Agency Conflicts in Delegated Portfolio Management: Evidence from Namesake Mutual Funds

Transcription

1 Agency Conflicts in Delegated Portfolio Management: Evidence from Namesake Mutual Funds by Stephen P. Ferris Department of Finance 404 Cornell Hall University of Missouri Columbia Columbia, MO Tel: (573) and Xuemin (Sterling) Yan* Department of Finance 427 Cornell Hall University of Missouri Columbia Columbia, MO Tel: (573) * We thank W.D. Allen, John Howe, Mike Sykuta, John Zimmerman, George Wood, and seminar participants at the Contracting and Organizations Research Institute (CORI) of University of Missouri Columbia for helpful comments. All remaining errors are ours.

2 Agency Conflicts in Delegated Portfolio Management: Evidence from Namesake Mutual Funds Abstract Namesake funds provide a unique sample for studying the two agency conflicts existing within a mutual fund. The first is between the fund company and its shareholders while the second occurs between the fund company and the portfolio manager. A typical namesake fund manager sits on his fund s board, frequently as the chairman, is the majority owner of the fund company, and has significant investments in the fund he manages. We find that namesake funds are more tax-efficient, consistent with the idea that managerial ownership helps align the interests of managers with that of shareholders. Because of fewer career concerns, we observe that namesake fund managers herd less while assuming greater unsystematic risk. We find limited evidence that namesake fund managers outperform their benchmarks and peers. Namesake funds do not attract greater investor cash flows after controlling for performance. Finally, our results suggest that the boards of namesake funds are ineffective since these funds charge higher fees and display less consistency with their stated investment objectives. These results offer support for the SEC s proposal of an independent board chairman accompanied by a three-quarters majority of independent directors. Keywords: mutual funds; agency conflict; namesake funds JEL Codes: G23; G34 1

3 Agency Conflicts in Delegated Portfolio Management: Evidence from Namesake Mutual Funds 1. Introduction Over the last year, the numerous scandals in the mutual fund industry often accompanied by large settlements resulting from legal and regulatory action have brought tremendous attention to this industry. 1 With nearly one of every two American households invested in mutual funds and the aggregate value of the industry s holdings over 7 trillion dollars (Donaldson, 2004), such attention is hardly unexpected. Although there are numerous instances of late trading, market timing, self-dealing, and other abuses, the overarching issue facing the mutual fund industry is one of mitigating and managing its agency conflicts. Indeed, we believe that the well-publicized scandals and wrongdoing of mutual funds are the result of agency conflicts that have been inadequately addressed by both the industry and its regulators. There are two sets of agency conflicts operating within a mutual fund that generate the problems currently challenging the industry. These conflicts are graphically presented in Figure 1. The first of these is the conflict between the fund company, which is interested in generating fee income, and its shareholders, who desire to maximize the after-tax returns from their investment. The second conflict within the industry is more subtle, and consequently has received less attention from the media and regulators. This is the conflict between the fund company and the fund manager. This occurs because of career concerns by the fund manager that might incent them to make investment decisions inconsistent with shareholder wealth maximization. 1 Morningstar at com for instance, maintains a current listing of regulators charges, settlements and fund company statements in response to allegations. 2

4 This study will examine the nature and impact of these agency conflicts by using a special type of mutual fund: namesake funds. A namesake fund is a fund who shares its name with its managers. An example of a namesake fund is that of Baron Asset fund and the Baron Growth fund, both of which are run by Ron Baron, a former equity analyst. Indeed, namesake funds are typically founded and managed by former money managers or analysts. We select namesake funds for our analysis because the nature of namesake funds allows a unique examination of the various forces latent within the two agency conflicts that plague mutual funds. One such force is the board structure of namesake funds. The board structure of these funds might not be aligned with shareholder interests. Unlike other mutual funds, the portfolio manager of a namesake fund is almost always a director of the fund, and in many cases, also serves as board chair. Indeed, for our sample of namesake funds, over 98% of the portfolio managers serve as fund directors while 73% chair the board. This phenomenon of the portfolio manager simultaneously serving as the board chair provides the individual with an unusual ability to dominate both the fund s strategic and investment decisions. These decisions might not necessarily be consistent with the interests of shareholders. Further, our analysis of this feature of namesake funds provides evidence regarding the usefulness of the recent proposal by the SEC to require an independent board chair for mutual funds accompanied by a three-quarters majority of independent directors. 2 Another force shaping the nature of this first agency conflict is that of managerial ownership. In the case of namesake funds, however, portfolio managers have two ownership interests. The first ownership interest of the portfolio manager is in the fund itself. Prior studies 2 William Donaldson, current chair of the SEC, argued in Congressional testimony (8 April 2004) that such independence was an important part of the SEC s efforts to restore overall accountability to the fund board. 3

