excess cash. In other states of the market, the picture reverses. During such times, high excess cash firms outperform their low excess cash peers. In the best state of the market (‘High’), the spread in returns of high and low ECM portfolios amounts to 1.08% per month. This is consistent with more abundant investment opportunities present during times of economic expansion. High ECM firms have readily available resources to take advantage of such opportunities, while those with low excess cash either cannot afford to make similar investments, or may be forced to obtain funds though costly external financing. 16 2.4 Excess Cash, Investment, and Profitability If high excess cash holdings do in fact proxy for growth opportunities, as the empirical results pre- sented thus far suggest, it is natural to ask whether high excess cash firms invest more in the future than do their peers with lower holdings. In this section, I demonstrate that investment increases with the level of excess cash for up to ten years following portfolio assignment. However, I find no relationship between excess cash and future profitability, hinting at a possibility of overinvestment by high excess cash firms. For each of the five excess cash quintiles, Figure 2.2 presents average ratios of investment to total assets measured in the year of portfolio assignment and in each of the following ten years.20 The five groups report comparable levels of investment in the year of the sort, but the differences in investment among them become striking beginning the following year. The conclusions are similar whether I use all firms (Panel A) or consider just those that survive for the entire ten years (Panel B). Future investment increases dramatically with the level of excess cash. One year after portfolio assignment, high excess cash companies invest, on average, an amount equal to 13.8% of their assets, while comparable numbers for the middle and low groups are 10.7% and 9%, respectively. What is even more intriguing is that this shock to investment decays very slowly. Indeed, in each of the following ten years, average investment of the top group exceeds that of the low excess cash firms. Five years following the sort, high excess cash companies invest, on average, 11.9% of assets while firms in the bottom group invest just 9.6%. Only ten years after portfolio assignment do the differences in investment activity between the two groups revert to Year 0 level. In related research, Riddick and Whited (2009) use theory and simulation to demonstrate that in the presence of positively correlated income shocks, firms that generate high cash flow find it more valuable to invest this cash flow rather than keep it as savings.21 Empirically, Riddick and Whited (2009) focus on cash flows, rather than cash levels, and confirm that firms with high cash flows tend to save less. By contrast, I focus on companies with different excess cash levels and observe that even when controlling for differences in past cash flows, firms with unusually high levels of cash tend to invest more in the future than do firms with lower excess cash. The research questions addressed in Riddick and Whited’s 2009 work and in this paper are different, but the findings of the two papers are nonetheless related. This can be seen by recognizing that while high cash flow firms tend to save a smaller fraction of their cash flow, they also tend to hold a higher fraction of assets as cash. Evidence of a positive relationship between cash flow and cash level is provided in Table 2.1 of this paper and in Table IV of Opler et al. (1999). Thus, a positive relationship exists among cash flow, cash level, and future investment and the observations of Riddick and Whited (2009) are consistent with the findings of this paper. High cash flow firms tend to hold more cash and invest more in the future.22 20For ease of exposition, the figures use ECM quintiles rather than deciles. The results are qualitatively similar when deciles are used. 21Gamba and Triantis (2008) relax several assumptions of Riddick and Whited (2009) and find that cash flow is frequently used to increase a firm’s cash holdings (i.e., positive propensity to save). 22The findings of this section also relate to the work of Gopalan et al. (2010) who document a positive correlation 17 How profitable are the investments that high ECM firms undertake? Figure 2.3 depicts the average return on assets of each excess cash quintile. Regardless of whether I use all firms (Panel A) or study just those that are present in the sample for the entire ten years (Panel B), the conclusion is similar. There is no monotonic relationship between excess cash holdings and future profitability. In fact, firms in the high excess cash group are, on average, the least profitable in each of the ten years following portfolio assignment.23 These findings can be interpreted as indicative of overinvestment by high excess cash companies and can be viewed as evidence of suboptimal cash holdings. The long-term profitability of high excess cash firms suffers due to the costs of holding cash and overinvestment. The earnings of companies in the bottom group are low due to cash shortfalls, but profitability of firms in the middle group is the strongest as these companies choose cash levels that are neither too low nor excessive. Indeed, firms in the third quintile typically report the highest return on assets during the ten years following portfolio assignment. Table 2.8 summarizes average profitability, investment activity, leverage, and cash holdings of high and low excess cash groups during the ten years prior to and after the year of portfolio inclusion. The differences in profitability and investment of the two groups are stable during the ten years leading up to Year 0 and become more pronounced beginning in Year 1. Low excess cash groups are, on average, more levered than their high excess cash peers in each year. Their leverage increases slightly during the ten years prior to portfolio assignment. Alternatively, cash holdings exhibit very interesting dynamics both before and after portfolio inclusion. The average cash-to- assets ratio of the high ECM firms increases monotonically each year τ , from 0.19 in τ = −10 to 0.34 in τ = 0, and then monotonically declines to 0.17 in τ = 10. The dynamics of cash holdings of the low ECM group are exactly opposite. Their cash-to-assets ratio falls from 0.09 in τ = −10 to 0.02 in τ = 0 and then rises to 0.07 in τ = 10. These patterns in cash holdings, coupled with the investment dynamics, are consistent with the idea that low excess cash firms either lack investment opportunities or lack liquid resources to take advantage of such opportunities, while high excess cash firms gradually build up their cash reserves and then use their savings for investment purposes. 2.5 Conclusion This paper documents a positive relationship between corporate excess cash holdings and future stock returns. Firms with high excess cash outperform their low excess cash peers by 0.40% per month. Neither market nor three- and four-factor asset pricing models can explain this difference in returns. Contrary to the intuition that cash is particularly valuable in market downturns, I find between firm asset liquidity and stock liquidity. Among other things, they demonstrate that this relationship is weaker when deployment uncertainty is high, which happens when a manager is expected to transform liquid assets such as cash into illiquid assets such as investments. 23The positive link between excess cash and future 12-month returns is particularly interesting given the lack of a relationship between excess cash and profitability over the following decade. However, in unreported results, I find that starting two years following portfolio assignment, stocks of high excess cash firms do not perform significantly differently from those of low excess cash firms. 18 that in such times, stocks of firms with high excess cash perform worse than those of companies with lower levels. Although cash is less risky than assets in place, I show that high excess cash firms have larger market betas. Finally, I find that future investment activity is strongly and positively related to excess cash, with differences in investment persisting for up to ten years, but I observe no significant relationship between excess cash and future profitability. Thus, the empirical evidence suggests that firms build cash reserves in anticipation of future investment. These firms have or are acquiring growth options, as is reflected by their higher market betas. They are, therefore, riskier than their low excess cash peers and earn higher returns. During market downturns, growth options of high excess cash firms become less valuable, as is reflected in their lower returns during these times. However, during expansions, these companies have readily available resources to take advantage of investment opportunities. In the future, high excess cash firms exercise their growth options as is evidenced by their dramatically higher investment spending over the following years. Some findings of this paper are puzzling and warrant further research. In particular, it is interesting that high excess cash firms exhibit poor accounting performance over the course of a decade following portfolio assignment. If overinvestment is the reason for the poor profitability of such companies, the results of this paper raise questions regarding proper use of resources by the firms with large excess cash balances and, more generally, about the ability of managers to pick optimal levels of cash holdings. 19 Table 2.1: Determinants of Cash Holdings Slope t-stat Intercept -2.370 -51.243 MB 0.122 17.754 Size -0.074 -14.249 CPX -1.984 -11.201 WC -1.345 -13.456 LTD -1.881 -18.572 RD 0.519 2.653 CF 0.726 4.691 σIND 4.873 11.184 Div 0.006 0.155 R2 23.413 This table reports the results of the cross-sectional regressions used to estimate excess cash measures. Excess cash for firm i as of the end of month t is estimated as the residual it from the cross-sectional regression Ciτ = γ0t + γ1tMBiτ + γ2tSizeiτ + γ3tCPXiτ + γ4tWCiτ + γ5tLTDiτ + γ6tRDiτ + γ7tCFiτ + γ8tσ IND iτ + it, where C is the log of ratio of cash to total assets less cash; market-to-book ratio MB is measured as the book value of assets, less the book value of equity, plus the market value of equity, divided by assets; Size is the log of real (adjusted by CPI) assets; CPX is the ratio of capital expenditures to assets; WC is the ratio of net working capital calculated without cash to assets; LTD is the ratio of long-term debt to assets; RD is the ratio of research and development expenses to sales; CF is defined as operating income before depreciation less interest less dividends less taxes divided by total assets; and σIND is the mean of standard deviations of CF over 10 years for firms in the same 2-digit SIC industry. Regressions also include a dividend dummy, Div, and industry dummies based on Kenneth French’s 17 industry definitions. Each cross-sectional regression uses all firms that have fiscal year ends between t−11 and t. τ refers to the fiscal year ending between t−11 and t. All variables with the τ subscript thus use the most recent data for firm i. Reported are average coefficients of December cross-sectional regressions, corresponding t-statistics, and average adjusted R2 values. Sample period is 1960-2006. 20 Table 2.2: Characteristics of ECM Deciles ECM Cash β BM Size ROA CF Debt Accr Ag Low 0.017 0.861 -0.444 -0.336 0.057 -0.014 0.217 -0.027 0.140 2 0.031 0.900 -0.364 0.099 0.099 0.029 0.229 -0.019 0.147 3 0.045 0.928 -0.367 0.245 0.106 0.035 0.233 -0.017 0.151 4 0.064 0.942 -0.375 0.272 0.107 0.035 0.236 -0.024 0.157 5 0.087 0.954 -0.376 0.263 0.106 0.033 0.227 -0.025 0.187 6 0.113 0.971 -0.397 0.300 0.112 0.037 0.215 -0.028 0.142 7 0.147 0.993 -0.394 0.306 0.113 0.036 0.205 -0.029 0.160 8 0.193 1.005 -0.413 0.232 0.109 0.032 0.189 -0.032 0.153 9 0.258 1.037 -0.449 0.093 0.097 0.019 0.178 -0.034 0.173 High 0.420 1.024 -0.551 -0.469 0.042 -0.036 0.154 -0.040 0.232 High-Low 0.403 0.162 -0.107 -0.133 -0.015 -0.021 -0.063 -0.013 0.093 [19.08] [5.37] [4.44] [5.28] [2.58] [5.18] [12.26] [3.8] [4.56] This table reports selected average characteristics of each excess cash measure (ECM) decile to which firms are assigned as of the end of each calendar year τ . Cash is the most recently available ratio of cash to total assets; τ is the beta obtained from market model regressions using daily data from year τ with one lead and lag of market excess return; BM is the log of book-to-market ratio measured as in Davis et al. (2000); Size is the log of real (adjusted by CPI) assets; ROA is operating income before depreciation over assets; CF is operating income before depreciation less interest less dividends less taxes over total assets; Debt is measured as the ratio of long-term debt to long-term debt plus market value of equity; Accr, Accruals, is calculated as [(change in current assets - change in cash) - (change in current liabilities - change in short-term debt - change in taxes payable) - depreciation expense] / average total assets; and Ag is asset growth defined as the ratio of total assets to lagged total assets minus one. t-statistics for the difference between High and Low values are reported in square brackets. The sample period is 1960-2006. 21 Table 2.3: ECM Decile Portfolio Returns Period Low ECM2 ECM3 ECM4 ECM5 ECM6 ECM7 ECM8 ECM9 High High-Low 1960-2006 0.852 0.939 1.060 1.105 1.107 1.185 1.176 1.249 1.255 1.253 0.401 [3.66] [4.03] [4.53] [4.72] [4.64] [4.89] [4.88] [5.13] [4.99] [4.68] [4.19] 1960-1982 0.913 0.949 1.085 1.123 1.060 1.113 1.127 1.200 1.255 1.242 0.330 [2.60] [2.69] [3.10] [3.20] [2.96] [3.19] [3.34] [3.58] [3.67] [3.54] [3.51] 1983-2006 0.794 0.929 1.036 1.088 1.152 1.254 1.223 1.296 1.255 1.262 0.469 [2.57] [3.02] [3.31] [3.48] [3.63] [3.72] [3.55] [3.67] [3.41] [3.14] [2.85] This table reports average raw value-weighted returns, in percent per month, and the corre- sponding t-statistics for different excess cash measure (ECM) deciles as well as for the difference between deciles of high and low ECM for different time periods. Stocks are first sorted into quintiles based on market betas, and then into ECM deciles within each beta quintile. At the beginning of each month t, an investment is made in the stocks that were assigned to a particular ECM decile as of the end of month t − 5, and the position is held without rebalancing for the following 12 months. 22 Table 2.4: Fama-MacBeth Regression Results ECM β BM ME Ag Accr I CF Debt RU12 Issue (1) 0.084 [3.78] (2) 0.105 -0.117 0.177 -0.151 [5.34] [1.14] [3.46] [3.24] (3) 0.093 -0.668 [4.24] [6.06] (4) 0.067 -2.254 [2.95] [7.42] (5) 0.067 -1.411 [2.93] [3.83] (6) 0.084 0.171 [3.93] [0.29] (7) 0.088 0.350 [4.25] [1.70] (8) 0.079 0.310 [3.81] [1.97] (9) 0.079 -0.600 [3.57] [4.59] (10) 0.074 -0.061 0.085 -0.182 -0.410 -1.784 -0.097 1.157 -0.119 0.149 -0.328 [3.69] [0.68] [1.94] [4.46] [4.08] [6.17] [0.32] [2.71] [0.71] [1.10] [3.42] This table reports the results of Fama-MacBeth 1973 regressions. Every month stock returns in month t, in percent, are regressed on ECM, excess cash measure; β is beta obtained from market model regressions using daily data from t − 16 to t − 5 with one lead and lag of market excess return; BM, log of book-to-market ratio, measured as in Davis et al. (2000); ME, log of market capitalization measured as of the end of t − 1; Ag, asset growth, defined as the ratio of total assets to lagged total assets minus one; Accr, Accruals, calculated as [(change in current assets - change in cash) - (change in current liabilities - change in short-term debt - change in taxes payable) - depreciation expense] / average total assets; I, Investment, defined as capital expenditures plus acquisitions less sale of property, plant and equipment, divided by total assets; CF, cash flow, computed as operating income before depreciation less interest less dividends less taxes over total assets; Debt, estimated as the ratio of long-term debt to long-term debt plus market value of equity; RU12, 12-month (t−12 to t−1) compounded return; and Issue, measured as Ln[MEt−1/MEt−36] − RU36, where MEt is market capitalization as of the end of month t, and RU36 is the log 3-year buy-and-hold return ending in month t − 1. Reported are average coefficients and t-statistics. Accounting data is taken from annual report for the fiscal year ending between t − 16 and t − 5. ECM is computed as of the end of month t − 5. Sample period is 1960-2006. 23 Table 2.5: Time Series Characteristics ECM MKT HML SMB UMD AGF ACCRF DEBTF MEAN 0.40 0.93 0.47 0.22 0.83 -0.83 -0.69 0.18 STD 2.27 4.35 2.85 3.16 3.91 3.82 2.57 3.62 SHARPE 0.18 0.21 0.16 0.07 0.21 -0.22 -0.27 0.05 SKEW 2.39 -0.47 0.02 0.54 -0.65 -1.35 -0.28 -1.19 KURT 5.19 2.02 2.61 5.77 5.70 6.41 2.17 12.57 MIN -12.58 -22.54 -12.40 -16.79 -25.06 -28.41 -11.34 -31.72 MAX 25.66 16.56 13.85 21.96 18.39 10.85 9.57 14.29 Correlation Coefficients ECM 1.00 0.23 -0.47 0.34 0.14 -0.24 -0.17 -0.54 MKT 0.23 1.00 -0.41 0.30 -0.07 0.14 0.13 -0.12 HML -0.47 -0.41 1.00 -0.28 -0.12 -0.23 -0.19 0.67 SMB 0.34 0.30 -0.28 1.00 0.01 -0.30 0.01 -0.13 UMD 0.14 -0.07 -0.12 0.01 1.00 0.01 -0.09 -0.21 AGF -0.24 0.14 -0.23 -0.30 0.01 1.00 0.57 0.01 ACCRF -0.17 0.13 -0.19 0.01 -0.09 0.57 1.00 0.01 DEBTF -0.54 -0.12 0.67 -0.13 -0.21 0.01 0.01 1.00 This table reports selected time series characteristics of the monthly returns (in percent) of portfolio of high minus low excess cash measure deciles (ECM), value-weighted market index (MKT), as well as value (HML), size (SMB), momentum (UMD), asset growth (AGF), accruals (ACCRF), and Debt (DEBTF) factors. Returns of AGF, ACCRF, and DEBTF factors are calculated by taking a value-weighted long position in the decile of stocks with the highest Ag, Accr, and Debt measures, respectively, and an offsetting short position in the decile of stocks with the lowest values. Reported are averages, standard deviations, Sharpe ratios, skewness, kurtosis, as well as lowest and highest monthly returns. The bottom eight rows report correlation coefficients. Sample period is 1960-2006. 24 Table 2.6: Unconditional Risk Adjustment Intercept Mktrf HML SMB UMD AGF ACCRF DEBTF R2 (1) 0.401 [4.19] (2) 0.344 0.119 5.06 [3.67] [5.57] (3) 0.518 -0.003 -0.325 0.162 26.16 [6.10] [0.15] [10.13] [5.84] (4) 0.468 0.003 -0.313 0.162 0.049 26.73 [5.36] [0.16] [9.66] [5.84] [2.32] (5) 0.281 -0.144 5.66 [2.96] [5.90] (6) 0.298 -0.150 2.70 [3.05] [4.08] (7) 0.462 -0.340 29.21 [5.74] [15.27] (8) 0.330 -0.124 -0.042 -0.338 34.67 [4.10] [5.03] [1.13] [15.82] (9) 0.291 0.053 -0.160 0.104 0.020 -0.115 -0.096 -0.229 42.59 [3.67] [2.76] [3.84] [3.76] [1.05] [4.30] [2.68] [7.87] This table reports the results of unconditional regressions of returns (in percent per month) from High minus Low excess cash measure (ECM) portfolio (value-weighted returns are used) on market excess return (Mktrf), value (HML), size (SMB), momentum (UMD), asset growth (AGF), accruals (ACCRF), and leverage (DEBTF) factors. Mktrf, HML, SMB, and UMD are from Kenneth French’s data library. Returns of AGF, ACCRF, and DEBTF factors are calculated by taking a value-weighted long position in the decile of stocks with the highest Ag, Accr, and Debt measures, respectively, and an offsetting short position in the decile of stocks with the lowest values. Reported are regression coefficients, t-statistics, and adjusted R2 values. Sample period is 1960-2006. 25 Table 2.7: ECM Decile Portfolio Returns Conditional on Market State Mkt Low ECM2 ECM3 ECM4 ECM5 ECM6 ECM7 ECM8 ECM9 High High-Low Low -5.961 -6.135 -6.076 -6.096 -6.181 -6.101 -6.146 -6.024 -6.165 -6.274 -0.313 [13.30] [13.74] [13.64] [13.88] [14.23] [13.69] [14.09] [13.50] [13.66] [12.81] [1.78] 2 -1.689 -1.501 -1.496 -1.521 -1.475 -1.581 -1.424 -1.316 -1.423 -1.546 0.143 [7.68] [7.27] [7.75] [7.54] [7.32] [7.47] [7.04] [6.23] [6.23] [6.62] [1.03] 3 1.276 1.408 1.522 1.687 1.667 1.685 1.572 1.545 1.618 1.588 0.312 [5.52] [6.84] [7.22] [8.25] [7.87] [7.50] [7.72] [8.33] [6.93] [6.43] [2.26] 4 3.888 3.745 4.129 4.124 4.154 4.387 4.406 4.443 4.672 4.660 0.772 [14.84] [15.87] [17.80] [17.78] [16.58] [15.19] [14.39] [13.05] [12.60] [10.60] [2.67] High 6.686 7.115 7.159 7.269 7.305 7.470 7.409 7.535 7.507 7.768 1.082 [18.26] [20.65] [20.76] [22.30] [19.79] [21.98] [21.62] [21.19] [20.51] [17.63] [4.20] This table reports average value-weighted returns, in percent per month, and the corresponding t-statistics for different excess cash measure (ECM) deciles as well as for the difference between deciles of high and low ECM for each market return group. To determine market return quintiles, months from January 1960 to December 2006 are assigned into 5 groups based on market return in that month. Sample period is 1960-2006. 26 Table 2.8: Annual Profitability, Investment Activity, Cash Holdings and Leverage Around Inclusion into High or Low ECM Groups High ECM Low ECM Rel Yr ROA Inv Leverage Cash ROA Inv Leverage Cash -10 0.151 0.100 0.172 0.188 0.142 0.097 0.215 0.087 -9 0.147 0.103 0.169 0.194 0.137 0.097 0.215 0.086 -8 0.140 0.107 0.169 0.199 0.134 0.097 0.218 0.084 -7 0.137 0.108 0.169 0.206 0.131 0.101 0.216 0.081 -6 0.130 0.110 0.168 0.213 0.128 0.103 0.217 0.079 -5 0.123 0.113 0.169 0.221 0.124 0.106 0.217 0.074 -4 0.112 0.115 0.170 0.231 0.119 0.108 0.221 0.068 -3 0.104 0.115 0.170 0.246 0.114 0.107 0.223 0.063 -2 0.095 0.116 0.169 0.266 0.106 0.108 0.225 0.057 -1 0.085 0.116 0.166 0.293 0.098 0.109 0.227 0.046 0 0.070 0.119 0.166 0.339 0.078 0.107 0.223 0.024 1 0.061 0.138 0.169 0.278 0.097 0.090 0.239 0.043 2 0.069 0.128 0.175 0.250 0.106 0.091 0.247 0.051 3 0.079 0.123 0.181 0.231 0.110 0.092 0.249 0.056 4 0.079 0.121 0.186 0.218 0.115 0.093 0.251 0.060 5 0.083 0.119 0.189 0.206 0.117 0.096 0.254 0.063 6 0.094 0.115 0.192 0.196 0.120 0.096 0.257 0.065 7 0.099 0.113 0.193 0.188 0.120 0.096 0.259 0.067 8 0.096 0.110 0.198 0.182 0.122 0.096 0.263 0.069 9 0.098 0.110 0.203 0.174 0.123 0.096 0.262 0.070 10 0.106 0.106 0.205 0.169 0.126 0.097 0.261 0.073 This table reports average return on assets (ROA), calculated as is operating income before depreciation over assets; investment-to-assets ratios (Inv), defined as capital expenditures plus acquisitions less sale of property, plant and equipment, divided by total assets; leverage, estimated as the ratio of long-term debt to long-term debt plus market value of equity; and cash-to-assets ratios (Cash) of high and low excess cash measure (ECM) quintiles in each of the 10 years preceding and following inclusion in the corresponding ECM portfolio. Sample period is 1960- 2006. 27 Figure 2.1: Time Series of High Minus Low ECM Portfolio Returns  ! " ! # $% & ' ( & ) ' * + , $ - . ( & $/ 0 , $ / ! ! 1  !" # $% &' () #* $' +$ (% #,-./.. #,0./.. ###./.. ##0./.. ##-./.. ##1./.. 023. 0234 0235 026- 0263 025. 0254 0255 022- 0223 -... -..4 -..5 7/#89(%:;<# !"# $%&'(=  ! " ! # $% & ' ( & ) ' * + , $ - . ( & $/ 0 , $ / ! ! 1  ! "# $ %& '( ) ###*+** ###*+,* ###-+** ###-+,* ###.+** ###.+,* -/0* -/01 -/02 -/3. -/30 -/2* -/21 -/22 -//. -//0 .*** .**1 .**2 4+#5'6'78&9:%# !"#$%&'() This figure plots in Panel A average raw monthly returns (in percent) of the portfolio that is long the decile of high excess cash firms and short the decile of low excess cash firms. Panel B shows log of cumulative monthly return of this portfolio. Sample period is 1960-2006. 28 Figure 2.2: Excess Cash and Investment.  ! " ! # $% & ' ( & ) ' * + , $ - . ( & $/ 0 , $ / ! ! 1  ! "# $% & #! %'% (' )$ $# %$ *+ ,% -( ****./0 ***10/0 ***11/0 ***12/0 ***13/0 ***14/0 0 1 2 3 4 5 6 7 8 . 10 )/*)99*:-;&$ <#,;$*:(99(=-!>*?(;%@(9-(*)$$->!&#!% A(= BC-!%-9#*2 BC-!%-9#*3 BC-!%-9#*4 D->E  ! " ! # $% & ' ( & ) ' * + , $ - . ( & $/ 0 , $ / ! ! 1  ! "# $% & #! %'% (' )$ $# %$ *+ ,% -( ****./0 ***10/0 ***11/0 ***12/0 0 1 2 3 4 5 6 7 8 . 10 9/*:;<"-"(<$*=!>? @#,<$*A(>>(B-!C*D(<%E(>-(*)$$-C!&#!% F(B G;-!%->#*2 G;-!%->#*3 G;-!%->#*4 H-CI This figure plots average investment-to-assets ratios (in percent) for each excess cash quintile during the year of portfolio assignment and the subsequent ten years. Investment is defined as capital expenditures plus acquisitions less the sale of property, plant, and equipment divided by total assets. Panel A uses all firms, while Panel B uses just those that survived for the entire ten years. The sample period is 1960-2006. 29 Figure 2.3: Time Series of High Minus Low ECM Portfolio Returns  ! " ! # $% & ' ( & ) ' * + , $ - . ( & $/ 0 , $ / ! ! 1  !" #$ %& '% &( )) !" ) &&&&*+, &&&&-+, &&&&.+, &&&&/+, &&&0,+, &&&00+, &&&01+, &&&02+, &&&03+, , 0 1 2 3 4 * - . / 0, (+&(55&67$8) 9!:$)&6'55';7%<&='$">'57'&())7<%8!%" ?'; @#7%"75!&1 @#7%"75!&2 @#7%"75!&3 A7<B  ! " ! # $% & ' ( & ) ' * + , $ - . ( & $/ 0 , $ / ! ! 1  !" #$ %& '% &( )) !" ) &&&**+, &&&*-+, &&&*.+, &&&*/+, &&&*0+, &&&*1+, , * - . / 0 1 2 3 4 *, 5+&6#$787'$)&9%:; <!=$)&>'::'?8%@&A'$"B':8'&())8@%C!%" D'? E#8%"8:!&- E#8%"8:!&. E#8%"8:!&/ F8@G This figure plots average return on assets (in percent) for each excess cash quintile during the year of portfolio assignment and the subsequent ten years. Profitability is defined as operating income before depreciation divided by total assets. Panel A uses all firms, while Panel B uses just those that survived for the entire ten years. The sample period is 1960-2006. 30 Bibliography Almeida, H., Campello, M., and Weisbach, M. S. (2004). The cash flow sensitivity of cash. Journal of Finance, 59(4): 1777–1804. Anderson, C. W. and Garcia-Feijóo, L. (2006). Empirical evidence on capital investment, growth options, and security returns. Journal of Finance, 61(1): 171–194. Bates, T. W., Kahle, K. M., and Stulz, R. M. (2009). Why do us firms hold so much more cash than they used to? Journal of Finance, 64(5): 1985–2021. Berk, J. B., Green, R. C., and Naik, V. (1999). Optimal investment, growth options, and security returns. Journal of Finance, 54(5): 1553–1607. Bernardo, A. E., Chowdhry, B., and Goyal, A. (2007). Growth options, beta, and the cost of capital. Financial Management, 36(2): 5–17. Carhart, M. M. (1997). On persistence in mutual fund performance. Journal of Finance, 52(1): 57– 82. Carlson, M., Fisher, A., and Giammarino, R. (2004). Corporate investment and asset price dy- namics: Implications for the cross-section of returns. Journal of Finance, 59(6): 2577–2603. Carlson, M., Fisher, A., and Giammarino, R. (2006). Corporate investment and asset price dynam- ics: Implications for seo event studies and long-run performance. Journal of Finance, 61(3): 1009– 1034. Chen, S. S., Ho, L. C., and Shih, Y. C. (2007). Intra-industry effects of corporate capital investment announcements. Financial Management, 36(2): 125–145. Cooper, M. J., Gulen, H., and Schill, M. J. (2008). Asset growth and the cross-section of stock returns. Journal of Finance, 63(4): 1609–1651. Daniel, K. and Titman, S. (2006). Market reactions to tangible and intangible information. Journal of Finance, 61(4): 1605–1643. Davis, J. L., Fama, E. F., and French, K. R. (2000). Characteristics, covariances, and average returns: 1929 to 1997. Journal of Finance, 55(1): 389–406. Dimson, E. (1979). Risk measurement when shares are subject to infrequent trading. Journal of Financial Economics, 7(2): 197–226. 