Three essays on empirical asset pricing and systematic ambiguity

Abstract

This dissertation consists of three independent chapters on empirical asset pricing and systematic ambiguity. The first chapter examines the optimal executive compensation policy when a manager is ambiguity-averse. Standard principal-agent models of optimal contract design poorly explain existing practices of executive compensation, in particular the prevalence of stock- based compensation. This paper addresses this inadequacy by analyzing the optimal con- tract for an ambiguity-averse manager in a continuous-time moral hazard model. We study the impact of ambiguity aversion on the optimal structure of managerial compen- sation plans. The model predicts that an ambiguity-averse manager undertakes less risky projects and exerts lower effort than a risk-averse manager. The optimal contract for the ambiguity-averse manager therefore contains a larger fraction of the firm’s equity to mitigate ambiguity aversion and to provide stronger incentives. The paper compares com- pensation plans consisting of stocks or stock options. The findings of the paper reveal that stock option holdings in the optimal contract have advantages over stocks in terms of incentive costs to shareholders and additional risk-taking incentives. The second chapter presents a hedge fund portfolio choice model for an investor facing ambiguity. The investor faces both idiosyncratic hedge fund ambiguity and aggregate mar- ket (stock market or macroeconomic) ambiguity. The optimal hedge fund asset allocation model reveals that firstly an investor tends to reduce her allocation to risky assets under ambiguity, hedge funds or stocks alike, and secondly only systematic ambiguity is priced in equilibrium. Moreover, the more directional (strongly correlated with the market) the hedge fund strategy is, the lower the optimal allocation to hedge funds by an investor who is sensitive to ambiguity. The theoretical model derives the equilibrium two-factor capi- tal asset pricing model under ambiguity (ACAPM) that demonstrates how the ambiguity factor is priced in financial markets. It contributes to the alpha versus alternative beta debate by postulating the hypothesis that expected hedge fund returns embed a risk pre- mium for systematic ambiguity exposure. In the empirical section, we measure ambiguity as the cross-sectional dispersion in survey-based macroeconomic forecasts for growth in the Industrial Production Index and in stock market forecasts for S&P 500 Index returns, and we construct the systematic ambiguity factors from the universe of S&P 500 stocks. We estimate ambiguity betas for long/short equity hedge fund strategies and document sig- nificant ambiguity exposures, especially for directional long/short equity hedge funds. We compare the out-of-sample performance of portfolios constructed according to the hedge fund alphas’ ranking with and without systematic ambiguity exposures and find that the former outperform the latter. The third chapter addresses an anomaly in convertible bond underpricing by investi- gating the role of convertible arbitrage hedge funds as liquidity providers to the convertible bond market. Prices of convertible bonds are sensitive to unexpected demand shocks gen- erated by convertible arbitrage hedge funds. This paper investigates whether the price pressure created by innovations in hedge fund demand can account for the systematic mis- pricing of convertible bonds.