5 of corporate managers suggest that managerial ownership helps to align the interests of managers with those of shareholders (see e.g., Morck, Shleifer, and Vishny (1988) and McConnel and Servaes (1990)). Namesake fund managers have significant investments in the funds they manage. Based on information gathered from the Securities and Exchange Commission (SEC), nearly 70% of namesake fund managers own more than $100,000 in the funds they manage, and over 90% of these managers own more than $100,000 in the funds belonging to their fund family. 3 Further, in an industry study, Sanders (1998) identifies twelve fund managers who significantly invest with shareholders. Nine out of these twelve managers are namesake fund managers. 4 Complementing the portfolio manager s direct ownership in the fund is his intangible ownership attributable to reputation. Namesake fund managers have significant reputational capital at risk because their names are directly associated with the fund. This combination of tangible and intangible ownership interests in the fund, can help to align the interests of the portfolio managers with those of shareholders. The second ownership interest held by the portfolio manager is in the fund company. Unlike other fund managers, namesake managers hold a majority interest in the fund company. This moves the interests of the portfolio manager away from those of the fund s shareholders and towards those of the fund company. Although the namesake manager has two ownership interests, these interests are not typically comparable in magnitude. The portfolio manager s investment in the fund company is 3 Starting from 2002, the SEC requires funds disclose the dollar range of director investments in the fund and the family of funds. The range runs from $1 to $10,000, $10,001 to $50,000, $50,001 to $100,000 and more than $100, The nine namesake fund managers are Ron Baron, Michael Fasciano, George Mairs, Ron Muhlenkamp, David Schafer, Robert Torray, Roy Papp, David Dreman, and Chris Davis. The remaining three managers are Richard Lawson of Weitz Series Hickory, Martin Whitman of Third Avenue Value, and David Williams III of Excelsior Value and Restructuring. 4

6 usually much larger than that in the fund he manages. Consequently, we believe that the net incentive effect of the portfolio manager s ownerships will be to align his interest with that of the fund company rather than that of the shareholders. Namesake funds also provide a unique opportunity to study the second agency conflict within the mutual fund industry: that between fund managers and their employing fund companies. More specifically, we use this sample of namesake funds to examine how career concerns affect fund managers investment decisions. Chevalier and Ellison (1999a) analyze how manager behavior might be affected by a manager s desire to avoid losing his job. They show that career concerns of younger portfolio managers cause them to be more risk averse in selecting their fund s portfolios. In particular, younger managers tend to take on lower unsystematic risk and deviate less from the typical behavior than their older counterparts. Namesake fund managers, however, are their own bosses. They will not be easily terminated from their positions. Almost 80 percent of these managers are the founders of the fund companies, and nearly all of them are Chief Executive Officers (CEO) and/or Chief Investment Officers (CIO) of the firm. These two agency conflicts manifest themselves in a variety of the fund s characteristics and operations. More specifically, we examine how agency conflicts effect the fees, style consistency, tax efficiency, herding behavior, risk levels, return performance, and new investor cash flow of namesake funds. An analysis of namesake funds on these dimensions provides us with useful insights regarding the relative strength of factors that either mitigate or exacerbate agency conflicts. Such evidence will be useful to the public debate on how best to reform the mutual fund industry. 5

7 The factors contributing to these agency conflicts exert different, and sometimes contrary influences on the mutual fund. By analyzing the fee structure of these funds, we gain insight into how the boards of these funds balance the desire for higher fee revenue against a shareholder preference for lower fees and higher net returns. Further, this analysis will provide insight into the relative strengths of the conflicting incentives resulting from the portfolio manager s two ownership interests. Our analysis of the style consistency of these funds allows us to determine how a board chair that serves as portfolio manager balance individual investment desires and style against the publicly stated investment strategy. Since the boards of namesake funds are insider-dominated, we expect to find less effective governance manifested in greater style drift and higher fees. By examining the tax efficiency of namesake funds, we investigate the incentive effects of ownership. Because the portfolio managers of namesake funds typically have significant amounts of their own capital invested in these funds, we hypothesize that namesake funds will be more tax efficient than other funds. Because of their majority ownership in the fund company, these portfolio managers have less concern about career progression. Consequently, it is likely that managers of namesake funds will herd less and assume higher levels of unsystematic risk relative to other managers. An important manifestation of the two agency conflicts present within the mutual fund industry is the extent to which they influence fund performance. To examine this issue, we estimate two measures of performance for namesake funds. The first is realized stock returns while the second is risk-adjusted returns. The relative performance of namesake funds is unclear because of the conflicting influences of high levels of managerial ownership and a board that lacks independence. Finally, we examine the ability of the namesake fund to attract new investor 6

8 cash flow. Because the managers of namesake funds have a history of success as analysts and managers, such individuals have star power. Hence, we hypothesize that they have a greater ability to attract new investment capital than the managers of other funds. We organize our study as follows. In the following section, we describe our data and method of sample construction. Section three examines board effectiveness by analyzing the fee levels and style consistency of namesake funds. In section four, we discuss our findings regarding the tax efficiency of namesake funds. In section five we present our results from an analysis of the herding and portfolio risk profile of the portfolio managers for these funds. Section six contains our results from an analysis of the return performance of our sample funds while section seven presents our findings regarding investor cash flows. We conclude in section eight with a summary and a brief commentary. 2. Data and Sample Construction 2.1 Sample Construction Our primary data source for mutual funds is the Center for Research in Security Prices (CRSP) Survivor Bias-Free Mutual Fund Database. The sample period is Following many prior studies of mutual funds, we focus on diversified domestic equity funds. In particular, we exclude international funds, sector funds, and balanced funds. We also exclude index funds because they are not actively managed. 5 We use the ICDI fund objective code to assign each fund one of three investment objectives: Aggressive Growth (AG), Long-term Growth (LG), and Growth and Income (GI). 5 We identify index funds by searching for index in fund names. To ensure accuracy, We hand check all selected funds and delete spurious ones. 7

