Alternative-Investment Mutual Funds, Part 1: Sources of Return

For more than a decade hedge fund practitioners and academic research have shown that the main part of hedge funds’ returns corresponds to their capturing of risk premia (beta) rather than market inefficiencies (alpha). This understanding has led to various implications for the hedge fund industry and for investors. Some firms have been led to construct investable benchmarks for hedge funds on the basis of an analysis of the underlying risk factors and a subsequent replication of the corresponding risk premia. In this article we apply the same techniques to analyze alpha versus beta returns for alternative-investment mutual funds, e.g., long/short equity funds and multi-strategy funds.

We follow the work of Fung and Hsieh[1] and Jaeger and Wagner[2] in choosing risk factors (hereafter, asset-based style [ABS] factors) and we also use their statistical techniques. We will discuss the economic reasons certain factors are chosen, but will not discuss the statistical techniques used. The interested reader is referred to the articles noted above to learn more about the statistical methods.

ABS factors have interesting characteristics. They are investible products such as the S&P 500 and the GSCI and are therefore bias-free because only market prices are used. We do not use alternative-investment fund returns in constructing ABS factor returns so we can translate alternative-investment fund risk into a conventional setting. In other words, we can take alternative-investment fund risk and directly relate it to the risk of investable securities such as stocks and bonds—conventional asset classes with which investors are more familiar. The use of ABS factors, then, brings alternative-investment fund risk factors much closer to conventional risk factors.

In terms of performance evaluation ABS factors allow us to answer the following questions. Let’s suppose alternative-investment fund returns are not risk-free. Few if any of us are willing to believe that alternative alpha is risk-free. What then are the risks inside alternative alphas? The answer to this question tells us about the systematic risk factors inherent in alternative-investment fund strategies. Put differently, it tells us about alternative-investment fund betas. In terms of risk management, through these ABS factors we can relate alternative-investment fund risks to risks of conventional assets. Finally, in terms of measuring the expected risk and return, we can now directly relate the assessment of risk and return on alternative-investment funds in a framework consistent with our opinions on such traditional investments as stocks, bonds, and interest rates.

The ABS factors we chose are in Table 1.  Our choice of factors is based on the work of the authors above. We note to the right of each instrument the function/risk the factor covers.

 Table 1. Asset-Based Style Factors

For both the Value factor and Size factor exposures, we use Ken French’s data for HML and SMB respectively.  We did so because neither of the MSCI ETFs has 5 years of history.  For the Credit Risk factor calculation, we used the 7-10 Year Treasury Bond (IEF) and the Investment Grade Corporate Bond ETF (LQD)  Finally, for the trend following factor, we used the Newedge index as the S&P Systematic Global Macro ETF (SGMI) does not have the necessary five years of weekly history. And except for the Gabelli fund, all other risk factors are either ETFs or futures-based.

Our data set is the weekly returns for the past five years for each of the factors above and the Lipper alternative-investment classifications: Absolute Return, Alternative Credit, Equity Market-Neutral, Event-Driven, Global Macro, Long/Short Equity, and Multi-Strategy. Managed Futures is not included, since only one managed futures mutual fund has five years of history.  All our data comes from either Datastream or Lipper for Investment Management. 

To give a brief preview of our results, risk factors can explain from 53% to 79% of the alternative-investment weekly return variance. For six of the seven fund classifications (Absolute Return, Alternative Credit, Event-Driven, Global Macro, Long/Short Equity, and Multi-Strategy) the risk factors explain 69%-79% of the return variance.  And we can tentatively say that only the Equity Market Neutral classification has a statistically significant positive alpha.  All the other alternative-investment classifications have an alpha of zero.  This is in comparison to many long-only mutual funds, which typically have negative alphas.

In the next article we will review the results of our alternative-investment fund risk decomposition in more detail.