Recent advances in explaining hedge fund returns: Implicit factors and exposures

Abstract

We survey articles covering how hedge fund returns are explained, using largely non-linear multifactor models that examine the non-linear pay-offs and exposures of hedge funds. We provide an integrated view of the implicit factor and statistical factor models that are largely able to explain the hedge fund return-generating process. We present their evolution through time by discussing pioneering studies that made a significant contribution to knowledge, and also recent innovative studies that examine hedge fund exposures using advanced econometric methods. This is the first review that analyzes very recent studies that explain a large part of hedge fund variation. We conclude by presenting some gaps for future research.

Introduction

A large and growing body of literature has investigated hedge fund performance attribution through the use of implicit or statistical factor models (e.g. Akay et al., 2013, Billio et al., 2012, O'Doherty et al., 2015). Investors want to know what is behind hedge fund return variation and what to expect from different hedge fund strategies or funds with different styles. Investors need to be familiar with the principles that enable them to understand hedge fund performance behavior. Although many studies describe the role of factors or exposures of hedge funds in delivering excess returns to investors, there is no survey that summarizes and discusses the results. This issue creates confusion to investors who do not have a clear picture or a holistic interpretation of the dynamics of hedge fund performance attribution.

The present study therefore closes an important gap. The aim of this study is to survey the literature and investigate the hedge fund return generating process within implicit or statistical factor models. This is the first survey and synthesis of older literature to yield a historical perspective, along with a survey in more detail of recent innovative studies that depict advances in hedge fund performance attribution.¹ Hence readers will have an integrated view and a deeper understanding of hedge funds. Our findings both make life easier for hedge fund investors and highlight opportunities for further research.

Our main conclusions are that early studies (e.g. Sharpe, 1992), through the use of Principal Component Analysis and Common Factor Analysis (the most common statistical approach) dealt mainly with linear factor models, giving weight to the asset categories or where the fund manager invests. They depicted a static representation of hedge fund performance attribution. Then there was a development toward non-linear models that tried to explain hedge funds' performance as option portfolios (e.g. Agarwal and Naik, 2004, Fung and Hsieh, 2004). Nevertheless, in recent years there have been several studies (e.g. Bali et al., 2014, O'Doherty et al., 2015, Patton and Ramadorai, 2013) that use more advanced models regarding hedge fund exposures. They confirmed previous studies that hedge funds have nonlinear returns, but they moved further and showed how these nonlinear exposures change over time according to financial conditions. Different strategies frequently have different exposures. However, there are a few exposures that are valid for virtually every hedge fund strategy (equity market, volatility, liquidity). Furthermore, systematic risk and more specifically macroeconomic risk have a significant role in explaining hedge fund performance for nearly all strategies. Higher moment factors provide extra explanatory power to the models.

This paper makes a number of important contributions to the understanding of the hedge fund literature. First, it covers a significant gap by presenting a survey that summarizes and discusses studies examining hedge fund performance attribution within statistical factors and exposures. It demonstrates a historical perspective by combining earlier and more recent innovative studies, discussing their strengths and weaknesses. Therefore the reader is able to look at the dynamic nature of the literature explaining hedge fund returns. This study assists investors in their asset allocation process in two ways: it facilitates a deeper understanding of what is behind hedge fund return variation and it also enables investors to know what to foresee from funds with different strategies or fund styles. Last but not least, we have identified some gaps for future research. An example is the absence of a unified framework that takes into consideration the comprehensive macroeconomic environment along with the internal structure of the hedge fund industry in explaining returns, or identifying the proportion of alpha affected by each of the underlying factors.

In Section 2 we depict different general approaches in measuring the performance of all hedge fund strategies.² Section 3 briefly reviews earlier linear studies. Section 4 reviews in detail the most recent nonlinear models within the bottom-up, up-bottom, and alternative modeling approaches, as we describe later. In the final Section 5 we summarize the key conclusions and reveal some gaps that should be addressed in future research.