In this paper, we investigate the ability of a modified RBC model to reproduce asymmetries observed for macroeconomic variables over the business cycle. In order to replicate the empirical skewness of major U.S. macroeconomic variables, we introduce a capacity constraint into an otherwise prototypical RBC model. This constraint emerges due to the assumption of kinked marginal costs of utilization, where the kink is located at a utilization rate of 100 percent. We find that a model with a suitably calibrated cost function reproduces the empirical coefficients of skewness remarkably well.