Abstract: | We present a trend‐based alternative to the standard first‐order autoregression model in persistence of profits studies. This is motivated by reservations over the interpretation of the standard model, and rests on a different concept of dynamic competition. A nine‐category taxonomy of long‐run persistence stereotypes is developed. Structural time series estimates are presented for a sample of UK companies. We find the null of long run competitive equilibrium not rejected in nearly a third of cases, but non‐eroding persistence to be present in around 60%. |