Hedge funds: Performance and risk exposure in times of crises
Master thesis Finance
Colby Harmon 10070168 Thesis supervisor: R. Vlahu
Statement of Originality
This document is written by, Colby Harmon who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Inhoudsopgave
1. Introduction 4 2. Hedge funds history 7 3. Literature on hedge fund performance 9 3.1 Evolution of performance measures 9 3.2 Studies on performance of mutual and hedge funds 10 4. Strategy performance 16 4.2 Strategies 16 4.1 Financial crisis impact on hedge fund performance 18 4.3 Strategy alpha 20 4.4 Portfolio allocations 21 4.4.1 Cross correlations 21 4.4.2 Portfolio allocations of hedge funds with other assets 22 4.4.3 Shortcomings of portfolio allocation 24 5. Data 26 5.1 Choice of database 26 5.2 Strategy indices 28 5.3 Seven factor model 29 5.4 Biases 30 5.5 Hypotheses 31 6. Methodology 37 6.1 The seven factor model 37 6.2 Risk adjusted performance 38 6.3 Regression 39 7. Results 40 7.1 Risk adjusted performance evaluation 40 7.2 Time variation 41 7.2.1 Strategy results 43 7.2.2 Strategy returns during two financial crises 47 7.3 Caveats 51 7.4 Robustness test 52 8. Conclusion 54 APPENDIX 621. Introduction
Hedge funds are mildly regulated investment vehicles with a great flexibility when it comes to trading (Edelman et al., 2012). By following highly sophisticated investment strategies they promise to deliver absolute returns to their investors, even in times of extreme market volatility. They believe this because they have the ability to de-correlate with the financial markets (Fung et al., 2008). This promise has increased the interest in the hedge fund industry. The total assets under management in 1990 were estimated to be around $39 billion (Heinz et al., 2009). In 2015 the total amount of assets under management in the hedge fund industry reached nearly $3.2 trillion (Prequin Global, 2016).
Figure 1. Estimated Hedge Fund Industry Assets. (Hedge fund research, 2015).
The financial turbulence in 2008 drove down the total assets under management by almost 25 percent as can be seen in figure 1. This sharp decline was unexpected and led to decline in the popularity of the hedge fund industry. The breaking of the hedge funds promise for delivering high returns caused the decline in popularity. Also when investors wanted to withdraw their money from the hedge funds, the hedge funds would not allow them to do so. In times of extreme market volatility hedge fund managers are allowed to enforce a redemption restriction (Ding et al., 2008). This however is not something that is excessively discussed when attracting new investors. So investors were shocked and feared for their money.
The financial market turbulence offered to a number of hedge fund strategies opportunities to capitalize on the market downturn through their ability to short. Together with the declining numbers of hedge fund competitors, a possible profitable situation arose for those flexible enough to dynamically change their investment strategy (Brunnermeier and
than the market in times of financial turmoil. Credit Suisse and Tremont (2008) show that even in previous times when the markets show extreme volatility, hedge funds can survive and perform. Looking at Figure 1 the hedge fund industry did not only survive the subprime crisis, it picked up where it left off by growing to new heights.
The aim of this thesis is to analyze the performance of individual hedge fund strategies as well as to give insights into their risk-return profiles during crises. When an investor has the choice of a range of hedge fund investment strategies it is useful to have this knowledge to aid in the decision of which strategies to add to his portfolio, especially in times of uncertainty. An understanding of the individual hedge funds strategies returns and their risk exposure is vital. The complex characteristics of these hedge fund strategies is however not always clear, because they are not obliged to report and specify their asset positions (Amin and Kat, 2001). In order to identify and analyze the hedge fund returns and asset positions, Fung and Hsieh (2004) developed a useful method using the identification of systematic risk factors. Using this method an understanding of the hedge fund returns can be linked to market risk factors.
The question of how different hedge fund investment strategies performed during the subprime and Dotcom crises is answered in this thesis. Hedge fund strategies returns are analyzed and linked to observable market risk factors (Fung and Hsieh, 2004). These risk factors, or asset based style factors, pinpoint and quantify the hedge funds characteristics and risks. This provides an insight to the investment exposures of hedge funds. The comparison of risk factors that the ten hedge fund investment strategies are exposed to during the crises shows us the dynamic investment capabilities of hedge fund managers. Also it shows the different risk exposures of each fund so these can also be compared. Strategy outperformance can be measured and identified by applying these risk factors (Fung and Hsieh, 2004).
Ten hedge fund investment styles are obtained from the Credit Suisse Hedge Fund database and are analyzed using the seven-factor model introduced by Fung and Hsieh (2004) to gauge the strategies performance during the Dotcom and subprime crises. The use of traditional asset prices allows the seven-factor model to explain a substantial amount of the systematic risk a portfolio is exposed to. The use of indices and the knowledge about a strategies risk exposure can help investors and managers to design portfolios to manage risk preferences. Hereby saving the investors time by not forcing them to evaluate individual hedge fund investment transactions in order to gauge the risk (Fung and Hsieh, 2004).
The literature generally focuses on funds of hedge funds (FoF’s) (e.g., Fung et al. 2004, 2008; Edelman et al., 2012), which is an aggregate of individual hedge funds, and not
on specific hedge fund strategies as is done here. The risk factors that affect the performance of ten individual hedge fund styles are identified and analyzed. Further, a larger time frame is taken here, compared to previous research, in order to get better insights over the dynamics of hedge fund performance. The period 2004-2010 is analyzed and split into two groups that show two different economic conditions. The period 2004-2007 is the time of the subprime bubble inflation, followed by the 2007-2010 period, which is the subprime bubble aftermath. Also, an analysis is performed of hedge fund performance and risk factor exposure during two different types of crises. The Dotcom crisis is characterized as an equity crisis (Brunnermeier and Nagel, 2004) and the subprime crisis as a credit crisis (Brunnermeier, 2008). This sheds light on the dynamic trading strategies of hedge funds they employ to take advantage of current market environments.
The thesis starts with a brief history of hedge funds and its origins in section 2. Section 3 shows an overview of previous literature. Section 4 looks at individual hedge fund performance and the ten strategies that are going to be analyzed are defined. Section 5 presents the data. Followed by the methodology in section 6. Section 7 shows the results. Concluding remarks are in section 8.
2. Hedge funds history
Albert Winslow Jones coined the name hedge fund in 1949 by devising a strategy that included the short selling of stocks. Using a combination of short selling stocks and buying stocks with leverage, a technique that was then seen to increase risk, he managed to hedge against market risk and thus was able to reduce the risk exposure of investments (Ebbing, 2009). By structuring his company such that it was exempt from The Investment Company Act by the SEC he was able to create more diverse investment strategies than the ones available at that time. Market volatility had less effect on his long-short fund and rewarded superior stock selection (Caldwell, 1995). In the 60’s his fund was still performing well and other hedge fund managers started to copy his investment style. The stock market boom of the late 1960’s induced a change in strategies (Liang, 1999). The short changed into a long-bias strategy in order to reap the benefits of the rising market. This change would lead to a further growth of the industry.
