Fractal Co-integration of RMB Exchange Rate and China´s Stock Price

Nonlinear co-integration method is discussed. According to the generalized fractal co-integration relationship, one form of nonlinear co-integration is proposed based on GPH (Geweke, Porter-Hudak) method. Using the tests of the long-memory characteristic in finance time series, based on daily price series of RMB exchange rate and Shanghai stock markets from July 22, 2005 to April 24, 2009, empirical analysis of their relationship is performed with linear and nonlinear fractional co-integration method. As the results shown, Shanghai stock market and RMB/USD exchange rate series both show the characteristic of long-memory which implicates that China´s stock markets and RMB exchange rate are non-efficient. Furthermore, the empirical results ague that there tend to be a long-run linear equilibrium relationship and exist bidirectional causal relationship between China´s stock market and RMB exchange market.