Essays in asset pricing

The first chapter of this thesis documents interesting facts about firms whose quarterly earning announcements closely meet the most recent analyst consensus forecast, and investigates possible explanations. These firms enjoy higher long-lasting future returns, tend to be larger and are followed by more analysts, whose forecasts have a smaller dispersion. While the proportion of past quarters when forecasts are met is positively related to future returns, the proportion of past quarters when forecasts are strictly beaten is negatively related to future returns. Return differentials based on past earning surprises are not easily explained by standard risk factors. A model in which investors ignore fundamentals as long as news are good shows that firms who anticipate investors' reaction to earnings performance might have incentives to manage their earnings in order to avoid negative surprises, and that closely matching the forecasts might help them with this objective. Firms managing earnings might perform better in the long run despite their medium term objective.;The second chapter examines the effects of exogenous changes in the performance fees paid from terminal investors such as households to intermediaries managing their assets, on endogenous variables such as the risky asset volatility and risk premium, in the context of a dynamic equilibrium asset pricing model. The importance of the household behavior is highlighted by comparing two settings, one in which households are naive and perceive the investment opportunities as constant, and one in which households are more sophisticated and perceive the true opportunity set. Performance fees are convex in the actual performance, and have a component proportional to performance, as well as a component capturing the convexity. When households are naive, a higher proportional fee coefficient implies higher volatility and risk premium. Moreover, a higher convex component has little impact on the volatility, but reduces the risk premium. In contrast, when households are sophisticated, a higher proportional fee component has little impact on the economy dynamic, but a higher convex component increases stock volatility and risk premium. Empirically, hedge fund performance fees are related to market returns and volatility in ways consistent with naive households.