Faculty Scholarship 1994 - Present

Co-Movements of NAFTA Stock Markets

The North American Free Trade Agreement (NAFTA) was signed on December 17, 1992. This agreement between the U.S., Canada, and Mexico created the world�s biggest single market by eliminating all trade and investment barriers among the three countries. The NAFTA had to be carefully crafted to be compatible with other agreements, particularly the General Agreement on Tariffs and Trade (GATT), to which all three members were included as parties (NAFTA, Article 101). NAFTA took effect on January 1, 1994. It immediately lifted tariffs on the majority of goods produced by the three nations. It also provides for gradual elimination, over the course of 15 years, of most remaining barriers to cross-border investment and to the movement of goods and services between the territories of the three countries. Unlike the E.U., the NAFTA does not create full economic integration or adopt a common external trade policy among its members. All of the NAFTA countries have a free market economy. The U.S. has the largest economy of the three consisting of private companies with modern technological advances. Canada has a market economy similar to the U.S. Mexico has the least developed economy of the NAFTA members. It contains both outdated and modern industries with an increasing emphasis on private enterprise. Mexico continues to work on building an adequate infrastructure to help modernize the economy. The U.S.�s GDP is over 10 times greater than both Canada and Mexico and also has the largest growth rate in GDP. The GDP is slightly higher for Canada but Mexico has a higher growth rate. The GDP per capita is over $40,000 in the U.S. followed by $35,000 in Canada. Mexico�s GDP per capita is significantly lower at $9,600. The makeup of the GDP per country among agriculture, industry, and services are similar across countries although the U.S. has a lower industrial sector and higher service sector relative to the other two economies. Mexico has the highest inflation rate. Mexico also has the lowest unemployment rate although it is reported that a large percentage of workers in Mexico are underemployed. The co-movements of the world�s national equity market index returns have long been a popular research topic in the finance literature. Low correlation between national stock markets is often presented as evidence in support of the benefit of global portfolio diversification. Although the co-movements of the world�s stock markets have been studied extensively, the co-movements of stock markets in a common market or free trade area have not received sufficient attention. Our study provides empirical evidence on the co-movements of the NAFTA stock markets by using rolling correlation analysis and Granger causality tests. We use the Morgan Stanley Capital International (MSCI) weekly U.S.-dollar stock market indexes in the study. The index data are drawn from the Datastream database for the June 1995-May 2005 period. The weekly index returns are computed as the natural log difference in the indexes, ln (Ii,t/ Ii,t-l). The weekly average index returns and the volatilities of the NAFTA stock markets are compared. Rolling correlation analysis is used to study the time-varying correlation patterns between pairs of NAFTA stock markets and the portfolio diversification implications of the long-run trends. The Granger causality test is used to study the predictability and the weak-form efficiency of each stock market�s weekly index returns. The returns data indicate that the Mexican stock market has the highest average weekly return during the June 1995-May 2005 period. The Canadian stock market has the second highest average weekly return and the U.S. stock market has the lowest average weekly return during the same period. The standard deviation figures indicate that the Mexican stock market has the highest volatility, the Canadian stock market has the second highest volatility, and the U.S. stock market has the lowest volatility. The Canadian stock market has the highest return per unit of volatility risk, the U.S. stock market has the second highest return per unit of volatility risk, and the Mexican stock market has the lowest return per unit of volatility risk. For the global investor, the expected portfolio diversification benefit of investing in different countries depends on the expected correlation between the stock markets of the countries. However, recent studies demonstrate that correlation coefficients between global stock markets can give an idea about the degree of uncertainty involved regarding the expected correlation coefficients. We use the rolling correlation analysis technique to study the time-varying correlation between the U.S., Canadian, and Mexican weekly stock market index returns during the June 1995-May 2005 period. Yearly rolling correlation coefficients between the weekly index returns of pairs of stock markets are computed by rolling the sample period ahead one week at a time starting with the first year. Specifically, the latest weekly return observation is added while the earliest observation is deleted. The rolling correlation analysis results show that the correlation of the U.S. stock market with the Canadian and Mexican stock markets is increasing during the June 1995-May 2005 period. It implies that the portfolio diversification benefit to U.S. global investors with these two NAFTA stock markets has been decreasing. The correlation coefficient between the Canadian and Mexican stock markets decreased sharply from the 1995-1998 period to the 2002-2005 period. This implies that the portfolio diversification benefit of investing in each other�s stock markets has increased for Canadian and Mexican investors. An independent variable X Granger-causes changes in dependent variable Y, if Y can be better forecasted with past values of X and Y, than just with past values of Y alone. The causality in the Granger sense does not imply a cause and effect relationship, but one of predictability. In several recent studies, the Granger-causality technique is used to determine if some national stock market index returns can be used to predict the future returns of other national stock market indexes. We use the Granger test to study the linkages between the U.S., Canadian, and Mexican stock markets during the June 1995-May 2005 period. Our Granger causality test results with weekly returns data indicate that all three NAFTA stock markets are weak-form efficient, i.e., the past returns of none of the three stock markets can be used to predict its future returns. Neither the past returns of the U.S. stock market can predict the future returns of the Canadian and Mexican stock markets nor the past returns of the Canadian and Mexican stock markets can predict the future returns of the U.S. stock market. The past returns of the Mexican stock market cannot predict the future returns of the Canadian stock market. However, the past returns of the Canadian stock market can be used to predict the future returns of the Mexican stock market at the conventional five-percent significance level.