This paper applies the methodology of Bai and Ng (2002, 2004) for decomposing large panel data into systematic and idiosyncratic components to both returns and turnover. Combining the methodology with a generalized-least-squares-based principal components procedure, we demonstrate that this approach works well for both returns and turnover despite the presence of severe heteroscedasticity and non-stationarity in turnover of individual stocks. We then test the duo-factor model of Lo and Wang's (2000), which is based on mutual fund separation. Our results indicate that trading due to systematic risk in returns can account for as much as 73% of all systematic turnover variation in the weekly time-series and 76% in
