George A. Martin

Understanding Hedge Fund Performance: Research Issues Revisited—Part II

In this article, the authors briefly review the potential market factors affecting various hedge fund strategies. The market factors affecting hedge fund returns are based on the underlying securities held and the trading philosophy behind each strategy. A simplified multi-factor model is used to capture the market factors driving various hedge fund strategies, and the actual performances of various hedge fund strategies relative to these market based factors are presented. The market factors are also used to illustrate the relative performance of various hedge fund strategies conditional on extreme values of these market factors. The empirical relationship between the measured market factors and various hedge fund strategies over time is also explored. Results show that while for style pure indices of hedge fund performance the factor relationships remain fairly constant, for diversified funds of funds, which may dramatically change their strategy, the factor loadings that explain fund of fund returns change. These results imply that for funds of funds, which emphasize market timing, their benefit in terms of systematic asset allocation is also time varying.

Understanding Hedge Fund Performance

This article is the first of a two-part review of various micro (fund) and macro (strategy) factors that impact fund performance. In this part, the authors concentrate on reviewing some of the micro (e.g., fund level) elements driving fund performance. Specifically, they analyze the impact of fund characteristics such as survivor bias, age, and size on the funds performance. In some cases, the results presented in this article seem to question previous results such as the impact of stale prices on return relationships, survivor bias, the use of historical hedge fund index data, and incentive fees. In other cases, the results support previous research on issues such as performance persistence, lock-ups, and a number of other micro/fund effects.

The Benefits of Hedge Funds: Asset Allocation for the Institutional Investor

Academic and practitioner research has previously focused on the performance of various hedge fund strategies as stand-alone investments and as part of traditional stock and bond portfolios. In this article evidence is presented on the actual drivers of return of the various strategies and how traditional style-based performance analysis and asset allocation frameworks (e.g., mean/variance return/risk optimization) can be used to determine the appropriate allocation to hedge funds.

The Impact of Leverage on Hedge Fund Risk and Return

As in the case of traditional investments, hedge funds are often compared on a risk adjusted basis. Risk adjustment is of particular importance for hedge fund analysis since two funds may differ solely by leverage, such that they may differ on an absolute return basis, but be similar on a risk adjusted basis. While leverage should theoretically not affect the level of risk adjusted return within a strategy, it is possible that funds which use higher levels of leverage may in fact trade differently than funds using lower levels of leverage. In this article, the effect of leverage on hedge fund risk and return is analyzed. In brief, results are presented on the level of leverage used in various hedge fund strategies. Results are also provided on the degree to which leverage above or below the median fund leverage results in superior or inferior risk adjusted performance within a particular hedge fund strategy. Overall the results show that while different hedge fund strategies may use different amounts of leverage, within a particular hedge fund strategy, there is little evidence of a significant difference between risk adjusted performance of funds with above-median and below-median leverage.

The Long-Horizon Benefits of Traditional and New Real Assets in the Institutional Portfolio

This article analyzes the potential role of an expansive set of real asset classes in reducing inflation risk in the portfolios of long-horizon institutional investors. The author proposes a simple model of the evolution of asset returns that can be parameterized by key variables, such as 1) the sensitivity of asset returns to expected and unexpected changes in inflation, and 2) the degree of persistence in inflation. Each of the variables is a significant determinant of the long-horizon inflation properties of assets. Using his model and research, coupled with research available in the academic literature, Martin provides insight into the viability of various real asset classes as potential hedges for inflation.

Naive and Optimal Diversification for Managed Futures

he past decade has witnessed a dramatic increase in the use of alternative investment vehicles T such as hedge funds and managed futures products as additions to traditional stock and bond portfolios. Academic and practitioner literature (Chance [1994]; Schneeweis [ 19961) shows that the returns of hedge funds and managed futures have a low correlation with traditional investment vehicles such as stocks and bonds. This low correlation is due to the wide variety of markets traded by such alternative investment vehicles as well as their different trading methods (e.g., the ability to go long and short or use high leverage), and it provides a major reason for the use of these investments. Specifically, the differing investment styles and investment areas enable commodity trading advisors (CTAS) of managed futures to create portfolios that offer risk and return tradeoffs that are not available through traditional stock, bond, or even hedge fund investments. Unfortunately, the differences in investment styles and markets traded between CTAs and traditional investment vehicles often hide the extent to which managed futures offer risk characteristics similar to traditional stock and bond investments and the degree to which investment management concepts that work for stock and bond portfolios are also true for CTA-based asset portfolios. In this article, results are presented, first, on the number of randomly or naively” selected and equally weighted CTAs that must be included in a portfolio for that portfolio to accurately track a variety of benchmark CTA and asset indexes [ e.g., S & P 500 and Goldman Sachs commodity index (GSCI)]. Second, because historical managed futures performance is often used to determine ex post optimal CTA selections and weightings in investor portfolios, the robustness of the level of inclusion of managed futures in mean-variance optimal multiasset class portfolios is tested. Results show the potential benefits of managed futures investments in both naive and optimal portfolio determination.’ It is shown that the impact of random diversification on naively constructed (randomly chosen and equal weighted) CTA portfolios and the impact of CTA investment in mixed stock or bond portfolios are, as expected, similar to that shown to exist for traditional stock and bond investment. That is, depending on the homogeneity of the sample, between five and fifteen CTAs generally are required for the CTA portfolio to achieve a lower variance bound and to track 6 6