9 We identify namesake funds by comparing fund names and manager names. We first obtain all funds whose fund name and manager name share at least one word. We then exclude those funds where the common word is investment, team, capital, or management. To focus on agency issues, we also exclude those funds whose managers are sub-advisors. The managers of these funds are not employees of the fund companies; rather, they are hired by the fund companies to manage the funds as subcontractors. To ensure accuracy, we hand check all selected funds. In total, we identify 219 namesake funds. For our sample of namesake funds, we also hand collect background information about their managers and directors from various sources including fund prospectuses, statements of additional information (SEC form 485B), and fund websites. 2.2 Characteristics of Namesake Funds Table I presents the summary statistics for our sample of namesake funds. The second column presents the mean values for a number of namesake fund characteristics. For comparison, the third column presents the corresponding means for all other equity mutual funds. Namesake funds are on average smaller than other equity funds. An average namesake fund has total net assets (TNA) of $ million, while an average equity fund has $ million in total assets. Namesake funds also reside in smaller fund families. The fund family to which a namesake fund belongs has an average market share (by TNA) of 0.09%, compared to that of other equity funds at 1.08%. The average turnover rate of namesake funds is 56.67%, substantially lower than that of other equity funds, whose annual turnover is in excess of 76%. The average expense ratio of namesake funds is 1.24%, which is 27 basis points higher than that of other equity funds. This difference in expense ratios, however, can not be explained 8

10 by 12b-1 fees. The average 12b-1 fee of a namesake fund is 24 basis points, 12 basis points lower than that of other equity funds. The average total load charged by namesake funds is 1.16%, which is much lower than the 2.72% charged by other equity funds. Only 31.8% of the namesake funds charge a load fee, while over 54% of other equity funds require a load fee. This suggests an important difference in the distributional channels between namesake funds and other equity funds. Namesake funds are more likely to sell shares directly to investors while other equity funds are more likely to use brokers. The average tenure for namesake fund managers is 5.98 years, nearly 2 years longer than managers of other equity funds. 6 But this is to be expected. Namesake funds are usually familyowned and family-managed, and hence their managers are less likely to leave the fund or lose their jobs. We will examine the characteristics of namesake fund managers in greater detail in the following subsection. Namesake funds hold slightly more cash and less stock than other equity funds. On average, namesake funds hold 8.93% of their assets in cash and 87.99% in stocks, compared to 6.95% cash and 89.71% stocks for other equity funds. Further, we note that a higher percentage of namesake funds are Growth and Income funds compared to other equity funds. 2.3 Characteristics of Namesake Fund Managers Table II presents the characteristics of namesake fund managers. Unlike their counterparts in other equity funds, nearly all namesake fund managers (98.32%) are directors of the funds that they manage. Almost three quarters of these managers are chairman of their boards. Approximately 79% of the namesake fund managers are also founders of the fund 6 The average manager tenure (5.98 years) is less than the average fund age (8.05 years) for two reasons. First, some funds are managed by children or grandchildren of the founders of the funds. Second, the manager tenure data are missing for some older funds. These managers tend to have a longer-than-average tenure. 9

11 companies. Managers of namesake funds also hold significant ownership in the fund companies, and all of them are CEO and/or CIO of the fund companies. Almost all of the namesake fund managers have relevant prior experience. Panel B reports that approximately 65% of these managers were formerly portfolio managers. Another 20% or so were analysts or directors of research in money management firms. Approximately 12% of these managers were formerly vice presidents, managing directors, or CIOs of financial institutions. Interestingly, 2.38% of these managers are former university professors. Significantly, namesake fund managers have sizable investments in their funds and fund families. For example, 68.22% of these managers invested $100,000 or more in the funds they manage. Perhaps more revealing, 91.20% of these managers invested $100,000 or more in their fund family. We collect the above ownership information from SEC form 485B. The SEC requires that funds disclose director investments in the fund and fund family in terms of dollar ranges. 2.4 Governance Characteristics of Namesake Funds In this section, we describe the board composition of namesake funds. The boards of directors for namesake funds lack independence because their managers sit on the boards, often as the chairmen. Consequently, an analysis of these boards is central to our understanding of the agency conflicts that exist within a mutual fund. Panel A of Table III presents our findings concerning board composition. The average number of directors in a namesake fund is 6.9. Approximately two thirds of the directors are independent. Using a sample of funds in the 50 largest fund families, Tufano and Sevick (1997) report that there are on average 8.7 directors in a board, and 6.2 of them are independent directors. Compared to funds contained in Tufano and Sevick s sample, namesake funds have 10

