Imperfectly Competitive Financial Markets

The first part of this dissertation, titled "Strategic Behavior and Asymmetric Information in Financial Markets", studies the effects of changes in the precision of both public and private information in financial markets in which traders are not price-takers but act strategically. Two different mechanisms of price formation are considered. The first one is a mechanism with market orders and competitive market makers. The second one is based on limit orders and market clearing. Under both regimes, the disclosure of more public information increases the expected profits of liquidity traders at the expense of privately informed agents. These results are potentially changed in two cases: when the acquisition of private information is costly and when the disclosure requirements are not uniform across firms and we allow for discretionary liquidity traders. The implications for price volatility, trading volume, incentives to produce private information, efficiency of associations of investors and mechanism design are explored. In the second part (a joint work with Murugappa Krishnan) titled "Insider Trading and Asset Pricing in an Imperfectly Competitive Multi-Security Market", we study a multi-security financial market in a correlated environment with asymmetric information and imperfect competition, in which market makers learn about each return from every order flow, even as an informed trader manipulates what they can learn. Our model is a generalization of a single-security model by Kyle. In contrast to a previous analysis by Admati under perfect competition, where the effect of a correlated environment is only to generate various ambiguities, strategic behavior restores various theoretical regularities, and can "neutralize" all of the correlatedness arising from the structure of returns and liquidity noise. Even with imperfect private information, strategic behavior helps generate an equilibrium with simpler structure, which is valuable for applications, especially for justifying traditional event study procedures even when there is private information.