عملکرد صندوق های تامینی و صندوق سرمایه گذاری مشترک بازارهای نوظهور

Use of short selling and derivatives is limited in most emerging markets because such instruments are not as readily available as they are in developed capital markets. These limitations raise questions about the value added provided by hedge funds, especially compared to traditional mutual funds active in these markets. We use five existing performance measurement models plus a new asset-style factor model to identify the return sources and the alpha generated by both types of funds. We analyze subperiods, different market environments, and structural breaks. Our results indicate that some hedge funds generate significant positive alpha, whereas most mutual funds do not outperform traditional benchmarks. We find that hedge funds are more active in shifting their asset allocation. The higher degree of freedom that hedge funds enjoy in their investment style might thus be one explanation for the differences in performance.

Institutional investors and high-net-worth individuals have put significant amounts of money into hedge funds, seeking high returns as well as diversification benefits promised by hedge fund managers (see Fung et al., 2008). Due to the absence of reliable data, academic literature on hedge funds in the 1990s was restricted to descriptive analysis and relatively simple performance metrics (e.g., Fung and Hsieh, 1997, Fung and Hsieh, 1999 and Ackermann et al., 1999). However, as more information and data have become available, more sophisticated techniques from quantitative finance have been used to analyze hedge funds. One important stream of this literature has developed multifactor performance measurement models (Fung and Hsieh, 2001 and Agarwal and Naik, 2004) that identify the sources of hedge fund returns and separate the risk premiums from different investments (beta) and the alpha that hedge fund managers provide. Recent literature shows that classical, linear performance measurement models often cannot capture the dynamic trading strategies in the different asset classes and markets that many hedge funds pursue (Agarwal and Naik, 2004 and Capocci and Hübner, 2004). Moreover, hedge funds employ a variety of trading strategies, so analyzing all hedge funds using only one performance measurement framework that does not consider the characteristics of the specific strategies is of limited value. Hedge-fund-style specific performance measurement models are needed so as to capture the differences in management style (Fung and Hsieh, 2001, Fung and Hsieh, 2004 and Agarwal and Naik, 2004). In this paper, we use recent innovations from performance measurement literature (Agarwal and Naik, 2004, Fung and Hsieh, 2004 and Fung et al., 2008) to analyze the performance of emerging market hedge funds. We define “emerging markets” as those countries or areas of the globe that are in the process of rapid growth and industrialization, such as China, India, and Latin America, as well as many eastern European and southeastern Asian countries. These markets exhibit significant growth opportunities, but also high political and economic risks, making emerging markets more volatile than mature markets (De Santis and İmrohoroğlu, 1997). A main difference between emerging market hedge funds and other hedge funds is that use of short selling and derivatives was relatively limited in the previous two decades because such instruments were not as readily available as they are in developed capital markets.1 These limitations raise questions about the value added provided by these funds, for example, compared to traditional long-only mutual funds. Emerging market hedge funds have been analyzed as one among many strategies in hedge fund performance measurement literature such as Fung and Hsieh, 1997, Fung and Hsieh, 2001, Agarwal and Naik, 2004 and Capocci and Hübner, 2004. However, all these authors do not analyze these funds in detail or try to extract the main differences between these funds and other hedge funds.2 This is somewhat surprising, especially given the relative importance of emerging markets in the hedge fund industry.3 Further the underlying factors, such as emerging market stock and bond indices, are – at least recently – more readily available than for other hedge fund strategies which involve more complex arbitrage strategies. Our analysis will show that appropriate factor models can be derived much more easily for emerging market hedge funds than for other hedge funds. Among the few authors who focus on emerging market hedge funds are Sancetta and Satchell (2005). However, they analyze only a small sample of 15 emerging market hedge funds over a relatively short period (60 months). Furthermore, their aim is to apply a new test statistic for market timing on a data sample. More recently, Strömqvist (2007) analyzes the skills of emerging market hedge fund managers. Her focus is on comparing emerging market hedge funds with other hedge fund strategies, while our focus is on comparing emerging market hedge funds with mutual funds active in this market. Abugri and Dutta (2009) analyze whether emerging market hedge funds follow a pattern similar to that reported for advanced market hedge funds after 2006. The focus of this paper also differs from this analysis, in that we compare hedge funds and mutual funds active in emerging markets, while these authors analyze whether emerging market hedge funds are comparable with hedge funds that are active in advanced markets. Furthermore, we analyze individual hedge fund data; Abugri and Dutta (2009) consider hedge fund indices.4,5 The aim of this paper is to fill a gap in literature by providing a broad evaluation of the performance of emerging market hedge funds and mutual funds. We build upon insights from both the hedge fund and mutual fund literature and analyze six factor models, some of which are representative of recent innovations in this growing field of research. For comparison purposes, we start with the classical single-factor (1) Capital Asset Pricing Model (CAPM) and then extend our analysis to more complex multifactor models, including, (2) Fama and French (1993), (3) Carhart (1997), (4) Fung and Hsieh (1997), and (5) Fung and Hsieh (2004). All these models are useful in identifying the risks underlying hedge funds and mutual funds, but they cannot account for the specific characteristics of emerging market hedge funds. We thus employ emerging market risk factors to set up our sixth model: an emerging market asset class factor model. In our analysis we compare the performance of hedge funds not only with traditional benchmark indices, but also with traditional mutual funds that have an investment focus in emerging markets. Most studies only consider either hedge funds or mutual funds; we analyze both investment vehicles active in this growing market.6 Our analysis builds upon the Center for International Securities and Derivatives Markets (CISDM) database, which is one of the largest hedge fund databases ever analyzed for this purpose. It contains data on 566 hedge funds which have an emerging market focus. Additionally, we select 1542 mutual funds active in emerging markets from the Thomson Financial Datastream database. The analysis covers the years 1995 through August 2008, which is advantageous for three reasons. First, the results will not suffer as much from the survivorship and backfilling biases that plague much of the older hedge fund research.7 Second, this period contains bull as well as bear markets, allowing us to analyze the performance of emerging market hedge funds in different market environments; many other studies are limited to the analysis of bull markets.8 Third, the analyzed time period contains some critical events for emerging market hedge funds, such as the Asian crises in 1998 and the technology bubble in 2000. We consider these events in detail in our analysis of structural breaks, subperiods, and market environments. Our main findings can be summarized as follows. (1) Hedge fund returns and alphas are much higher than those of traditional mutual funds. (2) Some hedge funds outperform traditional benchmarks, whereas most mutual funds tend to underperform traditional benchmarks. (3) In bad or neutral market environments, hedge funds outperform mutual funds while generating the same returns in good environments. Overall, our analysis indicates that emerging market hedge funds perform better than their traditional competitors. We also discuss potential reasons for the performance differences, i.e., higher flexibility, liquidity risk, lower regulation, and technical problems such as return smoothing. The remainder of the paper is organized as follows. Section 2 covers the methodology, i.e., the six performance measurement models we use in the empirical part. Section 3 presents our data and discusses how we deal with the several data biases inherent in hedge fund data. In Section 4 we present our empirical findings, and we conclude in Section 5.

