Abstract: We examine whether the collective trades and holdings of various market participants contain information about stocks. We use machine learning to capture nonlinearities and interactions in the relation between trades and holdings and future stock returns. The resulting predictor generates a long-short portfolio with significant out-of-sample alpha, forecasts future firm fundamentals, and assigns stocks on the right side of most anomalies. A factor model based on our predictor achieves higher Sharpe ratio than existing models. Our findings suggest that combining the trades and holdings of multiple participants and accounting for nonlinearities and interactions provides valuable information for price discovery.

Victor DeMiguel’s bio

Victor DeMiguel is a Professor at London Business School, which he joined in 2001 after earning a PhD from Stanford University and an MS from Universidad Politecnica de Madrid. Victor’s main research interest is portfolio selection and asset pricing in the presence of parameter uncertainty and market frictions. His papers have been published in top journals such as The Journal of Finance, The Review of Financial Studies, The Journal of Financial Economics, Management Science, and Operations Research. One of his most popular papers is “Optimal Versus Naive Di-versification: How Inefficient is the 1/N Portfolio Strategy”, which received the Best Paper Award from the Institute for Quantitative Investment Research and has more than 4,000 citations in Google Scholar. Victor serves as an Associate Editor of the journals Management Science and Operations Research and an external consultant to asset-management firms such as SYZ and Goldman Sachs.

Join Zoom Meeting

https://us06web.zoom.us/j/82659728268?pwd=V5mGsmLPTXhsvAdCf3mEzNPETViy6r.1

Meeting ID: 826 5972 8268

Passcode: 252155