In the beginning of the 1970’s the market downturn caused many closures of hedge funds, making it a very unpopular investment vehicle. However, new wind was blown into the industries sails with the hedge fund of Julian Robertson. In the time period 1980-1990 his fund had the largest returns in the industry. It is reported that the investors compound rate of return was 43% (Fung and Hsieh, 1999). The success of the Quantum Fund, which gained significant value due to their shorting strategy during the Asia crisis, added to the hedge fund popularity again (Brown et al., 1999).
The strong performance of hedge funds in the 80’s was followed by a decrease in the industry in the 90’s. This was mainly because of the immense losses incurred due to incorrect speculations during the Russian debt crisis of 1998. Long Term Capital Management (LTCM) followed suit by failing in 1998 through speculating on the convergence of European interest rates (Edwards, 1999). However, as a result of the debt crisis in Russia, spreads between private and government bonds increased in an abnormal fashion. This gave LTCM enormous liquidity problems and it incurred losses over 90 percent of its total value (Edwards, 1999).
After the Dotcom-crisis in 2001 the industry size grew rapidly and the number of funds doubled within the five years following (Fung and Hsieh, 2004). The popularity of the hedge fund kept on growing until the most recent crisis, when the entire industry had difficulties to keep their head above water.
The fundamentals set by Jones in 1949 still form the basis for a lot of hedge funds. Hedge funds reduce overall risk by taking different (long and short) positions in an asset. Since then however the industry has grown and different strategies have evolved from Jones’ base principle (Fund and Hsieh, 1999). These strategies each have their own definition of what they exactly entail. Ridley (2004) comments that the common strategy of shorting is used for stocks, bonds and commodities. But different types of investments that don’t require short selling of assets can also be classified as hedge funds. He provides a list of common hedge fund characteristics, which include alternative and absolute returns, capital preservation, specialist incentives and future expectations.
However, the most commonly used definition of hedge funds is provided by Connor and Woo (2003) who describe it as a privately organized pooled investment vehicle that is not easily accessible to the general public. Hedge funds are able to take a variety of positions in a number of different assets. This is something that their clients are not able to do. Their clients usually are wealthy individuals, pension funds and private banks (Connor and Woo, 2003). This is because hedge funds usually have a very high minimum investment. Furthermore, additional large fees have to be paid to the managers. The average life expectancy of a hedge fund is no longer than 10 years, even if they perform well.
3. Literature on hedge fund performance
This part starts with an overview of the evolution of the performance measures of returns used in previous papers. Followed by an overview of studies that look at the similarities and differences to hedge fund and mutual fund returns compared to the market benchmark.
3.1 Evolution of performance measures
The most commonly used CAPM based model was Jensen’s alpha (1968) together with Sharpe’s reward to variability ratio (1966). There has been an increase in interest in multi-factor models because of more recent literature on cross-sectional variations in stock returns. By examining the cross-sectional variations of average stock returns on U.S. stock, Fama and French (1998) and Chan et al. (1996) show that there is almost no relation to Sharpe’s beta (1964) or the capital asset pricing model. Instead they classify other factors that have strong explanatory power over the cross-section of average returns (Capocci and Hübner, 2004). A number of these explanatory factors are the leverage, company size, price – earnings, book-to-market, dividend yield and the momentum factor (Elton et al., 1996). Aside from these previous cross-sectional models other multi-factor models have also been introduced, such as the Fama and French three-factor model (1993), Carhart four-factor model (1997) and the international Fama and French model (1998).
Nevertheless there have been studies performed in recent years that give rise to some doubt on the usefulness of the previously mentioned models. By including the factors company size and book-to-market equity the Fama and French three-factor model (1993) gets better results than the classic capital asset pricing model. However, research by Kothari and Warner (2001) shows that the three-factor model finds significantly abnormal results when none actually exist.
The Carhart four-factor model (1997) is based on the Fama and French three-factor mdoel (1993), but it in includes an additional factor introduced by Jagadeesh and Titman (1993), namely the one-year momentum anomaly. According to Capocci and Hübner (2004) this model represents the market equilibrium by looking at four risk factors. Furthermore they show that by adding these additional factors the model has a higher explanatory power over the cross-section of average returns.
In Hedge fund literature a number of different models are used to evaluate performance. For example, by extending the asset class factor model introduced by Sharpe (1992), Fund and Hsieh (1997) find five dominant hedge fund investment styles. By also
using a style-based analysis on a multi-factor approach Scheeweis and Spurgin (1998) they follow in Fung and Hsieh’s (1997) footsteps of using risk factors to performance. Liang (1999) uses an extension of Fung and Hsieh’s model (1997) that is based on the funds characteristics and makes use of the Sharpe ratio. Agarwal and Naik (2002) suggest a general asset-class factor that consists of excess returns on passive option-based strategies. As well as on buy and hold strategies to benchmark the performance of hedge funds (more over in section 4.4.3). This interpretation is picked up by Fung and Hsieh (2004) and is now the leading model for estimating hedge fund performance.
Previous literature suggests the use of a multi-factor model, instead of the CAPM model, to do performance studies. Unfortunately there is no unanimously accepted performance model. Currently the most popular model in measure performance in the hedge fund industry is the Fung and Hsieh (2004) seven-factor model. The risk factors defined in this model have the highest explanatory power in the available databases (Wermers, 2011).
3.2 Studies on performance of mutual and hedge funds
Hedge funds gained an increased level of interest because of their growing prominence in the financial markets (Liang, 1999). Regardless of this increasing interest there has only been limited research done that compares hedge funds to for example the market index or mutual funds. This is because of the difficulties to gather hedge fund data due to their private characteristics (more over in section 5). So in an attempt to understand hedge funds more we compare them to something more familiar like mutual funds (Agarwal and Naik, 2005). Both of the funds are investment vehicles but the strategies they use are not the same.
Mutual funds generally use a passive buy and hold strategy in which they take long positions in highly liquid assets. The funds returns are often compared to a benchmark index. Hedge funds on the other hand use more dynamic strategies. Both long and short positions in illiquid assets are taken, as well as having an absolute return target (Agarwal and Naik, 2005). This contrast shows the difference in the risk return characteristics of the two investment vehicles.
Generally speaking, performance studies that compare hedge to mutual funds can be divided into two groups. They either conclude or deny that hedge funds have significantly higher realized returns than funds that use a passive strategy (Capocci and Hübner, 2004). Ackermann et al. (1999) find that with this outperformance there is also an increased volatility for hedge funds compared to mutual funds. But just as Liang (1999) they also
conclude that although hedge funds outperform mutual funds, they under perform compared to the market equity index. However, when comparing the Sharpe ratios of hedge funds with the market indices Ackermann et al. (1999) find varied results. They address this difference as possibly being driven by their risk adjustments for systematic risk through betas. On the other side there have been a number of studies that show the weakness of the Sharpe-ratio as a risk adjusted performance measure. Amin and Kat (2003), Agarwal and Naik (2004) and Fung and Hsieh (1997, 2001) show that due to the dynamic trading strategies the payoffs to hedge funds are nonlinear.