12 smaller boards, but similar percentage of independent directors. The namesake funds have smaller boards in large part because they are smaller funds than those in the Tufano and Sevick s sample. 3. Fund Governance: Fees and Style Consistency The board of directors is central to effective mutual fund governance and the control of agency conflict between the fund and its shareholders. Indeed, the primary responsibility of the board of directors of mutual funds is to protect the interests of fund shareholders and thereby mitigate the agency conflicts that result from the divergent interests of shareholders and the fund. Specifically, the board decides whether to renew the investment advisory contract, votes on fee changes, approves any merger or liquidation, and monitors the specific investments within the portfolio to ensure that the adviser s investments comply with the fund s stated investment objectives. As noted previously, the boards of namesake funds are not independent. Almost all of the namesake fund managers sit on the board, and nearly three-quarters serve as chairman. Because of these characteristics of namesake funds, we hypothesize that their boards are less effective. We use two measures to assess the quality of governance for namesake funds. The first measure is fund fees. As argued by Tufano and Sevick (1997) and Del Guerico, Dann and Partch (2003), higher fees reflect less effective fund governance since the investment advisory contract which stipulates fund fees must be renewed annually by the board of directors. Consequently, we hypothesize that namesake funds are more likely to charge higher fees. This hypothesis is consistent with the presence of agency conflict despite ownership and reputational incentives to the contrary. 11

13 In Table I, we show that the average expense ratio of namesake funds is 27 basis points higher than other equity funds. It is possible, however, that this higher fee charged by namesake funds is due to the smaller size of these funds. Panel B of Table III presents our regression analysis of fund expense ratios, controlled for size and other fund characteristics. Consistent with casual empiricism, larger funds, older funds, and funds from larger families charge lower fees. After controlling for these factors, we still find that namesake funds charge significantly higher fees than other equity funds. Specifically, the coefficient of the namesake fund dummy variable is above 0.2, suggesting that the average expense ratio of namesake funds is more than 20 basis points higher than other equity funds. Our second measure of board effectiveness is style consistency. One of the responsibilities of the board is to monitor fund portfolios to ensure that they comply with the stated investment objectives of the fund. If a fund has a weaker board, then it is more likely to suffer style drift. Following Brown and Harlow (2002), we estimate style consistency using two proxies: the R-square and the tracking error of the regression of fund returns on factor returns. Panel C of Table III presents our results concerning style consistency. Consistent with the view that namesake funds have less effective fund governance, we find that the average R- squares are reliably lower, and the average tracking errors are reliably higher for namesake funds, when compared to other equity funds. For instance, the average R-square is 0.82 for namesake funds, and 0.89 for other equity funds. The average tracking error is 2.23% for namesake funds, and 1.71% for other equity funds. The differences are statistically significant at the 1 percent level. These results for style consistency and annual expense ratios provide strong evidence that the insider-dominated boards of namesake funds provide less effective oversight than those 12

14 of other funds. These results further suggest that independent boards are necessary in controlling the agency conflict that exists within a mutual fund between its shareholders and the fund. Our findings are also consistent with the SEC s claims that independent mutual fund boards will improve governance of these funds and thereby benefit shareholders. 4. Tax Efficiency Tax planning is an essential element of the investment process and will influence the magnitude of the after-tax returns earned by investors. But in mutual funds, shareholders have no control over the timing or the size of taxable distributions. Average fund managers have little incentive to improve the fund s tax efficiency because fund managers are typically evaluated based on their funds pre-tax performance. 7 But since the managers of namesake funds have significant amounts of their own capital at risk, we expect that their portfolio decisions are more tax efficient than those of other fund managers. Hence, an analysis of this issue will allow us to assess the incentive effects provided by ownership in the fund. To examine the tax-efficiency of mutual funds, we define a fund s total pretax return in the same manner as Bergstresser and Poterba (2002). Specifically, we express a fund s total pretax return as the sum of dividends, short-term capital gains, long-term capital gains, and undistributed capital gains. We define the fund s after tax-return as: R = d + g + g u (2) p s l + R a ( τ ) d + ( 1 τ ) g + ( 1 τ ) g + ( τ )u = 1 1 (3) d d s cg l ucg 7 The SEC now requires funds disclose after-tax returns based on assumed tax rates. However, the fund industry and media still focuses on pre-tax returns. 13

15 For our analysis, we assume that the marginal tax rate on dividends (τ d ) is 31%, and the marginal tax rate on long term capital gains (τ cg ) is 20%. Consistent with SEC policy, we set the tax rate on undistributed capital gains as zero (τ ucg =0) in calculating pre-liquidation after-tax returns. Since we only have data on total capital gain distributions, we assume all capital gains are long-term. 8 We define tax efficiency as follows: R a R p 0 Tax Efficiency = 0 if R < 0 and d + g + g > 0 (4) 1 if R a a if R a < 0 and d + g That is, we define tax efficiency as the ratio between the after-tax return and the pre-tax return when the after-tax return is positive. When the after tax return is negative, the tax efficiency is 1 if the fund does not have any taxable distributions and is 0 otherwise. Based on the above definition, tax efficiency is always between 0 and 1, where 1 represents the highest tax efficiency and 0 represents the lowest tax efficiency. Table IV presents our results for an analysis of namesake tax efficiency. Panel A presents the equally weighted and TNA-weighted average tax efficiency. Using either measure, we find that namesake funds have significantly higher tax efficiency than other equity funds. For example, using the equally weighted measure, the tax efficiency is 67.98% for namesake funds, but only 61.64% for other equity funds. The difference of 6.34% is statistically significant at the 1 percent level. Results for value-weighted averages are similar. To examine whether the difference in tax efficiency between namesake funds and other equity funds can be explained by other fund characteristics, we present the results from a regression analysis in Panel B. After controlling for fund size, age, and fund family size, we find s s + g l l = 0 8 Our results are robust to alternative choices of marginal tax rates and alternative ways to separate short-term versus long term capital gains. 14

