| dc.description.abstract | In this thesis, I discuss how asset prices are influenced by the decisions of heterogeneous investors. Asset prices are conventionally explained through a representative-investor whose risk-aversion fluctuates or who faces fluctuating fundamental uncertainty. Much progress has been made in describing how such an investor influences prices. Yet, such work poses considerable difficulties. In particular, empirical studies document that trade volume predicts asset prices and investors infer information from prices. Furthermore, the burgeoning household finance literature documents patterns in portfolio allocations across investors. While models with heterogeneous investors can address such shortcomings, more work is needed to understand them. In particular, little is known about how differentially informed investors learn in financial markets and how their opinions affect prices. I describe how disagreement affects volatility in my chapter one of my thesis. I also examine how asymmetrically informed investors learn from prices in chapter two. Finally, Joao Brogueira and I made a theoretical contribution in our published paper which is contained in chapter 3 of my thesis. I describe each chapter briefly. In chapter one, I provide conditions under which disagreement about dividend growth forecasts amplifies stock market volatility, in line with empirical evidence. In a frictionless economy with two Epstein-Zin investors, I model disagreement as exogenous heterogeneity in beliefs: one investor is pessimistic, the other is not. I show that disagreement amplifies volatility only if investors switch beliefs, that is if an investor is only temporarily optimistic. If instead one investor is permanently pessimistic, prices are less volatile than dividends, and higher disagreement lowers volatility âĂŤ in contradiction with evidence. Finally, I provide empirical support for switching beliefs among investors, using cross-sectional data from the Survey of Professional Forecasters. In chapter two I discuss the relationship between trade volume and stock market returns. There is substantial evidence that high trading volume predicts low returns, both in the cross-section but also across several years. To permit information-based trade among asymmetrically informed investors, economic models conventionally include noise traders. However, these models cannot explain the observed relationship between trade and returns. As noise traders demand random quantities, they generate a too volatile trade volume compared to the empirical low-frequency variations. I argue in “Trade Volume, Noise Traders and Information Acquisition with Neural Networks” that neural networks can be used to describe the empirical evidence. I first characterize elementary properties of neural networks. I then show in a model of trade among differentially informed investors, that neural networks permit information inference from prices at arbitrary precision but that information asymmetry can persist even without noise traders. Finally, I outline how such models might be able to explain why trade volume predicts excess years ahead. Finally, chapter 3 contains a paper I wrote together with Joao Brogueira. Our note presents a proof of the existence of a unique equilibrium in a Lucas (1978) economy when the utility function displays constant relative risk aversion, and the logarithm of dividends follow a normally distributed autoregressive process of order one with positive autocorrelation. We provide restrictions on the coefficient of relative risk aversion, the discount factor and the conditional variance of the consumption process that ensure the existence of a unique equilibrium. | en |