31 Dittmar, A. and Mahrt-Smith, J. (2007). Corporate governance and the value of cash holdings. Journal of Financial Economics, 83(3): 599–634. Fama, E. F. and French, K. R. (1992). The cross-section of expected stock returns. Journal of Finance, 47(2): 427–465. Fama, E. F. and French, K. R. (1993). Common risk-factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1): 3–56. Fama, E. F. and MacBeth, J. D. (1973). Risk, return, and equilibrium - empirical tests. Journal of Political Economy, 81(3): 607–636. Faulkender, M. and Wang, R. (2006). Corporate financial policy and the value of cash. Journal of Finance, 61(4): 1957–1990. Foley, C. F., Hartzell, J. C., Titman, S., and Twite, G. (2007). Why do firms hold so much cash? A tax-based explanation. Journal of Financial Economics, 86(3): 579–607. Gamba, A. and Triantis, A. (2008). The value of financial flexibility. Journal of Finance, 63(5): 2263–2296. Gomes, J., Kogan, L., and Zhang, L. (2003). Equilibrium cross section of returns. Journal of Political Economy, 111(4): 693–732. Gopalan, R., Kadan, O., and Pevzner, M. (2010). Asset liquidity and stock liquidity. Washington University in St. Louis Working Paper. Harford, J., Mikkelson, W., and Partch, M. (2003). The effect of cash reserves on corporate investment and performance in industry downturns. University of Washington Working Paper. Haugen, R. A. and Baker, N. L. (1996). Commonality in the determinants of expected stock returns. Journal of Financial Economics, 41(3): 401–439. Hovakimian, G. (2009). Determinants of investment cash flow sensitivity. Financial Management, 38(1): 161–183. Jegadeesh, N. and Titman, S. (1993). Returns to buying winners and selling losers - implications for stock-market efficiency. Journal of Finance, 48(1): 65–91. John, T. A. (1993). Accounting measures of corporate liquidity, leverage, and costs of financial distress. Financial Management, 22(3): 91–100. Kim, C. S., Mauer, D. C., and Sherman, A. E. (1998). The determinants of corporate liquidity: Theory and evidence. Journal of Financial and Quantitative Analysis, 33(3): 335–359. McLean, R. D. (2009). Share issuance and cash savings. University of Alberta Working Paper. 32 Mikkelson, W. H. and Partch, M. M. (2003). Do persistent large cash reserves hinder performance? Journal of Financial and Quantitative Analysis, 38(2): 275–294. Morellec, E. and Nikolov, B. (2009). Cash holdings and competition. Ecole Polytechnique Federale de Lausanne Working Paper. Opler, T., Pinkowitz, L., Stulz, R., and Williamson, R. (1999). The determinants and implications of corporate cash holdings. Journal of Financial Economics, 52(1): 3–46. Palazzo, D. (2009). Firms cash holdings and the cross-section of equity returns. Boston University Working Paper. Pinkowitz, L., Stulz, R., and Williamson, R. (2006). Does the contribution of corporate cash holdings and dividends to firm value depend on governance? A cross-country analysis. Journal of Finance, 61(6): 2725–2751. Pinkowitz, L. and Williamson, R. (2004). What is a dollar worth? The market value of cash holdings. Georgetown University Working Paper. Riddick, L. A. and Whited, T. M. (2009). The corporate propensity to save. Journal of Finance, 64(4): 1729–1766. Simutin, M. (2009). Excess cash and mutual fund performance. University of British Columbia Working Paper. Sloan, R. G. (1996). Do stock prices fully reflect information in accruals and cash flows about future earnings? Accounting Review, 71(3): 289–315. Titman, S., Wei, K. C. J., and Xie, F. X. (2004). Capital investments and stock returns. Journal of Financial and Quantitative Analysis, 39(4): 677–700. Yan, X. M. (2006). The determinants and implications of mutual fund cash holdings: Theory and evidence. Financial Management, 35(2): 67–91. Zhang, L. (2005). The value premium. Journal of Finance, 60(1): 67–103. 33 Chapter 3 Excess Cash and Mutual Fund Performance 3.1 Introduction Cash holdings of mutual funds can differ dramatically even for seemingly comparable funds.24 For example, at the end of 2007 close to one tenth of U.S. actively managed mutual funds with a growth objective held more than 10% of their total net assets in cash. For another tenth of the funds, this number was below 0.4%. Such striking differences in cash positions of funds competing with each other and pursuing the same objective are puzzling, yet the sources of these differences and the effects they have on future fund performance have received limited attention in the literature.25 In this paper, I document new important determinants of mutual fund cash holdings and study how cash balances in excess of the level needed to conduct normal operations (“excess cash”) impact fund performance. I emphasize excess cash because, as a discretionary amount, it has the potential to capture information about otherwise unobservable fund characteristics that affect fund performance. Information captured by excess cash may reflect, among other things, stock-picking skills, market-timing abilities, the investment opportunity set of the manager, and managerial expectations about liquidity needs of the fund. I define excess cash both (i) empirically, as the residual from cross-sectional regressions of cash- to-total net assets ratio on fund characteristics, and (ii) theoretically, as the difference between actual cash position and the target balance predicted by a model of optimal fund cash holdings that I develop. Using either definition, I find that while raw cash relates only weakly to future fund returns, funds with high excess cash holdings outperform those with low excess cash by statistically significant and economically important 2% per year. After standard risk-adjustment (e.g., controlling for the three factors of Fama and French, 1993) this difference in returns reaches nearly 3% annually. To understand why high excess cash funds outperform their low excess cash peers, it is helpful to recognize that fund cash holdings are affected by exogenous flows, which include withdrawals, deposits and dividends, and by endogenous managerial decisions about purchases and sales, which 24A version of this chapter will be submitted for publication. Simutin, M. (2010) Excess Cash and Mutual Fund Performance. 25The two main exceptions are Chordia (1996) and Yan (2006) who study the link between cash holdings and a number of fund characteristics. Yan additionally focuses on the relationship between aggregate cash holdings of mutual funds and future market returns. 34 in turn affect expenses incurred by the fund. Because I control for the differences in recent fund flows in defining excess cash, it is unlikely that the positive relationship between excess cash and fund performance is due to fund flow shocks. Instead, I conjecture that it is attributable to managerial decision to adjust the fund’s cash holdings. Adjustments to cash positions may reflect (i) managerial proficiency at controlling transaction costs of the fund, (ii) the manager’s stock-picking abilities and investment opportunities, (iii) managerial market-timing skills, and (iv) the manager’s aptitude at anticipating future fund flows. I develop these four hypotheses in detail and find empirical evidence supporting each conjecture. I first explore whether high excess cash proxies for the ability to control fund expenses. I develop a model of costly stock trading which suggests that relative to a manager who either invests all sales proceeds immediately and/or who transacts more frequently than is optimal, a cost-minimizing manager tends to carry a higher cash balance. The intuition behind this result is straightforward: to reduce price pressure, a cost-minimizing manager has to make more trips to the market when purchasing an illiquid stock than when selling a liquid stock. As a result, he carries excess cash during the course of adjusting portfolio composition. The model can thus justify the positive link between high cash positions and performance: managers carrying greater cash balances may be doing so as a result of their efforts to minimize transaction costs, and therefore they outperform their low excess cash peers. Consistent with the model, I find that future fund expenses decline with excess cash. I also consider the hypothesis that excess cash proxies for manager’s stock-picking abilities. Cash tends to earn a lower return than equities, and therefore unskilled managers may prefer to remain fully invested in stocks to attempt to match benchmark returns. On the other hand, a skilled manager who cannot presently find any attractive investment opportunities may carry a higher cash balance. In the future the manager will invest the excess cash as such opportunities become available.26 It is thus natural to expect that shares bought by high excess cash funds outperform those purchased by their low excess cash counterparts. I explore stock purchases and sales by mutual funds and find that high excess cash funds do in fact purchase stocks that significantly outperform purchases of the low excess cash group. Additions to the positions already held by high excess cash funds outperform those of the low excess cash group by 2% per year. The relationship between excess cash and future fund performance is thus consistent with superior ability of high excess cash fund managers to identify undervalued stocks that generate higher future returns. Interestingly, stocks sold by the high excess cash funds also outperform those sold by their low excess cash peers, suggesting that low excess cash fund managers are more skilled at identifying overvalued stocks. The managers of such funds may purposefully carry low excess cash because 26One can conjecture that skilled manager will allocate excess cash into stocks or exchange-traded funds while waiting for better investment opportunities. However, buying opportunities can arguably be more easily found following market dips, and thus not only will the this allocation fall in value due to the dip but it may also suffer as the manager tries to convert it back into cash. It is also important to distinguish this concept of identifying attractive investment opportunities from a related idea of timing the overall market. For example, Warren Buffett, whose company always carries a high cash position, is routinely praised for his ability to make successful investments in individual companies, but he has repeatedly denied that he attempts to time the market. 35 they are convinced that they can raise funds to cover cash shortfalls by disposing of those shares that are likely to underperform in the future. Thus, excess cash does not just relate to broad stock-picking skills, but specifically proxies for the ability to identify overvalued or undervalued equities. Alternatively, managers with high (low) excess cash may realize that they will need to reduce (increase) their cash positions to some target level, and may thus find it optimal to invest in applying the stock buying (selling) skills, which is reflected in the performance of their future trades. I also explore whether the positive relationship between excess cash and fund performance additionally relates to market-timing abilities. A skillful market-timer will naturally build up the fund’s cash position prior to a market downturn, but will at the same time carry a low cash balance before a period of strong market performance. If market-timing skills are mainly concentrated in the ability to predict market downturns, market timing may explain the strong performance by high excess cash funds relative to the low excess cash group. I use the traditional techniques of Treynor and Mazuy (1966) and Henriksson and Merton (1981) to find that market-timing skills are worse for the low excess cash group. As a result, the portfolio that is long the high excess cash funds and short the low excess cash group exhibits positive but statistically insignificant market timing. However, I find strong evidence of market-timing skills among funds with volatile excess cash holdings: managers of such funds actively adjust their excess cash positions in response to their changing expectations about future market returns. Finally, the positive link between excess cash and future fund performance may also relate to liquidity needs of the fund, in particular to future fund flows. If a manager fails to anticipate fund outflows and does not have sufficient cash on hand to meet such outflows, he will be forced to liquidate some of his shareholdings at a potentially disadvantageous time and price.27 Thus, it is conceivable that the difference in future performance of high and low excess cash funds relates to the superior ability of managers of high excess cash funds to anticipate fund outflows. I find that the difference in performance of top and bottom excess cash groups is particularly pronounced when future fund flows are low and is somewhat weaker when fund flows are high. This evidence is consistent with the notion that low excess cash funds do not carry sufficient cash on hand to cover outflows and are likely forced to liquidate some of their holdings, damaging fund performance. The high excess cash group, on the other hand, is well positioned to meet fund outflows and generates better returns. To determine whether the positive relationship between excess cash and mutual fund perfor- mance relates to the superior ability of the high excess cash fund managers to anticipate fund outflows, it is natural to explore whether high excess cash closed-end funds outperform their low excess cash peers. Unlike their open-end counterparts, closed-end funds rarely issue or retire shares, and shares are not normally redeemable until the fund liquidates. Thus, uncertainty about fund 27In the model of optimal cash holdings that I develop, managers take into account expected fund flows when determining their cash balances, and thus there are no differences in managerial abilities to anticipate flows. Empir- ically, however, managers can certainly differ in such abilities; in particular, some managers may be able to forecast the level of fund outflows more accurately than others by making use of unobservable fund characteristics. 36 flows do not motivate closed-end funds to carry cash balances. Empirically, I find no relationship between excess cash of closed-end funds and their performance. Central to my analysis is the definition of excess cash. To calculate excess cash, I thoroughly explore the determinants of cash holdings of mutual funds. Chordia (1996) and Yan (2006) show that fund cash holdings relate to factors such as load fees, fund size, level and volatility of fund flows, and prior performance. I complement their work by identifying additional important determinants of fund cash holdings. In particular, I show that funds holding riskier stocks, as proxied for by the average market beta of their shareholdings, carry more cash. This evidence can be interpreted as consistent with the notion that mutual funds manage the overall risk of their portfolio by adjusting their cash position: managers who have a preference for holding riskier stocks decrease total fund risk by carrying more cash. I also find that funds holding illiquid stocks tend to hold more cash. The cost of selling their holdings to cover unexpected cash shortfalls is potentially large for such funds, justifying the need to maintain larger cash balances. I additionally show that funds with higher return gap, the difference between realized fund returns and the returns on a passive portfolio of fund’s recently reported holdings (Kacperczyk et al., 2008), carry less cash. On the whole, compared to the determinants of fund cash holdings studied in the prior literature, the characteristics I consider explain three times more cross-sectional variation in cash positions. The rest of the paper proceeds as follows. Section 3.2 describes the data and summary statistics. In Section 3.3, I explore the determinants of fund cash holdings. Section 3.4 provides details of excess cash estimation and documents the positive relationship between excess cash and future fund performance. In Section 3.5, I present a model of optimal cash holdings and study the link between excess cash defined relative to a model-based target level and fund performance. Section 3.6 analyzes the sources of the positive relationship between excess cash and fund performance. Section 3.7 concludes. 3.2 Data and Summary Statistics 3.2.1 Data and Sample I obtain fund returns, cash holdings, investment objectives, fees, total net assets (TNA), and other fund characteristics from the Center for Research in Security Prices (CRSP) Survivor-Bias-Free Mutual Fund Database. I use Wharton Research Data Services (WRDS) mflink file to merge this database with Thomson Financial Mutual Fund Holdings, which contains information on fund stock portfolios.28 I restrict my analysis to diversified domestic equity mutual funds with aggressive growth, long- term growth, or growth and income objectives. I exclude international, balanced, sector, bond, money market, and index funds from the analysis.29 The CRSP database details fund asset com- positions including cash balances annually until the end of 1998 and quarterly thereafter, but as 28Wermers (2000) describes the databases and the merge procedure in great detail. 29Appendix A provides details of determining investment objectives of the funds. 37 Yan (2006) notes, the exact asset composition dates are not available prior to the 1990s. Further- more, the CRSP database does not report monthly total net assets prior to 1992, complicating calculations of level and volatility of fund flows, and certain variables (e.g., 12b-1 fees) are not reported prior to 1992. For these reasons, I focus my analysis on the 1992-2008 period. I limit my sample to funds with at least 50% of fund’s assets invested in equities, and to keep the focus on the funds that do not borrow heavily to invest, I require all funds to have positive cash holdings. I also exclude funds with TNA less then $15 million as Elton et al. (2001) show that the returns of such small funds tend to be biased upwards in the CRSP database. I additionally remove the first 18 months of returns for each fund in the sample to reduce the effect of an incubator fund bias documented by Evans (2006). Relaxing either of these restrictions does not qualitatively affect the results of this paper. Many mutual funds have multiple share classes, which typically differ only in fee structure (e.g., load vs. no load) and target clientele (e.g., institutional vs. retail). These share classes represent claims on the same underlying assets, have the same gross returns and the same cash and stock holdings; however, they are identified as separate funds in the CRSP database. For the purposes of this study, I combine such share classes into a single fund. In particular, I calculate TNA of each fund as the sum of TNAs of all share classes of that fund and define fund age as the maximum age of its share classes. For all other fund characteristics, I use the TNA-weighted average over the share classes. My final sample contains 17,242 fund-year observations representing 3,009 distinct funds. 3.2.2 Summary Statistics Table 3.1 presents summary statistics for selected fund characteristics. The average fund holds 5% of assets in cash, with median cash holdings of 3.3%. There are considerable cross-sectional differences in cash holdings: over the entire sample period, the average 10th percentile of holdings was just 0.66%, while the funds in the 90th percentile held over 11% in cash. Fund cash holdings have also been changing dramatically over time. Figure 3.1 plots the time series of average and median fund holdings. Assets held in cash have been steadily declining over the sample period: in early 1990s average (median) cash holdings amounted to nearly 10% (8%), while in 2007 the corresponding values were just 3.3% (2%). Exploring the reasons for this reduction in cash holdings over the last two decades is an interesting topic but is beyond the scope of this paper. Factors contributing to this decline likely include technological innovation in cash management, changes in risk of fund holdings or other fund characteristics, changes in risk preferences of managers caused by increased competition, and other factors.30 The average fund has $1.68 billion in total net assets, expense ratio of 1.29%, 12b-1 fee of 0.41%, front load of 1.4%, deferred load of 0.5%, and turnover of 83%. In a given month, an average 30It is interesting to note that corporate cash holdings have risen nearly two-fold over the same period. Bates et al. (2009) attribute this increase to higher volatility of cash flows and changes in firm characteristics. 38 fund receives a flow equivalent to 0.5% of its assets, although a median fund sees an outflow.31 Kacperczyk et al. (2008) show that return gap, the difference between realized fund returns and the returns on a passive portfolio of fund’s reported holdings, is an important determinant of future fund performance. I follow their methodology in calculating a 12-month return gap and find it to be marginally negative at −0.18% per year. A typical fund earns approximately 1.6% dividend yield, somewhat below the dividend yield of the U.S. stocks of 2.0% over the sample period. Market beta of the funds, βMktFund, calculated from market model regression using realized fund returns over the prior 12 months, is on average below one (0.96), which is due to the presence of low-risk assets such as cash in fund portfolios. Average market beta of fund holdings, βMktHold, at 1.05, is actually above one. There is considerable variation in fund betas: a tenth of the funds have loadings below 0.63 (0.70 when estimating betas of holdings), while market betas of another tenth of the funds exceeds 1.34 (1.46). I also estimate the liquidity beta of holdings, βLiqHold, to find that while average loading on the liquidity factor is close to zero, there are large cross-sectional differences in average liquidity of fund holdings.32 I also compute a measure of change in cash attributable to purchases and sales of stocks, PRCDS. To calculate this measure, I obtain contemporaneous (time t) fund holdings and holdings from six months ago, and for each fund compute it as [100 ·∑i pi,t−3 · (−Ni,t +Ni,t−6)] / [∑i pi,t−6Ni,t−6], where Ni,t is the number of shares of stock i held by the fund at time t and pi,t is the price of this stock at time t. PRCDS thus represents the dollar amount of inflows from sales of stocks less the dollar amount spent on purchasing new securities during the prior six months, scaled by the value of stock holdings at time t− 6. I assume that stocks are purchased and sold at the price prevalent at the end of month t− 3. The negative average PRCDS of −6.6 is mainly due to the fact that this measure does not account for fund flows but is strongly and negatively related to them. The bottom panel of Table 3.1 reports correlation coefficients. Cash holdings are negatively correlated with fund size, 12b-1 fees, deferred load, return gap, dividend yield of holdings, fund market beta, and net proceeds from stock sales and purchases. Cash is positively related to expense ratio, front load fee, turnover, past return, fund flows, volatility of fund flows, and market and liquidity betas of fund holdings. 31I estimate fund flows (FF) over N -month period ending in month t as FFNt = TNAt − TNAt−N (1 +Rt−N :t) TNAt−N , where TNAt is total net assets as of the fund at the end of month t and Rt−N :t is the fund return over the N -month period ending in month t. Berk and Green (2004) recommend using TNAt as the denominator to fully capture the percentage change in new funds. My empirical results are not sensitive to using this alternative estimation method. 32To calculate beta of the holdings, for each stock the fund holds I obtain market beta from the market model regression and liquidity beta from a two-factor model with market and Pastor and Stambaugh (2003) liquidity factors. βMktHold and β Liq Hold are weighted average loadings using the dollar value of investment in each stock as weights. I use prior 12 months of monthly data for estimation. Using Sadka (2006) liquidity factor instead of Pastor and Stambaugh factor does not affect the results of the paper. 39 3.3 Determinants of Fund Cash Holdings Cash holdings represent a substantial component of the mutual fund portfolios, and ample anecdotal evidence suggests that fund managers actively adjust their cash holdings in response to market conditions and investment opportunities.33 Yet, despite their importance, the determinants of mutual fund cash holdings have received little attention in the literature. To the best of my knowledge, the only two exceptions are Chordia (1996) who links cash holdings to fund loads and uncertainty about redemptions, and Yan (2006) who shows that fund size, fund fees, and other characteristics relate to fund cash holdings. In this Section, I complement their findings by documenting additional important determinants of fund cash holdings. Table 3.2 presents the results of cross-sectional regressions of fund cash holdings on a number of characteristics. Regression (1) shows that cash is negatively related to size, which is likely attributable to economies of scale. However, consistent with the findings of Yan (2006), controlling for the expense ratio in specification (2), there is a positive link between fund size and cash holdings. Regressions (3) shows that cash positions are related positively to fund expenses and negatively to 12b-1 fees. These two variables alone explain over 2% of cross-sectional variation in cash balances. Expenses are paid with cash on hand, leading funds with higher expenses to hold more cash. Jain and Wu (2000) and Barber et al. (2005) find that fund flows are positively related to marketing 12b-1 fees, and thus funds spending more on advertising tend to hold less cash. Barber et al. (2005) also observe that fund flows are higher for funds with lower front load fees. It is thus natural to expect that funds with high front load fees hold more cash to cushion against a potential cash shortfall. Deferred loads, on the other hand, discourage fund outflows, and it is natural to expect a negative relationship between deferred loads and cash holdings. Results of specification (4) are consistent with both of these observations, but the coefficient on deferred load fee is not statistically significant. Regression (5) shows that funds with higher turnover tend to hold more cash. Turnover is positively related to the expense ratio (see Table 3.1), which may in part explain this observation. Furthermore, as a fund turns over its portfolio, it may sometimes dip into its current cash holdings to finance purchases of new securities if it has not yet sold enough shares to obtain sufficient funds to make such purchases (i.e., in situations when buys occur prior to sells). This is arguably more likely to happen in high turnover funds, causing them to carry larger cash balances. Specification (6) confirms the finding of Yan (2006) that cash relates positively to past fund returns. This relationship is in part driven by the fact that fund flows follow past performance (e.g., Sirri and Tufano, 1998; see also Table 3.1), so funds with high returns receive higher inflows and temporarily hold more cash while deciding where to invest it.34 Related, regression (7) shows that 33For recent examples, see “Fund’s Extra Cash Holds Opportunities”, Wall Street Journal, April 8, 2009, page C13; “More Stocks Funds Declare Cash King”, Wall Street Journal, April 9, 2009, page C9; “Cash Regains Its Asset Status”, Barron’s, August 17, 2009, page 24; “Harvard, Yale Are Big Losers in ‘The Game’ of Investing”, Wall Street Journal, September 11, 2009, page A1. 