Do active portfolio strategies outperform passive portfolio strategies?

Do active portfolio strategies outperform passive portfolio strategies? Bachelor Thesis Finance Name Stella van Leeuwen ANR S765981 Date May 27, 2011 Topic Mutual Fund performance Supervisor Baran Duzce Word count 8259

Management Summary The question whether actively managed mutual funds outperform passively managed mutual funds has been subject to intense debate. The opinions are still divergent although a lot of research on this topic has been performed. This thesis answers the question through a descriptive and empirical research. By giving first a clear description on mutual funds in general and about the measures of performance, the literature is being reviewed about the question whether active portfolio strategies outperform passive portfolio strategies. Different research outcomes are discussed to come to the conclusion that there are actively portfolio strategies which can outperform the passive portfolio strategies. However, the active portfolio strategies also have higher expenses and fees. Most of the funds which have excess returns are not capable to cover these higher expenses with the excess return. The number of mutual funds which still have excess returns after expenses remains is even smaller. With the empirical it is then investigated that the active BNY Mellon Intermediate US Government Fund does not outperform the passive AIM S&P 500 Index Fund. Since only one active fund is being compared with one passive fund, the result that the active portfolio does not outperform the passive portfolio does not necessarily imply that in general active mutual funds do not outperform passive mutual funds. 2

Table of contents Page number Management Summary 2 Table of contents 3-5 Introduction 6-8 Chapter 1 Mutual Funds 9-23 1.1 Organization of Mutual Funds 9-10 1.2 Fee structure of Mutual Funds 10-14 1.2.1 Front-end load charges 11 1.2.2 Contingent deferred sale charges 11-12 1.2.3 Redemption fees 12 1.2.4 Exchange fees 12 1.2.5 12b-1 fees 12-13 1.2.6 Share classes 13-14 1.2.7 The fee-setting process 14 1.3 The Mutual Fund industry 14-16 3

1.4 Regulations of the Mutual Fund industry 16-17 1.5 Distribution Channels 18-19 1.6 Advantages of Mutual Funds 19 1.7 Disadvantages of Mutual Funds 20 1.8 Actively managed Mutual Funds vs. passively managed Mutual Funds 20-23 1.8.1 Active management of Mutual Funds 21-22 1.8.2 Passive management of Mutual Funds 22-23 Chapter 2 Measuring Mutual Fund performance 24-28 2.1 Jensen s alpha 24-25 2.2 Carhart s alpha 26 2.3 Conditional alpha approach 26-27 2.4 Sharpe Ratio 27-28 Chapter 3 Do active portfolio strategies outperform passive portfolio strategies? 29-35 3.1 Active portfolio strategies do not outperform passive portfolio strategies. 29-32 3.2 Active portfolio strategies do outperform passive portfolio strategies. 32-35 4

Chapter 4 Data and Methodology 36-39 4.1 Data description 36 4.2 Research Methodology 36-38 4.3 Limitations 38-39 Chapter 5 Conclusion 40-41 References 42-48 5

Introduction Mutual funds started growing explosively in the 1990 s and nowadays still represent an important part of the U.S. financial system. 1 (Investment Company Institute, 2011.) According to Gruber (1996), mutual funds are the second largest financial intermediary in the United States. The largest intermediaries are commercial banks. (Gruber, 1996) In 1990 there were 3079 outstanding mutual funds, this amount of funds more than doubled to 7581 in 2010. Also the total net assets invested in mutual funds have grown rapidly during that period. In 1990 the net assets of mutual funds amounted 1,065.19 billions of dollars. In 2010 the total net assets amounted 11,820.68 billions of dollars. 2 (Investment Company Institute, 2011) These numbers show that the number of mutual funds, as well as the value invested in mutual funds, have grown rapidly during these twenty years. As a consequence of this growth, the interest in the mutual fund market has grown a lot from academic perspective. There are two different approaches when it comes to the management of mutual funds. Active management always tries to beat the market by focusing on stock-picking, which means that management is continuously actively looking for opportunities in the market. The aim is to buy stocks that achieve a higher return than the average return in the market. 3 The costs of management are high and investors would only benefit from active management if the excess returns on actively managed portfolios are larger than the extra costs for management according to Shukla (2004). On the other hand, passive management believes in the efficient market hypothesis. (Basu, 1977) This means that security prices in an efficient capital market always fully reflect all available 1 http://www.ici.org/pdf/2011_factbook.pdf 2 http://www.ici.org/pdf/2011_factbook.pdf 3 http://www.investopedia.com/articles/mutualfund/07/active-share.asp 6