A number of studies have noticed a non-linear relationship between hedge fund returns and market returns. These studies are done by Fung and Hsieh (1997), Agarwal and Naik (2004) and Mitchell and Pulvino (2001) and propose a more refined method for studying neutrality. These studies all insist that hedge funds do have systematic risk, but that this risk can’t be noticed in the context of a linear factor model that is tested on the standard asset benchmarks (Agarwal and Naik, 2005).
Agarwal and Naik (2004) implement a piece wise linear model for the hedge fund returns as a function of market returns. By the use of buy-and-hold and option-based strategies they characterize the hedge funds systematic risk exposure. Agarwal and Naik (2004) use an option strategy in which they make a monthly trade in a short maturity high liquid put and call option on the S&P index. They find that the returns of the equity oriented hedge fund strategies exhibit non-linear behavior. This result is confirmed by Dor et al. (2003). Furthermore Agarwal and Naik (2004) find that next to the exposure to the equity market risk hedge funds have other significant risk exposures. These exposures consist of risk exposure to the size and value factors of Fama and Franch (1993) as well as the Carhart (1997) momentum factor.
However, only a few specific strategies (e.g., merger arbitrage funds; fixed income hedge funds; equity hedge funds) in the literature have been analyzed that start with looking at the underlying assets like stocks and bonds. This approach is used by Fung and Hsieh (2001), which they call an Asset Based Style (ABS) analysis. Using this approach they show that hedge fund strategies regularly generate option-like returns. They model the returns by looking at closely at a strategy named ‘trend following’. Just like option buyers, trend followers bet on big moves, and thus profit from a volatile market environment (Agarwal and Naik, 2005).
Fung and Hsieh (2001) posit that look-back1 straddles can explain the trend-following funds’ returns better than the standard asset indices. They demonstrate that the trend-following strategy return profile shows that there is a non-linear relationship between the equity market and trend followers. Fung and Hsieh (2002) provide validation that their results that trend following funds are exposed to systematic risk and are hence not market neutral.
Using an ABS approach Mitchell and Pulvino (2001) attempt to identify frequent return components by using apparent market risk factors. By analyzing approximately 5000 mergers over a 35-year time span to characterize the risks and returns in risk arbitrage funds. Risk arbitrage, also known as merger arbitrage, is a strategy that sells short the stock of the acquirer while also buying stock in the target firm. This strategy counts on the fact that the merger will be completed. The results show that the merger arbitrage funds’ returns have zero correlation with the market in a market upswing and that the correlation level is high when the market is in a downswing (Agarwal and Naik, 2005). The conclusion they draw is that the systematic risk in merger arbitrage funds, when the market is in a downturn, comes from the fact that mergers will be cancelled when the market is doing poorly.
Another paper that uses an ABS approach is Agarwal et al. (2005). They research the risk-return characteristics using data on Japanese and U.S. convertible bonds and their underlying stocks. A convertible arbitrage strategy consists of buying a portfolio of convertible securities and selling short the underlying stock, thus hedging the equity risk. Three trading strategies are used to try and explain the convertible arbitrage funds’ returns, namely positive carry, credit arbitrage and volatility arbitrage trading strategies. The positive carry strategy uses a long position in a convertible bond and a short position in the underlying stock, thus minimizing its credit and equity risk (Agarwal and Naik, 2005). By creating a long credit spread position, the credit arbitrage strategy tries to minimize its equity and interest rate risk and thus capture value from mispriced credit risk, which is inherent to convertible bonds. Volatility arbitrage seeks to capitalize on underpricing in the option in convertible bonds while minimizing both credit and interest rate risk (Agarwal and Naik, 2005).
A research into fixed income funds by Fung and Hsieh (2000) where they look at a different source of common risk factors. Here they find that this kind of fund is often exposed to yield spreads. The yield spread is the difference between the yields of the two bonds. Fung and Hsieh (2000) model “convergence trading” with options to explain the returns of these funds. This kind of trading speculates on the relative price between assets to converge. The
convergence strategy is the reversed of the trend following strategy (Agarwal and Naik, 2005). Trend following strategies try to profit from large up or down swings in a volatile market. When the price of an asset exceeds a certain level, a long position is taken. The entry and exit strategies of both trend following traders and convergence traders will be similar, but in different directions.
In an attempt to benchmark hedge fund returns, Fung and Hsieh (2004) us ABS-style factors in a model of hedge fund risk. Their model has a similar base as ones used in arbitrage pricing theory but makes use of dynamic risk factor coefficients. Seven asset based style factors provide a useful insight in the explanation of the monthly return variations. These seven factors help explain common risk factors in hedge funds. These risk factors include equity factors (equity market and its size), fixed income factors (bond market and credit spread) and trend following factors (bond, currency and commodity markets) (Agarwal and Naik, 2005). These risk factors together explain a large part of the returns in hedge fund portfolios.
In an examination of performance in funds-of-hedge funds (FoHFs), Fung et al. (2008) look at the performance, risk and capital formation in the hedge fund industry. Building on the seven-factor model introduced by Fung and Hsieh (2004) they look at the performance during a period from January 1995 to December 2004. Fung et al. (2008) find a high systematic risk exposure in FoHFs during this period.
Edelman et al. (2012) added another factor in their research of funds of hedge funds performance, namely the emerging market. The reason to include this factor was that emerging markets like India and China were experiencing a large growth in their economy and were offering nice equity returns during 2005-2010. The addition of this factor increased the explanatory power of the model introduced by Fung and Hsieh (2004).
Hedge fund managers and traditional fund managers regularly transact in the same markets. But both funds returns have different characteristics. The hedge fund return that can’t be explained by the exposure to systematic risk factors is the alpha (Agarwal and Naik, 2005). Mutual funds returns have a higher correlation with standard asset returns than hedge funds do. According to Fung and Hsieh (1997) this is because hedge fund managers are more skilled than traditional fund managers. Liang (1999) compliments to this finding by attributing positive unexplained returns for hedge fund groups to higher managerial skill. This however is not always the case, Fung et al. (2008) show that the alphas of different hedge fund styles seem to be cyclical and depend on the market environment.
Furthermore, Fung and Hsieh (2004a) empirically show that the equity long-short hedge fund strategies have significant alphas to both the conventional risk factors as well as the alternative ones. Fung et al. (2008) show that the alpha is actually insightful when looking at the quality of different FoHFs. They show that FoHFs that have positive and significant alphas are more likely to survive than those that don’t.
