Essays on hedge funds and portfolio delegation in the presence of an information differential.

Summary In this thesis we develop a model of the hedge fund industry in which the various fees charged (management and redemption fees) are endogenously determined in a competitive market. We empirically explore a first implication of the model which is that deteriorated skewness is caused by investment in illiquid assets. We then empirically explore two other implications of the model which are that the size of an AIF is decreasing with the volatility of liquidity withdrawals, and then that the return realized by an AIF is also decreasing with the volatility of withdrawals due to client liquidity needs. In the next chapter, we model hedge fund returns according to the Pareto Stable Law, which is better able to describe a skewed and leptokurtic distribution, and we then assess the impact of this modeling on risk and performance measures commonly used by hedge fund investors. In the fourth chapter, we propose an equilibrium model of an economy consisting of an investor and a fund manager where we use filtration coarsening techniques to account for the information differential and thus provide a new justification for the use of portfolio delegation. In the last chapter, we highlight the importance of modeling the evolution of an alternative fund according to a diffusion-jump process to correctly value a portfolio insurance contract managed according to the cushion method (CPPI). We show in a second step through the use of non-parametric statistical methods that the returns of hedge funds exhibit a level of jump activity similar to that present in equity indices, as well as the existence of a common factor between hedge funds and equity indices that corresponds to a systemic jump.