The contribution of this paper is twofold: In a first step, we develop an asset class factor model to describe the performance of hedge funds and mutual funds investing in emerging markets. Our results indicate that the market-related factors chosen for our model are much better at explaining the variation in emerging market returns than are non emerging market specific factor models presented in the literature and that they are slightly better than the emerging market specific model of Abugri and Dutta (2009). Our model explains a large proportion of the variation in both mutual fund and hedge fund returns. The second contribution of this paper is to employ various factor models to compare returns of hedge funds and mutual funds active in emerging markets. We find that hedge funds provide both higher returns and alphas than do traditional mutual funds. These findings are in line with other recent literature (Abel and Fletcher, 2004 and Strömqvist, 2007). In general, some hedge funds tend to outperform the benchmarks, but most traditional mutual funds do not. One possible reason could be more active management of hedge funds than of mutual funds. We find support for this hypothesis from the tests for structural breaks, the factor exposure, and from the analysis of the performance in different market environments. Regarding structural breaks, we only find significant breakpoints for hedge funds but not for mutual funds. This indicates that hedge funds are adjusting their risk taking while mutual funds are not. The factor exposure of hedge funds, which we reveal using a rolling regression, shows that hedge funds have a more volatile exposure, supporting the idea of a more active management. The analysis of different market environments shows that hedge funds provide to some extent downside protection in contrast to mutual funds that have a rather constant exposure to market movements. In conclusion, it seems that emerging market hedge funds are more active in shifting their asset allocation, probably since they are less restricted by their investors in investment style and policy. Furthermore, it is plausible that hedge fund style shifts have been especially pronounced in the most recent period (post 2006) since more alternative instruments, such as options and futures, are becoming available in emerging markets and hedge funds are not restricted in using them. It might thus also be that emerging market hedge funds now behave more like other hedge funds (see Abugri and Dutta, 2009), but we believe that additional research with more recent data is necessary to confirm this assertion, since the last, most recent subperiod analyzed is relatively short. However, investors need to be aware that (aside from the differences in their flexibility regarding asset allocation) there are numerous reasons which might be responsible for the performance difference between mutual funds and hedge funds, including the use of leverage, lock-up periods, and incentive fees for hedge fund managers. Lock-up periods are also a good example to emphasize the higher degree of freedom hedge fund managers enjoy in making investment decisions. For example, hedge funds might invest in illiquid positions and capture liquidity risk premiums, actions not allowed to traditional mutual funds (see Ding et al., 2009, for an analysis of liquidity in the hedge fund context). In case of illiquid investments, investors need to be aware that hedge fund managers might smooth their returns (see Getmansky et al., 2004), which might bias performance measurement results.26 Kouwenberg and Ziemba (2007) illustrate that incentive fees and manager’s own investment in the fund substantially affect the investment strategy of hedge fund managers. Both these elements are not widespread with traditional mutual funds. Furthermore, hedge funds are not subject to much regulation. Hedge funds in the United States are usually set up as limited partnerships, a legal form only lightly regulated, and hedge funds outside the United States are usually domiciled offshore, a practice that has both regulatory and tax advantages. All these advantages make hedge funds the more flexible investment scheme, both as to investment strategy and markets in which to invest. During the financial crisis hedge funds have been severely criticized and it is not clear whether future regulation in the financial services sector might diminish these regulatory advantages of hedge funds.27 Overall, it thus seems that a combination of technical problems (e.g., return smoothing) and economic advantages (e.g., higher flexibility and lower regulation) might account for the observed performance differences between hedge funds and mutual funds. The factor model developed in this paper can be put to a number of different uses. First, investors can use the model to identify well-performing funds in which to invest. Although past performance is not necessarily an indicator of future returns, investors heavily rely on past performance when making investment decisions (see Capon et al., 1996). Second, the model can be a tool for determining manager compensation as the model can detect whether a fund’s performance is mainly attributable to passive investment style or something more proactive. The model makes it possible to reward managers for only those returns superior to a specific benchmark, and thus attributable to the fund manager’s skill. Third, the model can be used for risk management as revealing the underlying assets will help identify the true risk of a fund. This might be especially relevant in identifying a drift in management style; catching any such changes early will help keep a portfolio both safe and profitable.