16 that namesake funds continue to remain more tax-efficient than other equity funds. The coefficient for the namesake fund dummy variable is slightly more than 5% in all regressions, only slightly smaller than the difference between average tax efficiencies of these two groups of funds. This result suggests that most of the difference in tax efficiency can not be explained by fund characteristics. We find that older funds are less tax-efficient, perhaps due to their large accrued capital gains. Controlled for age, larger funds are more tax-efficient. The size of the fund family is positively related to tax efficiency, but this relation is not statistically significant. 5. Career Concerns Our sample of namesake funds also allows us to examine the agency conflict that exists between the fund manager and the employing fund company. This conflict centers around how managerial career concerns influence the portfolio decisions of fund managers. Chevalier and Ellison (1999a) investigate how the behavior of mutual fund managers might be affected by their desire to remain employed. Chevalier and Ellison find that the probability of termination is a decreasing and convex function of performance. The probability of termination is sensitive to differences at negative excess return levels, but insensitive to differences at positive excess return levels. As a result, fund managers have an incentive to avoid unsystematic risk. Furthermore, the probability of termination is higher as a manager deviates more from the herd. This provides an implicit incentive for managers to herd. Consistent with the view that younger managers have more career concerns, Chevalier and Ellison find that younger managers take on less unsystematic risk than older managers, and are less likely to deviate from the herd than older managers. 15

17 But most namesake fund managers are founders and CEO/CIO of their fund companies. This suggests that namesake fund managers have few career concerns. Consistent with Chevalier and Ellison (1999a), we hypothesize that namesake fund managers assume greater unsystematic risk and herd less than other fund managers. 5.1 Measurement and methodology To estimate unsystematic risk and factor loadings, we estimate the following regression for each fund using data for the past 24 months, depending upon data availability: R t = α + m MKT + s SMB + h HML + u UMD + ε (5) t t t t t Following Chevalier and Ellison (1999a), we measure unsystematic risk by using the standard deviation of the error term ε t. Table V shows that namesake funds, on average, have a higher unsystematic risk. Consistent with Chevalier and Ellison (1999a), we measure the extent of herding by using the following six deviation measures: return deviation, unsystematic risk deviation, market beta deviation, SMB beta deviation, HML beta deviation, and UMD beta deviation. For example, we calculate market beta deviation for each fund as the absolute difference between fund market beta and the average market beta of all equity funds. 5.2 Findings for herding by namesake funds Table V presents our findings regarding herding by the portfolio managers of namesake funds. We find that namesake funds have significantly greater deviations from typical behavior than other equity funds, regardless of the deviation measure used. For example, the average return deviation is 2.33% for namesake funds, but only 1.84% for other equity funds. The average market beta deviation is 0.18 for namesake funds, but only 0.11 for other equity funds. These differences are statistically significant at the 10 percent level for return deviation, and at the 1 percent level for all five other deviation measures. 16

18 Overall, our results are consistent with the proposition that namesake funds have fewer career concerns, and hence take more unsystematic risk and herd less. These findings suggest that the agency conflict between the fund and fund manager is less acute for namesake funds relative to other funds. 6. Performance This section examines the extent to which the agency conflicts inherent in a mutual fund impact firm performance. In this sense, it answers the question: Do these agency conflicts affect shareholder value? We use both raw and risk-adjusted returns to determine how the return performance of namesake funds differs from those of other equity funds. 6.1 Raw return performance In this section, we compare the raw returns of namesake funds with that of other equity funds. We analyze both the net and gross raw return to namesake funds. Net return represents after-expense return, and is reported in the CRSP mutual fund database. We calculate gross return by adding back the expenses to net returns. We calculate excess returns by subtracting the value-weighted market returns from net or gross returns. Panel A of Table VI presents the timeseries average gross excess returns and net excess returns of TNA-weighted portfolios. The average gross return of namesake funds is basis points per month, and the average net excess return is 6.52 basis points. These results indicate that namesake funds outperform the market by 2.12% per year before expenses, and by 0.78% after expenses. Namesake funds also outperform other equity funds. The average gross excess return for other equity funds is 0.33 basis points per month and the average net excess return is basis points per month. This finding that equity mutual funds underperform the market by about the expenses 17

