Search results

In this article we propose a multivariate dynamic probit model. Our model can be viewed as a nonlinear VAR model for the latent variables associated with correlated binary time-series data. To estimate it, we implement an exact maximum likelihood approach, hence providing a solution to the problem generally encountered in the formulation of multivariate probit models. Our framework allows us to study the predictive relationships among the binary processes under analysis. Finally, an empirical study of three financial crises is conducted.

Bertrand Candelon is a professor in International Monetary Economics. He received a PhD from Universite Catholique de Louvain. After a postdoctoral fellowship at the Humboldt Universität zu Berlin, he joined University Maastricht, School of Business and Economics in 2001. He has written extensive works in the area of international finance, in particular on contagion and on the analysis of financial market co-movements. He is one of the founders of the Methods in International Finance Network.

This chapter sheds new light on the linkages between stock market fluctuations and business cycles in Asia. It shows that at cyclical frequencies stock markets lead business cycles by six months on average. China, Korea, and Taiwan constitute exceptions, as their real and stock market cycles are contemporaneously synchronized. The low level of maturity of these markets offers a potential explanation of this outcome. Furthermore, we find that the linkage also holds during phases of cyclical upswing and downturn, with the exception of China, where the financial market lags behind industrial production during expansions. Finally, for most of the countries (except Thailand and Malaysia), the linkage is also robust to the presence of financial crises.