information. It would therefore be impossible to beat the market and so there cannot be strategies of management which can beat the market. According to Browne (2000) passively managed mutual funds just try to track a benchmark which is mostly the S&P 500 index. Actively managed funds and passively managed funds have been researched a lot recently. Despite all this research, the opinions about the question Do active portfolio strategies outperform passive portfolio strategies? are still divergent. Gruber (1996) states that the average actively managed fund has negative performance compared to a set of indices. Also Malkiel (1995) and Shukla (2004) find that mutual funds underperform the market. Grinblatt and Titman (1989) concluded in their paper that investors cannot always benefit from superior abilities of management when purchasing the funds. Also Jensen (1967) argues that managers of mutual funds are not capable enough to predict security prices to outperform passively managed funds. However, Wermers (2000) argues that stock picking funds perform well enough to cover the costs associated with it which implies that active managed funds give better returns than passive managed funds. Also Petajisto (2010) finds evidence for the fact that the most active stock pickers outperform their benchmark indices, even after fees and transaction costs. This only holds for the most active stock pickers. In Ippolito s study of 1989, he documents a significant positive performance of mutual funds in the US market compared to the S&P 500 index. Grinblatt and Titman conclude in their research of 1988 that growth funds and aggressive growth funds have significantly positive gross returns on average compared with passive funds. To bring some clarity in the different definitions and measures, this thesis will be structured as follows: The first chapter will be a literature review on mutual funds. The definition, the 7

structure of the industry, the regulations and the fee structure will be explained. Furthermore the difference between passively managed funds and actively managed funds will be elaborated on, some facts about the industry will be given and the distribution channels are elaborated. Chapter two is about different performance measures of mutual funds. This is necessary to compare returns on actively managed funds with passively managed funds. In chapter three an overview is given of important papers which contain conclusions about the performance of active strategies in comparison with performance of passive strategies. The data methodology and the empirical research will be elaborated on in chapter four and this thesis will end up with the conclusion in chapter five. 8

Chapter 1 Mutual Funds 1.1 Organization of Mutual Funds A mutual fund can be defined as a facilitator between investors and funds. An investment company pools money from investors and uses this money to invest in different securities like stocks, bonds, money market instruments and other assets. Mutual funds are open-end investment companies which means that they invest in other companies and are able to adjust their investment criteria and fund size all of the time. The prices of open-end investment shares are based on the assets of the fund. 4 According to Mishkin and Eakins (2006) the policy of the fund is set by the Board of Directors and this Board of Directors oversees all activities of the fund. Managing the funds does not belong to the Board s tasks; the Board should only appoint a manager. The operating tasks of mutual funds are the responsibility of professional managers which make decisions about what securities to buy and when to sell them again. 5 A portfolio manager has different possibilities to spend the attracted money from investors on. By choosing where to invest the money in, the portfolio manager should take into account the objective of the fund. The objectives a mutual fund can have are aggressive growth, growth, growth and income, and income. The objective of the fund shows the investors what the main objective is and in that way investors can conclude in what type of investments the fund will mainly invest. This means that when the objective of the fund is aggressive growth, the investment company invests money especially in high growth stocks. (Budiono, 2009) A growth fund has the main objective to have growth on the long term. (Reints and Vandenberg, 1973) The objective of a growth-income fund 4 http://www.investopedia.com/terms/o/oeic.asp 5 http://www.investopedia.com/terms/f/fundmanager.asp 9

is to have long-term growth of capital while at the same time considering income and/or relative stability. (Reints and Vandenberg, 2009) A fund with the objective income takes into account the highest possible amount of income primarily. 1.2 Fee structure of Mutual Funds Managers are being paid for their work by a fee, which is a percentage of the fund s average assets under management. 6 High fees can indicate that a portfolio manager delivers good performance and hence the investors can have an incentive to invest in these funds because they believe the fund will perform better than other funds. (Elton, Gruber and Blake, 2001) According to Elton, Gruber and Blake (2001) funds which reward managers with an incentive fee, tend to attract managers with better stock picking abilities and managers which are more motivated to put a lot of effort into the management than funds with no incentive fees. Another advantage of using incentive fees is that the expense ratio is lower than with funds which do not use incentive fees. (Elton, Gruber and Blake, 2001) Important to know for investors is the fact that managers which are rewarded by incentive fees take higher risks than managers which are not rewarded by incentive fees according to Elton, Gruber and Blake (2001). Fees can differ between funds a lot, the fees range from 0.07 % of assets annually to 9.65%. (Luo, 2000) Therefore investors should take fees into account when choosing for a certain fund. It is also important for investors to judge whether the charged fees are valuable. This means that fees should lower expenses of the fund or should improve performance. (Dellva and Olson, 1998) 6 http://www.investopedia.com/terms/f/fundmanager.asp 10

