Neutrality of equity market neutral strategy to period and index
Authors
- First Online:
- Received:
- Revised:
DOI: 10.1057/jdhf.2012.13
- Cite this article as:
- Muhtaseb, M. & Colborn, R. J Deriv Hedge Funds (2012) 18: 301. doi:10.1057/jdhf.2012.13
- 65 Views
Abstract
This study examines the performance of equity market neutral (EMN) hedge funds by estimating alpha and beta, using monthly excess returns of the Fidelity Investments Spartan^{®} US Bond Index Fund, the S&P 500 index and two indices of EMN strategies. The study investigates how much exposure to systematic risk is experienced by EMN hedge funds and hence by the investors in such funds. In addition, the article ascertains the stability of the estimated betas of EMN hedge fund indices. Results show that EMN fund indices had no systematic (stock and bond market) risk before the subprime mortgage crises of 2008. During the crises, the DJ Credit Suisse EMN index showed a positive beta in excess of 0.4, yet the Greenwich Van Hedge EMN index showed a beta less than 0.1. Similarly, the DJ Credit Suisse EMN index had no bond market beta up until November 2008. The Greenwich Van Hedge EMN index had almost no exposure to the market. The results were sensitive to market conditions and the EMN index. DJCS and GVH both had positive stock market alphas, which turned insignificant around October 2008. Similarly, both indices had positive bond market alphas, which turned insignificant around August–November 2008.
Keywords
Equity Market NeutralperformanceindicesINTRODUCTION
The objective of this study is to investigate the performance behavior characteristics of indices representing the equity market neutral (EMN) hedge fund strategy over the period January 1995 – January 2012. The focus is on analyzing the exposure that EMN hedge funds have to US equity and bond markets.
The EMN strategy is to have minimal and ideally no exposure to the risk generated by the market. Its objective is to provide returns that are neither aided nor hindered by movements in the market. This does not mean that EMN strategies are not volatile or without absolute risk, but rather that the volatility is not due to volatility in the market. In addition, EMN funds claim to produce returns that have less risk than the returns generated by the market, and in fact, in terms of risk, are more like bonds rather than equities.
Equity market neutral historical growth AUM
Year | $ billions |
---|---|
1997 | 3.02 |
1998 | 5.75 |
1999 | 7.54 |
2000 | 11.41 |
2001 | 14.30 |
2002 | 14.39 |
2003 | 23.83 |
2004 | 26.98 |
2005 | 31.91 |
2006 | 42.98 |
2007 | 48.74 |
2008 | 32.60 |
2009 | 33.30 |
2010 | 30.10 |
1st qtr 2011 | 32.93 |
2nd qtr 2011 | 31.80 |
3rd qtr 2011 | 31.74 |
4th qtr 2011 | 28.12 |
1st qtr 2012 | 28.54 |
By using an EMN hedge fund index as a proxy for the performance of a ‘normal’ or ‘garden variety’ EMN hedge fund and the S&P 500 index as a proxy for the broad-based US equity markets, the beta for the EMN hedge fund is estimated using simple linear regression. The data cover the period January 1995 –October 2012, which encompasses periods of bull, bear and crises markets. The betas of two independent EMN hedge fund indices, Dow Jones Credit Suisse and Greenwich Van Hedge, are estimated by regressing each index's returns against the monthly returns of the S&P 500 index. If the beta for the EMN hedge fund indices is close to or equal to zero, then by definition we conclude that exposure to systematic risk is not present or is negligible. Linear regression on the data set over the entire period shows a static snapshot of the EMN indices’ beta estimate over a 17-year period. However, when beta is estimated for the indices using monthly returns for rolling 60-month periods, a more dynamic situation emerges. A critical outcome is the stability of betas of the indices over time.
This article also analyzes the claim that EMN hedge funds produce equity-like (level of) returns with bond-like risks (standard deviation). By using the Fidelity Investments Spartan® US Bond Index Fund^{1} as a proxy for the US bond market, comparison of the two indices highlights the similarities and differences.
