The persistence of profit: a new empirical interpretation

Inferences are drawn about the true coefficients which correspond to sample estimates of a persistence of profit model fitted over a large number of firms. This is done by generating simulated sampling distributions for the estimators over various distributional assumptions. Profits seem to be stationary for all firms, with an average short run persistence coefficient of 0.59, higher than most previous estimates. Long run profit rates differ between firms, although by less than is suggested by direct observation of variations in mean profit rates calculated over time. Short run persistence appears to be inversely related to unsystematic variation in profit.