According to Mishkin and Eakins (2006) different fees are charged by mutual funds. Fees can be divided into two categories: ongoing yearly fees and transaction fees (loads). 7 According to Chordia (1996), loads compensate brokers or other salesperson for selling the fund. The ongoing yearly fees cover costs for management, administration costs and the costs for marketing. The different fees a fund can charge are front-end load charges, deferred sale charges, redemption fees, exchange fees and 12b-1 fees. (Ferris and Chance, 1964) Some funds include all these charges, some funds charge just a few of them and there are also funds which do not charge any fees according to Dellva (1998) 1.2.1 Front-end load charges A front-end load is charged at the time of the initial purchase for the mutual fund. This amount is subtracted from the investment amount and therefore the size of the investment decreases. 8 The load will mostly decrease when the size of the investment is larger. (O neal, 2003) According to Dellva (1998) a front-end load is meant to let investors maintain their investment in the fund. Most companies who charge front-end loads classify them as class A shares. 1.2.2 Contingent deferred sale charges A contingent deferred sales charge is another term for deferred loads. It applies for the first years of ownership, it decreases each year and then disappears. (Mishkin and Eakins, 2006) According to Dellva (1998) the deferred sales charge protects underwriters from early redemption. This is 7 http://www.investopedia.com/university/mutualfunds/mutualfunds2.asp 8 http://www.investopedia.com/terms/f/front-endload.asp 11

done by recouping the high initial outlays in commissions to retail dealers over time through other fees and expenses. When the investor redeems the share, the percentage of the lesser of current net asset value is taken as load, or the original cost of the shares which are being redeemed. (O neal, 2003) Most of the time class B shares charge contingent deferred sales. 1.2.3 Redemption fees Furthermore there is a redemption fee; this is a percentage of the redemption price and it is imposed when shares are sold according to Ferris and Chance (1998). According to Dellva (1998) the redemption fee also protects underwriters from early redemption. 1.2.4 Exchange fees An exchange fee is being charged when money is transferred between funds within the same family. A family of funds contains a number of funds with all different objectives of investment. The advantage of such a family is that assets can be moved among funds in the family without costs or with low costs. (Fabozzi, Modigliani & Ferri, 1994) 1.2.5 12b-1 fees 12b-1 fees are charged to cover distribution costs as for example marketing costs and commissions (Dellva, 1998). The idea is that the efforts in marketing and the incentives for brokers to sell funds, increases the size of the fund and therefore economies of scale increase. 12

The costs for mutual funds can in that way decrease and thus investors have lower costs for their investment in mutual funds. (Dellva, 1998) The amount of the 12b-1 fee is subtracted from the net assets of the fund and the fee goes to the investment company. Class A shares as well as class B and class C shares charge their investors 12b-1 fees. 1.2.6 Share classes According to O neal (2003) the most important share classes are A, B and C classes. The classification holds for mutual funds which are sold by brokers. Share classes are claims on the portfolio of investments, and the different classes have different expense structures. Other differences between the classes can be found in the level of one-time load charges and also in that timing. Furthermore the 12b-1 fees differ between the classes. Investors are forced to decide which share class is the best one to choose. (O neal, 2003) Characteristics of A class shares are that they rely primarily on front-end loads and besides they have low 12b-1 fees. (O neal, 2003) In 2001 the average maximum front-end load was 5,2% on equity funds and 4,2% on bond funds. 9 B class shares combine contingent deferred loads with 12b-1 fees, these 12b-1 fees are higher. 10 And concluding, C class shares have small back-end loads and high 12b-1 fees. Over time class B shares become class A shares. For investors this is a good thing since the 12b- 1 fee of class A shares is lower than the 12b-1 fee of class B shares. The Securities and Exchange Commission contains two rules which have an impact on the share classes discussed above. The 9 http://www.ici.org/pdf/per09-03.pdf 10 http://www.ici.org/pdf/per09-03.pdf 13

12b-1 rule permits mutual funds to charge fees for distribution costs. Before the 12b-1 rule was established, these costs were covered by charging loads to investors. According to O neal (2003), the 18f-3 rule allows mutual funds to offer different mutual fund share classes. 1.2.7 The fee-setting process According to the Investment Company Act, independent directors have a role in the process of fee-setting. All advisory and the distribution contracts of management must be approved by the majority of the directors, and thus also their associated fees. (Tufano and Sevick, 1997) Managers of mutual funds come up with a proposal about the fee being charged and the independent directors should give their approval. There are some factors that directors take into account when determining fees. They should take into account the services the fund provides, but also the fees which comparable mutual funds charge, the performance of the fund and operating expenses. (Smith, 1994) 1.3 The Mutual Fund industry The largest mutual fund market can be found in the United States, from figure 1 it can be seen that the United States account for 48% of the total mutual funds in the world. Europe has the second largest mutual fund market. When we look at the type of mutual funds, it can be seen from figure 1 that domestic equity funds hold for one third of the mutual funds in the U.S. With that percentage it is the largest type of mutual fund in the U.S. industry. 14

