VALUE FUNCTION APPROXIMATION IN THE PRESENCE OF UNCERTAINTY AND INEQUALITY CONSTRAINTS - AN APPLICATION TO THE DEMAND FOR CREDIT CARDS

We present an algorithm for approximating the solution to discrete-tim e stochastic dynamic programs with inequality constraints. The algorit hm exploits the state preference approach to choice under uncertainty to reduce the dimensionality of the state space, and polynomial approx imation to improve computational efficiency. It is particularly useful in financial applications involving liquidity and credit constraints and more than one asset. The algorithm provides information about the derivatives of the value function, and therefore risk aversion, along with asset demands.