Has Machine Learning Defeated Trend Following Strategies in the Chinese Futures Market
Liang (2024) studies momentum and trend following across several asset classes using a transformer over 2015–2023. The abstract mentions a Sharpe ratio, excess returns, profits and transaction costs but not a benchmark, an out-of-sample test, multiple testing or the size of the universe.
“Our findings indicate that machine learning models, particularly MLP, significantly outperformed traditional strategies, achieving higher annualized excess returns and Sharpe ratios.”
AssetMulti-assetfutures
StrategyMomentummachine learning
TechniqueTransformerrecurrent neural network
GeographyChina
Also stated: 2015–2023, Sharpe ratio, Excess return, Profit, Price and volume, costs
Missing from abstract: benchmark, out-of-sample test, 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 neighbourhoodMomentum and trend following by year
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 | |
|---|---|---|---|---|---|
| Enhancing Time Series Momentum Strategies Using Deep Neural Networks by Lim et al. | 2020 | ||||
| Artificial Intelligence and Multi-Indicator Confluence Frameworks for Cryptocurrency Trading Signal Optimisation: Evidence from Major Digital Assets by Kaodui et al. | 2026 | ||||
| Under the Veil of Time Series Momentum: Return Decomposition Approach by Jin and Ren | 2026 | ||||
| Multi-Asset A/A+ Momentum Fair Value Gap Strategy by Bindra | 2026 | ||||
| Volatility Targeting Is Trendy: How Trend Following Explains Alpha in Volatility-Managed Strategies by Hood and Raughtigan | 2024 |