Figure 1: Percentage of total net assets, year-end 2010. Source: Investment Company Institute, 2010. Mutual Fund Fact Book, 50 th ed. (Washington, DC:ICI) 11 According to Wermers (2000), mutual funds with the growth objective can be placed in the most popular sector of the mutual fund industry. This can be explained by the fact that there were relatively high returns of growth stocks in the last twenty years. Another trend in the mutual fund market is that the mutual fund industry has become an industry which invests fully in common stock, while beforehand the investments were especially to bonds and cash. Another trend which can be seen from the industry is the fact that the trading activities have been doubled from 1975 to 1994. This is partly caused by the fact that the transaction costs have decreased a lot. 12 For investors there are a lot of possibilities to choose for a mutual fund. This wide range of choices has come up with increasing competition among funds. Funds are stimulated to perform well since in general mutual funds which perform in a good way, have higher sales than mutual funds performing worse. (Shapiro, 1964) An important characteristic of the mutual fund industry 11 http://www.ici.org/pdf/2011_factbook.pdf 12 http://mutualfunds.about.com/od/mutualfundbasics/a/mutualfunds_regs.htm 15

is the competition in promotion of sales through winning and stimulating efforts of sales dealers. (Hutner, 1964) 1.4 Regulations of the Mutual Fund industry The mutual fund industry is an extensively regulated industry. Investors are protected against fraud by these regulations. Mutual funds have to follow some operating standards, obey antifraud rules and disclose a complete overview of information to investors. (Budiono, 2009) Investors must have the opportunity to see information about the experience of directors but also about what knowledge these directors have and about the funds held by the director. (Mishkin and Eakins, 2006) Mutual funds are registered with the Securities Exchange Commission (SEC) and the commision monitors the funds compliances with the use of different acts. The most important ones are the 1940 Investment Company Act, 13 the Securities Act of 1933 (Carol, 1989) and the Securities Act of 1934. 14 The first regulation for mutual funds was the Securities Act of 1933. The main goal of this act is to ascertain that for investors, information is available about securities which are for public sale. The act also contains regulations to prevent fraud and dishonesty from investment companies. 15 In the second regulation for the mutual fund industry, the act of 1934, the authority and power of SEC is being regulated. Security trading should be investigated by the SEC and the SEC provides accounting and financial standards on brokers and dealers. (Anderson and Ahmed, 13 http://mutualfunds.about.com/od/mutualfundbasics/a/mutualfunds_regs.htm 14 http://mutualfunds.about.com/od/mutualfundbasics/a/mutualfunds_regs.htm 15 http://mutualfunds.about.com/gi/o.htm?zi=1/xj&zti=1&sdn=mutualfunds&cdn=money&tm=1923&gps=237_55 8_1276_837&f=11&su=p649.6.336.ip_&tt=2&bt=0&bts=1&st=32&zu=http%3A//www.sec.gov/about/laws.shtml% 23secexact1934 16

2005) Also the 1940 Investment Company Act contains several regulations on the mutual fund industry. According to the 1940 Investment Company Act, investment companies should be a domestic corporation or entity taxed as a corporation. This rule prevents personal holding companies to qualify for the favorable tax treatment of income under the act. (Anderson and Ahmed, 2005) According to Anderson and Ahmed (2005) to make sure that the majority of the Board of Directors is financial independent of the fund, the 1940 Act holds that at least 40% of the Boards members should be non-officers or advisors to the fund. Furthermore the 1940 Act also provides some policy guidelines. To prevent that non-investing activities contribute to fund s revenues, at least 90% of the investment companies gross income must be passive income like dividends, interest and gains from the sale of securities. Another requirement of the Act is that at the end of each quarter the fund must have at least 50% of its assets in cash, government securities, securities of other regulated investment companies or other financial assets. Also every single security should amount at maximum 5% of its total assets and an investment in any company should not represent more than 10% of the securities of that company. Besides, a mutual fund should limit the amount of investments in securities of two or more controlled companies in the same business. According to the Investment Company Act, investment companies have to provide information twice a year to their investors about their performance and about their current portfolio holdings. (Frank, Poterba, Shackelford & Shoven, 2004) This allows investors to monitor management companies. 17

1.5 Distribution channels In the mutual fund industry various channels exist through which mutual funds are sold from investment companies to consumers. Basically there is a distinction between direct-sold funds and broker-sold funds. (Chalmers and Tufano, 2009) The direct-sold funds are directly sold from the investment company to the consumers and these are no-loads funds. (Hortaçsu & Syverson, 2003) However, many investors choose and sell their mutual funds with the use of intermediated channels, the broker-sold funds which are load-funds. There are different kinds of brokers. The main distribution channels through which broker-sold mutual funds are sold are: the advice channel, the retirement plan channel, the supermarket channel and the institutional channel. 16 The advice channel provides especially investment advice and ongoing assistance by financial advisors. The retirement plan channel consists of contribution plans in which employers provide mutual funds for purchase through payroll deductions. 17 In the supermarket channel brokers offer mutual funds which come from a large number of fund sponsors. Finally, the institutional channel is being used by institutional investors like businesses, foundations and financial institutions. Transactions are done directly with the mutual funds or with the use of third parties. 18 According to Christoffersen (2006) the three main channels are captive brokers, unaffiliated brokers and no-load investment. Captive brokers represent just one fund family in contradiction to unaffiliated brokers which do not represent a particular fund family. No-load investments do not come up with brokers. (Christoffersen, 2006) Investors pay brokers or financial advisors for the selection of the fund and for their advice. The broker-sold funds ask a 16 http://www.ici.org/pdf/per09-03.pdf 17 http://www.ici.org/pdf/per09-03.pdf 18 http://www.ici.org/pdf/per09-03.pdf 18

