Market volatility and hedge fund returns in emerging markets

Abstract

In this article, we estimate several augmented Treynor and Mazuy (1966) models to examine the performance of hedge fund index returns in four different emerging market regions. In our estimations we match the fund returns with the regional emerging market equity and bond index data, which is a research approach that is pioneered by Fung et al. (2002). Whether market volatility affects the hedge fund returns or not is one of the main questions that we ask in the article. Our results reveal that stock and bond market volatility do not have a significant impact on fund returns for the most part, which is a result that is robust to various measures of volatility. Among the four regions we examine, only the emerging market hedge funds in the Global market yield statistically significant positive alphas that is robust and sizable. We also find no evidence for market timing skills in these emerging market hedge fund returns.

Keywords:

• hedge funds
• emerging markets
• volatility
• stocks and bonds
• alpha
• market timing

JEL Classification::

• G11
• G15

Notes

¹ In most studies on hedge fund performance, hedge funds operating in the emerging markets have been treated as a single group classified by the strategy that they employ and this strategy is called ‘emerging markets’.

² The CISDM database reports hedge fund data for other geographical regions including Asia/Pacific (including Japan), Middle East and Africa, etc. We focus only on the four main geographical regions that have the most number of hedge funds listed.

³ Using actual returns did not significantly change the results of our analyses.

⁴ We tried various specification of the ARIMA model and confirmed that an ARIMA(0,1,3) specification in fact fits the realized index volatility quite well.

⁵ Autoregressive and Moving Average (ARMA) terms are actually not needed in the conditional mean equation for all regions except Asia/Pacific (excluding Japan). For this particular region, an AR(2) specification was required to remove remaining serial correlation in the stock index return data.