AI-Ranked China A-Share Strategy Combining Factors and Trend Timing

Summary

This strategy note describes a China A-share approach that blends model-based stock ranking with technical timing. It begins with a manually defined pool of roughly 100–300 large-cap names associated with the CSI 150 universe, then ranks candidates using quantitative signals. Timing filters include MACD strength, price above its 25-day average, and relative returns over short and longer windows. The author also describes turnover and large-order money-flow inputs, plus an index MACD condition as a market risk trigger.

Portfolio rules include five holdings, five-day rotation, a per-stock allocation cap of 10%, and exits at stated profit and loss thresholds. The post explains the design choices but supplies no visible performance series, benchmark comparison, test period, transaction-cost assumptions, or validation of the model’s predictive value; the backtest section is effectively empty. The proposed stock universe, factor behavior, and market regime may limit generalization. The article’s claims about robustness should therefore be treated as hypotheses, especially because model selection, survivorship, and execution effects are not discussed.

Key ideas

• The approach ranks stocks within a restricted large-cap universe using an AI model and quantitative factors.
• MACD, a moving-average condition, and relative returns are used to time entries or filter candidates.
• Turnover and large-order money flow are included to represent liquidity and buying pressure.
• A CSI 300 MACD condition acts as a broad-market risk trigger.
• The portfolio rules specify five holdings, five-day rotation, an allocation cap, and profit and loss exits.
• The document describes design choices but provides no usable backtest results or validation details.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.