If managerial skill would lead to superior returns by hedge funds, then one would expect to see that these funds outperform year after year. Thus showing persistence in their returns. This is important for investors, because investing in hedge funds requires a lock-up period2, implying that hedge fund investors need enough information on the persistence of the performance before investing (Agarwal and Naik, 2005). When managers have more flexibility in their investment strategies you would expect to see more persistence in the performance of hedge funds. According to Agarwal and Naik (2005) this flexibility comes from greater restrictions on the outflows of capital and increased lock-up periods. The literature is split into two camps. Papers that conclude there is no persistence in performance and papers that show persistence over short or long term horizons. Brown et al. (1999) find no evidence of persistence using a sample of offshore hedge funds. Also Boyson and Cooper (2004) find non-existing persistence, over short and long horizons, in hedge fund strategies when they are selected only on past performance. But when funds are chosen based on manager tenure and past performance, they find persistence over the shorter horizon. On the other hand, Agarwal and Naik (2000) and Chen (2004) both find hedge fund persistence, except only for the short time horizon. When looking at a longer time period there is no persistence to be observed. Also Edwards and Caglan (2001) studied the persistence in hedge fund performance over an eight-year time span in the 1990’s and found evidence for persistence up to a two-year horizon. Using a different methodology than Brown et al (1999) and Agarwal and Naik (2000), Bares et al. (2003) find that there is evidence of short-term persistence, but that this quickly disappears when the holding period increases. The general message in all these papers is that investors need to be very vigilant when counting on past performance when selecting hedge funds for longer-term investment horizons.
During the Internet bubble of 2001 each hedge fund was affected in their own way. Brunnermeier and Nagel (2004) look at the type of stocks held by hedge funds in times of the technology bubble on NASDAQ. They find that the majority of the hedge funds were holding
2 “A lock-up period is a window of time in which investors of a hedge fund are not allowed to redeem or
large amounts of overpriced technology stocks. Ofek and Richardson (2003) attribute the rise and fall of the technology stock prices to the effects of short sale restrictions and heterogeneity. However, just before the prices of these stocks collapsed hedge funds reduced their exposure to these stocks. This led to an outperformance of their stock holdings compared to the benchmarks (Agarwal and Naik, 2005). The results found by Brunnermeier and Nagel (2004) imply that rational investors tend to ride price bubbles due to arbitrage limits and predictable investor sentiment. This goes against the efficient market view of rational speculation introduced by Friedman (1953) and Fama (1965) in which rational speculators would attack the price bubble effectively exerting a correcting force on the price.
4. Strategy performance
The aim of this thesis is to get a better understanding of the performance of individual hedge fund strategies. Therefor the ten main hedge fund strategies are analyzed. In order to do so market based outperformances and risk factors are identified. For investors and portfolio managers it is attractive to know how various strategies are affected by risks in order to make investment decisions. Most importantly it is useful to have an understanding of how different strategies are affected in turbulent economic times as different strategies will be affected in different ways. The recent debt crisis forms an ideal platform to analyze this. Previous research performed by Fung and Hsieh (2004) focuses on aggregate hedge fund indices. This however does not give us any insights in the particular hedge fund strategies (Liang, 2000). This thesis extends earlier research by specifically analyzing individual hedge fund strategies during the most recent debt crisis.
4.1 Strategies
The characteristics with respect to individual hedge funds such as performance and risk are analyzed in this thesis. In order to do that the hedge fund strategies will be defined in the following section. The subsection following the definitions of hedge fund strategies will contain the definition and relevance of the outperformance measure, the alpha.
In the hedge fund industry there are ten strategies that are most commonly used. Credit Suisse Hedge Fund database follows these strategies and provides a detailed description of each strategy. Based on Credit Suisse Hedge Fund Database and Hauser (2005) all strategies are defined here.
1. Convertible Arbitrage: the goal of this strategy is to benefit from purchasing
convertible securities and shorting the underlying stock when a pricing error is made in the conversion of the security. When the speculations on the spreads between the prices of convertible securities and non convertible securities increase the arbitrageur makes a profit.
2. Equity Long-Short or Equity Hedge: with this strategy market risk is minimized by
holding long and short positions in equity and equity derivatives and diversifying across industries or regions. This is based on the hedging principles introduced by Albert Winslow Jones. Holding a portfolio of short positions can reduce market exposure caused by long positions, thus this strategy profits form changes in stock spreads.
3. Equity Market Neutral: these funds usually take both long and short positions in
market. They look to take advantage of investment opportunities that are particular to a specific group of stock, while maintaining a diversified exposure over a broad group of stocks determined by industry, sector or region for example.
4. Event Driven: this kind of strategy usually invests in different asset classes seeking
to make a gain from possible mispricing securities associated with specific corporate or market events. These events can include: bankruptcies, mergers, reorganizations, spinoffs and litigations as well as more kinds of corporate events. These funds invest in various kinds of derivatives such as equities, options and fixed income instruments.
5. Dedicated Short Bias: more short positions are taken compared to long positions.
Returns are earned by maintaining a net short exposure (more dollars short than long) in long and short equities.
6. Emerging Markets: involves investing in equity or debt in developing markets
across the world. Each country has its own investing guidelines. In a few countries it is prohibited to short sell or the derivative market is not fully developed. This makes hedging a difficult practice so investments in long securities are most common.
7. Fixed Income Arbitrage: these funds try to make profits by taking advantage of
inefficiencies and prices inconsistencies between similar fixed income securities. These funds hedge out their exposure to the market and interest rate risk by trying to limit the volatility. For example when a bond is perceived to be undervalued a manager should buy it and short sell a similar perceived overpriced bond.
8. Global Macro: is a trend following strategy that focuses on identifying extreme
price valuations and uses leverage to bet big on expected market movements. The strategy looks at international factors such as employment, inflation and political stability. Speculations are made and positions are taken on the future movements of stock and bond markets, exchange rates and commodity prices.
9. Managed Futures: this type of fund (also known as Commodity Trading Advisors
or CTAs) focuses its investments on financial and commodity futures globally. With this they make bets based on anticipated movements with long and short positions.
10. Multi-Strategy: are funds that are know for their capability of allocating capital
among a number of hedge fund strategies in order to take advantage of anticipated opportunities. Regardless of the direction of movement in the interest rate, currency or equity market, managers try to maintain positive returns through the diversification of capital. This reduces risk and can help smooth the returns to investors. The most typical strategies that are used by multi-strategy funds are equity long short and arbitrage funds.
4.2 Financial crisis impact on hedge fund performance
To get a better picture of hedge funds and their individual strategy performance it is interesting to see the impact of past financial crises on their returns. Table 1 shows the returns of each strategy in percentages in times of financial turmoil. As can be seen in the table the returns of the hedge funds vary greatly between different time periods.
