Essays on the impact of incomplete information

This dissertation presents three essays that examine how incomplete information, parameter uncertainty, diverse beliefs and learning affect investors' optimal portfolio allocations, trading activities and asset prices.; The first essay examines the characteristics of asset returns and trading volume when the investors have incomplete information and hold diverse beliefs about new information. It develops a rational dynamic competitive model of trade in which the investors have heterogeneous prior beliefs about the risky payoff of an asset and about the forecasts of the risky payoff. The investors trade for both informational and speculative reasons. It finds that the optimal demands of an investor for the risky asset contain two components: a "traditional" component that seeks to exploit the investor's perceived mispricing and a speculative component that capitalizes on the investor's beliefs about payoff forecast in the following period. It also finds that expected return is lower when the investors are more uncertain about the future forecasts, and that trading volume is positive even when there is no price change. Furthermore, it shows that asset supply can influence the relation between trading volume and price volatility.; The second essay examines the impact of uncertainty about dividend or earning growth on utility and short-run asset prices under a fully rational setting. It shows that expected utility and asset prices are not necessarily lower in an incomplete information economy than in a complete information one. Uncertainty about growth rates may induce risk averse agents to believe that less information is better; they perceived that they are better off with less precise estimates of growth rate. As a result, an agent who holds more pessimistic beliefs or is more uncertain about growth rate is less willing to collect information that may improve her assessment of growth, even when information is costless. Information does not have any social value.; The third essay examines the impact of uncertainty about expected return on the dynamic portfolio decisions of an investor. It shows that opportunity to learn about a risky asset's expected return induces the investor who is more (less) risk tolerant than a logarithmic-utility investor to hold more (less) of the risky asset than she otherwise would if there were no learning. The precise role of learning in portfolio decisions also depends on the investor's prior uncertainty about expected return and her investment horizon. (Abstract shortened by UMI.)...