Information transmission and incentives in markets

This dissertation comprises of three independent essays studying information transmission and incentives in financial and non-financial markets. The first essay develops a structural model of stock markets that accounts for several stylized facts on the relationship between stock returns and trading volume. In this model the rate of public and private information arrival are probabilistic, the latter depending on the informed trader\u27s ability and effort. The ability is uncertain and stochastically changing over time. Time series properties of the model include contemporaneous correlation between volatility and volume, both unconditionally and conditional on the current information set, and autocorrelation in volatility. When short sales constraints are included, there is a positive correlation between stock returns and volume. Secondly, the essay studies traders\u27 incentives under delegated portfolio management. The second essay studies firms incentives in credit markets over industry evolution. As industry evolves, the number of producers first increases and later rapidly falls. Sometimes even seventy percent of the prevailing firms exit in such a shakeout. Other stylized facts are that the price of the output initially falls and the quantity produced by each producer increases, both at a decreasing rate, and stabilize after the shakeout. Even though the number of firms is increasing in growing markets, these markets are also characterized by higher exit rates than the mature markets. In this essay I develop a model of reputation formation in credit markets which, when applied to growing markets, gives rise to these phenomena. The third essay (joint work with Illtae Ahn) studies community enforcement in a private information, random matching setting, where buyers privately network for information and sellers have a short term incentive to supply low quality. We show that high quality can be sold in a sequential equilibrium when there are M sellers and buyers, even when each buyer periodically observes only $N\sp*(M)$ other buyers\u27 and sellers\u27 games where $0\u3c{\rm lim}\sb{M\to\infty}N\sp{*2}/M\u3c\infty.$ We show that when such networking is costly and M is large, low quality is supplied with positive probability in any Nash equilibrium. For this case, we characterize conditions for a sequential equilibrium in which both high and low quality are supplied