34At the same time, if funds with better past performance expect higher future flows, they may decide to hold less cash. The positive relationship between past returns and cash holdings observed empirically, however, indicates that 40 funds with high past flows tend to hold more cash. Prior research has used 12-month fund flow as a determinant of cash. While statistically important when used as the only explanatory variable, it becomes insignificant once more recent fund flows are taken into account. Regression controlling for lagged 1-, 6-, and 12-month fund flows shows that it is the more recent flows that are more important in explaining cash holdings (t-statistic on one-month flow is 6.78 compared to 2.58 for six-month, and 1.74 for 12-month flows). Managers have sufficient time to invest most of the cash inflow that happened over the previous year, but those inflows that occurred most recently may not yet be fully invested. Consistent with the findings of Yan (2006), past fund flow volatility in specification (8) is positively related to cash holdings. In the sample studied in this paper, however, the relationship is not significant at conventional levels. Regression (9) combines the variables that prior researchers found to relate to mutual fund cash holdings. Each of the regressors except deferred load fee is statistically significant, but jointly they explain only 5% of the cross-sectional variation in cash holdings. Interestingly, volatility of fund flows relates negatively to cash holdings in this multivariate specification. I next consider how managerial skill relates to cash holdings. It is natural to conjecture that skilled managers capable of identifying profitable investments, tend to generate better returns and have lower fund outflows, and thus carry less cash than poorly skilled managers. Kacperczyk et al. (2008) suggest that return gap, the difference between realized fund returns and the returns on a passive portfolio of fund’s reported holdings, may reflect managerial abilities, and I use this measure as a proxy for skill. Consistent with the argument above, regression (10) shows that return gap relates significantly and negatively to fund cash holdings. In fact, return gap alone explains 3% of the cross-sectional variation in cash positions. Specification (11) shows that funds whose portfolio of stocks earns a higher dividend yield hold less cash. Mutual funds receive dividend payments throughout the year but make payments to their shareholders only infrequently. Thus, higher cash flows from dividends received by funds holding higher yielding stocks represent a form of protection against cash shortfalls, and such funds allocate a smaller fraction of their assets to cash. Regression (12) illustrates that fund beta, calculated from the market model using realized fund returns over the previous 12 months, relates negatively to cash holdings. Cash is a component of the fund’s overall portfolio, and it is not surprising that funds with more cash are less risky as proxied for by market beta. Fund beta is an important determinant of cash holdings, explaining 2.4% of variation in cash positions among mutual funds. Average market beta of shareholdings (rather of the fund) is another important characteristic affecting fund cash holdings. Regression (13) shows that funds with risky stock portfolios hold more cash. This can be interpreted as evidence of funds managing average beta of their holdings. If a manager chooses to hold a portfolio of high-beta stocks, he will at the same time tend to hold more cash to decrease the risk over the fund’s overall portfolio. Regression (13) shows that fund’s liquidity beta also relates positively to fund cash holdings.35 It may be costly to adjust the this effect is weak. 35Using different proxies for liquidity of the holdings, Yan (2008) observes a similar relationship. His focus, however, 41 composition of illiquid stocks quickly in case of sudden withdrawals, leading the funds holding such stocks to carry more cash. Specification (14) studies the relationship between fund cash holdings and proceeds from share sales less spending on share purchases during the previous six months (‘proceeds’). If no new money flowed into the fund and no withdrawals were made, higher proceeds would translate into higher cash holdings. However, in presence of fund inflows that are invested by the manager, proceeds may relate negatively to cash holdings. Regression (14) shows that this is indeed the case: without controlling for other determinants of cash holdings, there is a negative relationship between proceeds and cash holdings, which is in part attributable to a negative correlation between fund flows and proceeds (see Table 3.1). Only when other fund characteristics are controlled for does the proceeds measure turn positive. The last three regressions combine important determinants of fund cash holdings. Specification (15) illustrates that controlling for fund return runup, fund flow over the previous month, and fund beta explains a comparable fraction of cross-sectional variation in cash holdings than a set of variables of regression (9) studied thus far in the literature. Regression (17) that uses the full set of explanatory variables explains three times as much variation in cash holdings as regression (9), illustrating the importance of the determinants of cash holdings that I document. Coefficients on all variables except loads, turnover, return runup, 12-month fund flow, and proceeds are significant. Excluding these variables in regression (16) results in an adjusted R2 that is two percent lower, which motivates me to keep them in the regression used to define excess cash. 3.4 Excess Cash Holdings and Fund Performance In this Section, I describe the methodology used to estimate excess cash and discuss the charac- teristics of funds with different excess cash measures. I next study the relationship between fund excess cash holdings and future returns. I define performance measures used in the analysis, and show that while raw cash is unrelated to future returns, funds with higher excess cash earn greater returns in the future. 3.4.1 Excess Cash Estimation Methodology To define excess cash holdings of mutual funds, I use the last specification of Table 3.2 that combines all of the considered fund characteristics and that achieves the highest adjusted R2. At every point in time when the data on fund cash holdings are available (annually prior to 1998 and quarterly thereafter), I estimate the following cross-sectional regression: CASH = γ0 + γ1LNTNA + γ2EXP + γ3FL + γ4DL + γ512B1 + γ6TURN + γ7RU12 + γ8FF1 + γ9FF6+ γ10FF12 + γ11σFF + γ12RG + γ13DY + γ14βMktFund + γ15β Mkt Hold + γ16β Liq Hold + γ17PRCDS + ε, (1) is on the impact of liquidity on the link between fund size and fund performance. 42 where time and fund suffixes are suppressed for brevity. CASH is the percentage of fund total net assets held in cash; LNTNA is log of total net assets; EXP is the expense ratio; 12B1 is actual 12b-1 expenses; FL and DL are front and deferred loads; TURN is fund turnover ratio; RU12 is the 12-month fund return runup; FF1, FF6, and FF12 are prior 1-, 6-, and 12-month fund flows; σFF is the volatility of one-month fund flows over the previous 12 months; RG is the Kacperczyk et al. (2008) annual return gap; DY is fund dividend yield; βMktFund is market beta of the fund; β Mkt Hold and βLiqHold are market and liquidity betas of fund holdings; and PRCDS is proceeds from fund stock sales less stock purchases, scaled by dollar value of all stock holdings.36 I define excess cash for a given fund as the residual ε from this regression and assign funds into quintiles on the basis of this value.37 The results of this paper are robust to reasonable alternative excess cash definitions. In fact, in settings where the determinants of cash include lagged fund flow (FF1), past returns (RU12) and lagged fund beta (βMktFund), the positive relationship between excess cash and future fund performance emerges. Appendix B provides more details and discusses the results obtained using a simplified excess cash definition. 3.4.2 Characteristics of Excess Cash Portfolios Table 3.3 presents average (in Panel A) and median (in Panel B) characteristics of funds in different excess cash groups. As is natural to expect, funds with higher excess cash hold a higher fraction of total net assets in cash: while funds in the highest quintile hold on average 11.6% of assets in cash, the comparable figure for funds in the lowest group is just 1.5%. The remaining fund characteristics are used as regressors in explaining fund cash holdings. It is thus not surprising that there is no monotonic relationship between excess cash and any of the variables, regardless of whether averages or medians are considered. Several characteristics exhibit a U-shaped relationship with excess cash (e.g., expense ratio or fund flows), but for all characteristics average values of the top and bottom groups are comparable. 3.4.3 Performance Measures To explore the relationship between fund excess cash holdings and future performance, I examine raw returns of the funds and consider several factor-based performance measures, which I now describe. 36All independent variables are winsorized at 1% and 99% in each cross-section. 37I can alternatively calculate excess cash holdings using a fixed effects model. In untabulated results, I find that the empirical conclusions of this paper are similar under this estimation approach. However, I later focus on the predictability of fund performance, and it is more appealing to use a method that calculates excess cash at a given time using only data available up to that point. Thus, I report the results based on the cross-sectional regression approach. 43 3.4.3.1 Market Model The first measure I consider is the market model alpha, estimated as the intercept αMi from regres- sion Rit = αMi + β M i RMt + εit, where Rit is the excess return of each of the five excess cash fund groups, or the difference in returns between high and low excess cash quintiles, and RMt is market excess return. 3.4.3.2 Fama-French Three-Factor Model I next complement the market model with the value and size factors, and estimate the Fama and French (1993) 3-factor performance measure as the intercept from regression Rit = αFFi + β M i RMt + β HML i HMLt + β SMB i SMBt + εit, where HML and SMB are value and size factors. 3.4.3.3 Carhart Four-Factor Model To adjust for momentum in stock returns (Jegadeesh and Titman, 1993, I next consider the Carhart (1997) four-factor model Rit = αCARi + β M i RMt + β HML i HMLt + β SMB i SMBt + β MOM i MOMt + εit, where MOM is the momentum factor.38 3.4.3.4 Multifactor Model with Liquidity Factors The analysis of the determinants of cash holdings indicates that liquidity may be an important factor affecting fund cash levels. To adjust for potential differences in liquidity of funds in different excess cash groups, I complement the Carhart four-factor model with either Pastor and Stambaugh (2003) or Sadka (2006) liquidity factor obtained from WRDS. 3.4.3.5 Ferson-Schadt Conditional Model Ferson and Schadt (1996) show that commonly used unconditional performance measures may be unreliable if risk premiums or betas are time-varying. They propose a model based on conditional performance that uses a pre-determined set of conditioning variables. As a robustness check, I consider the following conditional performance regression Rit = αFSi + β M i RMt + β HML i HMLt + β SMB i SMBt + β MOM i MOMt + ∑ F βFi (ZF,t−1RMt) + εit, 38I obtain value, size, and momentum factors from Kenneth French’s data library, http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data library.html. 44 where ZF,t−1 is the demeaned value of the macroeconomic variable F at t− 1. Following previous studies, I include the following macroeconomic variables: dividend yield of the S&P 500 index, term spread (difference between 10-year Treasury note and three-month Treasury bill), default spread (difference between rates on AAA and BAA bonds), and the three-month Treasury bill rate.39 The intercept αFSi from this regression is the conditional performance measure. 3.4.4 Future Fund Performance To study the relationship between excess cash and future performance, I estimate excess cash at the end of each month t when cash holdings data are available and assign funds into quintiles on the basis of excess cash. I hold the resulting five TNA-weighted portfolios for 12 months beginning in month t + 4. I skip three months between excess cash estimation and beginning of holding period to ensure that all data required for excess cash calculation (e.g., fund holdings) are publicly available.40 The choice of 12-month holding period is motivated by the fact that prior to 1999 cash holdings are observed only annually.41 The first estimation of excess cash in my sample happens at the end of 1992, and as a result, the return series start in April 1993. Prior to 1999, when cash holdings are available on an annual basis, no portfolios overlap, whereas starting in 1999, during any given month in quarter τ , the quintile q portfolio contains funds that were assigned to this group as of the end of quarters τ − 2 through τ − 5. 3.4.4.1 Raw Cash and Future Performance I first show that there is no significant relationship between raw cash and future returns. To do this, I assign funds into quintiles on the basis of raw, rather than excess cash. Table 3.4 presents future raw and risk-adjusted returns of the five resulting groups. Consistent with the observations of Yan (2006), regardless of the performance measurement approach, there is no link between cash holdings and future returns.42 In particular, funds in the low cash group earn on average 0.45% per month during the holding period compared to 0.47% for high-cash funds. Similarly, low-cash group earns an average Carhart four-factor alpha of −0.17% compared to −0.12% generated by the high-cash quintile. Conditional Ferson-Schadt alpha of the low and high-cash funds are −0.18% and −0.11%, respectively. In no case is the difference between returns of high- and low-cash quintiles significant. Negative average alphas across all cash groups are consistent with voluminous prior literature documenting poor risk-adjusted performance of actively managed mutual funds (e.g., 39Dividend yield is computed using CRSP files. Data on Treasury and corporate bond rates are obtained from the Federal Reserve, http://research.stlouisfed.org/fred2. 40Empirical relationship between excess cash and future returns is marginally stronger if instead I start holding the portfolios in month t+ 1 immediately following excess cash calculation. 41Considering shorter holding horizons using post-1998 data (when cash holdings are available on a quarterly basis) results in similar, and generally stronger relationship between excess cash and future fund performance. 42It appears that funds in the middle quintile (CASH3) earn the highest raw and risk-adjusted returns in the future. This can be interpreted as evidence of fund performance suffering when its cash holdings deviate from some average level. I find that assigning funds into groups on the basis of their absolute deviation from demeaned cash holdings leads to a negative, although insignificant relationship between this deviation measure and future returns. 45 Gruber, 1996; Carhart, 1997; Wermers, 2000; and others). 3.4.4.2 Excess Cash and Future Performance Table 3.5 studies the relationship between excess cash and future returns. Regardless of which performance measure is used, there is a strong positive relationship between fund excess cash holdings and future returns. The difference between raw returns of high and low excess cash groups reaches 0.18% monthly (0.51% vs. 0.32%). Controlling for exposure to the market, value, and size factors, I find that the difference in Fama-French alphas amounts to 0.23% (−0.04% vs. −0.26%) per month. Accounting additionally for either momentum and liquidity factors results in the difference in alphas of high and low excess cash group of around 0.20%, whereas the difference in the conditional Ferson-Schadt performance measures stands at 0.22% (−0.04% vs. −0.26%). For each performance measure considered, the difference between high and low excess cash funds is both statistically significant (t-statistics between 2.42 and 3.04) and economically meaningful (difference in annual returns between 2% and 3%). It is also interesting to point out that although alphas of each excess cash group are negative, they are statistically indistinguishable from zero for the top two quintiles. Figure 3.2 plots cumulative abnormal returns (based on the Ferson-Schadt performance mea- sure) of the excess cash quintiles in event time during five years following portfolio assignment. Several observations related to this figure are particularly interesting. First, the difference in per- formance persists over the entire five-year period: the gap between cumulative abnormal returns of high and low excess cash groups actually widens during the first three years following portfolio assignment, and then appears to stabilize. Second, low-cash firms perform remarkably poorly: their abnormal returns average around −0.13% per month over the course of five years. Such perfor- mance may be attributable to costs associated with cash shortfalls. Funds with low excess cash may be forced to sell their holdings at a cost and at a disadvantageous time to raise cash to meet withdrawals or satisfy other outflows (e.g., fund expenses). Fund performance suffers, causing more withdrawals, further damaging fund performance, and thus possibly trapping low excess cash funds in a punishing cycle. Finally, it is also interesting to point out that two top excess cash quintiles perform comparably well, with the top group edging slightly ahead. The third quintile underper- forms the top two over the first three years, and then surpasses the second highest quintile in the fifth year.43 To verify that the differences in performance between high and low excess cash funds are not limited to a particular time period, Figure 3.3 shows the time series of cumulative abnormal returns (based on the Ferson-Schadt conditional performance measure) from a portfolio that is long the high excess cash funds and short the low excess cash group. The plot illustrates the steadily increasing cumulative returns from this long-short position. The outperformance of the high excess cash group 43Given the results described here, it is natural to ask whether low excess cash funds are more likely to shut down, merge with another fund, or be taken over. In unreported results, I find no relationship between excess cash and the likelihood of such events happening. This observation is in line with the finding in the prior literature that investors fail to flee the worst performing mutual funds (e.g., Sirri and Tufano, 1998). 46 is particularly pronounced in 1999, near the peak of the dot-com bubble. 3.4.4.3 Excess Cash and Future Performance: Fama-MacBeth Regressions Table 3.6 further confirms the robustness of the positive relationship between excess cash and fund performance by presenting the results of Fama-MacBeth 1973 regressions of annual fund returns from months t + 4 to t + 15 on selected fund characteristics, including excess cash, measured at the end of month t. Regression (1) confirms that excess cash relates positively to future fund performance, whereas specification (2) shows that raw cash is unrelated to future returns. The next three regressions demonstrate that return gap relates positively, while fund size and marketing and distribution 12b-1 expenses are related negatively to future performance.44 Including them jointly alongside excess cash in regression (6) does not affect the significance of the excess cash measure. Given that raw cash is unrelated to future fund performance while excess cash relates to it positively, it is interesting to ask which fund characteristics must be controlled for to achieve a statistically significant relationship between cash holdings and future fund returns. To determine this, I consider cash in combination with selected fund characteristics as regressors. Specifications (7) through (9) show that including last month’s fund flow, 12-month runup, or fund beta as explanatory variables in addition to cash, increases the t-statistic on cash but the coefficient remains insignificant. However, as long as one combines fund beta with either fund flow or runup, the coefficient on cash holdings becomes statistically significant. Thus, to obtain a positive relationship between cash and future fund performance, it is particularly important to control for fund beta, and additionally for either recent fund flow or fund performance. In particular, in regression (12) that includes cash, fund flow, runup, and fund beta as explanatory variables, the t-statistic on cash holdings coefficient reaches 2.52. 3.4.4.4 Excess Cash and Future Performance: Conditioning on Prior Market Return Following a market downturn, the financial press abounds with stories of managers who adjusted their cash position in response to poor recent market returns (see footnote 33). Surprisingly often, such moves by the managers tend to be portrayed as savvy even though the managers adjusted their cash holdings only after the period of low market returns, rather than prior to it. In untab- ulated results, I find that this tendency to increase cash holdings following a market downturn is particularly pronounced among unskilled managers (those with lower return gap). For the purposes of my analysis, this implies that following a downturn, the portfolio of high excess cash funds is likely to include the funds run by such managers who are likely to continue to underperform (see Kacperczyk et al. (2008)). In such times, the returns of high excess cash portfolio will be dragged down due to the fact that it includes these poorly-performing funds. To account for this, I do not assign funds into excess cash groups if the prior 12-month market return was negative, and otherwise do not alter the procedure used to construct excess cash portfo- 44None of the other variables I considered were significantly related to future fund performance at conventional levels, and I do not include them in Table 3.6. 47 lios.45 Table 3.7 reports the resulting future raw and risk-adjusted returns of different excess cash groups. Compared to the findings presented in Table 3.5, where funds are assigned into groups without conditioning on prior market performance, the difference in future returns between high and low excess cash funds is approximately 50 percent larger. For example, the difference in the Ferson-Schadt conditional alphas reaches 0.32% per month (t-statistic of 3.95). What is even more interesting, alphas of the high excess cash group are in no case negative, regardless of which perfor- mance measure is used. Thus, the relationship between excess cash and future fund performance is particularly pronounced when high excess cash portfolio is less likely to include unskilled managers who increase their cash position in response to poor prior market returns. 3.5 A Model of Mutual Fund Cash Holdings In this Section, I develop an infinite-horizon model of mutual fund cash holdings and show that cash position in excess of the level predicted by the model relates positively to future fund performance.46 Consider a manager who enters period t carrying a fraction Ct of total net assets in cash and decides what fraction Xt of net assets to sell and buy, Xt = Amount sold per unit of TNA −Amount bought per unit of TNA. Cash-to-total net assets ratio in the following period is affected by this decision and shock FFt+1, a fund flow per unit of total net assets: Ct+1 = Ct +Xt + FFt+1. Fund flows per unit of total net assets can be thought of as a combination of mutual fund flows unrelated to purchase and sale decisions of the manager, and may incorporate new deposits, with- drawals, as well as flows from dividends: FFt+1 = µFF + σFF εFFt+1, where εFFt+1 is standard normal. Manager’s objective is to choose Xt to minimize the present value of the sum of opportunity, adjustment, and shortage costs associated with carrying cash.47 Per-period cost of carrying cash is ρCt + 1 2 λX2t + a Ct . 45I alternatively consider not including into a high excess cash portfolio those funds that increased their cash holdings by more than half following a 12-month period of negative market returns, and obtain similar results. 46The model in its current form is not designed to explain the empirically observed relationship between excess cash and future fund performance but rather to show that this relationship is robust to defining excess cash relative to a model-based target level. 47Cost-minimization goal of the manager is appropriate in a competitive mutual fund industry, where managers do not possess stock-selection, market-timing or other skills. The manager maximizes total assets under management by minimizing fund’s costs. 48 The first component of this amount, ρCt, captures the opportunity cost of holding cash. Equities tend to earn a higher return than cash holdings, and ρ can be thought of as average return foregone by carrying cash rather than investing in higher-yielding assets. The second component, 12λX 2 t , is the adjustment cost associated with buying or selling shares. Parameter λ may reflect both actual trading expenses and market impact costs. Finally, a/Ct captures the shortage cost, which decreases with cash holdings of the fund. When cash position is particularly low, any outflow shock can result in a high cost as the fund scrambles to obtain the necessary cash to cover the outflows. Additionally, holding particularly low cash balance may result in lower flexibility to act quickly to take advantage of profitable investment opportunities that might arise. Bellman equation associated with this problem is Vt(Ct) = min Xt { ρCt + 1 2 λX2t + a Ct + βEt [Vt+1(Ct+1)] } , where β is the discount factor. The model clearly abstracts from other potentially important factors affecting optimal cash positions, but it captures the key costs associated with holding cash and allows to obtain an estimate of a target cash level, and thus to compute a measure of excess cash that does not rely on regression estimation. To solve the model, I make a simplifying assumption that adjustment and shortage cost pa- rameters, λ and a, are equal across funds with similar risk characteristics. More specifically, to estimate λ and a, in each cross-section I first assign funds into ten groups on the basis of their lagged market beta, βFundMkt , and for each group i obtain the empirical estimates of the remaining parameters {ρi, µFF,i, σFF,i, βi}. Specifically, I compute the opportunity cost parameter ρi as the product of average market beta of the funds in group i and quarterly risk premium of 1.5%. Pool- ing all fund-quarter observations of each beta group, I obtain the empirical estimates of the mean and volatility of fund flows and use the resulting values as µFF,i and σFF,i parameters. Finally, I approximate the quarterly discount factor as βi = 1/1.015 = 0.985.48 For each of the ten beta groups, I solve the model numerically using the set of parameters {ρi, µFF,i, σFF,i, βi} and arbitrary values λi and ai to obtain a policy function Xt(Ct). I then compute the distance between vectors of empirical and model-predicted policy decisions: Di = max ( abs ( Xmodeli −Xempiri )) , where for group i containing N fund-quarter observations, Xmodeli is the N × 1 vector of model- predicted policy decisions given cash holdings Ci of this group, and X empir i is the N × 1 vector of empirically observed decisions taken by the fund managers of group i.49 The values {λi, ai} that 48I focus on the quarterly horizon because more fund-quarter observations are available in my sample than fund-year observations. This restricts the sample to the 1998-2008 period, which is also appropriate given that cash-to-total net assets ratios are not stationary prior to 1998 (see Figure 3.1). 