The Asian currency crisis did not greatly affect the hedge fund industry. Most of the strategies actually profited from the devaluation of the Asian currencies. A scarcity of foreign exchanges led the value of Asian equities and currencies to hugely decrease. Brown Goetzmann and Park (1998), note that even though emerging market funds had a large exposure to the foreign currencies, they did not cause the crisis. Corsetti, Pesenti and Roubini (1999) show that the root of the problem lies with the account imbalance and the quantity and quality of the bank over lending. Economies were globally affected as a consequence of inadequately adapted financial sectors. However, according to Hedge Fund Research (2009), the hedge fund industry was almost completely able to de-correlate from the equities market. By doing so they gained from the disturbance in the market. All except the emerging market strategies, which took a big hit.
The 1997-1998 period was an unstable one for the emerging markets hedge fund style. It was characterized with lower corporate profits, increasing default rates, loose fiscal policies and increasing foreign currency debt (Credit Suisse, 2007). 1998 is also known for the year that Long Term Capital Managent (LTCM) nearly went bankrupt and had to be bailed out. LTCM incurred immense losses after the Russian government defaulted on their sovereign currency bonds (Edwards, 1999). Heinz et al. (2009) sees similarities between the Russian sovereign crisis and the sub prime crisis because the market volatility increased while the stock markets and the market liquidity decreased around the globe.
Table 1. Performance, in the form of absolute returns, of individual hedge fund strategies during past financial crises can be seen below. Provided by Credit Suisse (2008). Analysis of their database: Credit Suisse Tremont
Index LLC.
The technology equity bubble bursting hugely increased the interest in the hedge fund industry. The industry performed much better than the equities market. Brunnermeier and Nagel (2004) found that this outperformance is due to the fact that hedge funds were riding the bubble. Hedge fund managers figured out that their exposure to high priced technology stocks in their portfolio were too large. As a consequence they reduced their exposure in already highly priced technology stocks and switched their investments to technology stocks whose price were on the up rise.
It is noticeable in Table 1 that the equity market neutral strategy has shown a constant performance. Capocci et al. (2005) find significant persistence in the outperformance in bullish market circumstances, yet no underperformance in bearish markets. The fact that hedge fund managers are able to use both long and short positions in various asset classes has allowed them to de-correlate from financial trends in global market decline in pursuit of delivering positive returns in bear markets (Credit Suisse, 2008).
However, Credit Suisse (2008) notes that analyzing historical periods of market disruptions gives an insight into how hedge fund strategies perform during severe market downturn. It is very difficult to say how these strategies will perform in future due to the ever-changing economic conditions. For this reason it is always advisable to maintain a well diversified portfolio to protect against possible downturns.
4.3 Strategy alpha
Strategy risks are selected as benchmarks in this thesis. If this is the case, following Fung et al (2008), the alpha then explains the average return of a strategy after it has been compensated for its exposure to their systematic risk factors. In hedge fund studies the alpha is the measure of outperformance of the strategy. According to Kosowski, Naik and Tea (2007), the alpha shows if a certain strategy adds value over the systematic bets it takes on risk factors. By means of analyzing the alphas of all strategies over the course of the two most recent crises, the systematic outperformance of these strategies can be determined.
Schneeweis (1999) and Berk and Green (2004) are also studies that look at hedge fund alphas, but they attribute the discovered alphas to managerial skill. That, they argue, is because these are returns amassed on fund specific risks that are not connected to systematic risks. Future performance and the liquidation of funds are also impacted by the alpha. For this reason, they should be analyzed when making an investment decision.
In a research on FoHF’s, Fung et al (2008) looked at alpha returns between 1995 and 2004. They found that only one fifth showed positive alphas that were significant at the 5 percent level. Furthermore, they find that the positive alpha funds have a larger and steadier capital inflow, which in turn increases the amount of invested capital in time. Contrary to funds with positive and significant alphas are the beta only funds. This beta measures the sensitivity of the performance of the different strategies to a risk factor. The capital flows (in and out) of these funds exhibit evidence of the investors return chasing behavior. In this paper Fung et al. (2008) show that the likelihood of liquidation of funds that have positive significant alphas is lower than that of the beta-only funds. The increase in the amount of capital within the hedge fund industry has lead the magnitude of the alpha to decline according to Berk and Green (2004). They confirm that investors who have learnt about previous performance drive down alpha returns and consequentially invest their capital in superior funds, leaving no alphas remaining in the equilibrium. Funds that have delivered positive alphas are not expected to deliver positive alphas in the future according to Fung et al. (2008). They also show evidence of the reverse being true, that funds that historically have not produced positive alphas are expected to do so in the future.
Eling (2009) finds contrastingly that previous research gives mixed results about the alphas in an examination of 25 studies. The reason for these mixed results, he says, comes from the differences in methodologies used and that the significance of these alphas are largely dependent on these same methodologies. He finds that alpha outperformance exists for up to six months, after which the levels of significance decline. This finding coincides
nearly all of the previous studies looked at. The alpha itself, he concludes, is not connected to the performance measure model, yet it does depend on the strategy followed.
There is no uniform answer with regard to the alpha. Identifying the alphas for the numerous strategies during the previous two crises is thus interesting. The necessary research will be conducted in this thesis.
4.4 Portfolio allocations
Investors have a lot of options when constructing their portfolio. When they include only hedge funds in their portfolio it can be very valuable to look at the diversification of risk possibilities within the portfolio. To analyze this it is good to investigate the cross-correlations of the hedge fund strategies (section 4.4.1). In section 4.4.2 an optimal portfolio of hedge fund strategies in combination with other assets is discussed. The shortcomings of portfolio allocation are discussed in section 4.4.3.
4.4.1 Cross correlations
This thesis looks at different risk return profiles of hedge fund trading strategies with other assets. Yet in the decision of what hedge funds strategies to combine in a portfolio the cross correlations should be considered (Adrian, 2007). Cross correlations depict the co movement of hedge fund returns relative to their volatility. If hedge fund strategies within a portfolio are highly correlated this increases the risk of the portfolio. On the other side, when the correlations are lower, diversification benefits are higher (Credit Suisse and Tremont, 2008).
Hedge fund strategies tend to have low cross correlations because of their different investment styles. Credit Suisse (2008) and Adrian (2007) find that this is due to their exposure to different assets classes. Agarwall and Naik (2004) find that most hedge fund strategies perform poorly in economic downturn, which implies that their correlations tend to move together.
Table 2. Cross correlations of hedge fund strategy returns 1994-2006. (Adrian, 2007)
Adrian (2007) finds an average cross correlation of 40 percent of the hedge fund strategies with the Credit Suisse Hedge Fund Index during this period. High strategy correlations entail high underlying assets correlations. When the market is hit by an economic shock, this could lead to an industry crisis. Amin and Kat (2001) show that many hedge fund strategies show systematic exposure to the traditional equity markets. However they find that the Convertible Arbitrage and Equity Market Neutral investment strategies are exceptions and show relatively low equity market risk exposure. This does not mean that these strategies’ returns are not exposed to systematic market risk factors. Yet that there is an unclear picture of what the systematic risk factors are for these strategies.