The next section presents literature review. The hypotheses for market risk neutrality and performance are developed in the section after that. The subsequent section presents both the methodology and the data used for this research. The empirical results are discussed in the penultimate section. The final section discusses the conclusions drawn from the research.
LITERATURE REVIEW
Research focused on hedge fund performance generally falls into a few broad categories. The first category encompasses studies comparing and contrasting hedge fund performance to that of various indices, such as Ackermann et al (1999), Liang (1999) and Capocci and Hubner (2004). The second category offers comparisons of hedge fund performance to that of mutual funds, as studied by Ackermann et al (1999) and Liang (1999), who conclude that hedge funds, as a group, perform better than mutual funds on a risk-adjusted basis. Finally, research to analyze, define and categorize hedge fund investment styles and performance. Edwards and Caglayan (2001) find that market neutral hedge fund strategies as well as event-driven and macro hedge fund strategies are the only ones to provide protection to investors in down trending markets.Ennis and Sebastian (2003) find that after the stock market peaked in March of 2000 hedge funds in general did not provide protection to investors; rather it was the managers’ good timing of the market. Muhtaseb (2003 and 2006) discusses the merits of hedge funds like investing versus long only investing.
The strategy aims at balancing long and short positions to ensure zero or negligible market exposure and consequently a performance pattern that is independent of market movements (Lhabitant, 2002). As more factors are hedged away, the opportunity set for the manager to add value is reduced. In theory, a perfectly hedged portfolio should yield the risk-free rate minus transaction costs. In practice, as the portfolio is not hedged against some residual factors, the strategy could yield returns above the T-bills, but with much less risk than unhedged positions and with a potentially lower correlation to the market (Lhabitant, 2002). EMN hedge funds have provided returns approximately equal to those of broad indexes while assuming much less risk than a portfolio of common stocks (McCrary, 2005). Market neutral investing has a major disadvantage with respect to other long/short strategies. It requires an increased number of transactions to maintain the hedged (no exposure to the market) position. These transactions are costly, and they have potential consequences in terms of undesirable tax effects for hedge fund investors (Lhabitant, 2002).
There are several motivations for this research. First, several studies suggest that market neutral hedge funds represent a large percentage of the hedge funds operating in the industry. Capocci and Hubner (2004) and Capocci et al (2003) report that as much as 28.3 per cent of hedge funds employ a market neutral strategy. Second, most hedge funds in general do have exposure to the market. Results from several research articles that focus on all hedge funds as a group may not fully apply to EMN hedge funds or other individual strategies. EMN hedge funds have a specific objective, which is to deliver alpha while simultaneously hedging (removing) the fund's exposure to the general market. Capocci et al (2003) studied a database of 2894 individual funds and proposed that EMN hedge funds should be further studied due to their unique and interesting characteristics. A growing number of researchers are looking at individual hedge fund strategies for the focus of their research in order to understand the unique characteristics of the various individual strategies. Fung and Hsieh (2002) focus on fixed income arbitrage funds in their research, Mitchell and Pulvino (2001) concentrate their studies on risk arbitrage, and Gatev et al (1999) study relative value hedge funds on a stand-alone basis. Further investigation of the EMN hedge fund strategy on a stand-alone basis provides for a greater understanding of the strategy's individual particularities. Navone (2001) looked at the diversification benefits of the addition of EMN hedge funds to a portfolio of mutual funds and found that the inclusion of EMN hedge funds in a portfolio would significantly improve the risk-return tradeoff, but did not look into the neutrality of the funds. Finally, few researchers have studied the performance of hedge funds through bear markets. The 17-year period of data employed in this study includes the bear markets starting in March 2000 caused by the bursting of the Internet bubble, the post-9/11 market and the collapse of the sub-prime markets in the later part of 2007.
Brooks and Kat (2001) found that EMN hedge funds did not exhibit a high correlation with the equity markets. Other strategies showed a low and typically negative correlation to the bond markets, but EMN funds were an exception. They suggest that this could be due in part to the high degree of leverage employed in the EMN strategy. They also report that the correlations of all hedge fund strategies among each other are high in all but the EMN strategy, reasoning that other strategies share a high degree of similarity of exposure to systematic risk. In addition, EMN funds exhibit standard deviations that are lower than the bond markets but mean returns that are higher.
