| Since the pioneering work of Kyle (1985), most models analyzing strategic trading with private information assume that all traders share a common prior belief about the underlying distribution of the private information (signal). In this dissertation, I relax the common priors assumption and extend the work of Kyle (1985) by incorporating heterogeneous prior beliefs among traders. Specifically, it is assumed that all traders agree to disagree over their prior beliefs. Using market makers' belief as a benchmark, an informed trader may be overconfident or underconfident based on the trader's prior belief. This dissertation shows that heterogeneous prior beliefs and overconfidence, in particular, have profound effects on trading strategies, liquidity, volume, and prices in the markets.In the first essay, i.e., the first chapter, of this dissertation, I introduce heterogeneous prior beliefs into a duopoly, strategic trading model. The model shows that overconfidence enhances trading aggressiveness, magnifies trading volume, increases market depth, alleviates the asymmetry of information, and mitigates noise traders' losses. These results suggest that overconfidence may explain the observed high volume phenomenon in real financial markets. Interestingly, the model indicates that an overconfident, informed trader may survive better than a rational, informed trader. This result implies a rationality of overconfidence in the markets. But, the model also yields a Prisoners' Dilemma outcome when both informed traders become overconfident.In the second essay, i.e., the second chapter, of this dissertation, I develop a sequential auction model incorporating three motives for trade--asymmetric information, liquidity, and heterogeneous prior beliefs. The model helps to explain the closing end of the U-shaped intraday patterns in trading volume and price volatility, documented on the NYSE and other markets. In the continuous limit, the sequential auction equilibrium converges to the continuous auction equilibrium of Kyle (1985), except that at the last instant there is a large discrete, informed trade. Importantly, the discrete trade is generally bounded because it is still subject to a non-zero price impact, despite the fact that the asymmetry of information vanishes eventually at the end of trading. |