49I compute each element of Xempiri as Cji,t+1 − Cji,t − FF ji,t+1 −DY ji,t+1, 49 minimize the distance Di for group i are used in the later analysis as parameter values λ j i and a j i for each fund j that falls into this group. I next estimate the target cash level for each fund j and quarter t using the set of parameters{ ρjt , µ j FF , σ j FF , β, λ j it, a j it } , where ρjt is the product of quarterly risk premium of 1.5% and the lagged market beta of fund j computed using one year of data up to quarter t, µjFF and σ j FF are mean and volatility of fund flows of fund j, β = 0.985, and λjit and a j it are the adjustment and shortage parameters of fund j given that it belongs to beta group i as of quarter t.50 The model then yields a policy function and an estimate of the target cash level for each fund-quarter observation. The average model-predicted target cash position across all fund-quarter observations is 4.5% of total net assets compared to 4.0% cash-to-TNA ratio observed empirically for the set of funds used in the estimation.51 3.5.1 Model-Based Excess Cash and Future Fund Performance The empirical results presented in the previous Section highlight a positive relationship between mutual fund excess cash positions and future fund performance. Excess cash in that analysis is defined as the residual from a cross-sectional regression of cash holdings on a number of fund characteristics. Defining excess cash in such manner is very appealing, as it accounts for a wide number of variables that affect cash holdings. I now show that the key empirical results of this paper do not depend on this particular way of estimating excess cash by demonstrating that the positive relationship between excess cash and future fund performance is robust to defining excess cash as the difference between actual and model-based target cash holdings. More specifically, in each cross-section I assign funds into five groups on the basis of excess cash calculated as ECjt = Cjt − Ĉjt Ĉjt , where Cjt and Ĉ j t are actual and model-based target ratios of cash to total net assets of fund j at time t. I then calculate future raw and risk-adjusted returns of each of the five quintiles in the same manner as was done earlier. Table 3.8 shows a generally monotonic positive relationship between excess cash defined in this manner and future fund performance. Whereas raw and market-adjusted returns of high excess cash funds are only marginally higher than those of their low excess cash peers, the average difference between other measures of risk-adjusted returns of the two groups exceeds 0.20% monthly. For example, the conditional Ferson-Schadt performance measure is −0.25% for the low excess cash group but reaches −0.02% for the high excess cash funds. The difference in future performance of funds with different excess cash levels is thus not specific to the regression- where Cji,t is cash-to-TNA ratio of fund j which belongs to beta group i as of the end of quarter t, and FF j i,t+1 and DY ji,t+1 are the ratios of quarter t+ 1 fund flows and dividends to TNA. 50If fund j that belongs to beta group i at a given point in time has less than four quarterly observations available, I use as parameters the mean and volatility of the fund flows of the beta group, µFF,i and σFF,i, rather than of the fund itself, µjFF and σ j FF . 51Average model parameters used are {ρ = 0.015, µFF = 0, σFF = 0.1, β = 0.985, λ = 1.14, a = 0.00003}. 50 based definition of excess cash employed throughout this paper but is robust to defining excess cash relative to a target position predicted by a model of optimal cash holdings. 3.6 Sources of Relationship Between Excess Cash and Fund Performance In this Section, I explore the driving forces behind the difference in future performance of funds with different excess cash holdings. I link the positive relationship between excess cash and future returns to manager’s ability to control fund expenses, his stock-picking and market-timing skills, and managerial aptitude to anticipate fund flows. 3.6.1 Excess Cash and Ability to Control Fund Expenses I now consider a framework of transacting in shares of a stock in a setting with fixed and variable costs. The model suggests that relative to a manager who either invests all sales proceeds immedi- ately and/or who transacts more frequently than is optimal, a cost-minimizing manager will tend to carry a higher cash balance. The framework can thus justify the positive link between high cash positions and performance: managers carrying greater cash balances may be doing so in part to minimize transaction costs and as a result they outperform their low-cash peers. Consistent with the model, I show that future fund expenses decline with excess cash. I also demonstrate that excess cash is particularly valuable for large funds. 3.6.1.1 Model of Costly Stock Trading Consider a fund that holds nS shares in stock S, with current market price per share of pS . The manager would like to sell all of his holdings in S and invest the proceeds in nB shares of stock B, which trades at pB per share. Costs associated with either buying or selling ni shares of stock i at price pi per share are Fi + Vi(nipi)2, where Fi and Vi are fixed and variable costs, respectively.52 Suppose that the manager can only transact at discrete points in time, and for simplicity assume that price is not directly affected by manager’s decisions. The manager’s objective is to minimize the total cost associated with transacting in a given stock i: NiFi + N∑ r=1 Vi (nri pi) 2 , where Ni is the number of distinct trades the manager makes to either acquire or dispose of stock 52Trading costs (including price impact) tend to increase more than linearly with trade volume. See, e.g., Keim and Madhavan (1996) and the survey paper by Madhavan (2000). 51 i, and nri is the number of shares of stock i the manager buys or sells during his rth transaction. 53 Given that the manager will make Ni transactions in stock i and that total variable costs increase with the dollar value of shares bought or sold in a given transaction, the number of shares nri that will minimize the total cost is ni/Ni, where ni is the total number of shares of stock i that the managers would like to buy or sell.54 Thus the manager’s problem can be rewritten as min Ni Costi (Ni) = min Ni NiFi +NiVi ( ni Ni pi )2 = min Ni NiFi + 1 Ni Vi (nipi) 2 . The number of transactions that minimizes the total cost is N∗i = √ Vi Fi (nipi) if √ Vi/Fi (nipi) is an integer, or N∗i = arg min Ni∈ {⌊√ Vi/Fi(nipi) ⌋ , ⌊√ Vi/Fi(nipi) ⌋ +1 }Costi (Ni) otherwise, where bxc denotes the integer part of x. Thus the optimal number of transactions N∗i increases in variable cost Vi and decreases in fixed cost Fi. Consider now again the problem of selling shares in stock S and investing the proceeds in stock B. Given that the manager minimizes total transaction costs, the change in fund’s cash holdings associated with selling stock S and purchasing stock B during a period of time of length t is ∆Ct = ⌊ t N∗S + 1 ⌋( nS N∗S pS − FS − VS ( nS N∗S pS )2) − ⌊ t N∗B + 1 ⌋( nB N∗B pB + FB + VB ( nB N∗B pB )2) . If √ VS/FS < √ VB/FB, as for example might be the case if stock B is less liquid than S, then N∗S < N ∗ B. In other words, in such case the manager will take a longer time to purchase the desired amount of stock B than to sell his holdings in stock S, and as a results the change in cash unrelated to transaction costs will be non-negative at any point t. On the other hand, a non-optimizing manager who either invests all sales proceeds immediately and/or who transacts in the illiquid stock more frequently than is optimal will cause a change in cash that is not larger than the change in cash of a fund run by a cost-minimizing manager. If one additionally requires that the managers can only use fund cash reserves to cover fixed and variable costs but not to finance stock purchases directly, then the managers will use the proceeds 53This set-up implies that the manager does not face any costs of delaying his transactions (i.e., future costs are discounted at the rate of zero), but I assume that the manager prefers to conduct his transactions as soon as possible. A manager may prefer to conduct the transaction as soon as possible, for example, when he receives a signal about future performance of a stock. 54This can be readily seen by solving the problem min{nri }NiFi + ∑Ni r=1 Vi (n r i pi) 2 s.t. ∑Ni r=1 n r i = ni. The derivative of the associated Lagrangian with respect to the jth choice variable nji is 2Vin j ipi = λ, where λ is the Lagrange multiplier. This suggests that for every j and k, nji/n k i = 1, or n j i = n k i = ni/Ni. 52 from the sale of stock S to cover the purchase of stock B. As a result, a cost-minimizing manager will carry a higher or similar cash balance than a manager who invests the sales proceeds quicker, even when √ VS/FS ≥ √ VB/FB. Figure 3.4 shows cumulative changes in cash holdings under two scenarios: when fund manager buys a less liquid stock than he sells (in Panel A), or when he finances the purchase of stock B by proceeds from the sale of stock S (in Panel B). In either case, at any point in time, the cumulative change in cash unrelated to transaction costs is non-negative. By contrast, a corresponding change in cash of a fund run by a manager who invests all sales proceeds immediately will be strictly non-positive. High cash positions may thus proxy for managerial ability to control costs: cost-minimizing managers carry higher cash balances and generate better results than do those managers who make sub-optimal decisions by reinvesting the proceeds from sales of shares immediately or by otherwise transacting inefficiently. 3.6.1.2 Empirical Evidence The framework outlined above implies that managers who are better able to control their transac- tion costs may carry higher cash balances. It additionally suggests that managers who purchase less liquid stocks may also carry higher cash positions than do their peers holding more liquid shares. To test these conjectures empirically, I now explore how excess cash holdings relate to future fund expenses and future liquidity of shareholdings. Table 3.9 demonstrates a monotonic negative relationship between excess cash and several measures of future fund expenses. For example, high excess cash funds spend on average 0.21% of total net assets less on expenses during the twelve months following portfolio assignment than do their low excess cash peers (t-statistic of 4.94). Similar patterns emerge for 12b-1 marketing expenses, management fees, and turnover. Table 3.9 also shows that average liquidity of fund shareholdings declines (loading on the liq- uidity factor rises) with excess cash: a year following portfolio assignment, high excess cash funds hold stocks that are on average the least liquid than holdings of any other group. This suggests that high excess cash funds buy stocks that are considerably less liquid than those purchased by the low excess cash group. The relationship between excess cash and both future expenses and liquidity of shareholdings is particularly striking given that I control for the differences in expenses and liquidity when estimating fund excess cash holdings. If funds carry high excess cash in part due to lower liquidity of their shareholdings, it is natural to expect that excess cash is particularly valuable for large funds holding illiquid stocks. Small funds can adjust their shareholdings reasonably quickly without causing dramatic price pressure, whereas large funds – particularly those transacting in illiquid stocks – that try to minimize total trading costs may have to move slower and hold more cash as a results. To check this conjecture, I assign funds in each excess cash quintile into tertiles on the basis of their size, measured at the same time as excess cash. Within each excess cash-size group, I then sort funds into tertiles on the basis 53 of liquidity of their shareholdings, βHoldLiq . Consistent with the hypothesis, Table 3.10 shows that the difference in future returns between high and low excess cash funds is particularly pronounced for large fund with illiquid holdings. Regardless of the performance measure used, the difference in returns between high and low excess cash funds exceeds 0.50% per month for large illiquid funds, and is generally lower for smaller and more liquid funds. In light of the evidence presented in this section, the positive relationship between excess cash and fund performance can be interpreted as a link between excess cash holdings and the ability to control fund costs. Funds carry high excess cash as a result of minimizing total costs of transacting in stocks. In the future, high excess cash funds continue to manage their costs well, which contributes to their stronger performance relative to their low excess cash peers. 3.6.2 Stock Selection Holding high cash balances can impose a significant cost on fund performance. Wermers (2000), for example, estimates that non-stock holdings drag fund returns down by 0.7% per year. Thus, any manager who rationally decides to carry high excess cash needs to be able to make good investments to compensate for the lower return cash tends to earn relative to equity benchmarks. It is natural to conjecture that such managers tend to make better stock purchasing decisions in the future. Managers of low excess cash funds, on the other hand, will tend to be comprised of those who find it unfavorable to hold high cash balances. This group is likely to include at least two subsets of managers: those with poor stock purchasing skills, and/or those skilled at identifying poorly performing stocks. The latter subset includes those managers who are able to raise cash quickly and at a low cost by selling those of their holdings that are likely to generate low returns in the future.55 To check these conjectures, I now study the performance of stocks purchased and sold by different excess cash groups. I begin by comparing fund holdings at the time excess cash holdings are estimated with holdings a year later. I determine which stocks were bought and sold by each fund during this period, and note the number of shares acquired or disposed.56 The time of purchases and sales is not directly observable, and I assume that all transactions take place six months after excess cash calculation. I then calculate returns of the shares bought and sold over the following six-month period. I estimate both raw returns as well as style-adjusted performance, calculated following Daniel et al. (1997).57 55Low excess cash group can certainly include managers with good stock purchasing skills who are fully invested and do not anticipate better buying opportunities to arrive in the near future. Later in the section, I discuss the effects they may have on the empirical results presented here. 56To compute how many shares were bought and sold, I use CRSP files to adjust for events such as stock splits and stock dividends that affect the number of shares outstanding and share price. Visual examination of the Thom- son holdings database reveals that while the number of shares reported by most funds is split-adjusted, there are observations where the number of shares of a given stock held by a fund does not change despite the fact that the stock underwent, e.g., a 2-for-1 split. In such cases, I do not perform any adjustments when calculating the number of shares bought and sold. 57An alternative approach is to estimate alphas from short-window regressions for individual stocks, and calculate their average for each fund/time observation. However, such a procedure is likely to produce biased performance measures, as Boguth et al. (2009) demonstrate. This is particularly important given that most mutual funds follow momentum strategies. For example, Grinblatt et al. (1995) find that 77% of mutual funds are momentum investors. 54 Table 3.11 shows average future returns earned by stocks bought and sold by each excess cash group. I separate all stock purchases made by a given fund into additions to already existing positions (‘old’ buys), and investment in those stocks not held by the fund at the beginning of the period (‘new’ buys). I first calculate the average returns earned by buys and sells of each fund in each cross-section (weighted by the dollar amount spent on buys or earned from sells), and then obtain a TNA-weighted cross-sectional average of these returns for each excess cash quintile. Table 3.11 shows time series means of raw and style-adjusted returns. The ‘All Buys’ column of Table 3.11 shows that managers of high excess cash funds are consid- erably better at identifying investment opportunities. The stocks they buy earn on average 0.12% per month more than the shares purchased by the low excess cash funds. Interestingly, a lot of this outperformance is concentrated in the additions to the positions already held by the funds: future raw returns of the ‘old’ buys made by the high excess cash funds exceed those of the low excess cash group by 0.18% per month. The difference in returns of the ‘new’ buys is smaller but still sizeable, at 0.11% per month. Stronger outperformance by the ‘old’ buys rather than by the ‘new’ buys of the high excess cash groups is intriguing. It is possible that managers of high excess cash funds realize the attractiveness of their existing holdings, and carry excess cash because they plan to add to these positions in the near future.58 If the managers are able to invest an optimal amount in these stocks instantaneously, they may carry no excess cash; but they may be unable to do so, for example due to the risk of causing price pressure that will prevent them from buying at attractive prices. Thus, they carry excess cash and gradually invest it, focusing on investments in their existing holdings.59 The last two columns of Table 3.11 show that future performance of stocks sold by high excess cash funds is actually better than that of shares sold by the low excess cash group. The magnitude of the difference is substantial: 0.20% and 0.14% monthly for raw and style-adjusted returns, respec- tively.60 Thus, consistent with the argument developed at the beginning of this section, managers of low excess cash funds are particularly skilled at identifying shares that will perform poorly in the future. It is therefore likely that at least a subset of those managers who choose to hold low excess cash do so because they are convinced that they will be able to meet any cash shortfall by selling those of their stock holdings that are likely to perform poorly in the future. On the whole, the evidence presented here demonstrates that managers of high excess cash funds are skilled at stock purchasing (identifying undervalued stocks), whereas managers of low excess cash group are proficient at determining which stocks to sell (identifying overvalued stocks). The differences in returns of the buys and sells between the two groups are likely biased downwards 58Footnote 26 disusses why such manager may prefer to hold cash rather than exchange-traded funds or similar assets as they wait for buying opportunities. 59Lending further support to this argument, in untabulated results I find that relative to the low excess cash group, high excess cash funds spend proportionately more on purchases, in particular when adding to their existing holdings. 60’New’ buys earn higher returns than ‘old’ buys, which in turn earn higher returns than sells (e.g., 0.99%, 0.43%, and 0.19%, respectively, for the low excess cash group). This is likely due to the fact that ‘new’ buys tend to be past winners, that sells tend to be past losers (e.g., Grinblatt et al., 1995), and that not all transactions take place in the middle of the examined 12-month period. 55 by the presence in the low excess cash group of those managers who are skilled stock purchasers but who are already fully invested. The results presented above can also be interpreted as consistent with the notion that managers of high and low excess cash funds invest in development of different sets of skills. In particular, consider a setting where mutual fund managers can improve their ability to identify undervalued or overvalued equity by investing their effort into some form of a buy and sell ‘technology’. The amount of effort they can expend, however, is limited (e.g., by the number of hours they can work productively or by the number of stocks they can analyze), and the managers must decide how to allocate it between the two technologies. Managers holding unusually large cash balances may realize that they need to bring their cash position down to some target level. To do so, they need to determine which purchase decisions to make, and thus they have a strong incentive to invest in the ability to identify which stocks to buy. Low excess cash fund managers, on the other hand, will find investing in the ‘sell technology’ more appealing. As a result of these allocations of effort, managers of high excess cash funds make better purchase decisions, whereas those running low excess cash funds are more skilled at identifying which stocks to sell, which is consistent with the results presented here. The findings of this section are suggestive, but do not fully explain the positive relationship between excess cash and future fund returns. In particular, buys of high excess cash funds out- perform those of low excess cash group, but their sells also outperform. Thus, high excess cash funds improve their returns by making better buy decisions, whereas low excess cash funds do so by making better sell decisions. Yet, if performance of the fund is driven mainly by the purchase rather than sales decisions, the results of this section points to one of the potential sources of the positive link between excess cash and future fund performance. 3.6.3 Market Timing A mutual fund manager with market-timing ability may optimally increase the fund’s cash position prior to a market downturn. At the same time, he will decrease cash holdings prior to a bull market. Thus, it is unlikely that the positive relationship between excess cash and fund performance proxies solely for the manager’s ability to successfully anticipate major turns of the stock market. However, if market-timing skills are mainly concentrated in the ability to predict market downturns, market timing may at least in part explain the superior performance of high excess cash funds relative to the low excess cash group.61 Additionally, if mutual fund managers possess any market-timing ability, one would expect to find stronger evidence of it among the funds in the extreme excess cash groups. To test these conjectures, I begin by studying the market-timing skills of the funds in different excess cash groups. Table 3.12 presents the results of two commonly used approaches to test for 61Most studies investigating the market-timing ability of mutual funds find insignificant or significantly negative market-timing performance (e.g., Treynor and Mazuy, 1966; Henriksson and Merton, 1981; Chang and Lewellen, 1984; Henriksson, 1984; Cumby and Glen, 1990; and Becker et al., 1999). Jiang et al., 2007 use portfolio holdings data to find evidence of managerial market-timing ability at six- and nine-month horizons. 56 market timing. First, I follow Treynor and Mazuy (1966) and estimate the regression Rit = δ0i + δ1iRMt + δ2iR2Mt + ηit for each of the five excess cash quintiles and for the difference between high and low excess cash groups. Additionally, I use the Henriksson and Merton (1981) approach, estimating Rit = φ01 + φ1iRMt + φ2i max(0, RMt) + νit. Significantly positive coefficient on δ2i or φ2i can be interpreted as indicative of successful market- timing abilities. Contrary to the notion that funds in high and low excess cash groups possess better market- timing ability, coefficients δ2i and φ2i are actually the largest for funds in the second and middle excess cash quintiles. Portfolio of funds in both low and high excess cash groups exhibit insignificant market-timing skills: for example, when using the Henriksson-Merton measure, the coefficient φ2i is −0.06 (t-statistic of −1.40) for the low excess cash group and −0.01 (t-statistic of −0.27) for the high excess cash group. Managers of low excess cash funds have somewhat worse market-timing abilities than their high excess cash peers, and as a result, the portfolio that is long high and short low excess cash funds exhibits positive but statistically insignificant market timing when either measure is used. On the whole, though, it appears that the relationship between excess cash and future fund performance cannot be attributed entirely to the market-timing skills of the managers.62 A related and interesting question to ask is whether excess cash holdings of mutual funds predict market returns.63 The results presented above show that the returns of neither of the excess cash groups considered provide robust evidence of market-timing skills. Yet, it is possible that at least some subset of mutual fund managers possess market-timing abilities. The funds run by such managers are likely to have more volatile excess cash holdings, as the managers adjust cash positions in response to their changing expectations about market returns. Thus, if managers possess any market-timing skills, the evidence will be more pronounced among the funds with high volatility of excess cash holdings. To test this hypothesis, at the end of each quarter τ I calculate volatility of excess cash holdings as standard deviation of excess cash during the past 12 quarters (from τ − 11 to τ), requiring at least eight observations to be present.64 I assign funds into five groups on the basis of this measure, 62It may be tempting to interpret the statistically insignificant intercept from the Henriksson-Merton regression with the High-Low portfolio returns as the dependent variable as supporting the idea that the difference in returns between high and low excess cash funds in attributable to differential market-timing abilities. However, this intercept cannot be characterized as a performance measure; see for example, Boguth et al. (2009) or Ferson (2009). 63In untabulated results, I confirm the finding of Yan (2006) of no relationship between aggregate raw cash holdings and future market returns. 64To obtain a meaningful measure of volatility of excess cash holdings, I limit the sample to post-1998 data, when cash holdings are available quarterly. 