HYPOTHESES
Hypothesis 1:
EMN hedge funds are market neutral, and the neutrality is stable
where R_{EMN}=rate of return from the EMN index; R_{Free}=riskless rate of return proxied here by the T-bill rate; β=beta or systematic risk; R_{Index}=rate of return from the S&P 500 index or bond market index; and α=alpha or the manager's value added; if β=0 then the excess returns will equal alpha plus an error term. R_{EMN}=α+e; e=error term.
Therefore, EMN hedge funds would aim to have a beta that is as close to zero as possible. If the beta is zero, implying that the EMN funds do not earn returns by taking on market exposure, then the return that remains is alpha, which reflects the portfolio manager's ability to earn a return independent of the overall market movement.
Simple linear regression, run on the available monthly returns over the 17-year period of this study, shows a snapshot of the fund's beta over this time frame. The results (single regression model) of the entire 17-year period model are of limited value because they do not show how the beta of the strategy changes over time.
Hypothesis 2:
EMN hedge funds provide better returns than the broad-based US equity market indices
EMN hedge funds claim to provide returns that are better than the broad market indices. To evaluate this claim, the returns of the S&P 500 index and the Spartan US Bond Index Fund are compared on an absolute and risk-adjusted basis with those of the DJ Credit Suisse and Greenwich Van Hedge EMN indices.
Hypothesis 3:
EMN hedge funds exhibit risk characteristics that are closer to those of bonds than equities
EMN fund managers assert that the level of risk to which investors in their funds are exposed is comparable to that of the bond market more so than that of the stock market. By regressing the monthly returns of each of the two EMN indices against the monthly returns of the Spartan US Bond Index Fund, we estimate the bond market beta for each of the two EMN hedge fund indices. If EMN hedge funds were to have bond-like risk then this would be reflected in the significance of the beta estimate.
RESEARCH METHODS
The data^{2}
The EMN hedge fund index monthly returns are obtained from Greenwich Van Hedge Alternative Investments L.L.C. (www.greenwichai.com) and Dow Jones Credit Suisse (www.hedgeindex.com). The returns of both the Greenwich Van Hedge and Dow Jones Credit Suisse indices are updated and published monthly for the period January 1995 – January 2012. These indices are limited to hedge funds domiciled in the US and are reported ‘net of fees’. The EMN index returns are evaluated against the S&P 500 monthly returns and the Spartan US Bond Index Fund.
The S&P 500 index monthly returns for the period January 1995 – January 2012, the monthly returns of a 1-year T-bill, and the monthly returns of the Spartan US Bond Index Fund over the period July 1996 – October 2012 are used to observe the risk similarities with the EMN indices. These represent the returns of the broad US equity and bond markets and the risk-free rate of return. This time frame encompasses periods of both bull and bear markets, providing the opportunity to observe performance characteristics of the EMN strategy over the short-term, long-term, bear and bull markets.
This research uses the S&P 500 index as a proxy for the returns of the US stock market. The S&P 500 index and 1-year Treasury bill monthly rates of return are obtained from www.standardandpoors.com and the Federal Reserve (www.treasury.gov/resource-center/data-chart-center/interest-rates/Pages/TextView.aspx?data=billrates), respectively. The data for the Spartan US Bond Index Fund are available on www.Fidelity.com.
Risk exposure and measurement
Beta
Therefore, EMN hedge funds would aim to have a beta that is as close to zero as possible. If the beta is zero then the return that remains is alpha or the portfolio manager's ability to identify and exploit mispricing, timing and varying exposure opportunities.
Sharpe ratio
Developed by William F. Sharpe in 1964, the Sharpe ratio measures the level of excess reward per unit of total risk taken (Sharpe, 1964). Originally dubbed the reward-to-variability ratio by Sharpe, the ratio gained in popularity but the name did not, and it has since become known as the Sharpe ratio.
where: x is the index or fund; R_{x} is the average monthly rate of return of x; R_{f} is the best available rate of return of a ‘risk-free’ security (that is cash); and StdDev(x) is the standard deviation of r_{x}
Therefore, using the Sharpe ratio we can standardize the reward to risk of the various indices to determine the level of risk involved. Generally, a Sharpe ratio over 1 is considered good.
