On the extent of the market: a Monte Carlo study and an application to the United States egg market

This paper investigates the extent of the market, using a switching regimes model similar to those used in stochastic frontiers estimations. We started by performing a Monte Carlo simulation on our model, seeking to evaluate its performance in terms of correctly estimating the probability of integration of two markets. Our Monte Carlo results under the assumption of half-normal and exponential distribution of the errors, revealed that these two distributions predict almost correctly the probability of integration of two markets. The half-normal error distribution model tends to slightly underestimate the true probability of integration, while the exponential error distribution model tends to slightly overestimate the true probability of integration. We, finally, applied the model to the United States egg market using data from three highly productive states and one less productive state. The model predicts that, the markets pairs considered are integrated. That is, the four markets studied belong to the same economic market in the sense of Marshall. Further, based on our Monte Carlo study, we find that the true probability of integration of two given markets lies in between the half-normal model estimates and the exponential distribution model estimates.