What Factors Drive Trading around the World?

Abstract

This paper examines the implications of Lo and Wang’s (2000, 2006) mutual fund separation model in the cross-sectional behavior of global trading activity. It demonstrates that return-based factors work poorly around the world. On average across countries, market-wide turnover accounts for 37% of all systematic turnover components in individual stock trading, and two additional Fama and French (1993) factor turnovers increase the explanatory power by 23%. Similarly Lo and Wang’s (2000) turnovers account for on average only 64% of all systematic turnover components. Using this multifactor asset pricing-trading framework, a horse race is further performed to explore other factors in return which are important for explaining the common variation in turnover. All the return-based factors account for at most 67% of the common variation in trading, suggesting that stock pricing and trading volume may not be compatible around the world. In a cross-country analysis, the explanatory power of the return-based factor model varies substantially across countries and markets, with better performance for developed European markets and China. Surprisingly, in North America, Japan, and most emerging markets there are larger amounts of commonality in trading, usually higher than 47%, for reasons other than return motive.