Kai Cao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Kai Cao has a total of 1 co-authors. Here are the top co-authors by number of shared publications:
Haifeng You
Hong Kong University of Science & Technology (HKUST) - Department of Accounting
1 shared publication
Latest Publications
- Open Access
- 20 September 2020
- Published by Elsevier in SSRN Electronic Journal
Fundamental Analysis Via Machine Learning
We examine the efficacy of machine learning in one of the most important tasks in fundamental analysis, forecasting corporate earnings. Our analyses show that machine learning models, especially those that accommodate nonlinearities, generate significantly more accurate and informative forecasts than a host of state-of-the-art earnings prediction models in the extant literature. Further analysis suggests that machine learning models uncover economically sensible relationships between historical financial information and future earnings. We also find that the new information uncovered by machine learning models is of considerable economic significance to investors. The new information component of the machine learning-based forecasts is significantly associated with future stock returns. Stocks in the quintiles with the most favorable new information outperform those in the least favorable quintiles by approximately 70 bps per month, suggesting that the new information is not well understood by investors. Finally, insights from machine learning models are useful for improving the extant models.
- 20 September 2020
- Published by Elsevier in SSRN Electronic Journal
Open AccessFundamental Analysis Via Machine Learning
We examine the efficacy of machine learning in one of the most important tasks in fundamental analysis, forecasting corporate earnings. Our analyses show that machine learning models, especially those that accommodate nonlinearities, generate significantly more accurate and informative forecasts than a host of state-of-the-art earnings prediction models in the extant literature. Further analysis suggests that machine learning models uncover economically sensible relationships between historical financial information and future earnings. We also find that the new information uncovered by machine learning models is of considerable economic significance to investors. The new information component of the machine learning-based forecasts is significantly associated with future stock returns. Stocks in the quintiles with the most favorable new information outperform those in the least favorable quintiles by approximately 70 bps per month, suggesting that the new information is not well understood by investors. Finally, insights from machine learning models are useful for improving the extant models.