Evaluating policy feedback rules using the joint density function of a stochastic model