Spatial market integration in the presence of threshold effects

A large body of research has evaluated price linkages in spatially separate markets, Much recent research has applied models appropriate for nonstationary data. Such analyses have been criticized for their ignorance of transactions costs, which may inhibit price adjustments and thus affect tests of integration. This analysis utilizes threshold autoregression and cointegration models to account for a neutral band representing transactions costs. We evaluate daily price linkages among foul corn and four soybean markets in North Carolina, Nonlinear impulse response functions art used to investigate dynamic patterns of adjustments to shocks. Our results confirm the presence of thresholds and indicate strong support for market integration, though adjustments following shocks may take many days to be complete. In every case, the threshold models suggest much faster adjustments in response to deviations from equilibrium than is the case when threshold behavior is ignored.