Three Essays in Financial Economics

Abstract

The dissertation consists of three chapters. In the first chapter I study heterogeneous beliefs speculative bubbles in a setup in which investment is endogenous and affects information flow. I show that during periods of major technological innovations it is socially optimal to have a bubble in asset prices, because it mitigates learning externalities. However, the speculative bubble that arises in decentralized economy is not perfectly aligned with the optimal bubble level and typically is too small at the beginning of the bubble episode and too large close to the end.

The second chapter (joint with Martin Schmalz) presents a model of Bayesian learning about multiple parameters of firms' fundamentals, which explains asymmetries between upturns and downturns both in asset pricing and in corporate finance. Good performance in good times can be due to either desirable good idiosyncratic performance or to undesirable positive correlation with a market-wide factor. In contrast, good performance in bad times can either come from desirable good idiosyncratic performance or desirable negative correlation with a market-wide factor. As a result, in downturns prices react more strongly to news about fundamentals, stock picking earns higher returns and boards tend to fire CEOs more frequently.

The last chapter is an extension of the classical market microstructure model of Kyle (1985). Information about a security's fundamentals arrives much less frequently than trading occurs. As a result, the market maker, trying to infer from the trading volume whether the insiders have received new information, will react much stronger to larger orders than to smaller ones and the relationship between price concession and the size of the order becomes non-linear. Thus, the market can be very liquid for small orders and very illiquid for large ones.