A penalty function approach to occasionally binding credit constraints

Occasionally binding credit constraints (OBC) have recently been explored as a promising way of modeling financial frictions. However, given their highly non-linear nature, most of the literature has concentrated on small models that can be solved using global methods. In this paper, we investigate the workings of OBC introduced via a smooth penalty function. This allows us to move towards richer models that can be used for policy analysis. We show that in a deterministic setting the OBC approach delivers welcome features, like asymmetry and non-linearity in reaction to shocks. However, feasible local approximations, necessary to generate stochastic simulations, suffer from fatal shortcomings that make their practical application questionable.