Combining Stock Selection Models for Changing Chinese Market Regimes

Summary

This article describes combining two stock selection models designed for different Chinese equity market conditions: a rebound after an index decline, and a mild market pullback after stocks hit their daily price limit. Each model includes market environment features, so it selects stocks when its target conditions are present. If several models fit at once, the strategy ranks them by their historical backtest performance and gives priority to the stronger one.

The article illustrates possible inputs such as recent price range, limit-up activity, and the next-day average return of stocks that hit the limit. It reports that combining the two models reduced time spent out of the market and increased returns relative to using either model alone. It gives no numerical performance data or details on the backtest setup, so the claim cannot be assessed from the text. The author also notes that the combined strategy still has substantial idle periods and presents the material as an older implementation for study, not as a complete or current trading system.

Key ideas

• Build separate stock selection models for distinct market regimes.
• Include market environment features so each model acts only under its intended conditions.
• When multiple models qualify, prioritize selections using their historical backtest performance.
• The article reports less idle time and higher returns after combining two models, without providing numerical evidence.
• Additional models may be needed because the combined approach still spends substantial time out of the market.

Tags

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