A simulation-based analysis of parameter-stability tests using conditional cointegration models

This note conducts a simulation study of parameter-stability tests using conditional cointegrated vector autoregressive (CVAR) models. Monte Carlo experiments show that, in small samples, the stability tests based on conditional CVAR models under the assumption of weak exogeneity are more powerful than those based on a joint CVAR model; the reverse is observed, however, when the assumption does not hold. The overall assessment of the experiments leads to a practical procedure for testing the constancy of cointegrating parameters. JEL classi…cation: C32, C52.