19 is consistent with existing studies. For example, Chen, Hong, Huang, and Kubik (2003) report a gross return of 1 basis point and net return of -8 basis points for equity funds during Using median returns reveals less impressive performance for namesake funds. The median gross return is 4.92 basis points per month, and the median net excess return is negative at basis points. According to our estimates of median returns, namesake funds outperform the market by about 0.59% per year before expenses, but underperform relative to the market by about 0.5% per year after expenses. The median returns for all other equity funds are similar to their mean values. Examining fund performance across bull/bear markets reveals some interesting results. We use low frequency bull/bear market cycles. Specifically, and are classified as bull markets, and and are classified as bear markets. In total, there are 22 years of bull markets and 18 years of bear markets. Namesake funds produce superior returns in bear markets, but not in bull markets. Using average net excess returns, we find that in bear markets namesake funds outperform the market by basis points per month, and outperform peer funds by basis points per month. In contrast, namesake funds underperform the market by 1.74 basis points per month in bull markets, although they still outperform peer equity funds by 4.28 basis points per month. Using medians, namesake funds still outperform both the market and peer equity funds in bear markets, but by much narrower margins at 5.95 basis points and basis points, respectively. In bull markets, the median namesake funds not only underperform the market but also underperform peer equity funds by 4.09 basis points. 6.2 Risk-adjusted return performance 18

20 In the preceding section, we examine raw returns as a measure of fund performance. It is possible that namesake funds outperform the market and peer funds because they hold more risky portfolios. To account for this possibility, we evaluate fund performance using alpha, the intercept term in the regression of fund returns on risk factors. We focus on the Carhart (1997) four-factor model, which augments the Fama-French three-factor model with a momentum factor. We also use the CAPM and the Fama-French three-factor model and they produce similar results. The regression model is as follows: R t = α + m MKT + s SMB + h HML + u UMD + ε (6) t t t t t where MKT is the market factor, SMB is the size factor, HML is the value factor, and UMD is the momentum factor. We employ two different regression approaches: the portfolio regression approach and the fund regression approach. In the portfolio regression approach, we compute monthly TNA-weighted return of the namesake fund portfolio and then use it as the dependent variable in the above regression. In the fund regression approach, we use fund returns in the above regression and then report the average coefficient estimates. Panel B of Table VI presents the results for the portfolio regression approach. We find that namesake funds have an alpha of 3.51 basis points, outperforming both the market and other equity funds, which have an alpha of basis points. So after controlling for risk, we find that namesake funds still outperform their benchmarks and peers, although by a smaller amount than as measured by average returns. Namesake funds have slightly different factor loadings than other equity funds. Namesake funds have lower market beta than other equity funds (0.934 versus 0.976), consistent with our previous finding that namesake funds hold more cash than other equity funds. 19

21 Additionally, namesake funds tend to hold more small stocks, as reflected in their higher SMB beta. Namesake funds load slightly more on value stocks than other equity funds. Namesake funds have a lower momentum beta, suggesting that they do not chase returns as much as other equity funds. Using the fund regression approach produces qualitatively similar results. Panel C of Table VI summarizes the average coefficient estimates. We find that namesake funds have an average alpha of basis points, outperforming both their benchmarks and other equity funds, which have an average alpha of points. Namesake funds have smaller exposures to the market factor, value factor and momentum factor, but have greater exposure to the size factor, relative to other equity funds. Overall, we find some evidence that namesake funds outperform the market and peer funds. In particular, namesake funds demonstrate superior performance in bear markets. It should be noted, however, that the above results are somewhat sensitive to whether we examine mean or median returns. These results also mitigate our earlier findings of higher fees and greater style drift for namesake funds. 7. Investor cash flows It is possible that namesake funds are able to use their managers names as a marketing device to attract new investor cash flow since many of these namesake fund managers are star managers. Consequently, we hypothesize that namesake funds attract greater investor cash flows because of the star power of their managers. To test the ability of namesake funds to attract new cash flow, we compute annual fund flow as follows: 20

22 Fund Flow t TNA = TNA t + 1 t 1 ( + R ) TNA t t MGTNA t (7) where MGTNA is the total net asset acquired from mergers during the period. The above calculation assumes that fund flows occur at the end of the year. Alternative assumptions about the timing of the fund flow do not affect my results. Consistent with previous researchers such as Barber, Odean and Zheng (2002) and Elton, Griber and Blake (2003), we control for past returns, fund size, expenses, and 12b-1 fees in our analysis of fund flows. To examine the flow-performance relation, we sort all funds based on their past 1-year returns and then form five quintiles. Quintile 1 contains the worst performing funds while quintile 5 contains the best performing funds. Table VII presents the regression results. Fund flow is positively related to past performance. This can be seen through all coefficients on quintiles 2 5 dummy variables being positive and significant. Furthermore, within each model, these coefficients are increasing from quintile 2 to quintile 5. Consistent with prior studies (e.g., Chevalier and Ellison (1997)), we find that the flow/performance relationship is convex. For instance, in Model 1, we find that an average fund in the third return quintile attracts 4.1% more cash flow than an average fund in quintile 2. However, an average fund in quintile 5 generates 23.7% more cash flow than an average fund in quintile 4. In other words, the fund attracts more flow by moving from quintile 4 to quintile 5 than from quintile 2 to quintile 3. Consistent with Barber, Odean and Zheng (2002), we find that fund cash flows are positively related to 12b-1 fees. This suggests that advertising is effective in attracting investor cash flows. This is also consistent with the idea that investors pay more attention to direct fees like loads than less obvious costs like the 12b-1 fee. We find that larger funds generate lower cash flows, in part because we measure fund flows as percentages of the fund s TNA. 21