price for their service by charging fees; Chalmers and Tufano (2009) estimated the amount of money mutual fund investors spent on distribution channel fees at $15.2 billion in 2002. Different fees can be charged for the distribution channels. Front-end loads, back-end charges and 12b-1 fees are examples of fees charged for the distribution of funds. One of the advantages investors have who make use of brokers is the fact that brokers help to personalize the portfolio according to the risk tolerances a customer has. Furthermore, according to Chalmers and Tufano (2009) the broker can help the investor to save more money and help the investor to use its scarce time more efficiently. Brokers have to take into account certain regulations; for example brokers have to recommend suitable investments for their clients and they also have to put the client s interests ahead of their own interests. (Chalmers and Tufano, 2009) 1.6 Advantages of Mutual Funds Mutual funds provide several advantages for investors. One advantage is the fact that investors can hold a diversified portfolio without investing a large amount of money. Because a mutual fund pools money from many investors, the sum of their investments enables investment companies to hold a diversified portfolio. Liquidity of mutual funds is another advantage for investors. 19 For investors who do not have much time to spend or do not have knowledge about the industry another advantage arises with mutual funds. It is the fact that investors do not need to monitor the industry. The mutual fund is being managed by a professional manager which makes investment decisions and who monitors the portfolio. This is the largest difference between actively and passively managed funds according to Gruber (1996). 19 http://www.investopedia.com/university/mutualfunds/mutualfunds.asp 19

1.7 Disadvantages of Mutual Funds However there are also some drawbacks of mutual funds. An important disadvantage is the fact that mutual funds are not guaranteed by the U.S. Government, unlike bank deposits. 20 So investors can lose all their invested money if the fund performs badly. Mutual funds are also criticized for their high fees. (Gruber, 1996 and Wermers, 2000) The costs of professional management cannot be included in the price since funds are bought and sold at Net Asset Value. To let investors pay for professional management, investor companies can charge fees. It happens often that companies charge high fees but at the same time do not perform better than companies who do not charge high fees. Besides, according to Wermers (2000) mutual funds have higher trading costs than index funds. 1.8 Actively managed Mutual Funds vs. passively managed Mutual Funds In the mutual fund market there is a difference between actively managed funds and passively managed funds. Active management always tries to beat the market by focusing on stock-picking which means that management is continuously actively looking for opportunities in the market. The aim is to buy stocks that achieve a higher return than the average return in the market. 21 On the other hand, passive management believes in the efficient market hypothesis, which means that security prices in an efficient capital market always fully reflect all available information. (Basu, 1977) Therefore it would be unable to beat the market. Passive fund managers build up a 20 http://www.investopedia.com/articles/basics/03/041103.asp 21 http://www.investopedia.com/university/quality-mutual-fund/chp6-fund-mgmt/passive-active.asp 20

portfolio that reflects a certain market index or sector. The portfolio tries to contain the same risk and rate of return compared with the market index or sector. The portfolio will not change frequently as actively managed funds do. 1.8.1 Active Management of Mutual Funds According to Grinblatt and Titman (1989), to earn abnormal returns, a manager should have superior stock selection abilities. Grinblatt and Titman define abnormal returns as the intercepts from excess return regressions calculated with a benchmark that is mean-variance efficient from the perspective of an uninformed observer. With such a benchmark active managers can earn abnormal positive performance by trading securities in response to superior information. (Trinblatt and Titman, 1989) A lot of research has been done on actively managed mutual funds since it is very remarkable that actively managed funds have been growing very fast in the past twenty years, while at the same time their performances have been worse than those of index funds. (Gruber, 1996) One of the advantages of actively managed mutual funds is their way of pricing. These funds are sold at net asset value. The ability of management to provide superior performance is in this case not included in the price. The ability of management to provide superior performance is not included in the price of open-end mutual funds but is included in the price of closed-end mutual funds. (Gruber, 1996) The advantage of open-end funds where the ability is not included in the price is the fact that performance is predictable. And according to Gruber (1996), if the performance of mutual funds is predictable by some investors, the return on new cash flows should be better than the average return for all investors in these funds. 21

Actively managed mutual funds are often criticized for their high fees. Due to high fees, the expense ratio for a typical actively managed fund in the US is 140 basis points. Which is a lot compared with low cost funds which have an expense ratio between 10 and 20 basis points. (Malkiel, 2003) However according to Elton, Gruber and Blake (2001) management which is rewarded by incentive fees does perform better than management which is not rewarded by incentive fees. 1.8.2 Passive Management of mutual funds Besides active management of mutual funds, there is also the passive approach when it comes to the management of mutual funds. According to Malkiel (2003) the security markets seem to be efficient in general. Therefore it would be unable to beat the market. However there are some exceptions on this efficient market theory. There are some patterns which are predictable and sometimes there is not enough reaction on new information to affirm the efficient market hypothesis. (Malkiel, 2003) However these exceptions are most of the time small relatively to the amount of transaction costs which are needed to exploit the exceptional situations according to Malkiel (2003). Passive fund managers build up a portfolio that reflects a certain market index or sector. The portfolio tries to contain the same risk and rate of return compared with the market index or sector. Usually the S&P 500 index is used as a leading guide. (Hortaçsu and Syverson, 2003) The portfolio will not change as frequently as actively managed funds do. According to Hortaçsu and Syverson (2003), to accomplish the reflection of the S&P 500 index, mostly the same proportion of equity should be hold by the mutual fund as the index fund. 22

