Risk and Beta Anatomy in the Hedge Fund Industry

Abstract

Using a Bayesian time-varying beta model we explore how the systematic risk exposures of hedge funds vary over time conditional on some exogenous variables that managers are assumed to use in changing their trading strategies. In such a setting, we impose a structure on fund returns, betas and benchmark returns, developing a framework that could help explain how expected and unexpected hedge fund returns are correlated with systematic risk factors through the beta dynamics. Such a system also provides a useful way of, (a) inspecting how and through which channels systemic risk propagates over time; (b) evaluating the performance conditional on public information within a Bayesian context; (c) cloning hedge funds by the mean of beta replication; (d) monitoring the risk of hedge fund returns on a VaR-based context. Major findings of this work, based on the analysis of the CSFB/Tremont indices over the period 01/1994-09/2008, are that: (1) volatility, changes in T-bill, term spread and shocks in liquidity significantly impact on time variation of hedge fund betas; (2) increasing interdependencies in beta dynamics among hedge funds together with leverage levels and shocks in liquidity are the key factors underlying the dynamics of systemic risk; (3) conditional time variation in betas leads to conclude that hedge fund industry did not delivered excess returns over their own style benchmark; (4) replicating the risk/return characteristics of hedge funds through our beta modelling seems to do a good job, also delivering better performances in a risk-adjusted basis; (5) simulation-based exercises on VaR predictions prove that our technology could be a serious candidate in hedge fund risk monitoring systems.

Keywords: Bayesian analysis, Conditional time-varying beta, Hedge Funds, Performance, VaR

JEL Classification: C11, C13, G12, G13