Linear-quadratic games of resource depletion

In this paper we describe some methods for quantitatively analyzing dynamic, multiple agent models in which at least one agent takes into account his influence on the aggregate environment such as the degree of competitiveness in the petroleum industry. We confine our attention to models in which the agents solve stochastic, quadratic optimization problems subject to linear constraints. Discussion is presented in the context of a resource depletion example. The players are resource suppliers; the task is to compare their behavior when they play alternative dynamic games. Our approach to solving these models of dynamic games is first to deduce the stochastic Euler equations for each of the agents and then to simultaneously solve these stochastic Euler equations subject to the respective transversality conditions. The solution strategy involves factoring the characteristic polynomial of the system of stochastic difference equations. 22 refs.