Three essays in financial economics

(cont.) left on the table" due to underpricing in the IPO allocation is not capital the firm could have raised; instead, it is the empirical regularity associated with obtaining a high quality aftermarket, high equity valuation, and higher proceeds to the issuer. We examine the principal-agent problem in a simple continuous time framework when potential agents have heterogeneous priors. We find that the principal prefers agents with priors very different from his own. The principal will create a contract that includes side-bets to exploit gains from trade created by heterogeneous priors despite the distortionary effect on effort choice. In a semi-dynamic labor market, the principal can optimally choose to churn his employees to prevent them from learning about project profitability, even when agents' skills are increasing with job tenure. We develop several empirical predictions, and relate our model to the labor market in the financial industry.