Can Machine Learning Improve Value Investing? Evidence from Japanese High Book-to-Market Stocks
Value investing in equities over 2006–2020, studied by Mukai in 2026 with tree ensembles. The abstract mentions excess returns, profits, t-statistics and an out-of-sample test but not a benchmark, transaction costs, multiple testing or the size of the universe.
“This study examines whether the machine learning models, Elastic Net and XGBoost, can outperform Piotroski's (2000) F-score in a value-investing strategy focused on Japanese high book-to-market (B/M) firms.”
AssetEquities
StrategyValue investing
TechniqueTree ensemble
GeographyJapan
Also stated: Annual, 2006–2020, Excess return, Profit, t-statistic, Long-short, Fundamentals, out-of-sample test
Missing from abstract: benchmark, costs, multiple testing, universe
Institutional-grade audit of this strategy.
£5 — Request the audit- Audit of the paper claims accounting for costs, luck, crises, originality, geography, etc.
- Reproduction code integratable into Bloomberg, IBKR, etc.
- Bias-free datasets for backtesting included
- Delivered in 2 working days
The paper's neighbourhoodValue investing by year
FamilyValue investing106Momentum and trend following289Factor investing240Machine learning return prediction15318 more · 896
AssetEquities89
The family is counted across the whole universe. Asset, method and authors are counted inside the family.
| Similar papers | citation | same method | same asset | same authors | |
|---|---|---|---|---|---|
| Testing Fama-French Factors in a Thin Market: Evidence from NGX-Listed Banks by Agbroko | 2026 | ||||
| Managerial Expected Growth and Value Premium by Yuan | 2026 | ||||
| What Do Factor Models Really Price in the Tails? A Noise-Corrected Quantile Reassessment of the Fama-French Five-Factor Model by Wang | 2026 | ||||
| Growth Value Two-Factor Model by Yeh and Hsu | 2025 | ||||
| Dirty Business: Transition Risk of Factor Portfolios by Jagannathan et al. | 2025 |