4.4.2 Portfolio allocations of hedge funds with other assets
Hedge funds are able to reduce portfolio risk and therefor can complement traditional investment styles. This is possible because of hedge funds low or negative correlation with different assets. By granting investors access to dynamic portfolio management and alternative premiums, return improvements can be established (Ebbing, 2009). A number of studies, such as Brooks and Kat (2001), Kat (2002) and Lhabitant (2001), have found that hedge funds are different than the standard equity and bond returns. Hedge funds present diversification benefits as well as enhance portfolio management results in times when equities and bonds do not perform well (Amo et al. (2007), Credit Suisse and Tremont (2008), Staman and Scheid (2008)).
Credit Suisse and Tremont (2008) provide a good example of the benefits by analyzing the risk return profiles of it’s own hedge fund index in the period in the period of
financial turmoil in 2008. Their hedge fund index gained approximately 4 percent from July 2007 to June 2008. In this same period the S&P500 and the MSCI Global Index lost approximately 13 percent (Credit Suisse and Tremont, 2008). When comparing the volatilities of the indices over the same period, the Credit Suisse index showed a volatility of less than 5 percent while both equity indices were above 11 percent. The hedge funds index showed higher returns along with a lower volatility. This insinuates that hedge funds are alluring investment vehicles to have in a portfolio in times of economic turmoil.
A study conducted by Amin and Kat (2003) investigates the diversification benefits that could be reaped by combining hedge funds with traditional portfolios consisting of stocks and bonds. The initial mix consists of 50 percent hedge funds and 50 percent bonds. When a higher return is preferred, bonds are exchanged for stocks, while the allocation of hedge funds stay the same. At the point when all the bonds have been traded, the share of stocks will continue to increase at the expense of hedge funds.
Figure 2 Hedge fund optimal portfolio allocations.
Picture on the left depicts the optimal portfolio weight for stock, bonds and hedge funds on the y-axis. The standard deviation is shows on the x-axis. The picture on the right displays the mean return on the y-axis,
standard deviation on the x-axis and the skewness on the z-axis of the portfolio. Amin and Kat (2003).
The skewness3 increases as the share of hedge funds declines. Three differences are found when comparing the three-asset portfolio to a portfolio containing only equities and bonds. First, a higher mean is found in the three-asset portfolio for a given standard deviation, particularly when this standard deviation is low. Next, with the addition of hedge funds the portfolio skewness is reduced. Lastly, when share of the hedge funds within the portfolio is large, the kurtosis4 increases.
3 Skewness describes the asymmetry from the normal distribution and says something about the
length of its tails. A negative skewness implies longer left tails, this is the case for hedge funds.
4 A statistical measure that describes distribution around the mean; high kurtosis describes a
When considering portfolio allocation, higher losses in economic turbulence could come from ignored lower skewness and higher kurtosis. Generally speaking, when the equity markets are down, the hedge fund industry also suffers (Ebbing, 2009). Investors must weigh the choice of incorporating hedge funds to their portfolio by making a tradeoff between possible profits and losses.
In addition, Statman and Scheid (2008) confirm that hedge funds and equities share increased correlations with each other in times of economic disturbance. They found that hedge funds had approximately 90 percent correlations with the S&P500. Nevertheless, they also found that different asset classes, which were not correlated with the S&P500 before 2000, are now highly correlated because of the financial distress. Implying that all asset classes are increasingly correlated with equities in economic downturn. As Amin and Kat (2003) have shown, the inclusion of hedge funds within a portfolio can offer appealing mean-variance returns but increases the risk level. It eventually comes down to the investors risk appetite, rather than if it offers diversification benefits, to include hedge funds within the portfolio.
4.4.3 Shortcomings of portfolio allocation
Hedge funds are generally more complicated and riskier than traditional bonds and stocks. This leads to deficiencies with traditional portfolio allocation methods. The previous section has shown that hedge fund returns exhibit non normal distributions (skewness and kurtosis) because they also invest in illiquid assets. Traditional risk analysis, like Sharpe’s (1964) Capital Asset Pricing Model, which assumes linear asset payoff structures, should not be applied when analyzing the hedge funds industry (Agarwal and Naik, 2004).
Hedge fund strategies often exhibit non-linear option-like payoff structures. Agarwal and Naik (2004) find that the returns of a number of hedge fund strategies simulate put options on equity indices. This finding builds on previous research performed by Mitchell and Pulvino (2001) and Fung and Hsieh (2001), who found this resemblance for arbitrage and trend following strategies. The Multiple Objective Approach, developed by Lai (1991), is an option-based model that is able to capture the non-linear returns of hedge funds.
Agarwal and Naik (2004) explore conditional mean value at risk portfolios in order to analyze underlying hedge fund risk factors. These portfolios exhibit significant left tail risk. In their analysis Agarwal and Naik (2004), find that the use of previously mentioned traditional portfolio allocation methods (CAPM), underestimates potential losses in the tail by
54 percent. They argue that disregarding this tail risk can lead to excessive losses when the markets drop. Accordingly they appeal to the use of their model for portfolio allocation.
It is important for portfolio allocation to understand characteristics of hedge funds as well as the possible shortcomings of the methods being applied. This thesis will not seek to find optimal portfolios and therefor will not make use of any standard asset pricing models. Alternatively, by analyzing hedge fund performance, strategy risk factors are determined as well other risk related risk factors, as done by (Fung and Hsieh, 2004, 2008, Agarwal and Naik, 2004).
5. Data
5.1 Choice of database
Data concerning hedge fund strategies is obtained from the Credit Suisse Hedge Fund database, one of the first and largest data vendors in the industry. Hedge funds are private investment vehicles that were designed to avert regulation (Ackermann et al. 1999). For this reason, there is no agency that acts as a centrally regulated data vendor. The only way to obtain data on hedge fund performance is through independent private firms. Jaeger (2005) points out the divergences in hedge fund indices. The main difference is due to the fact that the databases have only a few funds in common, meaning that results may differ when using a different database. A hedge fund will only report their data to one data vendor, if it even reports anything at all.
The Credit Suisse Hedge Fund database is a database that supplies hedge fund strategies returns in the form of indices that have been developed to track these specific strategies’ performance. This database is a value-weighted hedge fund index that includes only funds compared to the use of separate accounts. Value-weighted indices, unlike equally weighted indices, more accurately represent investments because it automatically adjusts for corporate actions and changes in the share prices. This way the economic changes are reflected more accurately (Fung and Hsieh, 2004). This is good for examining hedge fund performance in times of economic turmoil and will be used in this thesis. The downsides in the use a value-weighted index, according to Fung and Hsieh (2004), are because of the unreliable reporting of the assets under management (AUM) by managers. A different problem is that large, well performing funds generally shut down because there is a maximum size for hedge funds. This will often times not be reported.
