Title: Predicting Future Earnings Trends Using Financial Factors and Machine Learning

Speaker: Jiang Zhaohui, Postdoctoral Researcher, School of Economics, Shandong University, PhD from University of York. His research focuses on the application of machine learning in asset pricing and behavioral finance.

Abstract: We evaluate a range of machine learning methods to predict the future direction of earnings across a large sample of 21,388 U.S. firms. Our approach utilizes custom-built asset pricing factors, which demonstrate greater predictive efficiency than traditional market price-related factors. We show that training models on one-year-ahead excess earnings direction, rather than solely on contemporaneous earnings direction, leads to improved out-of-sample performance. The proposed models achieve an accuracy of up to 70%, surpassing the performance reported in prior studies. Additionally, our results indicate that artificial neural networks significantly outperform both random forests and partial least squares regression. These findings offer valuable insights for stakeholders seeking to better identify and anticipate changes in firm profitability.

Date & time: 04 June 2026, 12:15 - 13:15

Venue: B321, Zhixin Building, Central Campus, SDU