Alpha
EMN hedge funds should produce positive alpha demonstrating the returns that are attributed to the manager's performance and not the market's influence.
Methodology
To test Hypothesis 1, the beta of EMN hedge funds are estimated against the monthly returns of the S&P 500 index over the 17-year period, January 1995 – January 2012. This research estimates the beta of the EMN hedge funds over rolling 60-month periods in order to determine how stable EMN funds are in terms of market risk exposure, as well as to observe the behavior in changing market conditions. Monthly returns are employed to calculate 60-month rolling betas.
Hypothesis 2 is tested by estimating the beta, alpha and standard deviation of the EMN hedge fund indices. This gives three different measures with which to evaluate the performance of EMN hedge funds against the broad market index to determine whether the funds in fact produce better returns than the broad stock market index.
Hypothesis 3 is tested by using the same regression methods as above, but instead the EMN indices risk measures are compared to those of the Spartan US Bond Index Fund. As the index fund used is the aggregate of all US bonds, there are investment grades, as well as high-yield or junk bonds, included in the index. Alpha and beta estimates over rolling 60-month periods as well as the returns of the EMN hedge funds are compared to those of a bonds index.
EMPIRICAL RESULTS
Regression analysis for the Dow Jones Credit Suisse index against the S&P 500 index using excess returns for all data points
Summary output | DJCS to S&P excess returns |
---|---|
Regression statistics | |
Multiple R | 0.292817853 |
R^{2} | 0.085742295 |
Adjusted R^{2} | 0.081238563 |
Standard error | 0.029260456 |
Observations | 205 |
ANOVA | |||||
---|---|---|---|---|---|
df | SS | MS | F | Significance F | |
Regression | 1 | 0.016299989 | 0.01629989 | 19.03805225 | 2.03842E-05 |
Residual | 203 | 0.173803376 | 0.000856174 | — | — |
Total | 204 | 0.1901033 | — | — | — |
Coefficients | Standard error | t-stat | P-value | Lower 95% | Upper 95% | |
---|---|---|---|---|---|---|
Intercept | 0.001669293 | 0.002049508 | 0.814484589 | 0.416321839 | −0.002371761 | 0.005710346 |
X variable 1 | 0.194818739 | 0.044649795 | 4.363261652 | 2.03842E-05 | 0.106781896 | 0.282855582 |
Regression analysis for the Greenwich Van Hedge index against the S&P 500 index using excess returns for all data points
Summary output | GVH to S&P excess returns |
---|---|
Regression statistics | |
Multiple R | 0.141080911 |
R^{2} | 0.019903823 |
Adjusted R–square | 0.015075763 |
Standard error | 0.011518044 |
Observations | 205 |
ANOVA | |||||
---|---|---|---|---|---|
df | SS | MS | F | Significance F | |
Regression | 1 | 0.000546917 | 0.000546917 | 4.122530255 | 0.043620141 |
Residual | 203 | 0.026931063 | 0.000132665 | — | — |
Total | 204 | 0.02747798 | — | — | — |
Coefficients | Standard error | t-stat | P-value | Lower 95% | Upper 95% | |
---|---|---|---|---|---|---|
Intercept | 0.004533163 | 0.000806578 | 5.620240112 | 6.24225E-08 | 0.002942818 | 0.006123509 |
X variable 1 | 0.035781197 | 0.01762272 | 2.030401501 | 0.043620141 | 0.001034148 | 0.070528246 |
These beta estimates are significant with a t-stat of 4.36 for DJ Credit Suisse and 2.03 for Greenwich Van Hedge. The R-squared is 8.5 per cent for DJCS and 2.0 per cent for GVH, indicating that more than 90 per cent of the risk of the equity market strategy index is not accounted for by the overall stock market movement. Over the long term, EMN as a strategy managed to deliver returns that have minute exposure to the stock market.
