Measuring the effects of geographical distance on stock market correlation

Abstract

Recent studies suggest that the correlation of stock returns increases with decreasing geographical distance. However, there is some debate on the appropriate methodology for measuring the effects of distance on correlation. We modify a regression approach suggested in the literature and complement it with an approach from spatial statistics, the mark correlation function. For the stocks contained in the S&P 500 that we examine, both approaches lead to similar results: correlation increases with decreasing distance. Contrary to previous studies, however, we find that differences in distance do not matter much once the firms’ headquarters are more than 40 miles apart, or separated through a federal border. Finally, we show through simulations that distance can significantly affect portfolio risk. Investors wishing to exploit local information should be aware that local portfolios are relatively risky.