Hedge funds risk and connectedness

Abstract

This paper focused on measuring the volatility spillovers and parametric Value-at-Risk (VaR) models for hedge fund strategies. In reference to the first, a volatility spillover index was created to measure the connectedness and relationships between strategies. Results suggest that during periods of distress, volatility spillovers tend to increase across strategies. Equity-Hedge and Event-Driven strategies seem to be the most connected, while the Equity-Market-Neutral strategy exhibits the least connectedness with other strategies. With regard the second, unconditional and conditional parametric VaR models under Gaussian and t-distributions were measured. In general, conditional GARCH-type models outperform their unconditional counterparts, particularly at the onset of the recent financial crisis. This suggests that GARCH-type VaRs are better in protecting against downside risk. Moreover, conditional Gaussian models are more suitable for mild periods, while those under the t-distribution effectively capture risk in extremely volatile period. Finally, to unravel the connection between the two measures, common hits between hedge fund strategies were examined. The results validate the results of the volatility spillover index.