Game theoretic models of price determination and financial intermediation.

The thesis is concerned with the evolution of prices in market economies. In the first chapter we analyze price competition among firms with limited capacities in the framework of the classical Bertrand Edgeworth model. For this model it is well known that a Nash equilibrium in pure strategies may not exist. We discuss the nature of this non-existence result. While enlarging the strategy space to include non-linear strategies in general does not suffice for existence in the simultaneous move game, the possibility of reactions to competitors' actions in a dynamic context may restore equilibrium. In the remaining part of the thesis we analyze intermediation in frictional markets. When market participants are informed only imperfectly about potential trading opportunities, search and negotiations may prove costly. In such markets intermediaries by publicly quoting prices can help to reduce the transactional costs of exchange. In chapter two we analyze the case, in which intermediaries have access to an information technology which informs the full market. We characterize equilibrium. The inability of market participants to coordinate market participation is reflected in a large variety of subgame perfect Nash equilibria. Using a refinement criterion we find that high valuation traders typically trade with intermediaries, while low valuation traders engage in search. Moreover, price competition among several intermediaries yields Walrasian outcomes. Nevertheless, the market exhibits the features of a natural monopoly. In chapter three we relax the assumption concerning the information technology. An intermediary's choice of an information network determines the size of his clientele and hence the probability of trading. In this case the size of the information network is viewed as a quality attribute by market participants and imperfect price competition among intermediaries obtains. We characterize the industrial structure of those markets as natural oligopolies. Consequently, there is no convergence to a fragmented industrial structure as the economy grows large. Still, as the largest competitors are of roughly equal size equilibrium allocations tend to be fairly competitive