According to Malkiel (2003) passive management is a successful strategy in an efficient market. The strategy is even successful when a market is not fully efficient. He also concludes his paper with the fact that the chance for investors to earn higher returns is larger for investors investing in passive managed funds than investors which invest in actively managed funds. 23

Chapter 2 Measuring Mutual Fund performance As the interest in mutual funds has grown a lot from academic perspective, the interest in how to measure mutual fund performance has also grown. According to Otten and Bams (2004) to judge performance of mutual funds, many models are available. Most authors use only one or at most two different models to judge performance of mutual funds. According to Otten and Bams (2004) when it comes to the measurement of risk-adjusted mutual fund performance there is a difference between unconditional models, which were the first models, and the conditional models. Conditional measures do not take into account the fact that managers can attain changing market information about expected returns and risk and adjust their portfolio on that basis. Jensen s alpha and Carhart s alpha are examples of unconditioned performance measures. On the other hand conditional performance measures take that factor into account by assuming that the factor loading on the market risk factor at time t is linearly related to a vector of instruments for the economic information set. (Cuthbertson and Nitzsche, 2004) An example is the extended model of Carhart. The following section contains an overview of the generally used models to judge performance. 2.1 Jensen s alpha The Capital Asset Pricing model is the simplest model to judge risk-adjusted performance of mutual funds. The model gives Jensen s alpha, which can be interpreted as the measure for superior performance compared with the market index. (Otten and Bams, 2004) Jensen s alpha 24

represents superior performance on the basis of the Capital Asset Pricing Model which can be stated as following: In the model represents the return on fund in period. represents the return on a market proxy portfolio and is the risk-free interest rate. According to Cuthbertson and Nitzsche (2004) a positive and significant alpha indicates superior performance of the portfolio. Because the CAPM is a single factor model, the performance can only be measured using one market index. According to Elton et al. (1993) to control for certain variables it can be valuable to add extra variables in the model. Besides, Fama and French came up with a three-factor model because that would be even more complete. The three-factor model is defined as following: In this model is the return on asset, represents the return on the risk-free asset, is the return on the market portfolio, is the return on the mimicking portfolio for the size factor and represents the return on the mimicking portfolio for the book-to-market factor. The extension of models continues by researchers and in that way models are being improved and different performance models arise. 25

2.2 Carhart s alpha An extension of the Capital Asset Pricing Model with Jensen s alpha is the Carhart measure according to Cuthbertson and Nitzsche. (2004) The extension is a four-factor model, and the extra variables include risk factors like size, book-to-market value and momentum. The model can be stated as following: The variables, and represent amounts of risk for size, book-to-market value and momentum comparable with portfolios. The risk factor size is being taken into account by the factor. Fama and French have found a negative relationship between stock returns and size. The conclusion they found is that higher returns are found by smaller firms. (Cuthbertson and Nitzsche, 2004) Also the fact that small firms face larger informational asymmetry then large firms is being included in the variable. The book-to-market factor in the model is represented by which means high minus low. According to Cutherbertson and Nitzsche (2004), firms with high book-to-market values have higher average returns than low book-tomarket value stocks. Besides, the factor is the momentum-risk factor. It represents the difference between the returns of portfolio s which performed high in the past and returns of poorly performed stocks in the past. 2.3 Conditional alpha approach Carhart s alpha and Jensen s alpha are unconditioned performance measures which do not take into account the fact that managers use information about the changing market to compose their 26

portfolio. Therefore Ferson and Schadt (1996) have come up with an extension of the CAPM, which includes multiple factors. The CAPM can then be rewritten as following: In this model Cuthbertson and Nitzsche define as follows: allows for the factor loading on the market risk factor at time t to be linearly related to a vector of instruments for the economic information set. Carhart came up with another conditional alpha measure: the Carhart four-factor model. In this model it is assumed that superior returns vary over time. This is not assumed in the model mentioned before. The model can be stated as following: 2.4 Sharpe Ratio Also the Sharpe ratio is widely used to compare actively managed funds with passively managed funds. This ratio, formerly called reward-to-variability ratio incorporates both risk and average return of funds according to Sharpe (1966). A high Sharpe ratio indicates good performance, so the higher the ratio, the better the performance is. The numerator R represents the difference between the fund s average annual return and the pure interest rate. It thus means the return provided to the investor for bearing the risk. The V, which is the denominator in the reward-to- 27

variability ratio, represents the standard deviation of the annual rate of return. The ratio thus measures the average return per unit of risk. The ratio is defined as following: In this formula represents the expected return on the portfolio, represents the risk-free interest rate and represents the standard deviation of the portfolio. The Sharpe ratio is very useful because it shows whether the returns of funds are caused by good management or by excess risk. 22 22 http://www.investopedia.com/terms/s/sharperatio.asp 28

