Interconnectedness in the Hedge Fund Sector: its Implications for the Performance and Risks in the Industry

This thesis investigates how the correlations between hedge fund returns are assosiated with the performance and risk in the hedge fund sector. We use the "absorption ratio" of Kritzman et al. (2011), which is based on PCA (principal component analysis), to gauge the level of correlations ("connectedness") in the sector, and we look at this metric as the proxy for the direct and indirect links between hedge funds. The research question was motivated by the fact that correlation between hedge fund strategies are currently the issue that concerns regulators and by the fact that there are evidence of hedge funds being prone to "crowding". Our main findings are the folowing. First, we document that the "absorprion ratio" is able to capture the dynamics of hedge fund returns, filtered for common risk factors. We relate this result to the phenomenon of hedge fund contagion documented in the literature. Second, we find that the risk factor based on "absorption ratio" has an explanatory power for the cross-section of hedge fund returns. The funds with greater exposure to AR earn on average 0.26% higher monthly return. When the factor is included in the model, it reduces hedge fund alpha by 20%. We aslo find that the factor earns significant risk-premium in the cross-section using the time-series/cross-sectional asset pricing test.