Finance, investment and growth: nonlinear time series evidence from 10 Asian economies

Abstract

This article employs financial development as the threshold variable to construct a threshold model and to re-examine the findings in Rousseau and Vuthipadadorn (2005) for 10 Asian economies. Our results show that in the high financial development regime, financial development could fuel economic growth, which is consistent with the conclusion in Rousseau and Vuthipadadorn (2005); however, there is no evidence shows that financial development could affect investment. In the low financial development regime, financial development has negative influence on investment growth, which is contradictory to the findings in Rousseau and Vuthipadadorn (2005); as to the relation between financial development and economic growth, the causality could not be supported by the data.

Notes

¹ Many researchers employed the threshold model to study similar topics. For instance, Tsay (1998), Berthelemy and Varoudakis (1995), De Gregorio and Guidotti (1995) and Huang et al. (2005), etc.

² Unit root test results of all the variables show that level values of all variables are I(1). Results of Johansen cointegration test show that except variables of Japan, the rest 9 countries have at least one cointegrating vector. Follow the multivariate linear test of Tsay (1998), we found that the relationship between the set of ΔI, ΔM1 and Δ(M2 − M1) and the set of ΔGDP, ΔM1 and Δ(M2 − M1) are nonlinear in eight countries, except for Malaysia and Pakistan. Because of these test results, we employ the TVAR model for Japan and the TVECM model for the rest seven countries. Please contact the authors for the complete test results.