57 and for each quintile j, compute aggregate excess cash at of the end of quarter τ as AECjτ = ∑ i∈j ECiτ · TNAiτ , where ECiτ and TNAiτ are excess cash and total net assets of fund i as of the end of quarter τ , and then estimate the regression RNM,τ+1 = ψ0j + ψ1jAECjτ + ς N j,τ+1, where RNM,τ+1 is the market return during the N -month period starting in the first month of quarter τ +1, N ∈ (1, 3, 6, 12). A significantly negative coefficient ψ1j can be interpreted as consistent with market-timing ability of group j, whereas a positive coefficient implies poor market-timing skills. The results of this regression are reported in Table 3.13. At each forecasting horizon, the bottom four groups exhibit either insignificant, or significantly negative market-timing abilities. However, aggregate excess cash holdings of the funds with high volatility of past excess cash holdings relate significantly and negatively to future three- and six-month market return.65 This evidence can be interpreted as consistent with skillful market timing by those mutual fund managers who frequently adjust their excess cash holdings. Such managers increase their excess cash position prior to market downturns and reduce it prior to a period of strong market returns. 3.6.4 Liquidity Reasons Edelen (1999) shows that fund flows have an adverse effect on fund performance. If a manager fails to anticipate fund outflows and does not have sufficient cash on hand to meet such outflows, he will be forced to liquidate some of his share holdings at a potentially disadvantageous time and price. Thus, it is possible that the difference in future performance between high and low excess cash funds may in part relate to superior ability of managers of high excess cash funds to anticipate fund outflows. To check this conjecture, I assign funds within each excess cash quintile into two groups on the basis of future realized fund flows. Table 3.14 demonstrates that excess cash proves particularly valuable for funds that experience low fund flows. For example, the difference between four-factor alphas of high and low excess cash funds is 0.30% per month when future fund flows are low, and just 0.10% when fund flows are high. This evidence provides support to the idea that at least a fraction of high returns earned by high excess cash funds is related to the fact that such funds are better positioned to meet future outflows. Low excess cash funds do not carry sufficient cash on hand to cover the outflows and are likely forced to liquidate some of their holdings, damaging fund performance. The high excess cash group, on the other hand, is more skilled at anticipating future fund outflows and as a result is well positioned to meet such outflows, generating better returns. 65Insignificant evidence of market timing at one- and twelve-month horizons is consistent with the findings of Jiang et al. (2007) who use mutual fund holdings data to study market-timing abilities. They find no evidence of market-timing skills at one- and twelve-month horizons, but observe positive timing ability and three- and six-month horizons. 58 An alternative liquidity-based explanation that may justify outperformance of high excess cash funds relates to liquidity of their shareholdings. Mutual fund managers may choose to carry more cash if they are holding a portfolio of illiquid stocks, and thus their stronger performance may be due to higher returns earned by their illiquid positions.66 To account for the differences in liquidity among excess cash groups, I control for exposure to a liquidity factor in defining excess cash. I also include Pastor and Stambaugh (2003) or Sadka (2006) liquidity factors when esti- mating performance. Inclusion of either factor in estimating fund abnormal returns in Table 3.5 has indistinguishable effect on the performance of either high or low excess cash groups. Thus, the difference in returns earned by high and low excess cash funds cannot be attributed to lower liquidity of stocks held by high excess cash funds.67 3.6.5 Excess Cash and Closed-End Fund Performance The results discussed above suggest that the positive relationship between excess cash and mutual fund performance may relate to the better ability of high excess cash fund managers to anticipate fund outflows. To determine if this may indeed be the case, it is natural to explore whether high excess cash closed-end funds outperform their low excess cash peers. Unlike their open-end counterparts, closed-end funds rarely issue or retire shares, and shares are usually not redeemable until fund liquidation. Managers of closed-end funds are thus free from concerns related to fund flows, and any motives for carrying cash balances are not tied to uncertainty about fund flows. A positive relationship between excess cash of closed-end funds and future fund performance may thus be interpreted as consistent with the notion that the similar relationship that I document for open-end funds is not driven by fund flow-related reasons. On the other hand, absence of such a relationship can be viewed as supportive of the idea that fund flows play an important role in the stronger performance of high excess cash open-end funds relative to their low excess cash peers. 3.6.5.1 Data To study the relationship between excess cash holdings of closed-end funds and their future per- formance, I begin by obtaining from CRSP the list of 608 closed-end funds that were in operation at some point between 1994 and 2008 (those with share code of 14).68 Using the COMPUSTAT files, I retrieve Central Index Keys (CIKs) for 572 of these funds. Closed-end funds may report their portfolio composition in several different filings with the Securities and Exchange Commis- sion (SEC): in N-30B, N-30D, and N-CSRS periodic reports mailed to fund shareholders, and in N-Q quarterly schedules of portfolio holdings. Out of the sample with valid CIKs, 537 funds have at 66Numerous studies document a negative relationship between stock liquidity and future returns. See, for example, Amihud (2002). 67In unreported results, within each excess cash quintile I assign funds into two groups on the basis of their exposure to liquidity factors (e.g., Pastor and Stambaugh, 2003, or Sadka (2006)), and study the returns of the resulting ten portfolios. The difference in returns between high and low excess cash funds is similar for high and low liquidity groups. Similar conclusion can be drawn from Table 3.10. 68Data in the Securities and Exchange Commission’s Edgar system are not widely available prior to 1994, which leads me to focus on the 1994-2008 period. 59 least one of such reports on file with the SEC. I download all such filings of these funds using SEC’s Edgar FTP server and hand-collect the data on fund objective, cash holdings, expenses, and net asset values. Unfortunately, only a minority of the closed-end funds in the sample have a domestic equity investment objective, while most others invest mainly in municipal or corporate bonds, and a number of funds focus on specific industries or on international markets. After restricting the sample to funds with at least 50% of their investments in U.S. equities, I arrive to a final sample of 54 funds or 833 fund-quarter observations.69 3.6.5.2 Summary Statistics Table 3.15 reports summary statistics for the sample. Closed-end funds hold on average considerably less net assets in cash (1.79%) than do their open-end counterparts (4.83% during the 1994-2008 period), suggesting fund flow concerns play an important role in determining cash holdings of open- end funds. Closed-end funds have on average less assets under management (average net asset value of 557 million), lower fund market beta, and somewhat higher average but lower median expenses than do open-end funds (see Table 3.1). A median closed-end fund has been in operations for 15 years and its shares trade at 17.5% discount to the per share net asset value.70 Cash holdings of the closed-end funds correlate positively with fund expenses and negatively with fund size, beta, and age, whereas correlations of cash with either recent fund returns or discount are statistically indistinguishable from zero. 3.6.5.3 Determinants of Closed-End Fund Cash Holdings To calculate excess cash holdings of closed-end funds, I begin by first exploring the determinants of their cash positions. Following the methodology used in analyzing open-end funds, in each cross- section I regress cash-to-net asset values of closed-end funds on a number of fund characteristics.71 Regression (1) of Table 3.16 shows that fund size plays a very important role in explaining fund cash holdings, with larger funds holding considerably less cash: average R2 of this regression exceeds 24% whereas a comparable number in the case of open-end funds was just 0.2% (Table 3.2). After controlling for size, cash holdings of closed-end funds relate negatively to fund market beta and positively to expenses and prior returns, which is consistent with what Table 3.2 shows to be the cases for open-end funds, although the coefficients on beta and runup are not statistically significant in the case of closed-end funds (regressions 2 through 4). Older closed-end funds hold considerably more cash, which may be due to such funds preparing to deregister. Somewhat surprisingly, I find a strong negative relationship between cash holdings and fund discount, which remains significant even after controlling for all other considered characteristics (regressions 5 and 6). 69Cash holdings in the first and third calendar quarters are available for very few funds, and I restrict analysis to using only data from the second and fourth calendar quarters. 70I measure closed-end fund beta and return runup in the same manner as for open-end funds. Fund discount is calculated as the difference between net asset value per share and market price per share, scaled by net asset value per share. 71The regressions are run semiannually – at the end of June and December – using the most recently available fund data provided that this data are not older than six months. 60 3.6.5.4 Excess Cash Holdings and Closed-End Fund Performance I define excess cash of the closed-end funds as the residual from cross-sectional regression (6) in Table 3.16. I then form quintile portfolios of the funds on the basis of their excess cash and calculate their average raw and risk-adjusted returns in the same manner as for the open-end funds. Table 3.17 shows that regardless of which performance measure is used, there is no statistically significant difference in performance of high and low excess cash funds. For example, the difference in Ferson- Schadt conditional alphas is just 0.02% monthly. While the sample is admittedly small, with on average less than 30 firms in each June and December cross-section, the lack of a relationship between closed-end fund cash holdings and future fund performance can be viewed as consistent with the notion that a positive relationship between cash positions and future performance observed for the open-end funds is at least in part due to fund flow reasons. If better performance by high excess cash open-end funds does in fact proxy for managerial ability to anticipate fund outflows, it is not surprising that there is no difference in future returns of high and low excess cash closed-end funds since fund flows do not present any concerns for their managers. 3.6.6 Concealing Portfolio Composition I now briefly explore an alternative and perhaps less conventional explanation for the positive relationship between excess cash and future fund performance. Suppose a manager has superior information indicating that a subset of the stocks his fund holds will perform exceptionally well in the future. The manager would like to build up his position in those stocks. However, he is concerned that building up the position too quickly will cause price pressure, preventing him from acquiring the shares at advantageous prices. He is further concerned that managers of other funds may be able to infer his information from observing his holdings. Such manager might find it optimal to adjust his share holdings prior to making his portfolio composition public. More specifically, to prevent others from inferring his information, he will reduce the holdings in the set of stocks about which he has superior knowledge. As a result, the manager will hold excess cash around the time of disclosing his portfolio composition. In the future, he will use excess cash to re-build it and continue to add to that subset of stocks, outperforming low excess cash funds. This setting can justify the positive relationship between excess cash and future fund perfor- mance. However, it is doubtful that most mutual fund managers will find it appealing to hide their portfolio composition by selling and then repurchasing certain stocks. Not only are there transac- tion costs associated with such strategy, but it is unlikely to provide a significant boost to fund performance for large funds. Yet, it may be more attractive for small funds who can adjust their positions quickly and may find it attractive to attempt to hide their portfolio composition. As Table 3.10 points out, however, the difference in returns between high and low excess cash funds is actually more pronounced for larger funds. It is therefore unlikely that the positive relationship between excess cash and future performance is related to the attempts of the managers of high excess cash funds to conceal their portfolio composition. 61 3.7 Conclusion This study documents a positive relationship between excess cash holdings of actively managed equity mutual funds and future fund performance. Funds with high excess cash – that is, with cash holdings in excess of the level predicted by fund characteristics or by a model of optimal cash holdings – outperform their low excess cash peers by over 2% per year. This difference in returns cannot be explained by the commonly used factor models, and in fact increases to nearly 3% annually after controlling for the three risk factors of Fama and French (1993). I explore potential sources of the difference in future performance, providing evidence that it is related to managerial stock-selection and market-timing skills, as well as to ability to anticipate fund flows and control fund expenses. I show that managers of high excess cash funds are skilled at making stock purchasing decisions: the stocks that they buy perform considerably better than those purchased by the low excess cash group. I also explore whether the positive relationship between excess cash and fund performance relates to market-timing abilities of the managers. I show that market-timing skills are worse for the low excess cash group. As a result, the portfolio that is long the high excess cash funds and short the low excess cash group exhibits positive although statistically insignificant market timing. I additionally find that managers of funds with volatile recent excess cash holdings appear to be successful in timing the market: these managers actively adjust their cash position in response to their changing expectations about future market returns, increasing it prior to market downturns and decreasing it before periods of strong market performance. I also link the positive relationship between excess cash and future fund performance to supe- rior ability of managers of high excess cash funds to anticipate fund outflows. The difference in performance of the top and bottom excess cash groups is particularly pronounced when future fund flows are low, and is somewhat weaker when fund flows are high, suggesting that low excess cash funds do not carry sufficient cash on hand to cover the outflows and are likely forced to liquidate some of their holdings, damaging fund performance. High excess cash group, on the other hand, is well positioned to meet fund outflows, and generates better returns. Finally, consistent with a model I develop, I show that high excess cash can also be interpreted as a proxy for the ability to control fund expenses, and find that managers of high excess cash incur significantly lower expenses in the future than do their low excess cash peers. To define excess cash, I complement prior research by documenting new important determinants of mutual fund cash holdings. In particular, I show that funds holding riskier, less liquid, or low dividend-paying stocks, as well as those run by managers with lower return gap (cf. Kacperczyk et al., 2008), carry more cash. Compared to the determinants of fund cash holdings studied in the prior literature, the characteristics I consider explain three times more cross-sectional variation in cash positions of the mutual funds. 62 Table 3.1: Summary Statistics CASH TNA EXP 12B1 FL DL TURN RU12 FF1 FF6 FF12 σFF RG DY βMktFund β Mkt Hold β Liq Hold PRCDS Mean 5.029 1,682 1.287 0.408 1.389 0.480 0.826 8.219 0.005 0.083 0.178 0.748 -0.184 1.557 0.962 1.048 0.002 -6.638 Median 3.341 232 1.225 0.367 1.005 0.000 0.621 7.150 -0.004 -0.017 -0.017 0.056 -0.157 1.482 0.936 1.005 -0.009 0.358 10th Pctl 0.662 16.7 0.774 0.188 0.000 0.000 0.159 -3.604 -0.033 -0.138 -0.241 0.005 -0.594 0.650 0.626 0.695 -0.301 -33.408 90th Pctl 11.293 3,455 1.889 0.747 3.184 1.646 1.682 21.103 0.044 0.310 0.916 1.566 0.200 2.504 1.338 1.461 0.299 17.498 Stdev 5.416 5,682 0.464 0.242 1.476 1.037 0.772 10.792 0.065 0.447 0.577 2.030 0.384 0.817 0.313 0.349 0.300 35.217 Correlations CASH 1.000 TNA -0.046 1.000 EXP 0.129 -0.423 1.000 12B1 -0.022 -0.038 0.513 1.000 FL 0.029 0.095 0.249 0.208 1.000 DL -0.019 -0.063 0.014 -0.058 -0.044 1.000 TURN 0.033 -0.114 0.198 0.067 0.007 0.009 1.000 RU12 0.060 0.092 -0.064 -0.040 -0.061 0.040 -0.002 1.000 FF1 0.086 -0.033 0.049 0.031 0.011 0.001 0.000 0.242 1.000 FF6 0.075 -0.048 0.073 0.057 0.022 0.011 0.027 0.233 0.541 1.000 FF12 0.078 -0.040 0.077 0.074 0.035 0.010 0.004 0.237 0.478 0.705 1.000 σFF 0.003 -0.157 0.071 0.023 0.057 0.006 0.102 0.050 0.243 0.475 0.562 1.000 RG -0.085 0.071 -0.167 -0.039 -0.030 0.011 -0.071 0.111 0.012 0.008 0.006 0.004 1.000 DY -0.020 0.037 -0.170 -0.005 -0.012 -0.013 -0.162 0.067 -0.008 -0.017 -0.013 -0.003 -0.006 1.000 βMktFund -0.135 -0.016 0.107 0.024 0.015 -0.014 0.132 0.019 0.020 0.034 0.057 0.030 0.030 -0.414 1.000 βMktHold 0.017 -0.012 0.166 0.035 0.014 0.009 0.187 -0.002 0.041 0.055 0.063 0.022 0.000 -0.476 0.698 1.000 βLiqHold 0.035 -0.002 0.003 0.003 -0.030 0.039 -0.018 0.167 0.013 0.008 0.023 -0.019 -0.015 0.078 -0.017 -0.127 1.000 PRCDS -0.078 0.039 -0.054 -0.012 -0.004 -0.010 -0.021 -0.216 -0.410 -0.628 -0.550 -0.296 -0.045 -0.021 0.025 -0.026 -0.038 1.000 This table reports summary statistics for fund characteristics. CASH is percentage of total net assets held in cash; TNA is total net assets (in million); EXP is expense ratio, in percent; 12B1 is actual 12b-1 expenses, in percent; FL and DL are front and deferred loads, in percent; TURN is fund turnover ratio; RU12 is the 12-month fund return runup, in percent; FF1, FF6, and FF12 are prior 1-, 6-, and 12-month fund flows; σFF is the volatility of one-month fund flows over the previous 12 months, scaled by 100; RG is annual return gap over the prior 12 months, in percent, calculated following Kacperczyk et al. (2008); DY is fund annual dividend yield, in percent; βMktFund is market beta of the fund, calculated from market model regression using realized fund returns over the prior 12 months; βMktHold and β Liq Hold are market and liquidity betas of the fund holdings, calculated using data from prior 12 months; PRCDS is proceeds from fund stock sales less stock purchases, scaled by dollar value of all stock holdings. To compute betas of fund holdings, for each stock the fund holds I obtain market beta from the market model regression and liquidity beta from a two-factor model with market and Pastor and Stambaugh (2003) liquidity factors. βMktHold and β Liq Hold are weighted average loadings using the dollar value of investment in each stock as weights. To calculate PRCDS, I obtain contemporaneous (time t) fund holdings and holdings from six months ago, and for each fund compute it as 100 · [∑i pi,t−3 · (−Ni,t +Ni,t−6)] / [∑i pi,t−6Ni,t−6], where Ni,t is the number of shares of stock i held by the fund at time t and pi,t is the price of this stock at time t. Statistics are calculated in each cross-section and then averaged. Sample period is 1992-2008. 63 Table 3.2: Determinants of Fund Cash Holdings LNTNA EXP 12B1 FL DL TURN RU12 FF1 FF6 FF12 σFF RG DY βMktFund β Mkt Hold β Liq Hold PRCDS R 2 (1) -0.109 0.002 [-6.34] (2) 0.035 1.508 0.016 [2.19] [18.30] (3) 1.828 -2.181 0.021 [14.63] [-11.02] (4) 0.104 -0.028 0.003 [3.69] [-1.27] (5) 0.238 0.004 [2.62] (6) 0.041 0.010 [4.25] (7) 7.666 0.676 0.209 0.013 [6.78] [2.58] [1.74] (8) 0.061 0.001 [1.64] (9) 0.225 2.342 -2.581 0.077 -0.047 0.270 0.031 1.124 -0.237 0.054 [13.81] [17.86] [-11.65] [2.70] [-0.81] [3.01] [3.11] [11.97] [-5.54] (10) -1.740 0.030 [-4.25] (11) -0.098 0.001 [-2.14] (12) -2.643 0.024 [-7.59] (13) 0.544 0.784 0.010 [2.38] [3.49] (14) -0.014 0.008 [-10.18] (15) 0.038 9.960 -2.729 0.047 [5.19] [9.43] [-6.72] (16) 0.181 1.713 -1.947 7.034 0.786 -0.113 -1.499 -0.394 -5.866 3.846 1.545 0.138 [9.68] [11.12] [-8.22] [6.20] [4.06] [-2.34] [-3.82] [-4.81] [-9.38] [9.15] [3.64] (17) 0.140 1.746 -2.100 -0.009 0.009 0.096 0.060 5.183 0.766 0.647 -0.228 -1.479 -0.383 -5.740 3.818 1.416 0.002 0.158 [6.95] [10.86] [-7.68] [-0.28] [0.10] [0.80] [1.59] [3.86] [4.54] [1.87] [-4.04] [-3.91] [-4.89] [-5.63] [8.57] [3.38] [0.70] 64 This table reports the results of cross-sectional regressions of fund cash holdings as a percentage of total net assets on fund characteristics. LNTNA is log of total net assets (TNA is in million); EXP is expense ratio, in percent; 12B1 is actual 12b-1 expenses, in percent; FL and DL are front and deferred loads, in percent; TURN is fund turnover ratio; RU12 is the 12-month fund return runup, in percent; FF1, FF6, and FF12 are prior 1-, 6-, and 12-month fund flows; σFF is the volatility of one-month fund flows over the previous 12 months, scaled by 100; RG is annual return gap over the prior 12 months, in percent, calculated following Kacperczyk et al. (2008); DY is fund annual dividend yield, in percent; βMktFund is market beta of the fund, calculated from market model regression using realized fund returns over the prior 12 months; βMktHold and β Liq Hold are market and liquidity betas of the fund holdings, calculated using data from prior 12 months; PRCDS is proceeds from fund stock sales less stock purchases, scaled by dollar value of all stock holdings. To compute betas of fund holdings, for each stock the fund holds I obtain market beta from the market model regression and liquidity beta from a two-factor model with market and Pastor and Stambaugh (2003) liquidity factors. βMktHold and β Liq Hold are weighted average loadings using the dollar value of investment in each stock as weights. To calculate PRCDS, I obtain contemporaneous (time t) fund holdings and holdings from six months ago, and for each fund compute it as 100 · [∑i pi,t−3 · (−Ni,t +Ni,t−6)] / [∑i pi,t−6Ni,t−6], where Ni,t is the number of shares of stock i held by the fund at time t and pi,t is the price of this stock at time t. Reported are average slope coefficients, corresponding t-statistics, and average adjusted R2 values. Sample period is 1992-2008. 65 Table 3.3: Characteristics of Funds in Different Excess Cash Groups CASH LNTNA EXP 12B1 FL DL TURN RU12 FF1 FF6 FF12 σFF RG DY βMktFund β Mkt Hold β Liq Hold PRCDS A. Means Low 1.521 5.569 1.442 0.398 1.808 0.511 0.875 8.517 0.004 0.103 0.227 0.772 -0.229 1.525 0.934 1.069 0.019 -9.318 EC2 2.104 5.756 1.339 0.407 1.954 0.435 0.795 7.299 0.000 0.050 0.144 0.628 -0.165 1.602 0.976 1.044 0.006 -4.403 EC3 3.111 5.706 1.300 0.408 1.981 0.432 0.818 7.681 -0.001 0.057 0.138 0.769 -0.165 1.571 0.990 1.050 -0.013 -3.981 EC4 4.972 5.662 1.336 0.427 1.948 0.414 0.862 7.476 -0.001 0.083 0.175 0.865 -0.169 1.513 0.998 1.058 -0.015 -6.210 High 11.621 5.626 1.383 0.407 1.899 0.494 0.876 8.456 0.004 0.097 0.206 0.822 -0.197 1.534 0.979 1.080 -0.002 -7.463 B. Medians Low 1.180 5.601 1.393 0.353 2.107 0.000 0.683 7.192 -0.005 -0.017 -0.008 0.051 -0.198 1.457 0.918 1.024 0.001 -0.459 EC2 1.783 5.713 1.296 0.365 2.324 0.000 0.641 6.712 -0.007 -0.029 -0.034 0.051 -0.150 1.570 0.947 1.003 -0.007 1.218 EC3 2.905 5.692 1.255 0.364 2.412 0.000 0.655 7.304 -0.006 -0.029 -0.034 0.059 -0.151 1.522 0.957 1.008 -0.023 1.342 EC4 4.815 5.708 1.302 0.391 2.304 0.000 0.673 6.334 -0.005 -0.018 -0.014 0.066 -0.160 1.447 0.962 1.013 -0.029 1.040 High 9.896 5.564 1.359 0.360 2.223 0.000 0.648 7.539 -0.003 -0.003 0.017 0.065 -0.187 1.428 0.945 1.042 -0.012 -0.964 This table reports average (in Panel A) and median (in Panel B) characteristics of funds in different excess cash groups. CASH is percentage of total net assets held in cash; LNTNA is log of total net assets (TNA is in million); EXP is expense ratio, in percent; 12B1 is actual 12b-1 expenses, in percent; FL and DL are front and deferred loads, in percent; TURN is fund turnover ratio; RU12 is the 12-month fund return runup, in percent; FF1, FF6, and FF12 are prior 1-, 6-, and 12-month fund flows; σFF is the volatility of one-month fund flows over the previous 12 months, scaled by 100; RG is annual return gap over the prior 12 months, in percent, calculated following Kacperczyk et al. (2008); DY is fund annual dividend yield, in percent; βMktFund is market beta of the fund, calculated from market model regression using realized fund returns over the prior 12 months; βMktHold and β Liq Hold are market and liquidity betas of the fund holdings, calculated using data from prior 12 months; PRCDS is proceeds from fund stock sales less stock purchases, scaled by dollar value of all stock holdings. To compute betas of fund holdings, for each stock the fund holds I obtain market beta from the market model regression and liquidity beta from a two-factor model with market and Pastor and Stambaugh (2003) liquidity factors. βMktHold and β Liq Hold are weighted average loadings using the dollar value of investment in each stock as weights. To calculate PRCDS, I obtain contemporaneous (time t) fund holdings and holdings from six months ago, and for each fund compute it as 100 · [∑i pi,t−3 · (−Ni,t +Ni,t−6)] / [∑i pi,t−6Ni,t−6], where Ni,t is the number of shares of stock i held by the fund at time t and pi,t is the price of this stock at time t. Excess cash is calculated as the residual from the cross-sectional regression (1) on page 42. Statistics are computed in each cross-section and then averaged. Sample period is 1992-2008. 66 Table 3.4: Fund Raw Cash Holdings and Future Performance Low CASH2 CASH3 CASH4 High High-Low R2 Raw 0.45 0.49 0.49 0.49 0.47 0.02 [1.31] [1.49] [1.45] [1.48] [1.51] [0.23] αM -0.15 -0.08 -0.10 -0.08 -0.08 0.07 0.111 [-2.75] [-2.00] [-1.76] [-1.89] [-1.00] [0.92] αFF -0.16 -0.10 -0.06 -0.09 -0.08 0.07 0.149 [-2.85] [-2.31] [-1.03] [-1.95] [-0.95] [1.12] αCAR -0.17 -0.12 -0.08 -0.10 -0.12 0.04 0.193 [-2.96] [-2.76] [-1.44] [-2.08] [-1.72] [0.76] αPS -0.17 -0.10 -0.07 -0.08 -0.11 0.06 0.187 [-2.93] [-2.33] [-1.20] [-1.77] [-1.49] [0.76] αSD -0.16 -0.11 -0.08 -0.09 -0.12 0.04 0.183 [-2.90] [-2.61] [-1.34] [-2.01] [-1.75] [0.69] αFS -0.18 -0.11 -0.06 -0.09 -0.11 0.07 0.245 [-3.36] [-2.57] [-1.22] [-1.92] [-1.49] [1.04] This table reports average raw and risk-adjusted returns, in percent per month, and the corresponding t-statistics for different raw cash quintiles (CASH) as well as for the difference between quintiles of high and low cash. At the beginning of month t + 4, an investment is made in the funds that were assigned to a particular cash group as of the end of month t, and the position is held without rebalancing for the following 12 months. Row labeled ‘Raw’ shows average unadjusted returns. Risk-adjusted returns are from market model (αM ), Fama-French (1993) 3-factor model (αFF ), Carhart (1997) 4-factor model (αCAR), 4-factor model with added Pastor-Stambaugh (2003) liquidity factor (αPS), 4-factor model with added Sadka (2006) liquidity factor (αSD), and the conditional Ferson-Schadt 1996) model (αFS). R2 is the adjusted R2 from regressions using as dependent variable the difference in returns between high and low cash funds. Returns are weighted by total net assets. Sample period is 1992-2008. 