Weekly Research Insights
When Combining Anomalies Beats Machine Learning | Seasonality in Returns Around Earnings Season | Decoding Economic News with Large Language Models
Hi there. In this week’s edition of Research Insights, I cover three topics with practical relevance for investors.
First, I explore how combining well-known anomalies can outperform complex machine learning models. Second, I discuss new research showing that stock market returns vary meaningfully across the earnings cycle. Third, I highlight a new domain-specific language model that extracts sentiment and macro signals from economic text.
