Post Nafta Integration of North American Stock Markets : Implications for Financial Decision Making

This paper investigates the degree of integration of in the North American equity markets in the post NAFTA period, 1994 to 2006 and two sub post NAFTA period: 1994-1999 and 2000-2006 using daily stock closing price(s) indices. The Johansen and Juselius (1990) method for determining the presence of co-integration vectors in a set of non stationary time series was applied to the natural logs of the daily closing stock prices of North American equity markets using the following representative indexes: US S&P 500; Canada S&P 500; and Mexico, IPC, MXX. We find that the correlation of returns between the US S&P 500 and the Canadian S&P 500 is statistically insignificant while the correlation between the US S&P 500 and the Mexican IPC, MXX is statistically significant at the 10 percent level. The results of the ADF unit root test indicate that each of the series is stationary in their log first differences. The Maximum-eigenvalue and Trace test indicate the presence of one co-integrating (equation) vector at the 5 percent level in the case of the full sample and also during the two sub-sample periods (all the sample data). These results lead us to the conclusion of the presence a long-run equilibrium relationship (co-integration) among the North American equity markets in the post NAFTA period. Also, there is an indication that the degree of integration of the three equity markets remained the same immediately following the NAFTA implementation (1994-1999) and also during 2000-2006 as there is no change in the number of cointegrating vectors although the estimated coefficients appear to be quite different for the two sub periods