However, a different database like the Hedge Fund Research (HFR) database uses an equal weighting for the funds within its index. The use of this method makes all funds equal in size for the index, so that no fund has more influence than the other (Ebbing, 2009). In doing so a few problems arise. Often in the hedge fund industry a small portion of the funds manage the largest part of the assets. Fung and Hsieh (2004) note that this causes winning funds to have a relatively insignificant contribution to the index. Furthermore, there is a large instant history bias because new funds are more likely to perform well and poorly performing funds shut down.
Table 3. Performance 1994-2016.
Average monthly returns, volatility and extreme returns of hedge fund strategies and the S&P500 between 1994-2016. Results are calculated based on monthly Credit Suisse Hedge Fund Database indices.
Table 3 shows that there is a large variation between the average returns and the standard deviations of the hedge fund strategies and the S&P500 in the period 1994-2016. The monthly average return from the Global Macro strategy was 0.91% while the Dedicated Short Bias had a negative monthly average return of -0.35%. The risk of a specific hedge fund trading strategy is measured by its standard deviation. The lowest risk is observed in the Multi-Strategy trading strategy with 1.45%, while the highest risk is observed with the Dedicated Short Bias trading strategy with 4.73%. This is slightly higher than the standard deviation observed in the S&P500 index, 4.3%. The period 1994-2016 has been characterized by a number of major economic crises, which can explain the high volatility of the market index. Also a large variability is observed when looking at the distribution of extreme returns of the different strategies. The largest monthly decline was experienced by the Equity Market Neutral fund with -40.45%, followed by the S&P500 with 16.94%. The Dedicated Short Bias experienced the largest monthly gain with 22.71%.
5.2 Strategy Indices
This thesis makes use of the Credit Suisse Hedge Fund Broad Index5 to analyze the
risk factors for the specific strategies. The Credit Suisse Hedge Fund Index provides the indices, which is the largest value-weighted index provider. Broad trends and patterns among strategies are analyzed by using these indices, such as Adrian (2007).
Credit Suisse Hedge Fund Index holds a number of requirements that must be met in order to be included in the database. Funds must hold at least 50 million AUM, with a proven minimum one-year track record as well as currently audited financial statements (Credit Suisse). The database tracks over 9.000 funds worldwide and no Funds of Funds.
In an analysis of the risk adjusted returns it is of importance to look at the standard deviation as a measure of risk. Since monthly returns are used here, they have a higher accuracy compared annual returns, but less accurate than weekly returns. On the other hand, a disadvantage of using weekly data is that quarterly or yearly results are used in order to calculate incentive fees. Thus the use of weekly returns is according to Ackermann et al. (1999) is fairly arbitrary. Also, the use of returns before 1994 are not seen as useable, according to Liang (2000), because of the survivorship bias.
Table 4. Descriptive statistics Credit Suisse Hedge Fund Strategy Indices 2004-2010
Variables based on 74 monthly returns between 2004-2010. For each strategy the mean, median, minimum, maximum and standard deviation of the returns are given in percentages. (Own analysis of CSHF).
5.3 Seven Factor Model
Fung and Hsieh (2004) constructed a model to analyze risk and return differences for hedge fund strategies. This model links the risks that the hedge fund strategies are exposed too with traditional assets. Following the seven factor model set up, seven risk factors are used to analyze and measure the outperformance of individual hedge fund strategies. The risk factors are based on monthly data.
The first equity risk factor is the Standard and Poor’s 500 (S&P500) value weighted index. The S&P500 is the most commonly used benchmark of the United States market because it most accurately shows the risk and return characteristics (Fung and Hsieh, 2004, 2006, Agarwall and Naik, 2004). The S&P500 index data is obtained from the CRSP database.
The second risk factor pinpointed by Fung and Hsieh (2004) is the spread between small and large capitalization stock returns (SC-LC). This spread is calculated by looking at the difference between the small (Russel 1000 index) and S&P500 stock returns index. The data on the Russel 1000 index risk factors was obtained from Bloomberg and the S&P500 data obtained from CRSP.
There are two bond-oriented risk factors in the model. The first of these fixed income risk factors is the monthly change in the United States Federal Reserve 10-year constant maturity yield or as Fung and Hsieh (2004) name it the change in the 10-year Treasury yields. The second factor considered is the spread between the 10-year Treasury yields and the Moody’s Baa bonds. This spread is referred to as the credit spread. For a small portion of hedge funds, these fixed income factors play a major role. The data for both these fixed income factors are obtained from Federal Reserve Bank of St. Louis database for research.
Fung and Hsieh (2004) pinpoint three other risk factors, namely currencies, bonds and commodities. They model each of these trend-following risk factors as a portfolio of lookback straddles. This lookback straddle is a combination of call and put options. The options give the owners the right to buy (call option) or sell (put option) at the assets lowest respectively highest price. The origin of the pricing of the lookback option can be found in Goldman, Sosin and Gatto (1979). The lookback straddle is made up out of both the lookback call and put. The maximum payout a trend-following strategy can obtain is that of the lookback straddle (Fung and Hsieh, 2001). This data is provided by Fung and Hsieh themselves.
Table 5. Descriptive statistics seven risk factors.
All variables are based on 74 monthly returns between 2004-2010. RSP500 is the S&P500 risk factor, RSCLC is the Small Cap minus Large Cap stocks, R10YR is the 10-year constant maturity yield, RCRDSPRD is the credit
spread risk factor. RPTFSBD, RPTFSFX and RPTFSCOM are the trend following portfolios of the bonds, foreign exchange and the commodities respectively. (Own analysis of CSHF)
5.4 Biases
A number of questions can be asked when using strategy indices while analyzing hedge fund performance. There are for example large differences in the way managers execute their strategies because there are no clear requirements of strategy implementation. This could have as a consequence that the index contains very diverse strategies, making it more difficult to compare their performance (Ebbing, 2009). A way to minimize this theoretical issue is to use one database. In this thesis the database of Credit Suisse Hedge Funds will be used. This database uses has a number of requirements that must be met in order to be selected for inclusion in the strategy index. Furthermore, comparing only the indices from a single database bypasses the previously mentioned issue. Also a number of practical biases arise when using strategy indices while performing an analysis on hedge fund performance.
Hedge funds have a relatively short life, this means that in hedge fund databases there are new entrants as well as a number of failed or dead funds. The survivorship bias occurs when failed funds are omitted, leading to an upwards bias in the performance. Ackermann et al. (1999) found this bias to affect returns by 0,013 percent per month. Consequent research performed by Brown and Ibbotson (1999) found this affect to be larger, namely 0,25 percent. Likewise, Fung and Hsieh (2006) find similar results to Brown and Ibbotson (1999). Although the bias exists, its effect is minimal. A possible explanation is that successful funds, that are no longer open to new investors, oftentimes stop reporting to the database. This introduces a new bias, which partially neutralizes the survivorship bias, namely the
self-selection bias (Fung et al., 2008). When the industry matures however the survivorship bias
Agarwall and Naik (2000) conclude that there is no need to make any adjustments for the survivorship bias.
