UC Berkeley

Financial intermediaries play an important role in the pricing of financial assets. For example, intermediaries may act on behalf of consumers in deciding how their wealth is invested, or they may act as providers of liquidity. This dissertation explores several ways in which intermediaries impact price informativeness, the transaction costs investors incur, and investor welfare.

In the first chapter, I examine how prices reveal information when intermediaries are informed. Using a model of repeated trade between a long-lived, informed, price-discriminating market maker and risk averse traders with endogenous hedging demands, I first show that traders are weakly better off trading with an informed dealer, as they may learn something about an asset's value in the process of transacting. Second, while long-term incentives can induce an informed market maker to honestly reveal information and increase risk-sharing, they also enable the market maker to hide her information and extract more rents, reducing price informativeness. This less desirable outcome dominates with respect to both the parameter space and a selection criterion. Finally, measures of market quality, such as the transient component of price volatility (illiquidity), may not accurately reflect welfare.

The second chapter discusses how relationships affect prices when intermediaries are concerned about adverse selection. When counter-parties trade in OTC markets, such as those for corporate bonds or derivatives, the lack of anonymity implies that future terms of trade can influence prices today. Using a model of repeated trade between an informed trader and uninformed market makers, I show that information asymmetry can affect the markups charged by dealers in two ways. First, for a given market structure (number of market makers), traders with more private information incur lower trading costs because dealers offer better terms to mitigate adverse selection. Second, even when dealers can not compete directly on price quotes, they compete indirectly by improving the informed trader's outside option, though this competition is imperfect. While repeated trade allows two given counter-parties to ameliorate adverse selection, the maximum number of dealers, and hence the total gains achievable, are limited by information frictions. An empirical implication is that the comparative statics of transaction costs only make sense conditional on market structure.

The third chapter considers the effect intermediaries have as financial advisors, and whether measures of their performance as mutual fund managers accurately reflect the value they add to an economy. Relative to the existing literature, I look at how the presence of mutual funds affects the price of the underlying asset in an economy. Once this pricing effect is accounted for, I show that standard measures of mutual fund performance may not accurately reflect whether fund management is welfare improving