Chapter 3 Do active portfolio strategies outperform passive portfolio strategies? 3.1 Active portfolio strategies do not outperform passive portfolio strategies Gruber (1996) finds it is interesting that in particular actively managed mutual funds have grown so fast recently, while their performance has been inferior to that of index funds in that same period. The data used for his research consists of al common stock funds listed in Wiesenberger s Mutual Funds Panorama at the end of 1984, with certain exceptions like foreign bonds, specialized stock funds and balanced funds. The period which is studied is 1985-1994. In his research he concludes that mutual funds offer a negative risk adjusted return. Investors can better invest in index funds and besides investors which invest in closed end funds will not pay as much as one dollar for every dollar under management. The reason why actively managed portfolios have grown so fast, although their performance has been inferior than index funds can be found in the fact that future performance is partly predictable by past performance. Investors, who know about this, can benefit from this predictability since their earnings on new cash flows are positive and above returns earned by average actively and passively managed funds. The reason why there is still money invested in funds that are predicted to perform bad in the future is the fact that not all investors can predict it. Malkiel (1995) set up a research to examine performance of mutual funds and the survivorship bias. Malkiel s research analyzes mutual-fund returns in the period 1971-1991 and all equity mutual funds are included in the data. He finds that mutual funds have underperformed the market in the period, not only after management expenses have been deducted but even gross 29

of all reported expenses except for load fees. Persistence in performance is besides other researchers, also confirmed by Malkiel. However it should be taken into account that that conclusion is influenced by survivorship bias and the relationship between past performance and future performance may not be robust. The general conclusion of Malkiel s research is that passive management is more favorable for investors than active management since active management on average does not perform better than passive management and active management implies more tax burdens for investors. Also Grinblatt and Titman (1989) try to answer the question whether professional portfolio managers have the ability to earn abnormal returns. The data which were used for this research consist of quarterly equity holdings of mutual funds in the period 1975-1984. This research also contains data of mutual funds which did not survive over time, which some other researches do not. The research indicates that the risk-adjusted returns of some funds were significantly positive. The research concludes that on average there is no positive abnormal performance of funds. However there are some funds which do earn significantly abnormal higher returns: growth funds, aggressive growth funds and funds with smallest net assets. These high returns are partly caused by active management. What is remarkably is the fact that these funds also have the highest expenses, so their actual returns, net of all expenses do not demonstrate abnormal performance. So for investors it is not possible to take advantage of the superior stock picking ability of active management. Another important research about the difference in performance of actively managed funds and passively managed funds has been done by Jensen (1967). Earlier researchers faced some problems with identifying risk, and measured performance relatively to each other. Jensen however comes up with an absolute measure of performance. The data of this research consist of 30

the returns on 115 open end mutual funds in the period 1955-1964. The data come from Wiesenberger s Investment Company. The conclusion of Jensen s research is that on average the mutual fund managers are not able to predict prices of stocks good enough to outperform passive strategies. Even for some individual funds there is not enough evidence to support the fact that superior performance shows up. In this conclusion also management expenses are taken into account. The above average predicting abilities are not high enough to cover the extra costs of management. In a research of 2004, Shukla also compares the returns on passively managed funds and actively managed funds. The sample consists of 1117 portfolio snapshots of 458 mutual funds in the period 1995-2002. Shukla calculates the excess return due to active management and compares that outcome with a return that would have been earned if the portfolio would not have been revised over time. Shukla concludes that the average excess return is not significantly. However there are some active managers who do perform superior but management of these funds also has higher expense ratios and therefore the investors do not benefit from this. The funds which do perform above average have small and concentrated portfolios and do not have the highest turnover. In a more recent study, Fama and French (2008) set up a research on mutual fund performance from the perspective of equilibrium accounting. This principle holds that if the mutual fund industry wins, in another industry there will be loss. So, one industry wins at the expense of another industry. The data which is being analyzed comes from the CRSP database and consists primarily of U.S. equity funds in the period 1984-2006. The conclusion is that there are funds which have information and thus have higher expected returns, but together with these better performing funds, there are also funds which have bad information and thus do not 31