A new bias emerges when a new fund enters the database. Entering a database is a possible way to appeal to new investors. A manager will only proceed with getting his fund included if its returns are alluring for investors. If the fund already has a track record, parts or all of its historical performance data will be ex-post added to the database. This is called the
instant-history or backfill bias and the returns in the database are upwardly biased. Fung and
Hsieh (2006) find this bias to be approximately 1.5 percent per year. The Credit Suisse database only makes adjustments on a going forward basis, meaning that it does not backfill historical data in order to avoid this bias.
Ackermann et al. (1999) report that there is a possible bias when funds stop reporting to a database in the periods prior to its liquidation. This bias is therefor also called the
liquidation bias. This bias, according to Ackermann et al. (1999), is around 0.7 percent a year.
The Credit Suisse database is aware of this bias and in order to minimize it, the database doesn’t remove funds in the process of liquidation in order to capture its negative performance before it stops operating.
Fung and Hsieh (2004) devised a way to avoid the possible database biases by using the returns of Funds of Funds (FoFs). They find that the index data obtained from FoFs, which invest in multiple funds, have fewer biases than the index data on individual funds. Furthermore, by netting out due diligence costs and the cost of portfolio construction these FoFs best reflect actual investments. The reduction of the self-selection bias comes from the fact that when a FoF invests in a non-reporting fund, the performance is not affected in the index. Historical performance is included of liquidated funds in the FoFs in order to minimize the survivorship bias. The backfill bias disappears when looking at FoFs because the history is not included. The only thing included is the performance of the fund over the period in which the FoF has invested in it.
5.5 Hypotheses
Previous studies by Ackermann et al. (1999) and Liang (1999) look at the performance of hedge funds. More specifically they compared the performance of hedge funds to the performance of mutual funds. The outcome of these studies is that hedge funds generally outperform mutual funds, but often times perform worse than the market. Examining previous research and looking at strategy specific characteristics, expectations will be formulated about the performance and also outperformance (alpha) of strategies.
The first hypothesis looks at the outperformance of the individual hedge fund strategies prior to the subprime crisis.
1. Convertible Arbitrage: This strategy offers arbitrage opportunities to investors by
having the ability to hedge against stock price movements. It capitalizes on the price inefficiency between the companies’ stock price and the convertible security (Credit Suisse and Tremont, 2009). The period before the Lehman bankruptcy is known to be a bull period. Therefor the price inefficiency is expected to be higher and an outperformance in the period prior to the crisis is expected.
2. Equity Hedge: This strategy is known to perform well in both bull and bear market
conditions. It’s general large long exposure to the equity market (Fung and Hsieh, 2004) and the bull period preceding the crisis fuels the expectation of an outperformance in this period
3. Equity Market Neutral: This strategy makes use of sophisticated models to estimate
securities’ price movements and to apply the information in order find returns that are not correlated with overall market movements (Credit Suisse, 2009). Because it is a market neutral strategy it may not be able to take full advantage of the bull market environment, but an outperformance in this period is expected.
4. Event Driven: A strategy that looks to take advantage of corporate events such as
mergers, takeovers and reorganizations. In times of economic prosperity there is a lot of available capital set to use in the hope of promising future prospects. In these times there are a lot of mergers and acquisitions. For this reason an outperformance in this period is expected.
5. Dedicated Short Bias: A strategy that generally holds a net short position in
equities. This makes the bullish market environment preceding the Lehman Brothers bankruptcy a difficult one for this strategy. Therefor an underperformance is expected.
6. Emerging Market: Compared to the economic conditions in the Asia currency crisis,
the market fundamentals are healthy as the bullish market is driving up corporate profits and exceeding expectations, default rates at very low levels and trade surpluses (Credit Suisse and Tremont, 2007). Therefor an outperformance is expected.
7. Fixed Income Arbitrage: A strategy employed by the well-known Long Term
Capital Management and that attempts to take advantage of pricing differentials in the fixed income security market (Fung and Hsieh, 2002). A proven outperforming strategy and therefor an outperformance is expected.
8. Global Macro: This strategy has the most flexibility when it comes to investments
diversification benefits (Credit Suisse and Tremont, 2008). Therefor an outperformance is expected in this period.
9. Multi Strategy: A strategy that takes advantage of its diversification capabilities in
order to consistently deliver returns, regardless of the direction of the movements in the equity, interest rate or currency markets. (Credit Suisse and Tremont, 2008). The de-correlation and diversification lead to the expectation of an outperformance in this period.
10. Managed Futures: This strategy invests tends to invest in equity, bond and
commodity futures in currency markets around the world. It has a very low correlation with stocks and bonds. Therefor it offers a risk reduction within a portfolio. Historically this strategy has performed well when the market volatility was high (Credit Suisse and Tremont, 2008). In the period before the Lehman Brothers bankruptcy an underperformance is expected.
Table 6. Period before crisis (2004-2007)
Null (H0) and alternative (H1) hypothesis for outperformance of all strategies. If H0:𝛼 > 0 is indicated in bold an
outperformance of the hedge fund strategy compared to the market is expected for the period 09/2004-09/2007. If H0: 𝛼 ≤ 0, no outperformance of the hedge fund strategy is expected.
The second hypothesis looks at the outperformance of the hedge fund strategies after the bursting of the subprime bubble.
1. Convertible Arbitrage: The ability to short sell plays an important role for this
strategy. The ban on short selling in September and October of 2007 will have a large impact on this strategy (Credit Suisse and Tremont, 2008). Also market risk (especially yields and debt duration) and credit risk exposure will make it difficult for this strategy. Credit Suisse and Tremont (2008) observe a flattening of the yield curve due to fears of inflation as well as direct credit exposure. Therefor an underperformance is expected in this period.
2. Equity Hedge: The majority of this hedge fund strategy performance can be
explained by the exposure to the stock market (equities) and the spread between small and
Strategy 𝐻! 𝐻!
1. Convertible Arbitrage 𝛂 > 𝟎 α ≤ 0
2. Dedicated Short Bias α ≤ 0 α > 0
3. Emerging Market 𝛂 > 𝟎 α ≤ 0
4. Equity Market Neutral 𝛂 > 𝟎 α ≤ 0
5. Event Driven 𝛂 > 𝟎 α ≤ 0
6. Fixed Income Arbitrage 𝛂 > 𝟎 α ≤ 0
7. Equity Hedge 𝛂 > 𝟎 α ≤ 0
8. Global Macro 𝛂 > 𝟎 α ≤ 0
9. Managed Futures α ≤ 0 α > 0