67 Table 3.5: Fund Excess Cash Holdings and Future Performance Low EC2 EC3 EC4 High High-Low R2 Raw 0.32 0.43 0.46 0.56 0.51 0.18 [0.97] [1.57] [1.36] [1.69] [1.62] [2.42] αM -0.25 -0.10 -0.11 -0.02 -0.02 0.23 0.072 [-3.91] [-1.97] [-2.71] [-0.33] [-0.58] [2.99] αFF -0.26 -0.09 -0.09 0.00 -0.04 0.23 0.112 [-3.93] [-1.57] [-2.07] [-0.09] [-0.64] [3.04] αCAR -0.26 -0.09 -0.12 -0.04 -0.05 0.20 0.140 [-3.73] [-1.57] [-2.69] [-0.89] [-1.16] [2.51] αPS -0.25 -0.08 -0.10 -0.03 -0.05 0.21 0.134 [-3.62] [-1.41] [-2.29] [-0.57] [-1.03] [2.51] αSD -0.26 -0.09 -0.12 -0.04 -0.06 0.20 0.131 [-3.73] [-1.46] [-2.56] [-0.77] [-1.17] [2.48] αFS -0.26 -0.09 -0.11 -0.04 -0.04 0.22 0.223 [-3.90] [-1.39] [-2.60] [-0.75] [-0.91] [2.93] This table reports average raw and risk-adjusted returns, in percent per month, and the corresponding t-statistics for different excess cash (EC) quintiles as well as for the difference between quintiles of high and low excess cash. Excess cash is calculated as the residual from the cross-sectional regression (1) on page 42. At the beginning of month t+4, an investment is made in the funds that were assigned to a particular excess cash group as of the end of month t, and the position is held without rebalancing for the following 12 months. Row labeled ‘Raw’ shows average unadjusted returns. Risk-adjusted returns are from market model (αM ), Fama-French (1993) 3-factor model (αFF ), Carhart (1997) 4-factor model (αCAR), 4-factor model with added Pastor-Stambaugh (2003) liquidity factor (αPS), 4-factor model with added Sadka (2006) liquidity factor (αSD), and the conditional Ferson-Schadt (1996) model (αFS). R2 is the adjusted R2 from regressions using as dependent variable the difference in returns between high and low excess cash funds. Returns are weighted by total net assets. Sample period is 1992-2008. 68 Table 3.6: Fama-MacBeth Regression Results EC CASH RG LNTNA 12B1 FF1 RU12 βMktFund (1) 0.097 [3.17] (2) 0.054 [1.22] (3) 1.023 [1.69] (4) -0.224 [-2.48] (5) -1.535 [-2.95] (6) 0.093 0.804 -0.236 -1.212 [3.21] [1.42] [-2.58] [-2.35] (7) 0.056 1.990 [1.60] [0.23] (8) 0.082 0.042 [1.85] [0.71] (9) 0.070 1.587 [1.74] [0.41] (10) 0.070 5.468 1.821 [2.00] [0.98] [0.48] (11) 0.076 0.101 3.512 [2.38] [2.13] [1.03] (12) 0.080 -0.784 0.102 3.595 [2.52] [-0.24] [2.13] [1.06] This table reports the results of Fama-MacBeth regressions. Fund returns from month t+4 to t+15 (in percent) are regressed on the following variables measured at the end of month t: EC, excess cash; CASH, percentage of total net assets held in cash; RG, return gap, in percent, measured following Kacperczyk et al. (2008) over the 12-month period ending in month t; LNTNA, log of total net assets (TNA is in million); 12B1, actual 12b-1 expenses, in percent; FF1, prior 1-month fund flows; RU12, fund return runup during the 12 months ending in month t, in percent; and βMktFund, fund’s market beta, calculated from market model regression using realized fund returns from t − 11 to t. Excess cash is estimated from the month t cross-sectional regression (1) on page 42. The regressions are run each period when cash holdings data are observed (annually prior to 1999 and quarterly thereafter). Reported are average coefficients and the corresponding t-statistics. Sample period is 1992-2008. 69 Table 3.7: Fund Excess Cash Holdings and Future Performance Conditional On Positive Market Runup Low EC2 EC3 EC4 High High-Low R2 Raw 0.32 0.56 0.56 0.63 0.58 0.26 [0.85] [1.49] [1.47] [1.70] [1.61] [3.02] αM -0.28 -0.05 -0.06 0.03 0.00 0.28 0.027 [-3.53] [-0.74] [-0.93] [0.51] [0.04] [3.33] αFF -0.29 -0.06 -0.05 0.04 0.03 0.32 0.072 [-3.51] [-0.97] [-0.74] [0.63] [0.42] [3.73] αCAR -0.28 -0.09 -0.07 -0.01 0.00 0.28 0.108 [-3.32] [-1.39] [-1.10] [-0.22] [0.02] [3.29] αPS -0.27 -0.07 -0.05 0.01 0.01 0.28 0.100 [-3.13] [-1.06] [-0.81] [0.11] [0.17] [3.21] αSD -0.28 -0.08 -0.07 -0.01 0.00 0.28 0.098 [-3.34] [-1.30] [-1.06] [-0.16] [0.00] [3.28] αFS -0.30 -0.08 -0.06 -0.02 0.02 0.32 0.213 [-3.54] [-1.32] [-0.96] [-0.42] [0.31] [3.95] This table reports average raw and risk-adjusted returns, in percent per month, and the corresponding t-statistics for different excess cash (EC) quintiles as well as for the difference between quintiles of high and low excess cash, conditional on the 12-month market return prior to cash measurement being non-negative. Excess cash is calculated as the residual from the cross-sectional regression (1) on page 42. At the beginning of month t+ 4, an investment is made in the funds that were assigned to a particular excess cash group as of the end of month t, and the position is held without rebalancing for the following 12 months. Row labeled ‘Raw’ shows average unadjusted returns. Risk- adjusted returns are from market model (αM ), Fama-French (1993) 3-factor model (αFF ), Carhart (1997) 4-factor model (αCAR), 4-factor model with added Pastor-Stambaugh (2003) liquidity factor (αPS), 4-factor model with added Sadka (2006) liquidity factor (αSD), and the conditional Ferson-Schadt (1996) model (αFS). R2 is the adjusted R2 from regressions using as dependent variable the difference in returns between high and low excess cash funds. Returns are weighted by total net assets. Sample period is 1992-2008. 70 Table 3.8: Model-Based Excess Cash Holdings and Future Performance Low EC2 EC3 EC4 High High-Low Raw -0.25 -0.25 -0.08 -0.16 -0.12 0.13 [-0.66] [-0.66] [-0.19] [-0.40] [-0.29] [0.81] αM -0.20 -0.20 -0.02 -0.11 -0.07 0.13 [-2.18] [-3.10] [-0.35] [-2.01] [-0.82] [0.89] αFF -0.27 -0.20 0.01 -0.08 0.02 0.29 [-4.24] [-3.82] [0.16] [-1.38] [0.28] [2.67] αCAR -0.28 -0.19 -0.02 -0.08 -0.02 0.26 [-4.32] [-3.63] [-0.25] [-1.45] [-0.25] [2.40] αPS -0.28 -0.18 0.02 -0.07 -0.01 0.27 [-4.29] [-3.38] [0.28] [-1.28] [-0.16] [2.44] αSD -0.23 -0.25 0.03 -0.08 -0.03 0.20 [-2.92] [-4.10] [0.34] [-1.06] [-0.28] [1.47] αFS -0.25 -0.19 0.00 -0.07 -0.02 0.23 [-4.14] [-4.32] [0.06] [-1.27] [-0.24] [2.18] This table reports average raw and risk-adjusted returns, in percent per month, and the corresponding t-statistics for different excess cash quintiles (EC) as well as for the difference between quintiles of high and excess low cash. Excess cash is computed as the difference between actual cash-to-total net assets ratio and model-based target cash-to- total net assets ratio, scaled by the target ratio. At the beginning of month t + 4, an investment is made in the funds that were assigned to a particular excess cash group as of the end of month t, and the position is held without rebalancing for the following 12 months. Row labeled ‘Raw’ shows average unadjusted returns. Risk-adjusted returns are from market model (αM ), Fama-French (1993) 3-factor model (αFF ), Carhart (1997) 4-factor model (αCAR), 4-factor model with added Pastor-Stambaugh (2003) liquidity factor (αPS), 4-factor model with added Sadka (2006) liquidity factor (αSD), and the conditional Ferson-Schadt 1996) model (αFS). Returns are weighted by total net assets. Sample period is 1998-2008. 71 Table 3.9: Future Expenses and Liquidity vs. Excess Cash 12B1 EXP MGMT TURN βLiqHold Low 0.396 1.144 0.564 0.589 -0.088 EC2 0.414 1.098 0.527 0.520 -0.117 EC3 0.398 1.075 0.536 0.468 -0.040 EC4 0.387 0.997 0.496 0.475 -0.016 High 0.358 0.934 0.444 0.450 0.089 High–Low -0.038 -0.210 -0.120 -0.139 0.177 [-1.59] [-4.94] [-3.67] [-2.71] [2.52] This table reports average future expenses and average liquidity of holdings of funds in each of the five excess cash groups. Excess cash as of month t is calculated as the residual from the cross-sectional regression (1) on page 42, and expenses and liquidity betas are measured as of month t+ 12. 12B1 is actual 12b-1 expenses, in percent; EXP is expense ratio, in percent; MGMT is the management fee, in percent; TURN is fund turnover ratio; and βLiqHold is the liquidity beta of the fund holdings, calculated using data from months t+ 1 to t+ 12. To compute betas of fund holdings, for each stock the fund holds as of month t+ 12, I obtain the liquidity beta from a two-factor model with market and Pastor and Stambaugh (2003) liquidity factors. βLiqHold are weighted average loadings using the dollar value of investment in each stock as weights. The bottom two rows show the difference between values for high and low quintiles, and the corresponding t-statistic. Sample period is 1992-2008. 72 Table 3.10: Fund Excess Cash Holdings and Future Performance Conditional on Fund Size and Liquidity of Holdings Low Excess Cash High Excess Cash High–Low Excess Cash High Liq Med Liq Low Liq High Liq Med Liq Low Liq High Liq Med Liq Low Liq A. Raw Returns Small 0.65 0.30 0.47 0.50 0.64 0.49 -0.15 0.34 0.02 [1.74] [0.89] [1.26] [1.46] [1.63] [1.45] [-0.79] [2.02] [0.08] Medium 0.44 0.27 0.51 0.58 0.46 0.70 0.13 0.19 0.19 [1.32] [0.82] [1.41] [1.72] [1.39] [1.66] [1.02] [2.46] [1.03] Big 0.19 0.17 0.10 0.38 0.53 0.66 0.09 0.36 0.55 [0.55] [0.51] [0.29] [1.16] [1.66] [2.07] [0.75] [3.54] [3.77] B. 4-Factor Alpha Small 0.07 -0.34 -0.11 -0.10 0.02 -0.09 -0.17 0.36 0.02 [0.39] [-5.21] [-0.68] [-0.91] [0.17] [-0.64] [-0.83] [2.73] [0.07] Medium -0.21 -0.28 -0.08 -0.03 -0.09 0.02 0.18 0.19 0.10 [-2.28] [-3.77] [-0.57] [-0.20] [-1.28] [0.15] [1.41] [2.34] [0.59] Big -0.35 -0.39 -0.41 -0.20 -0.06 0.09 0.14 0.33 0.50 [-3.94] [-4.40] [-3.57] [-1.58] [-0.68] [0.92] [1.20] [3.24] [3.74] C. 5-Factor Alpha (Pastor-Stambaugh) Small 0.03 -0.35 -0.07 -0.11 0.03 -0.07 -0.15 0.38 -0.01 [0.17] [-5.25] [-0.40] [-1.03] [0.22] [-0.49] [-0.70] [2.80] [-0.03] Medium -0.23 -0.27 -0.05 -0.03 -0.08 0.04 0.20 0.19 0.10 [-2.49] [-3.57] [-0.39] [-0.20] [-1.12] [0.28] [1.56] [2.28] [0.57] Big -0.35 -0.38 -0.39 -0.21 -0.05 0.11 0.15 0.34 0.50 [-3.95] [-4.25] [-3.38] [-1.59] [-0.52] [1.14] [1.21] [3.25] [3.72] D. Ferson-Schadt Conditional Alpha Small 0.10 -0.34 -0.17 -0.11 0.07 -0.05 -0.21 0.41 0.13 [0.59] [-5.16] [-1.07] [-0.95] [0.48] [-0.32] [-1.02] [3.16] [0.56] Medium -0.17 -0.32 -0.11 0.00 -0.08 0.06 0.17 0.24 0.17 [-1.91] [-4.52] [-0.86] [-0.01] [-1.13] [0.43] [1.30] [3.03] [1.07] Big -0.35 -0.36 -0.44 -0.19 -0.04 0.07 0.16 0.32 0.51 [-4.06] [-4.11] [-3.81] [-1.46] [-0.49] [0.78] [1.33] [3.15] [3.80] This table reports average raw and risk-adjusted returns, in percent per month, and the correspond- ing t-statistics for high and low excess cash (EC) quintiles as well as for the difference between quintiles of high and low excess cash, conditional on both size and liquidity of fund holdings. Excess cash as of month t is calculated as the residual from the cross-sectional regression (1) on page 42. Fund size is measured as of month t, and liquidity is computed as the average loading of fund’s holdings on the Pastor-Stambaugh (2003) liquidity factor as of month t. At the beginning of month t+4, an investment is made in the funds that were assigned to a particular excess cash/size/liquidity group as of the end of month t, and the position is held without rebalancing for the following 12 months. Row labeled ‘Raw’ shows average unadjusted returns. Risk-adjusted returns are from Carhart (1997) 4-factor model, 4-factor model with added Pastor-Stambaugh (2003) liquidity fac- tor, and the conditional Ferson-Schadt (1996) model. Returns are weighted by total net assets. Sample period is 1992-2008. 73 Table 3.11: Future Performance of Stocks Bought and Sold by Funds in Different Excess Cash Groups All Buys ‘Old’ Buys ‘New’ Buys Sells Raw DGTW Raw DGTW Raw DGTW Raw DGTW Low 0.85 0.31 0.43 0.02 0.99 0.41 0.19 -0.23 2 0.85 0.30 0.50 0.04 0.97 0.39 0.30 -0.18 3 0.97 0.41 0.60 0.15 1.08 0.51 0.28 -0.19 4 0.88 0.32 0.56 0.10 0.97 0.40 0.31 -0.15 High 0.98 0.43 0.61 0.16 1.09 0.52 0.39 -0.09 High-Low 0.12 0.12 0.18 0.15 0.11 0.11 0.20 0.14 [1.69] [2.14] [1.88] [2.55] [1.41] [1.73] [2.91] [3.91] This table reports average monthly returns, in percent, of stocks bought and sold between months t+ 1 and t+ 12 (inclusive) by funds assigned to each excess cash quintile at the end of month t. Returns are calculated separately for four categories: (i) All Buys, which includes all share purchases; (ii) ‘Old’ Buys, which includes additions to the stocks already held at the beginning of the period; (iii) ‘New’ Buys, which includes purchases of stocks not held at the beginning of the period; and (iv) Sells, which includes shares sold. Purchase and sale transactions are assumed to take place at prices prevalent at the end of month t+ 6. Reported are raw returns and style-adjusted returns (DGTW, calculated following Daniel et al. (1997), as well as t-statistics for the difference in returns of high and low excess cash groups (in square brackets). Sample period is 1992-2008. 74 Table 3.12: Market Timing of Excess Cash Groups Treynor-Mazuy Henriksson-Merton δ0i δ1i δ2i φ0i φ1i φ2i Low -0.233 0.955 -0.141 -0.154 0.986 -0.062 [-3.52] [69] [-0.69] [-1.67] [42] [-1.40] EC2 -0.134 0.968 0.238 -0.192 0.936 0.060 [-2.48] [86] [1.43] [-2.53] [49] [1.66] EC3 -0.114 0.965 0.056 -0.117 0.960 0.008 [-2.29] [93] [0.36] [-1.67] [54] [0.23] EC4 0.010 0.945 -0.145 0.015 0.960 -0.027 [0.21] [91] [-0.95] [0.43] [54] [-0.82] High -0.047 0.896 -0.059 -0.039 0.902 -0.011 [-0.78] [70] [-0.32] [-0.46] [42] [-0.27] High-Low 0.186 -0.059 0.082 0.115 -0.083 0.051 [2.58] [-3.92] [0.37] [1.14] [-3.25] [1.06] This table reports the coefficients and the corresponding t-statistics of the market timing regressions. Treynor-Mazuy and Henriksson-Merton specifications are Rit = δ0i + δ1iRMt + δ2iR2Mt + ηit and Rit = φ0i + φ1iRMt + φ2i max(0, RMt) + νit, respectively, where Rit is the excess return on a portfolio of five excess cash quintiles or the difference in returns of high and low excess cash quintiles, and RMt is the market excess return. Intercepts are in percent. Excess cash is computed as the residual from regression (1) on page 42. Sample period is 1992-2008. 75 Table 3.13: Market Timing Conditional on Volatility of Excess Cash Low Var(EC) Var(EC)2 Var(EC)4 Var(EC)4 High Var(EC) 1 month 0.105 0.014 0.017 0.004 -0.009 [2.11] [0.52] [1.09] [0.56] [-0.56] 3 months 0.121 0.107 0.066 0.055 -0.021 [1.14] [2.20] [2.18] [1.81] [-2.00] 6 months 0.013 0.171 0.076 0.059 -0.034 [0.09] [2.59] [1.76] [1.36] [-2.04] 12 months -0.034 0.219 0.005 0.007 -0.021 [-0.13] [1.80] [0.06] [0.09] [-0.67] This table reports slope coefficients and t-statistics from the regression of N -month mar- ket return beginning in month t+ 1, N ∈ (1, 3, 6, 12), on aggregate excess cash holdings of five portfolio that are formed on the basis of volatility of past excess cash holdings, Var(EC), which is computed for each fund at time t as standard deviation of excess cash holdings from twelve quarterly observation in months t − 33, t − 30, ..., t − 3, and t. Sample period is 1998-2008. 76 Table 3.14: Fund Excess Cash Holdings and Future Performance Conditional on Future Fund Flows Low EC2 EC3 EC4 High High-Low R2 A. Low Future Fund Flows Raw -0.02 0.22 0.15 0.21 0.21 0.23 [-0.04] [0.84] [0.42] [0.59] [0.69] [1.61] αM -0.62 -0.34 -0.44 -0.39 -0.30 0.32 0.143 [-5.75] [-2.88] [-4.78] [-5.03] [-2.72] [2.45] αFF -0.57 -0.40 -0.43 -0.39 -0.43 0.25 0.304 [-5.17] [-4.03] [-4.78] [-4.99] [-3.81] [2.12] αCAR -0.55 -0.30 -0.36 -0.34 -0.25 0.30 0.370 [-4.82] [-3.11] [-4.07] [-4.36] [-2.80] [2.52] αPS -0.50 -0.26 -0.32 -0.32 -0.20 0.30 0.366 [-4.44] [-2.70] [-3.65] [-4.07] [-2.37] [2.44] αSD -0.55 -0.29 -0.37 -0.35 -0.26 0.29 0.364 [-4.96] [-3.01] [-4.24] [-4.47] [-3.06] [2.45] αFS -0.51 -0.30 -0.37 -0.34 -0.26 0.25 0.423 [-4.49] [-3.14] [-4.10] [-4.37] [-3.06] [2.16] B. High Future Fund Flows Raw 0.67 0.78 0.74 0.86 0.77 0.10 [2.06] [2.32] [2.08] [2.52] [2.05] [1.02] αM 0.12 0.18 0.14 0.29 0.22 0.11 -0.005 [1.40] [2.11] [1.45] [2.91] [1.24] [1.08] αFF 0.07 0.23 0.21 0.32 0.22 0.15 0.175 [0.78] [2.54] [2.21] [3.20] [1.32] [1.65] αCAR 0.02 0.06 0.10 0.19 0.11 0.10 0.213 [0.22] [1.28] [1.13] [2.10] [0.17] [1.04] αPS -0.01 0.05 0.11 0.19 0.12 0.13 0.244 [-0.12] [1.17] [1.27] [2.15] [0.28] [1.47] αSD 0.02 0.08 0.11 0.20 0.12 0.10 0.219 [0.21] [1.43] [1.25] [2.28] [0.21] [1.09] αFS 0.01 0.09 0.13 0.21 0.12 0.11 0.296 [0.12] [1.52] [1.50] [2.48] [0.28] [1.30] 77 This table reports average raw and risk-adjusted returns, in percent per month, and the corresponding t-statistics for different excess cash (EC) quintiles as well as for the difference between quintiles of high and low excess cash, conditional on future fund flows. Excess cash as of month t is calculated as the residual from the cross-sectional regression (1) on page 42. Fund flow is for the 12-month period from t + 4 to t + 15. Within each excess cash quintile, funds are assigned into ‘Low’ or ‘High’ fund flow groups. At the beginning of each month t + 4, an investment is then made in the funds that were assigned to a particular excess cash / fund flow group, and the position is held without rebalancing for the following 12 months. Row labeled ‘Raw’ shows average unadjusted returns. Risk-adjusted returns are from market model (αM ), Fama-French (1993) 3- factor model (αFF ), Carhart (1997) 4-factor model (αCAR), 4-factor model with added Pastor-Stambaugh (2003) liquidity factor (αPS), 4-factor model with added Sadka (2006) liquidity factor (αSD), and the conditional Ferson-Schadt 1996) model (αFS). R2 is the adjusted R2 from regressions using as dependent variable the difference in returns between high and low excess cash funds. Returns are weighted by total net assets. Sample period is 1992-2008. 78 Table 3.15: Summary Statistics: Closed-End Funds CASH NAV BETA RUNUP EXP AGE DISC Mean 1.79 557 0.79 9.54 1.65 23.90 20.96 Median 0.06 358 0.79 9.12 1.02 15.12 17.50 10th Pctl 0.00 64.5 0.19 -7.76 0.41 4.16 -1.86 90th Pctl 5.24 1,354 1.38 26.90 3.34 68.87 49.64 Stdev 4.86 602 0.48 14.52 2.09 23.46 22.31 Correlations CASH 1.00 LNNAV -0.25 1.00 BETA -0.10 0.09 1.00 RUNUP -0.03 0.08 0.04 1.00 EXP 0.23 -0.33 0.01 -0.02 1.00 AGE -0.07 0.68 0.05 0.11 -0.28 1.00 DISC -0.01 -0.22 -0.06 -0.13 -0.16 -0.12 1.00 This table reports summary statistics for closed-end fund characteristics. The sample contains funds with at least 50% of their net assets invested in U.S. equities. CASH is percentage of total net assets held in cash; NAV is net asset value (in million); BETA is market beta of the fund, calculated from market model regression using realized fund returns over the prior 12 months; RUNUP is the 12-month fund return runup, in percent; EXP is expense ratio, in percent; AGE is fund age, in years; and DISC and discount of net asset value per share relative to market share price, in percent. Statistics are calculated semiannually in June and December cross-sections and then averaged. Sample period is 1994-2008. 79 Table 3.16: Determinants of Closed-End Fund Cash Holdings LNNAV BETA RUNUP EXP AGE DISC R2 (1) -0.023 0.242 [-5.89] (2) -0.023 -0.010 0.028 0.259 [-6.02] [-1.75] [0.58] (3) -0.025 0.816 0.057 0.307 [-5.48] [2.81] [6.12] (4) -0.024 -0.006 0.035 0.888 0.048 0.287 [-5.53] [-0.95] [0.71] [3.18] [5.42] (5) -0.027 -0.036 0.285 [-6.49] [-5.50] (6) -0.025 -0.004 0.028 0.707 0.044 -0.027 0.315 [-4.59] [-0.65] [0.56] [1.73] [4.48] [-2.29] This table reports the results of the cross-sectional regressions of closed-end fund cash holdings as a percentage of net asset value on fund characteristics. LNNAV is log of net asset value (NAV is in million); BETA is market beta of the fund, calculated from market model regression using realized fund returns over the prior 12 months; RUNUP is the 12-month fund return runup, in percent; EXP is expense ratio, in percent; AGE is fund age, in years; and DISC and discount of net asset value per share relative to market share price, in percent. Reported are average slope coefficients, corresponding t-statistics, and adjusted R2 values. Coefficient on AGE is multiplied by 100. Sample period is 1994-2008. 80 Table 3.17: Excess Cash Holdings and Future Performance of Closed-End Funds Low EC2 EC3 EC4 High High-Low R2 Raw 0.29 0.27 0.36 0.32 0.23 -0.06 [0.19] [0.36] [1.10] [1.31] [0.41] [0.36] αM 0.02 -0.01 0.08 0.04 -0.07 -0.09 0.021 [-0.68] [-0.49] [0.66] [1.22] [-0.79] [0.22] αFF -0.29 -0.20 -0.09 -0.10 -0.27 0.02 0.065 [-1.67] [-1.80] [-0.16] [0.36] [-1.95] [0.59] αCAR -0.13 -0.11 -0.07 -0.04 -0.12 0.01 0.059 [-1.02] [-1.15] [0.19] [0.66] [-1.25] [0.32] αPS -0.12 -0.09 -0.02 -0.06 -0.15 -0.04 0.066 [-1.07] [-1.12] [0.14] [0.84] [-1.21] [0.40] αSD -0.11 -0.11 -0.06 0.00 -0.09 0.03 0.060 [-0.96] [-1.00] [0.31] [0.78] [-1.14] [0.33] αFS 0.08 0.00 -0.03 0.02 0.10 0.02 0.074 [0.02] [-0.58] [-0.01] [1.22] [0.05] [0.01] This table reports average raw and risk-adjusted returns, in percent per month, and the corresponding t-statistics for different excess cash quintiles (EC) of closed-end funds as well as for the difference between quintiles of high and low excess cash. Excess cash is computed as the residual from cross-sectional regressions of cash-to-net asset value of closed-end funds on fund size, lagged market beta of the fund, prior 12-month return, expense ratio, age, and fund discount. At the beginning of month t+ 4, an investment is made in the funds that were assigned to a particular excess cash group as of the end of month t, and the position is held without rebalancing for the following 12 months. Row labeled ‘Raw’ shows average unadjusted returns. Risk-adjusted returns are from market model (αM ), Fama-French (1993) 3-factor model (αFF ), Carhart (1997) 4-factor model (αCAR), 4-factor model with added Pastor-Stambaugh (2003) liquidity factor (αPS), 4- factor model with added Sadka (2006) liquidity factor (αSD), and the conditional Ferson- Schadt 1996) model (αFS). R2 is the adjusted R2 from regressions using as dependent variable the difference in returns between high and low excess cash funds. Returns are weighted by net asset value. Sample period is 1994-2008. 81 Figure 3.1: Fund Cash Holdings  ! " # $ %& ' ( ) * + , %- . / . 0 1 % # 2 , % 3   !  !" #$ % &' () *+ " $$$$,-, $$$$.-, $$$$/-, $$$$0-, $$$$1-, $$$$2-, $$$$3-, $$$$4-, $$$$5-, $$$$6-, $$$.,-, $$$..-, .66/ .661 .663 .665 /,,, /,,/ /,,1 /,,3 /,,5 78!* 78()!* This figure plots average and median fund cash holdings as a percentage of total net assets. The sample contains U.S. open-end mutual funds with growth objective. Cash holdings between 1992 and 1998 are annual, and quarterly thereafter. 82 Figure 3.2: Cumulative Abnormal Returns of Excess Cash Groups  ! " # $ %& ' ( ) * + , %- . / . 0 1 % # 2 , % 3   !  !" !# $% &' () *+ ,- ." $# )/ (% !. ,0 )1 (. 2( ,% 34 35 36 37 38 39 3: 3; 3< = < >-,%?)@-##-A&,B)1-.C%-#&-)*DD&B,"(,% = <= ;= := 9= 8= 7= E-A F ; F : F 9 G&B? This figure plots cumulative abnormal returns of each excess cash quintile during five years following portfolio assignment. Abnormal returns are estimated from the conditional Ferson-Schadt 1996 performance regression Rit = αFSi + β M i RMt + β HML i HMLt + β SMB i SMBt + β MOM i MOMt + ∑ F βFi (ZF,t−1RMt) + εit, where Rit is the excess return of each of the five excess cash fund groups, RMt is market excess return, HML, SMB, and UMD are value, size, and momentum factors, and ZF,t−1 is the demeaned value of the macroeconomic variables F at t−1, which include dividend yield of the S&P 500 index, term spread, default spread, and the three-month Treasury bill rate. 83 Figure 3.3: Performance of High–Low Excess Cash Portfolios  ! " # $ %& ' ( ) ' * ( + , - % . ' / 0 ' 1 2 ' 3 % 4 - % 4   !  ! "# $ %& %' () *+ ,# -. /! 0& (' #1 ,) %0 /2 #3 ,0 4, /) 567 58 7 8 67 68 97 98 :7 :8 ;7 76 <6 == : 76 <6 == ; 76 <6 == 8 76 <6 == > 76 <6 == ? 76 <6 == @ 76 <6 == = 76 <9 77 7 76 <9 77 6 76 <9 77 9 76 <9 77 : 76 <9 77 ; 76 <9 77 8 76 <9 77 > 76 <9 77 ? 76 <9 77 @ 76 <9 77 = This figure plots log cumulative abnormal returns of a portfolio long the high excess cash funds and short the low excess cash funds. Abnormal returns are estimated from the conditional Ferson- Schadt 1996 performance regression Rit = αFSi + β M i RMt + β HML i HMLt + β SMB i SMBt + β MOM i MOMt + ∑ F βFi (ZF,t−1RMt) + εit, where Rit is the difference in returns between high and low excess cash quintiles, RMt is market excess return, HML, SMB, and UMD are value, size, and momentum factors, and ZF,t−1 is the demeaned value of the macroeconomic variables F at t − 1, which include dividend yield of the S&P 500 index, term spread, default spread, and the three-month Treasury bill rate. 84 Figure 3.4: Effects of Costly Stock Trading on Cumulative Change in Cash Panel A  !" #$ %& '# & "( !& ) *+ ,' #$ ( & & - . / / / & & & & & &/ & & &. / / / & & &0 / / / & & &1 / / / / 2 . / . 2 0 / 0 2 Panel B  !" #$ %& '# & "( !& ) *+ ,' #$ ( & & - . / / / & & & - 0 / / & & & - 1 / / & & & &1 / / & & & &0 / / 2 3" # ( " 4 5 '* # / 6 . / . 6 1 / 1 6 7 % 8* 3% &2 3 " # ( " 4 5 '* # & * ( 5( 9 8 5% 3 &2 3 " # ( " 4 5 '* # & * ( 5( This figure plots cumulative changes in cash implied by the model developed in Section 7.1. Figure in Panel A is plotted assuming that nS = 100 shares of stock S are being sold at price pS = 100 with variable cost VS = 0.00006 and fixed cost FS = 10, and nB = 1000 shares of stock B are being bought at price pB = 10 with variable cost VS = 0.00003 and fixed cost FS = 10. In Panel B, VS = 0.00003 and VB = 0.00006 and it is assumed that purchase of stock B can only be financed by sales of stock S, i.e., cash reserves can only be used to cover transaction costs. 85 Bibliography Amihud, Y. (2002). Illiquidity and stock returns: Cross-section and time-series effects. Journal of Financial Markets, 5(1): 31–56. Barber, B. M., Odean, T., and Zheng, L. (2005). Out of sight, out of mind: The effects of expenses on mutual fund flows. Journal of Business, 78(6): 2095–2119. Bates, T. W., Kahle, K. M., and Stulz, R. M. (2009). Why do us firms hold so much more cash than they used to? Journal of Finance, 64(5): 1985–2021. Becker, C., Ferson, W., Myers, D. H., and Schill, M. J. (1999). Conditional market timing with benchmark investors. Journal of Financial Economics, 52(1): 119–148. Berk, J. B. and Green, R. C. (2004). Mutual fund flows and performance in rational markets. Journal of Political Economy, 112(6): 1269–1295. Boguth, O., Carlson, M., Fisher, A., and Simutin, M. (2009). Conditional risk and performance evaluation: Issues, solutions, and new estimates of momentum portfolio alphas. Carhart, M. M. (1997). On persistence in mutual fund performance. Journal of Finance, 52(1): 57– 82. Chang, E. C. and Lewellen, W. G. (1984). Market timing and mutual fund investment performance. Journal of Business, 57(1): 57–72. Chordia, T. (1996). The structure of mutual fund charges. Journal of Financial Economics, 41(1): 3–39. Cumby, R. E. and Glen, J. D. (1990). Evaluating the performance of international mutual funds. Journal of Finance, 45(2): 497–521. Daniel, K., Grinblatt, M., Titman, S., and Wermers, R. (1997). Measuring mutual fund performance with characteristic-based benchmarks. Journal of Finance, 52(3): 1035–1058. Edelen, R. M. (1999). Investor flows and the assessed performance of open-end mutual funds. Journal of Financial Economics, 53(3): 439–466. Elton, E. J., Gruber, M. J., and Blake, C. R. (2001). A first look at the accuracy of the crsp mutual fund database and a comparison of the crsp and morningstar mutual fund databases. Journal of Finance, 56(6): 2415–2430. 