Tables 2 and 3 also illustrate alpha estimates for both the DJ Credit Suisse and the Greenwich Van Hedge indices. The alpha estimate for the DJ Credit Suisse index is not significant, yet the alpha of the Greenwich Van Hedge index is very significant. For the DJ Credit Suisse Index the estimated alpha is almost zero and the Greenwich Van Hedge index alpha is estimated at 0.45 per cent per month (t-stat is 5.62). This translates into an annualized alpha estimate of 5.4 per cent. On the basis of the results of Greenwich Van Hedge estimates, EMN hedge fund managers are producing positive annual returns to investors independent of the stock market's performance and in excess of the risk-free rate of return.
Return comparisons
Index returns for the period January 1995 – January 2012. Spartan US Bond Index Fund data cover the period June 1996 – January 2012
Total returns from January 1995 through January 2012 | ||||
---|---|---|---|---|
Data set | Average annual return (%) | Annual SD (%) | Average monthly return (%) | Monthly SD (%) |
Greenwich/Van Hedge Equity Market Neutral Index | 9.68 | 4.72 | 0.81 | 1.36 |
Dow Jones Credit Suisse Equity Market Neutral Index | 6.10 | 10.67 | 0.51 | 3.08 |
S&P 500 Index | 7.45 | 15.93 | 0.62 | 4.60 |
Spartan US Bond Index Fund | 6.04 | 3.51 | 0.50 | 1.01 |
Risk-free rate | 3.28 | 0.58 | 0.27 | 0.17 |
On a non-risk-adjusted basis, the GVH index outperformed the stock and bond markets. The DJCS index under-performed the stock market and performed as well as the bond market, over the period.
Excess returns comparison among all indices. Spartan US Bond Index Fund data covers the period June 1996 – January 2012
Data set | Average annual return (%) | Annual SD (%) | Average monthly return (%) | Monthly SD (%) |
---|---|---|---|---|
Greenwich/Van Hedge Equity Market Neutral Index | 5.58 | 4.02 | 0.47 | 1.16 |
Dow Jones Credit Suisse Equity Market Neutral Index | 2.81 | 10.57 | 0.23 | 3.05 |
S&P 500 Index | 4.17 | 15.89 | 0.35 | 4.59 |
Spartan US Bond Index Fund | 3.54 | 3.55 | 0.25 | 1.03 |
Risk-free rate | 3.28 | 0.58 | 0.27 | 0.17 |
Correlations of EMN indices with the S&P500 index and Spartan US Bond Index Fund
DJCS index | GVH index | S&P 500 | Spartan US Bond Index Fund | Risk-free rate | |
---|---|---|---|---|---|
Correlation between indices monthly excess returns | |||||
DJCS index | 1.0000 | 0.1890 | 0.2928 | −0.2206 | 0.1517 |
GVH index | — | 1.0000 | 0.1402 | 0.0886 | 0.2975 |
S&P 500 | — | — | 1.0000 | −0.2206 | −0.2332 |
Spartan US Bond Index Fund | — | — | — | 1.0000 | −0.1675 |
Risk-free rate | — | — | — | — | 1.0000 |
Interestingly, both EMN indices enjoy higher correlations with the risk-free rate. This reflects the common practice of EMN hedge fund managers of targeting a return objective expressed as cash rate (risk-free rate) plus an incremental rate of return (risk premium).
Table of confidence intervals for the mean excess returns both monthly and annualized
Data set | Annualized range of returns (95% confidence) (%) | Range of returns (%) | Monthly range of returns (95% confidence) (%) | Range of returns (%) | ||
---|---|---|---|---|---|---|
Confidence interval of excess returns | ||||||
Greenwich/Van Hedge equity market neutral index | 5.03 | 6.13 | 1.10 | 0.31 | 0.62 | 0.32 |
Credit Suisse/Tremont equity market neutral index | 1.37 | 4.26 | 2.90 | −0.18 | 0.65 | 0.84 |
S&P 500 index | 1.99 | 6.34 | 4.35 | −0.28 | 0.98 | 1.26 |
Spartan US Bond Index Fund | 2.96 | 4.12 | 1.16 | 0.08 | 0.42 | 0.34 |
Risk-free rate | 3.06 | 3.22 | 0.16 | 0.24 | 0.28 | 0.05 |
The ranges of excess returns of EMN hedge fund indices are greater than those of the bond market index yet smaller than those of the stock market index. The return ranges present a similar picture to that of the Sharpe ratios. The excess returns show that in general EMN strategy delivers returns at least as high as those of the stock market but with significantly lower risk, more comparable to that of the bond market.
