Abstract :
[en] The performance of a market timer can be measured through the Treynor and Mazuy (1966) model, provided the regression alpha is properly adjusted by using the cost of an option-based replicating portfolio, as shown by Hübner (2010). We adapt this approach to the case of multi-factor models with positive, negative or neutral betas.
This new approach is applied on a sample of hedge funds whose managers are likely to exhibit market timing skills. We stick to funds that post weekly returns, and analyze three hedge funds strategies in particular: long-short equity, managed futures, and funds of hedge funds. We analyze a particular period during which the managers of these funds are likely to magnify their presumed skills, namely around the financial and banking crisis of 2008.
Some funds adopt a positive convexity as a response to the US market index, while others have a concave sensitivity to the returns of an emerging market index. Thus, we identify “positive”, “mixed” and “negative” market timers. A number of signs indicate that only positive market timers manage to acquire options below their cost, and deliver economic significant performance, even in the midst of the financial crisis. Negative market timers, by contrast, behave as if they were forced to sell options without getting the associated premium. We interpret this behavior as a possible result of fire sales, leading them to liquidate positions under the pressure of redemption orders, and inducing negative performance adjusted for market timing.
Scopus citations®
without self-citations
5