86 Evans, R. B. (2006). Does alpha really matter? Evidence from mutual fund incubation, termination and manager change. Fama, E. F. and French, K. R. (1993). Common risk-factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1): 3–56. Fama, E. F. and MacBeth, J. D. (1973). Risk, return, and equilibrium - empirical tests. Journal of Political Economy, 81(3): 607–636. Ferson, W. E. (2009). The problem of alpha and performance measurement. Ferson, W. E. and Schadt, R. W. (1996). Measuring fund strategy and performance in changing economic conditions. Journal of Finance, 51(2): 425–461. Grinblatt, M., Titman, S., and Wermers, R. (1995). Momentum investment strategies, portfolio performance, and herding: A study of mutual fund behavior. American Economic Review, 85(5): 1088–1105. Gruber, M. J. (1996). Another puzzle: The growth in actively managed mutual funds. Journal of Finance, 51(3): 1044–1044. Henriksson, R. D. (1984). Market timing and mutual fund performance - an empirical-investigation. Journal of Business, 57(1): 73–96. Henriksson, R. D. and Merton, R. C. (1981). On market timing and investment performance ii: Statistical procedures for evaluating forecasting skills. Journal of Business, 54(4): 513–533. Jain, P. C. and Wu, J. S. (2000). Truth in mutual fund advertising: Evidence on future performance and fund flows. Journal of Finance, 55(2): 937–958. Jegadeesh, N. and Titman, S. (1993). Returns to buying winners and selling losers - implications for stock-market efficiency. Journal of Finance, 48(1): 65–91. Jiang, G. J., Yao, T., and Yu, T. (2007). Do mutual funds time the market? Evidence from portfolio holdings. Journal of Financial Economics, 86(3): 724–758. Kacperczyk, M., Sialm, C., and Zheng, L. (2008). Unobserved actions of mutual funds. Review of Financial Studies, 21(6): 2379–2416. Keim, D. and Madhavan, A. (1996). The upstairs market for large-block transactions: Analysis and measurement of price effects. Review of Financial Studies, 9(1): 1–36. Madhavan, A. (2000). Market microstructure: A survey. Journal of Financial Markets, 3(3): 205– 258. Pastor, L. and Stambaugh, R. F. (2003). Liquidity risk and expected stock returns. Journal of Political Economy, 111(3): 642–685. 87 Sadka, R. (2006). Momentum and post-earnings-announcement drift anomalies: The role of liquid- ity risk. Journal of Financial Economics, 80(2): 309–349. Sirri, E. R. and Tufano, P. (1998). Costly search and mutual fund flows. Journal of Finance, 53(5): 1589–1622. Treynor, J. L. and Mazuy, K. K. (1966). Can mutual funds outguess market. Harvard Business Review, 44(4): 131–136. Wermers, R. (2000). Mutual fund performance: An empirical decomposition into stock-picking talent, style, transactions costs, and expenses. Journal of Finance, 55(4): 1655–1695. Yan, X. M. (2006). The determinants and implications of mutual fund cash holdings: Theory and evidence. Financial Management, 35(2): 67–91. Yan, X. M. (2008). Liquidity, investment style, and the relation between fund size and fund performance. Journal of Financial and Quantitative Analysis, 43(3): 741–767. 88 Chapter 4 Conclusion In this thesis, I examine the relationship between excess cash holdings and future performance under two organizational structures. In the first essay, I establish a positive link between corporate excess cash holdings and future stock returns and provide evidence consistent with the notion that excess cash proxies for growth options. In the second essay, I document new important determinants of cash holdings of mutual funds, show that high excess cash funds outperform their low excess cash peers, and attribute this difference in performance to differences in managerial abilities. Both essays are motivated by the empirical observations that cash holdings often differ dramat- ically even among seemingly comparable companies as well as among mutual funds pursuing the same investment objective. Clearly, some amount of cash is needed to run day-to-day operations and some amount is a result of recent activities, such as dividend payments. But any cash position above the level predicted by an organization’s characteristics – “excess cash” – reflects manage- rial discretionary decisions. As such, excess cash proxies for unobservable characteristics – such as investment opportunities or managerial abilities – that may relate to future performance of a financial entity. This thesis is the first empirical research aimed to understand the link between excess cash and future performance. In the context of corporations, I present evidence that firms build cash reserves in anticipation of future investment. High excess cash firms have or are acquiring growth options, as is reflected by their higher market betas. They are, therefore, riskier than their low excess cash peers and earn higher returns. The difference in stock returns of high and low excess cash firms – 6% annually after standard risk adjustment – is statistically significant and economically important. In the future, high excess cash companies exercise their growth options as is evidenced by their dramatically higher investment spending over the following decade. Curiously, I find that high excess cash firms underperform during market downturns. While somewhat counterintuitive, this result is consistent with the notion that growth options of these companies fall in value during such times. During market booms, on the other hand, the value of the growth options rises, leading to better stock performance of the high excess cash firms. Some findings of this paper are puzzling and warrant further research. In particular, it is interesting that high excess cash firms exhibit poor accounting performance over the course of a decade following portfolio assignment. If overinvestment is the reason for the poor profitability of such companies, the results of this paper raise questions regarding proper use of resources by the firms with large excess cash balances and, more generally, about the ability of managers to pick optimal levels of cash holdings. The link between excess cash and corporate governance is an 89 equally interesting research venue to pursue. Whereas this area has received some recent attention (e.g., Dittmar et al., 2003; Dittmar and Mahrt-Smith, 2007; and Harford et al., 2008), focusing on excess cash may yield new insights. In the second essay of this thesis, I document a positive relationship between excess cash holdings of actively managed equity mutual funds and future fund performance. Fund with high excess cash outperform their low excess cash peers by nearly 3% annually after controlling for the three risk factors of Fama and French (1993). I link this difference in future performance to managerial stock selection skills, market-timing abilities, capacity to anticipate fund liquidity needs, and aptitude at controlling operating costs. More specifically, I show that managers of high excess cash funds are better skilled at identifying which stocks to purchase: the stocks bought by high excess cash funds outperform purchases by their low excess cash peers by nearly 2% per year. I demonstrate that while neither high nor low excess cash fund managers excel at market timing, the former group exhibits somewhat better market timing abilities. I also tie the positive relationship between excess cash and future fund performance to superior ability of managers of high excess cash funds to anticipate fund outflows by presenting two broad pieces of evidence consistent with such explanation. First, I show that the difference in performance between high and low excess cash funds is particularly pronounced during times with low fund flows. Second, I observe no relationship between excess cash holdings and future performance for closed-end funds, whose managers do not face exogenous fund flow shocks. Finally, consistent with a model I develop, I show that high excess cash can also be interpreted as a proxy for the ability to control fund expenses, and find that managers of high excess cash incur significantly lower expenses in the future than do their low excess cash peers. To define excess cash, I complement prior research (Chordia, 1996; Yan, 2006) by documenting new important determinants of mutual fund cash holdings. Compared to the determinants of fund cash holdings studied in the prior literature, the characteristics I consider explain three times more cross-sectional variation in cash positions of the mutual funds. In particular, I show that funds holding riskier, less liquid, or low dividend-paying stocks, as well as those run by managers with lower return gap (cf. Kacperczyk et al., 2008) carry more cash. The findings of this paper raise interesting questions about managerial abilities to determine optimal levels of cash holdings. Why do the managers of low excess cash funds hold very little cash? This decision may reflect manager’s over-confidence: expectations of good future fund performance, low fund outflows, and ability to adjust quickly should a cash shortfall occur. Managers of low excess cash funds are indeed successful at identifying which stocks to sell, but all other evidence points to their lack of important skills: these managers generate very low returns for as long as five years following portfolio assignment, make poor stock purchasing decisions, do not exhibit market- timing abilities, and incur higher expenses in the future. The performance of funds run by such managers particularly suffers during periods of low fund flows, suggesting that these funds may benefit from maintaining a larger cash cushion and, more broadly, from a more thoughtful cash holdings policy. 90 Bibliography Chordia, T. (1996). The structure of mutual fund charges. Journal of Financial Economics, 41(1): 3–39. Dittmar, A. and Mahrt-Smith, J. (2007). Corporate governance and the value of cash holdings. Journal of Financial Economics, 83(3): 599–634. Dittmar, A., Mahrt-Smith, J., and Servaes, H. (2003). International corporate governance and corporate cash holdings. The Journal of Financial and Quantitative Analysis, 38(1): 111–133. Fama, E. F. and French, K. R. (1993). Common risk-factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1): 3–56. Harford, J., Mansi, S. A., and Maxwell, W. F. (2008). Corporate governance and firm cash holdings in the us. Journal of Financial Economics, 87(3):535–555. Kacperczyk, M., Sialm, C., and Zheng, L. (2008). Unobserved actions of mutual funds. Review of Financial Studies, 21(6): 2379–2416. Yan, X. M. (2006). The determinants and implications of mutual fund cash holdings: Theory and evidence. Financial Management, 35(2): 67–91. 91 Appendix A Appendix to Chapter 2 A.1 Data Definitions Book equity used to calculate the book-to-market ratio, BM, is defined following Davis et al. (2000) as stockholders’ book equity plus balance sheet deferred taxes plus investment tax credit less the redemption value of preferred stock. If the redemption value of preferred stock is not available, I use its liquidation value. If the stockholders’ equity value is not on Compustat, I compute it as the sum of the book value of common equity and the value of preferred stock. Finally, if these items are not available, stockholders’ equity is measured as the difference between total assets and total liabilities. Size is calculated as the log of real (CPI-adjusted) total assets. Cash flow, CF, is operating income before depreciation less interest less dividends less taxes divided by total assets. Profitability is proxied for by the return on assets, ROA, calculated following Cooper et al. (2008) as operating income before depreciation over total assets. Debt is computed similar to Titman et al. (2004) as the ratio of long-term debt to long-term debt plus the market value of equity. Investment, I, is defined as capital expenditures plus acquisitions less sale of property, plant, and equipment divided by total assets. Accruals, Accr, is estimated following Cooper et al. (2008) as [(change in current assets - change in cash) - (change in current liabilities - change in short-term debt - change in taxes payable) - depreciation expense] / average total assets. Asset growth, Ag, is the ratio of total assets to lagged total assets minus one. Issue is measured following Daniel and Titman (2006) as Ln[MEt−1 / MEt−36] – RU36, where MEt is market capitalization as of the end of month t and RU36 is the log three-year buy-and-hold return ending in month t− 1. A.2 Alternative Definitions of Excess Cash The definition of excess cash as a residual from cross-sectional regression follows from the prior literature and is very appealing, as it accounts for a wide number of variables that affect corporate cash holdings. To address any concerns regarding the sensitivity of the results using this particular method of estimating excess cash, I demonstrate that the empirical conclusions of this paper are 92 robust to alternative measures of excess cash. In particular, I first propose a modified regression specification to estimate ECM and then discuss an approach that does not require running a regression to obtain excess cash. In untabulated results, I also confirm that the findings presented in this paper are robust to omitting variables that have been found to relate to future returns (e.g., market-to-book ratio) from the cross-sectional regressions used to define excess cash. A.2.1 Modified Regression Specification A potential concern with the regression used in Section 2.2 is that it scales or transforms the explanatory variables in a number of different ways. For example, some regressors are logs of levels (Size), some are scaled by assets (e.g., CF), while others are scaled by sales (RD). To ensure that the results of this paper are not driven by this particular specification, I explore an alternative approach that uses log transformations of all variables. More specifically, to obtain an excess cash measure for stock i in month t, I again use all of the stocks that have fiscal year ends between t−11 and t, but run a different cross-sectional regression each month t: lnCiτ = γ0t + γ1tlnMEiτ + γ2tlnAiτ + γ3tlnCPXiτ + γ4tlnWCiτ + γ5tlnLTDiτ + γ6tlnCFiτ + γ7tln(σINDiτ ) + it, where lnC is the log of cash level, lnME is the log of market equity, lnA is the log of real assets, lnCPX is the log of capital expenditures level, lnWC is the log of level of net working capital calculated without cash, lnLTD is the log of level of long-term debt, lnCF is the log of cash flow level, and ln(σIND) is the log of industry sigma. As before, I include dividend and industry dummies.72 Table A.1 presents the results of this modified regression specification. As one would expect, larger firms tend to hold higher levels of cash. Consistent with the results of Table 2.1, firms with larger capital expenditures, working capital, and long-term debt tend to hold less cash, while those with higher cash flows and greater industry sigma hold higher levels of cash. As in Table 2.1, cash holdings of dividend paying companies are not statistically different from those of non-dividend- paying firms. Table A.2 reports average value- and equal-weighted returns of the ten ECM portfolios formed in the same manner as described in Section 2.3, but using the alternative excess cash definition. The results are remarkably consistent with those presented in Table 2.2. Firms with high excess cash outperform those with low values by 0.37% per month (0.38% when equal-weighted returns are used). This amount is both statistically (t-statistics of 4.84 and 5.12 for value- and equal- weighted results) and economically significant. As before, this result is robust in subperiods, with 72I do not include research and development expenditures as an additional explanatory variable as many companies report a zero R&D level. Including this variable in a log form as is done with other regressors will dramatically (by more than 50%) decrease the sample size. In unreported results that include log R&D level as a regressor, I find that the conclusions of this paper still hold. 93 the difference in returns of high and low excess cash groups amounting to 0.44% from 1960-1982 and reaching 0.30% during the 1983-2006 subperiod (0.43% and 0.33%, respectively, when returns are equal-weighted). To ensure that excess cash does not simply proxy for other variables known to relate to future returns, I perform Fama-MacBeth 1973 regressions controlling for a number of firm characteristics that have been previously found to predict stock returns. The results of these regressions, presented in Table A.3, are in line with those presented in Table 2.4. Excess cash retains its stock return forecasting ability even after controlling for size, book-to-market ratio, asset growth, and other firm characteristics. In unreported results, I use the alternative excess cash measure introduced in this section to confirm the findings presented in other tables, but omit them for brevity. In particular, I find that high excess cash firms are riskier than their low cash peers, where risk is proxied for by market beta (as in Table 2.2). Neither market, three- and four-factor asset pricing models, nor models that include asset growth, accruals, and leverage factors can explain the difference in returns between high and low excess cash groups (as in Table 2.6). High excess cash firms perform worse than their low cash peers in down markets (as in Table 2.7), and future investment activity increases with, while future profitability is unrelated to, excess cash (as in Table 2.8 and Figures 2.2 and 2.3). A.2.2 Simplified Excess Cash Definition Estimation of excess cash as a residual from cross-sectional regressions is attractive because it controls for a number of variables that affect corporate cash holdings. However, one may be concerned that the empirical conclusions of this paper are sensitive to this estimation method. I now propose a simplified approach of computing ECM that does not rely on conducting regressions. I estimate this alternative excess cash measure for a given firm as the difference between the log of ratio of cash to total assets of this firm and the log of median ratio of cash to total assets of all firms in the same size decile and in the same two-digit SIC industry. Stocks are then assigned into ten portfolios in a manner similar to that outlined in Section 2.3. This method is very straightforward, although, unlike the approach employed throughout the paper, it clearly does not account for a number of other important determinants of cash holdings. The purpose of this simpler method is to demonstrate the robustness of the empirical results by demonstrating that even a less elaborate measure of excess cash retains the ability to forecast stock returns. Table A.4 presents average returns and the corresponding t-statistics for each ECM decile and for the difference between the high and low ECM groups. Consistent with the results of Table 2.3, firms with higher ECM earn greater stock returns in the future. Over the entire 1960-2006 period, the difference in value-weighted (equal-weighted) returns of high and low excess cash deciles amounts to 0.34% (0.44%) per month with a corresponding t-statistic of 4.13 (4.72). The difference in returns retains its statistical and economic significance in both subperiods considered, with value- weighted (equal-weighted) returns averaging 0.34% and 0.34% (0.36% and 0.50%) in the 1960-1982 and 1983-2006 subperiods, respectively. 94 To test the robustness of the positive link between the measure of excess cash discussed in this section and future stock performance, I perform Fama-MacBeth 1973 regressions to control for a number of variables related to future stock returns. Table A.5 delivers a message similar to that of Table 2.4. Excess cash holdings is a robust predictor of future stock returns even after controlling for market beta, book-to-market ratio, size, asset growth, accruals, investment, cash flow, leverage, momentum, and share issuance. The excess cash measure retains its significance in each of the specifications considered. As with the modified regression specification, in unreported results, I confirm the robustness of other empirical findings presented in the paper, but omit them for brevity. A.3 Results Obtained Using Equal-Weighted Returns The empirical findings of this paper are similar regardless of whether I use value- or equal-weighted returns. To keep the presentation focused, in the main body of the paper I study value-weighted returns. I now summarize the results obtained using equal-weighted returns. Table A.6 presents average returns of each ECM decile and of the portfolio long high excess cash firms and short the low ECM group. High excess cash firms robustly outperform their low excess cash peers: in the full sample, the difference in returns between top and bottom ECM deciles is 0.45% per month (t-statistic of 4.56). The difference in returns is also statistically and economically significant in the subsamples, amounting to 0.32% during 1960-1982 and reaching 0.58% during 1983-2006. The results of unconditional regressions of equal-weighted returns from the high minus low ECM portfolio on the commonly used four factors (market, value, size, and momentum) and the three factors I constructed (asset growth, accruals, and leverage) are shown in Table A.7. The results are similar to those presented in Table 2.6: profitability of the high minus low ECM strategy remains significant in each regression specification. Finally, Table A.8 confirms that high excess cash firms underperform their low-cash counterparts in the worst states of the market. The difference in equal-weighted returns of high and low ECM deciles is -0.17% during such times. During market upturns, on the other hand, high excess cash firms outperform the low excess cash group. During the best state of the market (‘High’), the difference in returns amounts to over 1% monthly. A.4 ECM Returns Conditional on Book-to-Market, Size, and Leverage In this Section, I confirm the robustness of the empirical relationship between excess cash and future returns by showing that returns of high excess cash firms exceed those of their low excess cash peers regardless of which book-to-market, size, and leverage group the firms belong to. Book- to-market ratio is commonly interpreted as a proxy for growth options, and the empirical results 95 of this paper are consistent with the notion that excess cash also proxies for growth opportunities. It is thus interesting to explore whether high excess cash stocks outperform their low excess cash peers in different book-to-market groups. It is also pertinent to explore the relationship between excess cash and future returns conditional on firm size: smaller firms arguably have more restricted access to capital markets than do larger ones, and it is natural to conjecture that excess cash is particularly valuable for smaller companies. Finally, in valuation settings, cash is often viewed as negative debt, and it is interesting to ask whether the link between excess cash and future returns depends on firm leverage. To explore whether high excess cash firms outperform the low excess cash group regardless of book-to-market, size, or leverage, I assign stocks into ECM deciles and independently sort them into tertiles on the basis of either book-to-market ratio, size, or debt. The results are similar regardless of whether value-weighted (Table A.9) or equal-weighted returns (Table A.10) are used, and I will focus the discussion on the value-weighted case. Panel A of Table A.9 assigns stocks into groups conditional on both book-to-market ratio and ECM. Regardless of which book-to-market group the firms belong to, high excess cash companies generate higher returns than do their peers with lower excess cash. The difference in high minus low ECM returns amounts to 0.45% per month for the low book-to-market firms and to 0.29% for the high book-to-market group. This confirms that the relationship between excess cash and future returns is not driven by differences in book-to-market ratios. Panel B illustrates that while the relationship between excess cash and future stock returns is particularly pronounced for smaller firms, it is also present for the larger ones. In particular, the difference in returns of high and low ECM groups amounts to 0.81% per month for the smallest stocks, and to a lower but sizeable 0.18% per month for their larger counterparts. This observation is consistent with the intuition that smaller stocks are considerably more sensitive to shocks to cash holdings and lack easy access to external financing: in times of economic downturns smaller stocks with low cash may run into trouble due to slowing business and lack of access to credit, while in times of economic expansion these firms may not have enough resources to take on profitable investment opportunities. Their high excess cash peers, on the other hand, are better suited to withstand economic hardships and to take advantage of the booming times when investment opportunities abound. Under the simplifying assumption of absence of transaction costs or other frictions associated with debt financing, cash can be viewed simply as negative debt. Panel C of Table A.9, however, points out that this characterization is not empirically accurate: future returns increase with the level of excess cash in each leverage group considered, with the difference in returns of high and low ECM deciles amounting to 0.27% for the least levered companies and to 0.46% for the firms with high debt ratios. 96 Table A.1: Determinants of Cash Holdings – Modified Regression Specification Slope t-stat Intercept 1.253 6.410 lnME 0.177 16.355 lnA 1.292 62.501 lnCPX -0.114 -7.988 lnWC -0.197 -24.769 lnLTD -0.210 -13.130 lnCF 0.051 5.034 ln(σIND) 0.230 9.041 Div 0.023 1.222 R2 73.187 This table reports the results of the modified cross-sectional regressions used to estimate excess cash measures. Excess cash for firm i as of the end of month t is estimated as the residual it from the cross-sectional regression lnCiτ = γ0t+γ1tlnMEiτ +γ2tlnAiτ +γ3tlnCPXiτ +γ4tlnWCiτ +γ5tlnLTDiτ +γ6tlnCFiτ +γ7tln(σINDiτ )+it, where lnC is log of cash level, lnME is log of market equity, lnA is log of real assets, lnCPX is log of capital expenditures level, lnWC is log of level of net working capital calculated without cash, lnLTD is log of level of long-term debt, lnCF is log of cash flow level, and ln(σINDiτ ) is the log of industry sigma. Regressions also include a dividend dummy, Div, and industry dummies based on Kenneth French’s 17 industry definitions. Each cross-sectional regression uses all firms that have fiscal year ends between t− 11 and t. τ refers to the fiscal year ending between t− 11 and t. All variables with the τ subscript thus use the most recent data for firm i. Reported are average coefficients of December cross-sectional regressions, corresponding t-statistics, and average adjusted R2 values. Sample period is 1960-2006. 