Chapter 1. Previous literature documents that mutual funds' flows increase more than linearly with realized performance. I show this convex flow-performance relationship is consistent with a dynamic contracting model in which investors learn about the fund manager's skill. My model predicts that flows become more sensitive to current performance after a history of good past performance. It also predicts that the effect of past performance on the current flow-performance relationship is weaker for managers with longer tenure. I consider an optimal incentive contract for money managers, and I provide an explanation for common compensation practices in the industry, such as convex pay-for-performance schemes and deferred compensation. In the optimal contract, flows become more sensitive to performance when the manager faces stronger incentives from the compensation contract. With learning, the manager's incentives become stronger after good performance, so that a manager exerts more effort when his assessed skill is higher. However, the relation between past performance and incentives becomes weaker over the manager's tenure. Using mutual fund data, I test the predictions of the model on the dynamic behavior of the flow-performance relationship, and I find empirical support for the theory. Chapter 2. We test through which channels quantitative easing affects the prices and issuance of securities. We exploit the announcement of the corporate bond purchase program by the European Central Bank, and we study the impact of the announcement on the cross section of European corporate bonds. We find that as the Central Bank increased the demand for bonds eligible for the program, eligible firms responded by substituting the issuance of ineligible bonds with the issuance of eligible bonds. As a result, bond prices were unaffected by their eligibility status, and all firms increased total issuance to the same extent. We show that monetary policy affected bond prices through a risk channel. Prices increased significantly more for bonds and firms that were exposed to higher levels of risk and uncertainty. However, risky firms did not issue more in response. Chapter 3. I develop a continuous-time game between a population of investors and an intermediary whose type is private information and whose portfolio allocation is neither observable nor contractible. I define and characterize a sequential equilibrium of the game and solve for a Markovian equilibrium where investors' posterior beliefs are the key state variable. In my model, demand for riskless assets undergoes dramatic changes that resemble the episodes of flight to safety observed during financial crises. I show that a risk-neutral intermediary chooses a portfolio that minimizes risk when beliefs are near the threshold below which the intermediary is terminated.