The life cycle of a hedge fund investment strategy

Hedge Fund Alpha: A Framework for Generating and Understanding Investment Performance (World Scientific, 2009) is a collection of papers edited by John M. Longo of the Rutgers Business School. In this post I am going to focus on one paper that I think should be of general interest, “From Birth to Death: The Lifecycle of a Hedge Fund Investment Strategy” by Longo and Yaxuan Qi.

The authors claim that the viability of a hedge fund is threatened if it has a single year of double-digit losses or two consecutive years of losses of any magnitude. They then analyze three well-known strategies—the small firm anomaly, momentum anomaly, and accrual anomaly—adjusting them to include a hedge component. Would following any of these single-factor strategies have enabled a hedge fund to survive the authors’ study periods?

The first hypothetical fund, the Hedged Size Anomaly Fund, goes long the smallest decile of firms in the CRSP universe, sells short the Russell 2000, and adds the rebate from the short sales. The authors use the 30-day T-bill to approximate this rebate. The returns of this hypothetical fund between 1981 and 2005 were never outright terrible, but between 1984 and 1990 they were, for the most part, negative; the average return over this period was -0.57%. The authors suggest that the fund would undoubtedly have gone out of business at some point during these seven lean years.

The second hypothetical fund is the Hedged Momentum Anomaly Fund, begun in 1993, the year of a comprehensive study of the momentum strategy, and ended in 2005. This fund goes long the top decile momentum portfolio of the firms in the CRSP universe, sells short the Russell 1000 Growth index, and again uses the 30-day T-bill to approximate the rebate. This fund had much more impressive returns overall, but it turned in a very poor performance in 1998. It lost more than 14%, a year in which the S&P 500 was up 28.6% and the Russell 1000 Growth gained 38.7%. The fund would have faced serious scrutiny, perhaps implosion, during this time.

The last hypothetical fund is the Accrual Anomaly Fund, begun in 1996 and ended in 2005. It follows a fundamentally based strategy that analyzes earnings quality. This fund buys a portfolio of low accrual firms and sells short a similar sized basket of high accrual firms, once again adding into its performance metrics the returns on the 30-day T-bill. This fund saw some exceptionally profitable years but also back-to-back double-digit losses (17.47% and 20.24%) in 1997 and 1998. So it might have folded during or after these grim years.

All three funds had respectable to very good average returns and CAGRs: (1) 9.56% and 8.82%, (2) 21.81% and 19.78%, and (3) 21.55% and 12.92%. But the returns were cyclical. An equally weighted portfolio of these three strategies, as should be evident to readers of recent posts in this blog, was superior in terms of buying survival time for the hedge fund: its average return was 17.33%, its CAGR 15.80%, and the only bad year was 1998, with a gross loss of 2.48%.

The authors conclude that if alpha is indeed cyclical, a hedge fund management company should either offer a multistrategy fund or run multiple funds under its management umbrella. (The authors don’t consider the possibility of successful regime switching.) Moreover, they argue, “hedge fund companies should act in a similar manner to [p]harmaceutical firms, creating a pipeline of products (i.e. hedge fund strategies) in the event that their primary product/strategy falters.” (p. 260)

The individual trader/investor would be wise to follow this advice.