Essays on Market Microstructure
Kyle, Albert S
MetadataShow full item record
This dissertation studies topics on market microstructure. The first chapter theoretically studies market manipulation in stock markets in a linear equilibrium. The second chapter empirically examines the presence of opportunities for liquidity arbitrage. The last chapter develops and examines a method to capture a co-movement of informed trading. In chapter 1, I study a theory of trade-based price manipulation in markets. I compare two different types of price manipulation studied in previous literature, uninformed and informed manipulation, in the same linear equilibrium model. I show that the presence of positive-feedback traders creates an incentive for the informed trader to bluff, but the opportunity is absent if a sufficient number of uninformed traders behave strategically. Numerical comparable statics results show that informed manipulation is more likely and more profitable when the noise trading is more volatile and that market efficiency could become worse under the presence of manipulation. A financial transaction tax can not prevent informed manipulation, but it reduces the liquidity of the market. Chapter 2 empirically investigates intra-day price manipulation in a stock market. My microstructure model is specifically designed to define the conditions under which a manipulation opportunity arises from the variation in liquidity as measured by price impact. My empirical analysis using data from the Tokyo Stock Exchange suggests that while there is a significant chance of uninformed manipulation across time and stock codes, it is not profitable enough to affect price fluctuations. Analysis of intraday price and trade sizes shows that the opportunity begins to disappear shortly. Chapter 3 studies contagion in a financial market by using a market microstructure model. We extend the Easley, Kiefer, and O'Hara (1997) model to a multiple-asset framework. The model allows us to identify whether the driving forces of informed trading common or idiosyncratic information events are. We apply the method to three groups of stocks listed on the New York Stock Exchange: American Depositary Receipts (ADRs) of developed and emerging countries, and blue chips. We find contagion among emerging-country ADRs during the Asian Financial Crisis of 1997, in the sense that informed trades were mostly driven by common information events.