The relationship between the real estate and stock markets of China: evidence from a nonlinear model

Abstract

The use of asymmetrical threshold cointegration tests is adopted in this study to investigate whether any significant relationship or asymmetric adjustment exists between the real estate and stock markets of China. Our results indicate the existence of a long run nonlinear relationship between the Shenzhen Composite Index and the Real Estate Price Index. In the short run, the Granger causality test favours the ‘wealth effect’ hypothesis; conversely, in the long run, the existence of the ‘credit price’ effect is discernible above a certain threshold value, whilst the ‘wealth effect’ is apparent below this threshold value, which implies a bidirectional feedback causal relationship. Our empirical results demonstrate that in the long run, the price transmissions between these two markets are nonlinear and asymmetric.

Notes

¹ As a general rule, the threshold value τ is unknown. The method adopted in Chan (1993) involves arranging the residual error terms in order, from small to large; the first and last 15% are then removed, and the middle 70% are selected. The sum of the squares of the residual errors is then minimized prior to the value of ω being determined.