23 Furthermore, funds in larger families attract more cash flows. We find only weak evidence that namesake funds attract more cash flows. In three of the four regressions, the coefficient of the namesake dummy is insignificant. In the fourth regression, the coefficient of the namesake dummy is positive and statistically significant at the 10 percent level. 8. Conclusion We argue in this study that the scandals currently plaguing the mutual fund industry are symptomatic of the two agency conflicts that exist within the mutual fund industry. Our sample of namesake funds represents a unique laboratory for studying these agency conflicts. They allow us to examine the widely publicized conflict that exists between the fund and its shareholders as well as the less prominent conflict that exists between the fund company and the fund managers. Further, because these funds are characterized by insider-dominated boards, they are very useful for testing the usefulness of the recent SEC proposal for independent boards and chairs. Our results suggest that the boards of these funds are ineffective and that agency conflict between shareholders and the fund is not mitigated with the oversight provided by an insiderdominated board. We find that average expense ratio of namesake funds is more than 20 basis points higher than those of other equity funds while they simultaneous demonstrate greater drift in their investment style. Theses results suggest that the arguments for greater board independence and oversight by the SEC have a basis in fact. But the characteristics of these funds that have resulted in weak boards and important agency conflict between the fund and its fund managers, have also produced less career concerns among portfolio managers. Because the fund managers have stature and standing in the 22

24 sponsoring firm, these managers are less inclined to herd less and to assume greater unsystematic risk. We conclude that namesake managers do not include career considerations in their portfolio investment decisions to the extent that managers of other funds are likely to do so. Further complicating a facile description of the extent of agency conflict within mutual funds are our results concerning performance. In many ways, performance is the measure by which we assess how significant are these agency conflicts. We find some evidence that namesake fund managers outperform their benchmarks and peers, especially in bear markets. We do caution however that these results are somewhat sensitive to the choice of a mean or median measure of average return. We also find that namesake funds are more tax-efficient, consistent with the idea that managerial ownership helps to align the interests of managers with that of shareholders. Namesake funds however do not attract greater investor cash flows after controlling for performance. 23

25 References Almazan, A., K. Brown, M. Carlson, and D. Chapman, 2003, Why Constrain Your Mutual Fund Manager? Forthcoming Journal of Financial Economics. Barber B., T. Odean, and L. Zheng, 2002, Out of Sight, Out of Mind: The Effects of Expenses on Mutual Fund Flows, Forthcoming Journal of Business. Bergstresser, D., and J. Poterba, 2002, Do After-Tax Returns Affect Mutual Fund Inflows? Journal of Financial Economics 63, Brown, K., and W., Harlow, 2002, Staying the Course: the Impact of Investment Style Consistency on Mutual Fund Performance, Unpublished Working Paper, University of Texas Austin. Brown, K., W. Harlow, and L. Starks, 1996, Of Tournaments and Temptations: An Analysis of Managerial Incentives in the Mutual Fund Industry, Journal of Finance 51, Carhart, M., 1997, On Persistence of Mutual Fund Performance, Journal of Finance 52, Chatzky, J., 2003, What s in a Name? Money, May, p.150. Chen, H., N. Jegadeesh, and R. Wermers, 2000, The Value of Active Mutual Fund Management: An Examination of the Stock holdings and Trades of Fund Managers, Journal of Financial and Quantitative Analysis 35, Chen, J., H. Hong, M. Huang, and J. Kubik, 2003, Does Fund Size Erode Performance? Liquidity, Organizational Diseconomies and Active Money Management, Unpublished Working Paper, Stanford University. Chevalier, J., and G. Ellison, 1997, Risk Taking by Mutual Funds as a Response to Incentives, Journal of Political Economy, Chevalier, J., and G. Ellison, 1999a, Career Concerns of Mutual Fund Managers, Quarterly Journal of Economics 114, Chevalier, J., and G. Ellison, 1999b, Are Some Mutual Fund Managers Better Than Others? Cross-Sectional Patterns in Behavior and Performance, Journal of Finance 54, Del Guerico, D., L. Dann, and M. Partch, 2003, Governance and Boards of Directors in Closedend Investment Companies, Journal of Financial Economics 69, Donaldson, William H., 2004, Review of Mutual Fund Industry Investigations, Federal Document Clearing House, Statement to Senate Committee on Banking, Housing and Urban Affairs, 8 April