97 Table A.2: ECM Decile Portfolio Returns – Modified Regression Specification Period Low ECM2 ECM3 ECM4 ECM5 ECM6 ECM7 ECM8 ECM9 High High-Low A. Value-Weighted 1960-2006 0.979 1.080 1.030 1.083 1.126 1.247 1.219 1.257 1.300 1.348 0.369 [4.40] [4.68] [4.55] [4.71] [4.95] [5.44] [5.27] [5.45] [5.57] [5.68] [4.84] 1960-1982 0.911 1.032 0.957 1.029 1.105 1.142 1.060 1.075 1.190 1.349 0.438 [2.63] [2.90] [2.76] [2.93] [3.17] [3.29] [3.10] [3.15] [3.57] [3.86] [5.40] 1983-2006 1.043 1.125 1.100 1.134 1.145 1.348 1.372 1.432 1.405 1.347 0.304 [3.69] [3.80] [3.75] [3.79] [3.88] [4.47] [4.38] [4.59] [4.30] [4.18] [2.38] B. Equal-Weighted 1960-2006 1.045 1.166 1.131 1.133 1.228 1.360 1.316 1.353 1.405 1.427 0.382 [4.50] [4.89] [4.80] [4.80] [5.16] [5.69] [5.42] [5.65] [5.79] [5.93] [5.12] 1960-1982 0.995 1.187 1.033 1.115 1.246 1.279 1.161 1.196 1.318 1.429 0.434 [2.70] [3.18] [2.84] [3.04] [3.35] [3.47] [3.16] [3.30] [3.67] [3.91] [4.94] 1983-2006 1.094 1.145 1.225 1.151 1.211 1.437 1.465 1.503 1.488 1.426 0.332 [3.80] [3.81] [4.04] [3.83] [4.03] [4.67] [4.57] [4.78] [4.53] [4.52] [2.78] This table reports average value-weighted (in Panel A) and equal-weighted (in Panel B) returns, in percent per month, and the corresponding t-statistics for different excess cash measure (ECM) deciles as well as for the difference between deciles of high and low ECM for different time periods. Excess cash for firm i is defined as the residual from a modified regression specification described in Appendix B.1. Stocks are first sorted into quintiles based on market betas, and then into ECM deciles within each beta quintile. At the beginning of each month t, an investment is made in the stocks that were assigned to a particular ECM decile as of the end of month t − 5, and the position is held without rebalancing for the following 12 months. 98 Table A.3: Fama-MacBeth Regression Results – Modified Regression Specification ECM β BM ME Ag Accr I CF Debt RU12 Issue (1) 0.109 [5.51] (2) 0.111 -0.175 0.136 -0.102 [5.87] [1.70] [3.79] [2.59] (3) 0.115 -0.945 [5.84] [7.96] (4) 0.076 -3.552 [3.75] [10.21] (5) 0.092 -1.845 [4.51] [4.73] (6) 0.115 0.339 [6.31] [0.47] (7) 0.114 0.532 [6.51] [2.47] (8) 0.100 0.546 [5.78] [3.51] (9) 0.110 -0.698 [5.95] [5.84] (10) 0.064 -0.062 0.075 -0.142 -0.472 -2.654 -0.372 1.316 0.057 0.377 -0.376 [3.18] [0.67] [2.25] [3.74] [3.92] [7.96] [1.06] [2.19] [0.33] [2.80] [4.17] This table reports the results of Fama-MacBeth 1973 regressions. Every month stock returns in month t, in percent, are regressed on ECM, excess cash measure defined as the residual from a modified regression described in Appendix B.1; β is beta obtained from market model regressions using daily data from t−16 to t−5 with one lead and lag of market excess return; BM, log of book- to-market ratio, measured as in Davis et al. (2000); ME, log of market capitalization measured as of the end of t− 1; Ag, asset growth, defined as the ratio of total assets to lagged total assets minus one; Accr, Accruals, calculated as [(change in current assets - change in cash) - (change in current liabilities - change in short-term debt - change in taxes payable) - depreciation expense] / average total assets; I, Investment, defined as capital expenditures plus acquisitions less sale of property, plant and equipment, divided by total assets; CF, cash flow, computed as operating income before depreciation less interest less dividends less taxes over total assets; Debt, estimated as the ratio of long-term debt to long-term debt plus market value of equity; RU12, 12-month (t−12 to t− 1) compounded return; and Issue, measured as Ln[MEt−1/MEt−36]−RU36, where MEt is market capitalization as of the end of month t, and RU36 is the log 3-year buy-and-hold return ending in month t−1. Reported are average coefficients and t-statistics. Accounting data is taken from annual report for the fiscal year ending between t−16 and t−5. ECM is computed as of the end of month t− 5. Sample period is 1960-2006. 99 Table A.4: ECM Decile Portfolio Returns – Simplified Excess Cash Definition Period Low ECM2 ECM3 ECM4 ECM5 ECM6 ECM7 ECM8 ECM9 High High-Low A. Value-Weighted 1960-2006 0.880 1.005 1.087 1.082 1.193 1.170 1.192 1.232 1.211 1.221 0.341 [3.80] [4.24] [4.55] [4.60] [4.83] [4.86] [4.68] [4.85] [4.83] [4.85] [4.13] 1960-1982 0.916 1.003 1.043 1.000 1.138 1.207 1.220 1.205 1.204 1.261 0.344 [2.60] [2.81] [2.90] [2.90] [3.15] [3.41] [3.50] [3.51] [3.59] [3.59] [4.15] 1983-2006 0.845 1.008 1.129 1.160 1.247 1.136 1.164 1.257 1.218 1.183 0.338 [2.78] [3.20] [3.56] [3.61] [3.67] [3.45] [3.13] [3.36] [3.27] [3.28] [2.40] B. Equal-Weighted 1960-2006 0.980 1.058 1.200 1.204 1.303 1.276 1.277 1.360 1.314 1.414 0.435 [3.99] [4.26] [4.76] [4.86] [5.03] [5.01] [4.75] [5.13] [4.96] [5.33] [4.72] 1960-1982 1.064 1.091 1.184 1.152 1.246 1.345 1.331 1.373 1.317 1.427 0.363 [2.81] [2.90] [3.11] [3.12] [3.30] [3.55] [3.55] [3.73] [3.67] [3.83] [3.87] 1983-2006 0.899 1.025 1.215 1.255 1.358 1.211 1.225 1.347 1.312 1.403 0.503 [2.83] [3.15] [3.64] [3.76] [3.81] [3.53] [3.18] [3.52] [3.37] [3.71] [3.22] This table reports average value-weighted (in Panel A) and equal-weighted (in Panel B) returns, in percent per month, and the corresponding t-statistics for different excess cash measure (ECM) deciles as well as for the difference between deciles of high and low ECM for different time periods. Excess cash for firm i is defined as the difference between log cash-to-assets ratio of that firm and log of median cash-to-assets ratio of all firms in the same size decile of the industry in which firm i belongs. Stocks are first sorted into quintiles based on market betas, and then into ECM deciles within each beta quintile. At the beginning of each month t, an investment is made in the stocks that were assigned to a particular ECM decile as of the end of month 100 Table A.5: Fama-MacBeth Regression Results – Simplified Excess Cash Definition ECM β BM ME Ag Accr I CF Debt RU12 Issue (1) 0.105 [4.55] (2) 0.113 -0.140 0.139 -0.156 [5.53] [1.36] [3.81] [3.30] (3) 0.109 -0.740 [4.80] [7.36] (4) 0.088 -2.348 [3.81] [8.08] (5) 0.099 -1.119 [4.12] [3.31] (6) 0.118 0.221 [5.92] [0.39] (7) 0.137 0.393 [7.05] [1.94] (8) 0.104 0.320 [5.26] [2.03] (9) 0.109 -0.616 [5.02] [4.90] (10) 0.090 -0.072 0.057 -0.181 -0.394 -1.872 -0.022 1.116 0.021 0.173 -0.411 [4.56] [0.79] [2.06] [4.56] [4.38] [6.76] [0.08] [2.91] [0.13] [1.30] [4.73] This table reports the results of Fama-MacBeth 1973 regressions. Excess cash for firm i is defined as the difference between log cash-to-assets ratio of that firm and log of median cash-to-assets ratio of all firms in the same size decile of the industry in which firm i belongs. Every month stock returns in month t, in percent, are regressed on ECM, excess cash measure; β is beta from market model regressions with daily data from t − 16 to t − 5 with one lead and lag of market excess return; BM, log of book-to-market ratio, measured as in Davis et al. (2000); ME, log of market equity as of the end of t− 1; Ag, asset growth, defined as the ratio of total assets to lagged total assets minus one; Accr, Accruals, calculated as in Cooper et al. (2008); I, Investment, defined as capital expenditures plus acquisitions less sale of property, plant and equipment, divided by total assets; CF, cash flow, computed as operating income before depreciation less interest less dividends less taxes over total assets; Debt, the ratio of long-term debt to long-term debt plus market value of equity; RU12, 12-month (t−12 to t−1) compounded return; and Issue, measured as Ln[MEt−1/MEt−36] − RU36, where MEt is market capitalization as of the end of month t, and RU36 is the log 3-year buy-and-hold return ending in month t − 1. Reported are average coefficients and t-statistics. Accounting data is from annual report for the fiscal year ending between t− 16 and t− 5. ECM is as of the end of month t− 5. Sample period is 1960-2006. 101 Table A.6: ECM Decile Portfolio Equal-Weighted Returns Period Low ECM2 ECM3 ECM4 ECM5 ECM6 ECM7 ECM8 ECM9 High High-Low 1960-2006 0.958 0.998 1.176 1.207 1.243 1.272 1.268 1.378 1.348 1.409 0.451 [3.76] [4.06] [4.76] [4.88] [4.93] [5.01] [5.03] [5.39] [5.18] [5.08] [4.56] 1960-1982 1.063 1.015 1.208 1.251 1.219 1.263 1.211 1.332 1.387 1.378 0.315 [2.82] [2.74] [3.25] [3.38] [3.21] [3.41] [3.36] [3.70] [3.82] [3.70] [3.04] 1983-2006 0.858 0.981 1.145 1.165 1.266 1.280 1.323 1.423 1.311 1.439 0.581 [2.50] [3.01] [3.50] [3.52] [3.78] [3.68] [3.74] [3.91] [3.52] [3.50] [3.50] This table reports average equal-weighted returns, in percent per month, and the corresponding t-statistics for different excess cash measure (ECM) deciles as well as for the difference between deciles of high and low ECM for different time periods. Stocks are first sorted into quintiles based on market betas, and then into ECM deciles within each beta quintile. At the beginning of each month t, an investment is made in the stocks that were assigned to a particular ECM decile as of the end of month t− 5, and the position is held without rebalancing for the following 12 months. 102 Table A.7: Unconditional Risk Adjustment: Equal-Weighted Returns Intercept Mktrf HML SMB UMD AGF ACCRF DEBTF R2 (1) 0.451 [4.56] (2) 0.399 0.108 3.87 [4.09] [4.87] (3) 0.571 0.012 -0.303 0.073 16.87 [6.12] [0.51] [8.61] [2.39] (4) 0.521 0.018 -0.291 0.072 0.050 17.39 [5.43] [0.79] [8.18] [2.37] [2.13] (5) 0.390 -0.060 0.90 [3.85] [2.47] (6) 0.418 -0.045 0.10 [4.09] [1.26] (7) 0.507 -0.309 22.35 [5.81] [12.77] (8) 0.469 -0.049 0.017 -0.306 22.61 [5.17] [1.78] [0.41] [12.63] (9) 0.436 0.055 -0.127 0.038 0.026 -0.063 0.005 -0.228 26.71 [4.69] [2.47] [2.97] [1.21] [1.17] [2.12] [0.14] [7.47] This table reports the results of unconditional regressions of returns (in percent per month) from High minus Low excess cash measure (ECM) portfolio (equal-weighted returns are used) on market excess return (Mktrf), value (HML), size (SMB), momentum (UMD), asset growth (AGF), accruals (ACCRF), and leverage (DEBTF) factors. Mktrf, HML, SMB, and UMD are from Kenneth French’s data library. AGF, ACCRF, and DEBTF factors are calculated by taking an equal-weighted long position in the decile of stocks with the highest Ag, Accr, and Debt measures, respectively, and an offsetting short position in the decile of stocks with the lowest values. Reported are regression coefficients, t-statistics, and adjusted R2 values. Sample period is 1960-2006. 103 Table A.8: ECM Decile Portfolio Equal-Weighted Returns Conditional on Market State Mkt Low ECM2 ECM3 ECM4 ECM5 ECM6 ECM7 ECM8 ECM9 High High-Low Low -5.941 -6.083 -6.010 -5.998 -6.028 -6.100 -6.078 -5.939 -6.053 -6.112 -0.170 [11.95] [12.70] [12.56] [12.30] [12.66] [12.57] [12.97] [12.32] [12.64] [11.85] [1.07] 2 -1.620 -1.348 -1.284 -1.485 -1.358 -1.442 -1.279 -1.196 -1.207 -1.308 0.313 [5.22] [5.14] [5.28] [6.13] [5.55] [5.82] [5.34] [4.63] [4.75] [4.85] [1.75] 3 1.520 1.430 1.668 1.867 1.848 1.844 1.698 1.800 1.712 1.762 0.242 [4.80] [5.47] [6.33] [7.04] [6.69] [6.56] [6.65] [7.51] [6.09] [5.91] [1.45] 4 4.194 3.984 4.327 4.400 4.461 4.645 4.623 4.665 4.876 4.926 0.732 [11.25] [12.86] [14.21] [14.41] [13.40] [13.90] [12.54] [11.78] [11.38] [9.93] [2.49] High 6.579 6.942 7.112 7.189 7.227 7.347 7.311 7.496 7.348 7.711 1.132 [15.15] [16.87] [17.53] [18.98] [17.07] [18.88] [18.50] [18.45] [17.92] [16.03] [4.42] This table reports average equal-weighted returns, in percent per month, and the corresponding t-statistics for different excess cash measure (ECM) deciles as well as for the difference between deciles of high and low ECM for each market return group. To determine market return quintiles, months from January 1960 to December 2006 are assigned into 5 groups based on market return in that month. Sample period is 1960-2006. 104 Table A.9: ECM Decile Portfolio Returns Conditional on BM, Size, and Debt: Value-Weighted Returns Low ECM2 ECM3 ECM4 ECM5 ECM6 ECM7 ECM8 ECM9 High High-Low A. Returns Conditional on Book-to-Market Low 0.382 0.587 0.734 0.840 0.833 0.776 0.910 0.973 0.986 0.835 0.453 [1.40] [2.21] [2.74] [3.16] [3.13] [2.86] [3.38] [3.58] [3.34] [2.75] [3.43] Medium 0.931 0.989 1.049 1.140 1.018 1.303 1.246 1.292 1.343 1.495 0.564 [4.18] [4.29] [4.50] [4.87] [4.29] [5.43] [5.14] [5.26] [5.54] [5.51] [4.10] High 1.344 1.257 1.434 1.363 1.559 1.565 1.501 1.659 1.541 1.636 0.292 [5.76] [5.49] [6.27] [5.87] [6.50] [6.43] [6.33] [7.05] [6.69] [6.75] [2.52] B. Returns Conditional on Size Low 0.482 0.810 0.990 1.256 1.084 1.136 1.154 1.320 1.324 1.293 0.811 [1.45] [2.60] [3.28] [4.01] [3.49] [3.60] [3.77] [4.16] [3.94] [3.82] [4.97] Medium 0.832 0.874 1.098 1.041 1.212 1.260 1.171 1.227 1.357 1.361 0.529 [3.14] [3.39] [4.17] [3.99] [4.62] [4.79] [4.31] [4.50] [4.95] [4.54] [3.52] High 0.957 1.000 1.029 1.059 1.040 1.143 1.194 1.233 1.166 1.135 0.178 [4.59] [4.66] [4.83] [4.93] [4.75] [5.25] [5.58] [5.79] [5.28] [4.89] [1.79] C. Returns Conditional on Debt Low 0.914 0.952 1.059 1.238 1.094 1.084 1.122 1.053 1.135 1.182 0.268 [3.57] [3.84] [4.27] [5.03] [4.40] [4.33] [4.43] [3.94] [4.16] [4.08] [2.00] Medium 0.914 0.987 1.115 1.155 1.164 1.315 1.195 1.353 1.428 1.436 0.522 [3.99] [4.34] [4.91] [4.97] [4.97] [5.51] [5.11] [5.78] [5.67] [5.18] [3.09] High 0.653 0.778 0.963 0.888 1.035 1.132 1.171 1.238 1.229 1.115 0.462 [2.59] [3.06] [3.79] [3.48] [3.91] [4.25] [4.49] [4.67] [4.64] [3.80] [2.98] This table reports average value-weighted returns, in percent per month, and the corresponding t-statistics for different excess cash measure (ECM) deciles conditional on book-to-market, size, and debt. Stocks are assigned into ECM deciles and are independently sorted into tertiles based on either the ratio of book-to-market, measured as in Davis et al. (2000), or on size, measured as CPI-adjusted total assets, or on debt, measured as the ratio of long-term debt to long-term debt plus market value of equity. Sample period is 1960-2006. 105 Table A.10: ECM Decile Portfolio Returns Conditional on BM, Size, and Debt: Equal-Weighted Returns Low ECM2 ECM3 ECM4 ECM5 ECM6 ECM7 ECM8 ECM9 High High-Low A. Returns Conditional on Book-to-Market Low 0.431 0.533 0.812 0.833 0.860 0.715 0.913 1.002 1.033 0.873 0.442 [1.43] [1.87] [2.78] [2.92] [3.04] [2.50] [3.21] [3.50] [3.38] [2.77] [3.16] Medium 0.972 1.012 1.110 1.226 1.071 1.334 1.290 1.336 1.383 1.635 0.663 [3.90] [4.17] [4.52] [4.92] [4.27] [5.25] [5.06] [5.15] [5.43] [5.60] [4.28] High 1.474 1.377 1.578 1.494 1.734 1.723 1.638 1.818 1.647 1.830 0.356 [5.93] [5.67] [6.62] [6.16] [6.83] [6.74] [6.55] [7.30] [6.84] [7.19] [2.69] B. Returns Conditional on Size Low 0.944 1.065 1.326 1.482 1.354 1.323 1.343 1.518 1.447 1.570 0.626 [2.78] [3.45] [4.36] [4.74] [4.32] [4.26] [4.38] [4.80] [4.39] [4.67] [3.66] Medium 0.901 0.892 1.122 1.065 1.278 1.300 1.197 1.336 1.396 1.448 0.547 [3.39] [3.48] [4.25] [4.09] [4.86] [4.97] [4.46] [4.90] [5.18] [4.84] [3.52] High 0.992 1.014 1.040 1.060 1.071 1.167 1.247 1.249 1.191 1.168 0.176 [4.60] [4.59] [4.80] [4.83] [4.79] [5.22] [5.68] [5.77] [5.34] [5.03] [1.78] C. Returns Conditional on Debt Low 1.004 1.085 1.210 1.356 1.255 1.177 1.219 1.128 1.282 1.346 0.342 [3.63] [4.14] [4.65] [5.26] [4.73] [4.52] [4.56] [4.08] [4.60] [4.43] [2.34] Medium 1.049 1.012 1.209 1.261 1.307 1.393 1.286 1.444 1.468 1.585 0.536 [4.19] [4.18] [5.03] [5.03] [5.26] [5.42] [5.20] [5.82] [5.55] [5.42] [2.94] High 0.816 0.884 1.044 1.003 1.155 1.199 1.276 1.347 1.356 1.249 0.433 [2.89] [3.28] [3.91] [3.70] [4.17] [4.36] [4.72] [4.90] [4.92] [4.13] [2.63] This table reports average equal-weighted returns, in percent per month, and the corresponding t-statistics for different excess cash measure (ECM) deciles conditional on book-to-market, size, and debt. Stocks are assigned into ECM deciles and are independently sorted into tertiles based on either the ratio of book-to-market, measured as in Davis et al. (2000), or on size, measured as CPI-adjusted total assets, or on debt, measured as the ratio of long-term debt to long-term debt plus market value of equity. Sample period is 1960-2006. 106 Bibliography Cooper, M. J., Gulen, H., and Schill, M. J. (2008). Asset growth and the cross-section of stock returns. Journal of Finance, 63(4): 1609–1651. Daniel, K. and Titman, S. (2006). Market reactions to tangible and intangible information. Journal of Finance, 61(4): 1605–1643. Davis, J. L., Fama, E. F., and French, K. R. (2000). Characteristics, covariances, and average returns: 1929 to 1997. Journal of Finance, 55(1): 389–406. Fama, E. F. and MacBeth, J. D. (1973). Risk, return, and equilibrium - empirical tests. Journal of Political Economy, 81(3): 607–636. Titman, S., Wei, K. C. J., and Xie, F. X. (2004). Capital investments and stock returns. Journal of Financial and Quantitative Analysis, 39(4): 677–700. 107 Appendix B Appendix to Chapter 3 B.1 Determination of Fund Objectives CRSP Survivor-Bias-Free Mutual Fund Database contains a number of codes that allow the de- termination of the investment objective of a given fund. These codes are assigned by different institutions and cover different time periods and different funds: Strategic Insight objective codes are widely available between 1992 and 1999, Lipper classifications are present from the end of 1999 through 2008, and Wiescat codes begin in early 2008. To obtain the sample studied in this paper, I include only funds with AGG, GRI, GRO, and ING Strategic Insight objective codes, those with G, GI, LCGE, MCGE, MLGE, and SCGE Lipper codes, and those with AGG, GCI, and GRD Wiescat codes. I identify as international funds all funds containing any of the words ‘International’, ‘Global’, ‘Emerging Market’, or ‘Non-US’ in their objective description. To identify index funds, I search for the words ‘index’ or ‘S&P 500’ in fund name. CRSP has recently added a flag variable identifying index funds. After all of the filters described here are imposed, only 78 funds in my sample are identified as index by this flag. It is not immediately obvious from exploring the fund names that those funds are indeed index, and I retain them in my sample. Excluding these funds has no qualitative effect on the empirical results. B.2 Simplified Excess Cash Definition Definition of excess cash as a residual from cross-sectional regression (1) on page 42 is very appeal- ing, as it accounts for a wide number of variables that affect corporate cash holdings. To address any concerns about the sensitivity of the results to this particular way of estimating excess cash, I now show that the main empirical conclusions of this paper are robust to an alternative measure of excess cash. More specifically, to obtain excess cash for fund i as of the end of month t, I run a simpler cross-sectional regression CASHit = γ0t + γ1tRU12it + γ2tFF1it + γ3tβMktFund,it + εit, where, as before, CASHit is the percentage of total net assets of fund i held in cash as of the end of month t; RU12it is the 12-month fund return runup ending at the end of month t; FF1it is fund flow during month t; and βMktFund,it is market beta of the fund estimated using realized fund returns from t− 11 to t. Specification (15) of Table 3.2 details the results of this estimation. 108 I assign funds into excess cash quintiles in the same manner described in Section 4, and evaluate future performance of each group. Table A1 shows that controlling for prior return runup, recent fund flows, and fund market beta is sufficient to generate a positive relationship between excess cash and future fund returns. B.3 Transition Probabilities To determine the persistence of excess cash, I estimate the transition probabilities of the excess cash groups, shown in Table A2. Excess cash is rather transitory: for example, the likelihood of remaining in the low excess cash for two more years is just 9.7%. Furthermore, given that a fund was assigned to the middle excess cash group at the end of a year, it is almost equally likely to be in either of the five groups a year later. High excess cash funds have the highest probability among the five groups to remain in the same quintile a year later, yet almost 20% of such funds move to one of the bottom two excess cash groups within a year. B.4 Temporary vs. Permanent Excess Cash An interesting question to ask is whether funds carry excess cash because of a shock to their cash holdings, or whether excess cash is a persistent fund characteristic. To address this issue, I examine the differences in future performance of funds in the high and low excess cash groups conditional on their excess cash holdings being either transitory or permanent. I assign a fund that falls into a high excess cash group at time t into a permanent group if it also belonged to either high or next-to-high excess cash quintile in at least two thirds of the observations over the previous three years, and otherwise assign it to a transitory group. I similarly separate low excess cash funds into permanent and transitory groups.73 This procedure results in approximately equal number of funds in both transitory and permanent (T and P) categories of high and low (H and L) excess cash quintiles. Table A3 reports average returns of each of the resulting four groups (LP, LT, HP, HT), as well as the differences in returns between them. Funds in both transitory and permanent high excess cash groups outperform funds in the low excess cash group. Interestingly, funds in transitory groups perform better than those in the permanent groups. For example, LT funds outperform LP funds by 16 basis points monthly, whereas HT funds generate 15 basis points per month more than do HP funds. 73I require the funds to have at least three valid excess cash observations in at least two of the previous three years. 109 Table B.1: Fund Excess Cash Holdings and Future Performance: Simplified Excess Cash Definition Low EC2 EC3 EC4 High High-Low R2 Raw 0.36 0.51 0.56 0.56 0.55 0.19 [1.13] [1.56] [1.64] [1.69] [1.72] [1.71] αM -0.18 -0.06 -0.03 -0.02 0.00 0.18 -0.006 [-2.35] [-1.06] [-0.56] [-0.33] [-0.03] [1.71] αFF -0.28 -0.08 0.03 0.01 0.00 0.28 0.193 [-4.59] [-1.63] [0.48] [0.14] [0.03] [3.27] αCAR -0.21 -0.08 -0.03 -0.02 -0.03 0.18 0.278 [-3.73] [-1.53] [-0.64] [-0.46] [-0.88] [2.16] αPS -0.20 -0.07 -0.03 -0.01 -0.02 0.18 0.277 [-3.55] [-1.25] [-0.50] [-0.31] [-0.64] [2.20] αSD -0.21 -0.08 -0.03 -0.02 -0.03 0.18 0.276 [-3.66] [-1.52] [-0.50] [-0.39] [-0.91] [2.08] αFS -0.22 -0.09 -0.02 -0.02 -0.03 0.19 0.294 [-4.05] [-1.62] [-0.38] [-0.49] [-0.74] [2.46] This table reports average raw and risk-adjusted returns, in percent per month, and the corresponding t-statistics for different excess cash (EC) quintiles as well as for the difference between quintiles of high and low excess cash. Excess cash of fund i as of the end of month t is calculated as the residual from cross-sectional regression CASHit = γ0t + γ1tRU12it + γ2tFF1it + γ3tβMktFund,it + εit, where CASHit is the percentage of fund total net assets held in cash as of the end of month t; RU12it is the 12-month fund return runup ending at the end of month t; FF1it is fund flow during month t; and βMktFund,it is market beta of the fund, calculated from market model regressions using realized fund returns from months t − 11 to t. At the beginning of month t + 4, an investment is made in the funds that were assigned to a particular excess cash group as of the end of month t, and the position is held without rebalancing for the following 12 months. Row labeled ‘Raw’ shows average unadjusted returns. Risk-adjusted returns are from market model (αM ), Fama-French (1993) 3- factor model (αFF ), Carhart (1997) 4-factor model (αCAR), 4-factor model with added Pastor-Stambaugh (2003) liquidity factor (αPS), 4-factor model with added Sadka (2006) liquidity factor (αSD), and the conditional Ferson-Schadt (1996) model (αFS). R2 is the adjusted R2 from regressions using as dependent variable the difference in returns between high and low excess cash funds. Returns are weighted by total net assets. Sample period is 1992-2008. 110 Table B.2: Excess Cash Groups Transition Probabilities Lowτ+1 EC2τ+1 EC3τ+1 EC4τ+1 Highτ+1 Lowτ 0.31 0.27 0.18 0.14 0.11 EC2τ 0.23 0.25 0.24 0.18 0.10 EC3τ 0.18 0.22 0.23 0.23 0.14 EC4τ 0.14 0.15 0.24 0.25 0.22 Highτ 0.09 0.10 0.12 0.23 0.46 This table reports transition probabilities of the five excess cash groups. Excess cash as of the end of calendar year τ is estimated from the cross-sectional regression (1) on page 42. Sample period is 1992-2008. 111 Table B.3: Fund Excess Cash Holdings and Future Performance: Permanent vs. Transitory High and Low Excess Cash Low Excess Cash (L) High Excess Cash (H) Differences in Returns Perm (P) Trans (T) Perm (P) Trans (T) HT–LT HP–LP LT–LP HT–HP HT–LP HP–LT Raw 0.09 0.39 0.56 0.73 0.35 0.47 0.29 0.17 0.64 0.18 [0.22] [1.03] [1.50] [1.94] [2.42] [2.68] [2.16] [1.21] [3.65] [1.30] αM -0.56 -0.21 -0.03 0.14 0.35 0.53 0.35 0.17 0.70 0.18 [-4.78] [-2.11] [-0.25] [1.21] [2.39] [3.07] [2.68] [1.18] [4.04] [1.29] αFF -0.56 -0.25 0.01 0.11 0.36 0.57 0.31 0.10 0.66 0.26 [-4.54] [-2.44] [0.09] [0.91] [2.34] [3.25] [2.27] [0.67] [3.73] [1.84] αCAR -0.45 -0.28 -0.08 0.09 0.36 0.37 0.17 0.17 0.54 0.19 [-3.79] [-2.64] [-0.75] [0.71] [2.31] [2.28] [1.35] [1.13] [3.05] [1.36] αPS -0.44 -0.27 -0.08 0.05 0.32 0.36 0.17 0.13 0.49 0.19 [-3.62] [-2.51] [-0.69] [0.44] [2.04] [2.19] [1.30] [0.87] [2.78] [1.31] αSD -0.46 -0.29 -0.09 0.10 0.39 0.38 0.17 0.18 0.56 0.20 [-3.88] [-2.76] [-0.78] [0.80] [2.49] [2.30] [1.32] [1.23] [3.22] [1.42] αFS -0.43 -0.27 -0.03 0.11 0.38 0.40 0.16 0.15 0.54 0.24 [-3.57] [-2.50] [-0.31] [0.92] [2.38] [2.45] [1.23] [0.98] [3.07] [1.67] This table reports average raw and risk-adjusted returns, in percent per month, and the cor- responding t-statistics for four excess cash groups: permanent (P) and transitory (T) low (L) and high (H) excess cash portfolios as well as for the difference in returns between these groups. A fund that falls into a high excess cash group at time t is assigned into a permanent group if it also belonged to either high or next-to-high excess cash quintile in at least two thirds of the observations over the previous three years, and is assigned to a transitory group otherwise. Low excess cash funds are similarly separated into permanent and transitory groups. Excess cash as of month t is calculated as the residual from the cross-sectional regression (1) on page 42. At the beginning of month t + 4, an investment is made in the funds that were assigned to a particular excess cash group as of the end of month t, and the position is held without rebalancing for the following 12 months. Row labeled ‘Raw’ shows average unadjusted returns. Risk-adjusted returns are from market model (αM ), Fama-French (1993) 3-factor model (αFF ), Carhart (1997) 4-factor model (αCAR), 4-factor model with added Pastor-Stambaugh (2003) liquidity factor (αPS), 4-factor model with added Sadka (2006) liquidity factor (αSD), and the conditional Ferson-Schadt (1996) model (αFS). Returns are weighted by total net assets. Sam- ple period is 1992-2008. 112 Bibliography Carhart, M. M. (1997). On persistence in mutual fund performance. Journal of Finance, 52(1): 57– 82. Fama, E. F. and French, K. R. (1993). Common risk-factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1): 3–56. Ferson, W. E. and Schadt, R. W. (1996). Measuring fund strategy and performance in changing economic conditions. Journal of Finance, 51(2): 425–461. Pastor, L. and Stambaugh, R. F. (2003). Liquidity risk and expected stock returns. Journal of Political Economy, 111(3): 642–685. Sadka, R. (2006). Momentum and post-earnings-announcement drift anomalies: The role of liquid- ity risk. Journal of Financial Economics, 80(2): 309–349. 113

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:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0071109/manifest

Comment

Related Items