Modeling the dependence among crude oil, stock and exchange rate: A bayesian smooth transition vector autoregression

Abstract

This study aims to investigate the relationship among West Texas Intermediate crude oil price which represents crude oil, Down Jones Industrial Index and exchange rate. We use smooth transition vector autoregression with Bayesian estimator to estimate the relationship among them due to the conventional VAR often faces with the over-parameterization problem and we can avoid using p-value in making the statistical inference. The results show that there is different relationship among variables in each regime. The impulse response indicates that most of variables will converge to their equilibrium within about 20 months. Finally, the spillover effect shows that three variables either influence or are influenced by other variables.