Predicting stock returns and improving out-of-sample forecasting for portfolio building

The top-tier Journal of Econometrics has published research on the development of a penalized two-pass regression with time-varying factor loadings - a procedure that resonates with existing structural approaches to big data and in panel econometrics. The empirical results indicate this method offers a better predictive performance of excess returns in asset and risk management and should help to improve the performance of time-varying portfolio allocation in asset selection. The paper is co-authored by Gaetan Bakalli (2021 GSEM Ph.D. Graduate), and GSEM Professors Stéphane Guerrier and Olivier Scaillet. It is published in the Journal of Econometrics, Themed issue on “Predictive financial modeling”.

Gaetan Bakalli and Olivier Scaillet received funding from the Swiss National Science Foundation, and Stéphane Guerrier received funding from the Swiss National Science Foundation and Innosuisse.

ABSTRACT

We develop a penalized two-pass regression with time-varying factor loadings. The penalization in the first pass enforces sparsity for the time-variation drivers while also maintaining compatibility with the no-arbitrage restrictions by regularizing appropriate groups of coefficients. The second pass delivers risk premia estimates to predict equity excess returns. Our Monte Carlo results and our empirical results on a large cross-sectional data set of US individual stocks show that penalization without grouping can yield to nearly all estimated time-varying models violating the no-arbitrage restrictions. Moreover, our results demonstrate that the proposed method reduces the prediction errors compared to a penalized approach without appropriate grouping or a time-invariant factor model.

The study is available open access: A penalized two-pass regression to predict stock returns with time-varying risk premia

> Click here to view the GSEM faculty’s publications in top-tier journals.