Neutrality to the bond market
Regression analysis for the Dow Jones Credit Suisse index versus the Spartan US Bond Index Fund using excess returns for the entire period
Summary output | DJCS to fidelity excess returns |
---|---|
Regression statistics | |
Multiple R | 0.226260881 |
R^{2} | 0.051193986 |
Adjusted R^{2} | 0.046037432 |
Standard error | 0.031186107 |
Observations | 186 |
ANOVA | |||||
---|---|---|---|---|---|
Df | SS | MS | F | Significance F | |
Regression | 1 | 0.0096557 | 0.0096557 | 9.927945 | 0.001900494 |
Residual | 184 | 0.1789535 | 0.0009726 | — | — |
Total | 185 | 0.1886091 | — | — | — |
Coefficients | Standard error | t-stat | P-value | Lower 95% | Upper 95% | |
---|---|---|---|---|---|---|
Intercept | 0.003792797 | 0.0023549 | 1.6106307 | 0.108974 | −0.000853186 | 0.0084388 |
X variable 1 | −0.709910772 | 0.2253067 | −3.150864 | 0.0019 | −1.154427483 | −0.2653941 |
Regression analysis for the Greenwich Van Hedge index against the Spartan US Bond Index Fund using excess returns for the entire period
Summary output | GVH to fidelity excess returns |
---|---|
Regression statistics | |
Multiple R | 0.019890768 |
R-square | 0.000395643 |
Adjusted R-square | −0.00503699 |
Standard error | 0.013007284 |
Observations | 186 |
ANOVA | |||||
---|---|---|---|---|---|
df | SS | MS | F | Significance F | |
Regression | 1 | 1.23216E-05 | 1.232E-05 | 0.0728271 | 0.78756662 |
Residual | 184 | 0.031130858 | 0.0001692 | — | — |
Total | 185 | 0.031143179 | — | — | — |
Coefficients | Standard error | t-stat | P-value | Lower 95% | Upper 95% | |
---|---|---|---|---|---|---|
Intercept | 0.004433368 | 0.000982176 | 4.5138244 | 1.134E-05 | 0.002495594 | 0.0063711 |
X variable 1 | 0.025359807 | 0.093972237 | 0.2698649 | 0.7875666 | −0.160041829 | 0.2107614 |
Again, the alpha estimates of the two EMN indices are not consistent. For the entire period, the DJ Credit Suisse index produced a monthly alpha estimate of 0.0038 or 0.38 per cent (t-stat 1.6) versus the Spartan US Bond Index Fund. As reported in Appendix B, the 60-month rolling alphas were positive and significant through October 2008. Results for the Greenwich Van Hedge index against the Spartan US Bond Index Fund produced monthly alpha estimates of 0.44 per cent (t-stat 4.51) monthly or 5.28 per cent annually. The alpha is sensitive to the prevailing market conditions. As shown in Table B1, beginning in August 2008, the 60-month rolling alpha lost statistical significance. Before August 2008, all the alphas of both EMN indices were statistically significant and sizable, ranging from 0.5 per cent to 1.0 per cent per month.
SUMMARY AND CONCLUSIONS
The primary objective of this article is to determine whether hedge funds’ claim of market neutrality and independence from market movements is true. This study also investigates the ability of EMN hedge funds to produce returns that are higher than the (stock and bond) market index, along with risk characteristics similar to those of a bond market index. The time span of the study encompasses both bull and bear markets, including the bursting of the dot com bubble, the post-September 11, 2001 markets, and most recently and ongoing the sub-prime mortgage meltdown.
