World and regional factors in stock market returns

Purpose – This paper aims to test the hypothesis that the national stock market returns are driven by a world factor, regional factors and idiosyncratic factors, and to measure the importance of each factor. Design/methodology/approach – The state-space model is applied to describe the sample returns and estimate a world factor, regional factors and idiosyncratic factors by Kalman filtering. Weekly and daily returns calculated from MSCI country indexes from January 1988 to December 2004 of 11 national stock markets in four regions, i.e. North America (the USA and Canada), South America (Brazil, Mexico and Chile), Europe (the UK, Germany and France), and Asia (Japan, Hong Kong, and Singapore) are used. Findings – The results support the hypothesis that national market returns are driven by a world factor, regional factors and idiosyncratic factors. National markets do not always respond mainly to the world factor; regional factors and idiosyncratic factors play important roles as well. They also respond to world news at a slower rate than regional news. Research limitations/implications – This paper does not identify the source or origins of news directly but the factors are assumed as random variables and are estimated under certain strict assumptions. Originality/value – This paper applies Kalman filtering to estimate a world factor and regional factors and test the importance of each factor directly, an extension of previous studies that mostly showed strong independence among markets.