26 Elton, E., M. Gruber, and C. Blake, 2003, Incentive fees and Mutual Funds, Journal of Finance 58, Glassman, J., 2003, When Fund and Manager Share a Name, Washington Post, January 19, 2003, H1 and H5. Grinblatt, M. and S. Titman, 1989, Mutual Fund Performance: An Analysis of Quarterly Portfolio Holdings, Journal of Business 62, Grinblatt, M. and S. Titman, 1993, Performance Measurement Without Benchmarks: An Examination of Mutual Fund Returns, Journal of Business 66, Gruber, M., 1996, Another Puzzle: the Growth in Actively Managed Mutual Funds, Journal of Finance 51, Haslem, J., 2003, Mutual Funds, Blackwell Publishing. Hendricks, D. J. Patel, and R. Zeckhauser, 1993, Hot Hands in Mutual Funds: Short-run Persistence of Relative Performance, , Journal of Finance 48, Jensen, M., 1968, The Performance of Mutual Funds in the Period , Journal of Finance 23, Kacperczyk, M., C. Sialm, and L. Zheng, 2003, On the Industry Concentration of Actively Managed Equity Mutual Funds, Unpublished working paper, University of Michigan. McConnel, J. and H. Servaes, 1990, Additional Evidence on Equity Ownership Structure, and Firm Performance, Journal of Financial Economics 27, Morck, R., A. Shleifer, and R. Vishny, 1988, Management Ownership and Market Valuation: An Empirical Analysis, Journal of Financial Economics 20, Nash, J., 2002, Did Success Spoil Ron Baron? Money, December 2002, Nanda, V., Z. Wang, and L. Zheng, 2003, Family Value and the Star Phenomenon, forthcoming Review of Financial Studies. Oster, C., 2003, How Mutual-Fund Timing Police Walk Beat, Wall Street Journal, September 26, Sanders, C., 1998, Taking Issue, Morningstar Investor, March, 22, Sirri, E. and P. Tufano, 1998, Costly Search and Mutual Fund Flows, Journal of Finance 53,

27 Tufano, P. and M. Sevick, 1997, Board Structure and Fee-Setting in the U.S. Mutual Fund Industry, Journal of Financial Economics 46, Wermers, R., 2000, Mutual Fund Performance: An Empirical Decomposition into Stock- Picking Talent, Style, Transactions Costs, and Expenses, Journal of Finance 55, Zheng, L., 1999, Is Money Smart? A Study of Mutual Fund Investors Fund Selection Ability, Journal of Finance 54,

28 Figure 1: Agency Conflicts in Namesake Mutual Funds Fund Company Interested in Maximizing fees Fund Shareholders Interested in Maximizing Risk-adjusted Returns Fund Company Interested in Maximizing Fees Fund Managers Have Career Concerns Agency Conflict #1 Agency Conflict #2 Between Fund Company and Fund Shareholders Between Fund Company and Fund Managers Board of Directors - Monitor Funds to Protect Interests of Fund Shareholders Managerial ownership In the Fund -Stock Ownership and Reputation Capital at Risk Help Align Interests Managerial ownership In the Fund Company - Exacerbate the Conflict Between Fund Company And Fund Shareholders Namesake Managers are Their Own Bosses and Have Minimal Career Concerns Boards of Namesake Fund lack independence Manager Sits on Board often as the Chairman Namesake Managers Invest Significantly in Their Funds Namesake Managers Have Significant Reputation Capital at Risk Namesake Managers are founders and Majority Owners of Fund Company Implications Less Herding Assume Greater Unsystematic Risk Implications Higher fees Less style consistency Higher tax efficiency Indeterminate Impact on Fund Performance

29 Table I Summary Statistics of Namesake Funds The sample period is The primary data source for mutual funds is the CRSP Survivor Bias- Free Mutual Fund Database. We include only actively-managed diversified domestic equity funds in my sample. In particular, We exclude international funds, sector funds, balanced funds, and index funds. We use the ICDI fund objective code and assign each fund one of three investments objectives: Aggressive growth (AG), Long-term growth (LG), and Growth and Income (GI). We identify namesake funds by comparing fund names and manager names. Item Namesake Funds Other Equity Funds Total Net Asset - $ million Fund Age years Turnover - % Expenses - % b-1 fee - % Total Load - % Percent of Funds with Load Manager Tenure years Market Share of the Fund Family - % Percent Cash Holding Percent Stock Holding Percent of Aggressive Growth Fund Percent of Long-term Growth Fund Percent of Growth and Income Fund

30 Table II Characteristics of Namesake Fund Managers The sample period is The primary data source for mutual funds is the CRSP Survivor Bias- Free Mutual Fund Database. We include only actively-managed diversified domestic equity funds in my sample. In particular, We exclude international funds, sector funds, balanced funds, and index funds. We identify namesake funds by comparing fund names and manager names. We hand collect the background information of namesake fund managers from fund websites, and SEC forms 485B and N-SAR. Panel A: Other Positions Held by Namesake Fund Managers Item Percent Percent of managers also board member of the fund Percent of managers also chairman of the board Percent of managers also founder of the investment advisory firm Percent of managers also significant owner of the investment advisory firm Percent of managers also CEO or CIO of the investment advisory firm Panel B: Prior Experience of Namesake Fund Managers Title Portfolio Analyst Director of Managing VP CIO Professor Manager Research Director Percent Panel C: Manager Ownership in Fund and Fund Family Percent of Managers Amount None >$0 >$10,000 >$50,000 >$100,000 Ownership in Fund Ownership in Fund Family