Filtering Chinese Stocks by Turnover, Price Change, and Large-Order Flow

Summary

This Chinese-language post presents a short-term stock screen using turnover, daily price change, auction return, trading volume, circulating market value, and large-order net flow. The main description calls for turnover between 3% and 12%, a negative product of daily price change and large-order net flow, and an auction return between -2% and 5%. The accompanying formula and Python example add filters for volume, market value, and eligible stock codes.

The screen is intended to avoid extreme auction moves and combine liquidity with order-flow information. The post cautions that it omits fundamentals and longer-term performance, creating quality and overfitting risks; it suggests adding valuation, company-size, market-context, or technical filters. It provides no backtest or returns evidence. Some stated thresholds and implementations are not fully consistent: for example, the prose and code express auction-return bounds differently, so the specification should be checked before use.

Key ideas

• The proposed screen combines turnover, daily price change, large-order net flow, and auction return.
• The formula and sample implementation add volume, circulating market value, and stock-code restrictions.
• The post frames the filters as a way to avoid extreme price moves while considering trading activity.
• The author warns that short-term price signals omit fundamentals and may overfit.
• The prose and sample code differ in some threshold details, and no performance validation is shown.

Tags

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