Skewness in Asset Returns: Does it Matter?

n much of traditional asset return analysis, an asset’s return distribution is assumed to be adequately described by its first two moments (expected return and variance) and its expected return described by its relationship with the assumed market portfolio.’ In fact, securities often exhibit additional return characteristics in which the probability of a security’s expected return in the extremes is often greater than that described by a simple normal distribution with equal expected return and variance. In short, an asset’s return distribution is often better described not only by its expected return and variance, but also by its higher moments of skewness or kurtosis. Academics have attempted to explain the role of skewness in an asset’s return distribution. In the early 1970s, academic literature contained theoretical models that described an asset’s expected return in three moments; that is, return, variance, and skewness (Jean [1971]). In addition to simple skewness, academics also have attempted to explain an asset’s return in terms of its co-skewness with the assumed market portfolio (Kraus and Litzenberger [ 19761) in which individuals overprice assets and lower expected returns to obtain assets with high positive co-skewness with a market portfolio (Schweser and Schneeweis [1980]). For traditional assets, however, research indicates a low level of stability in an asset’s skewness or co-skewness (Singleton and Wingender [1986]). Because when all else is equal, an investor will prefer positively skewed returns, the ability to use various derivative products (e.g., options) to create such skewed distributions is often presented as a selling point for active versus passive asset management (Zurack et al. [ 19971). The foundation of modern portfolio theory and the two moment 6.e.’ capital asset pricing) or three moment (multidimensional pricing), however, is that investors should be compensated only for the nondiversifiable risk associated with an asset. Thus, to obtain a downside truncated asset distribution (positive skewness), one must give up return either by paying directly to remove the downside returns (e.g., options, insurance) or by paying a higher price for assets that may have systematic positive co-skewness (Schweser and Schneeweis [1980]). This does not mean that skewness has no consequences for portfolio management, because investors rarely hold a fully diversified portfolio. However, if the risk associated with skewness is fully diversifiable or is fully priced, then does skewness really add anything to a fully diversified investor’s portfolio? In the area of alterna-

Who Invested with Madoff? A Flash Analysis of Funds of Funds

Financial fraud in general and Ponzi Schemes in particular continue to both fascinate and plague investors. A Ponzi scheme is often described as a securities fraud in which the promoter makes a false or misleading statement about his investment strategy and money from new investors is used to fund redemptions. When new investors cannot be found to offset investors asking to leave, current investors often are exposed to extreme loss of capital. This article examines whether funds of funds that invested with Madoff have any common characteristics. This question is of particular importance to the end investor, who has historically relied on the FOF industry to provide both manager selection and due diligence services. The results of this study should help investors and their consultants better identify those intermediaries that are less-well-poised to deliver investment management services with respect to hedge funds. The article analyzes a set of FOFs that were or were not invested with Madoff. The results show some relatively weak commonalities in the empirical characteristics of FOFs that invested in Madoff. Unfortunately, the apparent commonalities are insufficient in themselves to accurately and effectively predict which FOFs sought and acquired exposure to Madoff.

Who Invested In Amaranth?: A Flash Analysis of Fund of Funds

Throughout the second half of the month of September, investors and other participants in the hedge fund industry watched as Amaranth Advisors’ Multi-Strategy Funds suffered extraordinary losses of approximately 65% of capital on adverse bets in natural gas and other assets. This article identifies 80 fund of funds that were invested in Amaranth in September, 2006, utilizing direct or indirect, statistical methods. The results show that due to their exposures to Amaranth, fund of funds lost an average of 19.2M dollars and dissipated an estimated median of 2.1 years and an average of 5.7 years of their total alpha production. Investors in fund of funds should consider this a lesson in the importance of evaluating the overall investment processes of such fund of funds, and accordingly evaluate whether the value of all services provided by candidate fund of funds match or exceeds the value of expected fees.

Hedge Fund Incubation, Development, and Performance

An important theme in the history of the hedge fund industry is the disintermediation of the proprietary trading function of investment banks, and as such, the ‘privatization of the trading floor.’ Historically, many hedge funds were set up as independent businesses to pursue trading strategies originally pioneered by proprietary trading desks of large banks. This article presents new empirical information about the causal and associational consequences of some varying forms of institutional affiliation between hedge funds and larger investment organizations or service providers. Three types of relationships between hedge funds and outside organizations are explored and the risk, performance and exposure characteristics of each group are examined. The results show that the institutional commitment associated with seeding and operational support is positively correlated with risk-adjusted returns relative to the broader universe of hedge funds.