STORAGE DECISIONS IN INCOMPLETE MARKETS (MARKOV CONTROL, DYNAMIC EQUILIBRIUM)

This dissertation examines the role of market structure in determining the efficiency of storage and other economic decisions in a market economy. The examination pursues three main themes.;The second theme treats cases where the firm does not pursue the value-maximization objective. In this situation, we show that a futures market for storable commodities can restore storage decisions to optimality over a wide range of corporate objective functions.;The third theme consists of an application of the stock market optimality results to a model specialized to focus on petroleum storage decisions in an importing economy such as the United States.;The central efficiency result is developed in the context of an infinite-horizon stochastic dynamic programming problem. The underlying uncertainty is governed by a Markov process. Standard results in Markov control theory are used to demonstrate the existence and equivalence of solutions to two related problems: (1) economy-wide optimization guided by a central planner and (2) stock market equilibrium in a decentralized economy. We demonstrate the optimality of a stock market equilibrium in the case of homogeneous consumers in a general economic setting.;The first theme develops conditions which guarantee that a stock market equilibrium--even in incomplete markets--will be optimal when the technology decisions are made by a competitive firm whose goal is to maximize its own market value. In addition to regularity conditions, the essential assumption underlying this conclusion is that consumers can be treated as if they were homogeneous.