Purpose - The purpose of this paper is to investigate the relationships between stock market returns of 13 countries based upon monthly data spanning December 1987 to April 2007. Design/methodology/approach - Specifically, the principal component (PC) and maximum likelihood (ML) methods are used to examine any discernable patterns of stock market co-movements. Findings - Factor analysis provides evidence that stock returns in a number of Asian countries are highly correlated and, based on the resulting robust factor loadings, they form the first well-defined common factor. The paper also finds consistent results (based on both the PC and ML methods) suggesting that the stock market returns of developed countries are also highly correlated, and constitute our second factor. Practical implications - The paper concludes that, inter alia, geographical proximity and the level of economic development do matter when it comes to co-movements of stock returns and that this has important implications for financial portfolio diversification if the aim is to reduce systematic risks across countries. Originality/value - Very few previous studies have investigated the benefits from portfolio diversification by using the PC and ML methods.