We build a cross-sectional factor model for investors’ direct stockholdings, by analogy with standard time-series factor models for stock returns. We estimate the model using data on monthly direct equity holdings and trades using data from almost 10 million accounts in the Indian stock market. We find that household-specific factors such as the size of the equity portfolio and the number of stocks it contains are important predictors of individual stockholdings. Factors that capture the popularity, value, and size tilts of investors’ portfolios also help predict the stocks they hold. We use our new factor model to analyze the empirical equity “coholdings matrix”, which measures the degree to which stocks are jointly held in investor portfolios and thereby captures investor clientele effects.