A NUMERICAL ALGORITHM FOR SOLVING MODELS WITH INCOMPLETE MARKETS

We describe an incomplete markets model and outline a numerical soluti on algorithm to compute its steady-state equilibrium. The model depart s from the Arrow-Debreu world of complete contingent claims markets by assuming the presence of borrowing constraints. Solution methods that rely on approximations around some reference point or those that use the Euler equations in an interior solution are not suitable for the p resent model. Rather, we outline a numerical solution algorithm that d iscretizes the state space and the decision space and iterates on a nu merical procedure that obtains the decision rules and distributions of agent types for all cohorts in an attempt to find the steady-state eq uilibrium of the model. The algorithm is implemented using the Fortran compilers on the Minnesota CRAY X-MP and San Diego CRAY Y-MP/8-864.