The forte of EMN hedge funds is to remain neutral and independent from broad market influences. Estimating rolling monthly betas allows for a better understanding of the effects of market changes on the funds and the exposure as the market changes. Examining the neutrality of EMN hedge funds reveals that the EMN hedge fund strategies, over the long term, maintain an independent exposure to the (bond and stock) market up until the beginning of the economic recession (September 2008) triggered by the subprime mortgage crises.
The exposure that is observed is in certain periods statistically significant, yet very small, as shown in Tables A1 and B1. Before the subprime mortgage crises, rarely was the estimated stock of DJCS beta greater than 0.10. In November 2008, the stock market beta went up from 0.08 to 0.42. DJCS stock market beta did not decline subsequently, yet the beta of GVH returned to almost zero within 6 months and eventually turned statistically insignificant. Similarly, the bond market betas demonstrated comparable shifts in exposure. Dow Jones Credit Suisse 60-month rolling bond market beta went up from an insignificant 0.0051 in November 2008 to a huge and very significant −1.96 and stayed at that level through January 2012. Interestingly, Greenwich Van Hedge bond market beta was insignificant throughout the period, except in September 2008 when the beta was 0.33 and significant.
These results send an alert to EMN strategy investors. Although the strategy maintains a neutral exposure to the stock market, the neutrality diminishes significantly during a crises market environment, based upon the DJCS index. EMN strategy exposure to the bond market is also sensitive to the EMN index selected. Investors must carefully vet an EMN fund and the EMN index selected to evaluate risk exposures. The results do vary across the two indices, DJCS and GVH. Investors are advised to carefully evaluate the construction and constitution methods of the available indices to make sure that the indices conform to the true structure of an EMN strategy.
Investors with portfolios consisting solely of EMN hedge funds have experienced returns that on average outperformed the S&P 500 index, over the period of the study, based on the results of the GVH index. The DJCS index did not outperform the S&P 500 index on an absolute basis. On a risk-adjusted basis, both indices performed as well as the S&P 500 index (DJCS) or better (GVH).
Using monthly and annual percentage returns, this study also shows that EMN hedge funds do in fact perform better than the broad market over the entire period of study. This does not necessarily mean that an investment in an EMN hedge fund will generate returns that beat the market in any single year. The overall advantage of EMN hedge funds is that in periods of bear markets they produce absolute returns that are less negative than the market index. It is for this reason that investment in these funds should be for longer holding periods. EMN hedge funds are better utilized as an addition to a diversified portfolio to aid in periods when the market conditions are unstable.
In conclusion, the EMN hedge fund strategy does maintain a neutral or negligible exposure to systematic risk during normal and economic crises-free periods. The neutrality fluctuates as the market conditions change and the exposure to market risk increases during bear market periods. Although fund managers are producing positive alpha, the alpha produced decreased over the period studied. Stock market alphas turned negative and insignificant for DJCS and small yet positive for GVH. These findings agree with those of Capocci et al (2003) that EMN funds produce alpha while simultaneously hedging the exposure to the general market. Standard deviations of the EMN indices’ monthly excess returns are split above and below that of the bond market, which contrasts with the findings of Brooks and Kat (2001) that EMN hedge funds have higher excess returns and lower standard deviations than the bond markets. The EMN monthly excess returns are higher than those of the bond index, in line with the findings of Brooks and Kat (2001). Our results confirm and extend those of Lo and Mueller (2010), which demonstrate large returns during the mid-1990s that significantly decline over time. The average returns and Sharpe ratios of statistical arbitrage decline steadily from 1995 through 2007.
Alphas and Betas of DJCS and GVH against S&P500 index and Spartan US Bond Index Fund
Index | Betas exposures to Stock and Bond Markets | |
---|---|---|
Stock Market | Bond Market | |
DJCS | 0.195** | −0.71** |
GVH | 0.036** | 0.0264 |
Index | Alphas against Stock and Bond Market Indices | |
DJCS | 0.17% | 0.38% |
GVH | 0.45%** | 0.44%** |