improve their performance. In the research it is not possible to find out which funds have managers with stock-picking abilities and which funds do not. 3.2 Active portfolio strategies do outperform passive portfolio strategies Wermers (2000) shows the value of active management in his research on mutual fund performance. For this research Wermers combines two databases. One database contains quarterly snapshots from equity mutual funds in the period 1975-1994. And the second database contains monthly net returns, yearly returns and other fund s characteristics. Based on these databases he concludes that mutual funds hold stocks that outperform the market index (CRSP) on average. The average mutual fund holds a stock portfolio that beats the S&P 500 index. (Before transaction costs) In the research, stock holdings and the net returns level of mutual funds are taken into account separately. Because performance is divided into several components, a more accurate analysis of performance can be executed. The superior performance is largely caused by higher average returns due to certain characteristics and furthermore by stock picking talent of management. But when looking at performance on a netreturn level, the funds underperform market indexes. So mutual funds can beat the market by almost enough to cover expenses and transaction costs when only stock holdings are considered. But mutual funds holding cash and bonds lower the net returns of mutual funds. In his research, Wermers ignores the fact that actively managed funds have more tax burdens than passively managed funds. In a more recent paper Petajisto (2010) also found evidence for the fact that active managed funds outperform the benchmark indexes. This result even holds after fees and 32

transaction costs. Petajisto states that not all managers are the same; they differ in their way of managing and they can have different styles. These differences in management are very important for investors to take into account. On average, actively managed funds do not outperform their benchmark. However this is caused by some funds performing bad, but there are also some funds which do outperform their benchmark. Funds which take factor bets perform poor and also closet indexers perform poor. On the contrary, the most active stock pickers do outperform their benchmark, even after fees and expenses. Also concentrated funds perform well, but just enough to match the benchmark. The sample which is being researched consists of 2740 funds in the period 1980-2009, composed of different databases: S&P, Russell and Dow Jones Wilshire. Ippolito (1989) researched data of 143 mutual funds over the period 1965-1984. He finds that mutual funds, except for load charges, outperform index funds. However the higher riskadjusted returns have to compensate for the higher expenses and charges of the management. Only load funds earn high enough returns to cover the extra costs. Hendricks, Patel and Zeckhauser performed a research in 1993 on the persistence of mutual funds performance. Quarterly returns data over the period 1974-1988 is being analyzed from open-end, no-load, growth oriented equity funds. They conclude that managers have an advantage of persistence in mutual funds performance. Managers can adopt their strategy by selecting the top performing funds (based on the last four quarters). These top performing funds indicate that the fund will also perform well in the future and thus managers can significantly outperform the average mutual fund, at least the benchmark index. 33

In 1997 Daniel, Grinblatt, Titman and Wermers come up with a new measure to judge performance of mutual funds. They introduce benchmarks which are based on different characteristics of the funds. The characteristic timing measure can find whether managers can time their portfolio weightings on these characteristics and the characteristic selectivity measure detects whether managers can select stocks which perform better than the average stock with the same characteristics. The data consist of 2500 equity funds in the period 1975-1994. Their conclusion is that the average mutual fund does succeed in stock-picking, the stocks being picked outperform passive funds. However, this difference in performance is not large, only 100 basis points. And besides, the amount of excess return is almost equal to the amount of the management fee. This implies that investors do not gain from the excess return. Although the average excess return is small, there are some funds performing extremely well. So there are some active funds which perform better than passive funds. Also Grinblatt and Titman (1988) conclude that there are some active funds performing significantly positive. Growth and aggressive-growth funds have significantly positive gross returns on average. This positive gross return is partly caused by active management of the fund according to Grinblatt and Titman. For their research, quarterly equity holdings in the period 1975-1984 are being analyzed. What distinguishes their study is the fact that their data also contains data which do not survive the entire period. Therefore the results are not subject to a survivor-bias. In the 2008 study of Fama and French, they find that there are some managers of mutual funds which have stock-picking abilities, but these are offset by funds with bad information and inferior performance. Their conclusion also contains the statement that on average mutual funds underperform three-factor and four-factor benchmarks by the amount of fees and expenses. So, 34

on the one hand they support the fact that there are some funds outperforming their benchmark. However, on average mutual funds do not outperform their indices because higher excess returns go together with higher fees and expenses. Also Cuthbertson and Nitzsche find in their literature study of 2006 that there are 2-5% equity funds in the UK and US that outperform their benchmark. However their findings also include the fact that 20-40% of mutual funds performs very bad. The 2-5% top performing funds is mainly caused by stock picking talents and not by market timing. Especially aggressive growth funds tend to outperform their benchmarks. However, they also conclude that investors can hardly benefit from these excess returns due to higher fees and expenses. 35

Chapter 4 Data and Methodology 4.1 Data description The mutual fund data will be restrained from Wharton Research Data Services. The data of historical performance of open-ended mutual funds are from the CRSP Mutual Fund database. The evaluation period is from January 2001 until March 2006. These data are survivor-bias-free, which implies that also dead funds are included in the database. Most other databases about mutual funds contain a survivor-bias. This means that the results should be taken with care. The average return in the dataset is then higher than it actually is because good performing mutual funds survive most of the time and bad performing funds do not survive. The active portfolio which is being used is the BNY Mellon Fund Trust: BNY Mellon Intermediate US Government Fund; Investor Share. As benchmark the ISIIX passive AIM Stock Fund: AIM S&P 500 Index Fund; Institutional Class Share is being used. The returns are on a monthly basis. 4.2 Research Methodology The following values are computed from the CRSP dataset. Table 1 contains descriptive statistics on the active fund and the passive fund. 36