Decomposing systemic risk of the hedge funds industry: An approach based on Extreme Value Theory

Abstract

This master’s thesis studies the systemic risk of the hedge funds industry through a model developed and used by van Oordt and Zhou (2019a, 2019b). This model measures systemic risk by splitting it into two components: the Pure Tail Risk, defined as the relative tail exposure towards the whole financial system approximated by the S&P500, and the Systemic Linkage, which represents the probabilistic link of the extreme negative returns between the hedge funds and that proxy. Our first intended contribution resides in the explanation and the reconciliation of these two measures, the downside risk­related component and the tail dependence aspect, into one single metric. Our second contribution relates to innovations brought in the estimation method, which relies on Extreme Value Theory to overcome the scarcity of data and the low frequency of reporting inherent to the hedge funds databases. In order to do so, we implement a LASSO-­Generalized Pareto Regression for tail exposure explanations and use copula distributions for tail dependence measurement.Our work provides several new insights. We observe a significant positive correlation be­tween liquidity and systemic risk indicators, which highlights the link between shadow bank­ing, hedge funds and systemic risk. Moreover, a larger fund size and a longer advance notice act as systemic risk reducers, while the lockup period drives up the systemic threat. Driven by the Pure Tail Risk, the results show an increase of the systemic risk during period of stability such as the period of 2003 to 2006 and after the 2008 crisis. Coupled with the high commonal­ity of the industry in such periods, observed by Bussière et al. (2014) between 2003 and 2006, our findings highlight the potential threat that hedge funds bring to the stability of the financial system. We uncover a common dynamic behaviour of hedge fund strategies over time, even though Equity Hedge and Event Driven hedge funds show significantly higher values. This re­sults is all the more important given that we observe that the Equity Hedge strategy represents the largest fraction of the industry. Finally, our model stresses a shift of the hedge funds tail risk dynamic after the 2008 crisis, showing a genuine potential for future research.As a result, we believe that our efforts to identify and explain the systemic risk of the hedge funds industry and its components have led to insightful results, which point to an increasing need for an adapted regulation.