A number of assets do not trade publicly but are sold to a restricted group of investors who subsequently receive private information from the issuers. Thus, the holders of such privately placed assets learn more quickly about their assets than other agents. This paper studies the pricing implications of this “learning by holding”. In an economy in which investors are price takers and risk-neutral, and absent any insider trading or other transaction costs, we show that risky assets command an excess expected return over safe assets in the presence of learning by holding. This is reminiscent of the “credit spread puzzle”—the large spread between BBB-rated and AAA-rated corporate bonds that is not explained by historical defaults, risk aversion, or trading frictions. The intuition is that the seller of a risky bond needs to offer a “coordination premium” that helps potential buyers overcome their fear of future illiquidity. Absent this premium, this fear could become self-justified in the presence of learning by holding because a future lemons problem deters current market participation, and this in turn vindicates the fear of a future lemons problem

Affluent households can respond to taxation with means that are not economically viable for the rest of the population, such as sophisticated tax plans and international tax arbitrage. This article studies an economy in which an inequality-averse social planner faces agents who have access to a tax-avoidance technology with subadditive costs, and who can shape the risk profile of their income as they see fit. Subadditive avoidance costs imply that optimal taxation cannot be progressive at the top. This in turn may trigger excessive risk-taking. When the avoidance technology consists in costly migration between two countries that compete fiscally, we show that an endogenous increase in inequality due to risk-taking makes progressive taxation more fragile, which vindicates in turn risk-taking and can lead to equilibria with regressive tax rates at the top, and high migrations of wealth towards the smaller country.

An entrepreneur with limited liability needs to finance an infinite horizon investment project. An agency problem arises because she can divert operating cash flows before reporting them to the financiers. We first study the optimal contract in discrete time. This contract can be implemented by cash reserves, debt, and equity. The latter is split between the financiers and the entrepreneur and pays dividends when retained earnings reach a threshold. To provide appropriate incentives to the entrepreneur, the firm is downsized when it runs short of cash. We then study the continuous-time limit of the model. We prove the convergence of the discrete-time value functions and optimal contracts. Our analysis yields rich implications for the dynamics of security prices. Stock prices follow a diffusion reflected at the dividend barrier and absorbed at 0. Their volatility, as well as the leverage ratio of the firm, increase after bad performance. Stock prices and book-to-market ratios are in a non-monotonic relationship. A more severe agency problem entails lower price-earning ratios and firm liquidity and higher default risk.