Determinants of Hedge Fund Performance

This dissertation consists of three essays which focus on the determinants of hedge fund performance. The first essay defines two distinct styles of active portfolio management: alpha active and beta active. I develop measures of beta activity and find ample evidence that top beta active managers deliver superior out-of-sample performance. In addition, I find that beta activity measure successfully captures the time varying nature of beta exposures that can be interpreted as the common factor driving the long term out-of-sample predictive power of both Systematic Risk and R2. The second essay attempt to span the space of potential risk factors with exchange traded funds and replicate hedge fund return through selected ETF portfolio. I find the portfolio of clones created with my procedure provides better out-of-sample performance than the portfolio of “cloneable” hedge funds. In contrast, “non-cloneable” hedge fund portfolio reflects the hedge fund active management style supported by the superior risk-adjusted performance. The third essay investigates a new dimension of market timing activity. I decompose hedge fund excess return to alpha and beta return and find a strong monotonic mean reversal pattern in out of sample performances of portfolios sorted by beta return. I identify two types of managers: Multi-active Managers and Risk-writing Managers, and find the superior performance of bottom quartile beta return portfolio is mainly driven by Multi-active Managers. I find evidence that Risk-writing Managers exhibit the greatest total risk and beta risk, and generate returns through excessive risk taking. Multi-active Managers actively manage their market position that are reflective of their beliefs, continuously search for market opportunities and effectively adjust their beta positions to reflect their evolving market expectations