Unexpected Financial Ratio Beats Popular Metrics at Predicting Stocks
-
Academics used AI and machine learning to test 29,000 financial ratios as stock market predictors.
-
The best performing metric was an unexpected one related to declines in sales contributions from acquisitions compared to rental expenses.
-
This ratio may work because it exploits negative market sentiment and temporary setbacks.
-
The study found many ratios perform as well as widely used academic theories like book-to-market.
-
Proprietary trading firms could profit from this strategy but it requires careful ongoing management.