Screening Stocks with Recent Limit-Ups, MACD, and Price Amplitude

Summary

This post presents a Chinese stock screen built around price amplitude above 1, MACD below zero two days earlier, and more than two limit-up sessions within ten days. It frames the combination as a short-term technical and market-sentiment filter. The final selection description shifts the limit-up condition to a ten-day gain threshold above 20%, so the post does not maintain one fully consistent rule throughout.

The article identifies several weaknesses: the filter may overlook company fundamentals, its simple conditions may be prone to overfitting, and stocks that hit daily limits can behave unstably. It recommends adding fundamental quality criteria or using predictive models, and suggests using cumulative price gains instead of counting limit-up events. Formula and Python examples are included, but the code also differs from the prose in details such as how amplitude and the MACD condition are calculated. No backtest or performance evidence is presented, so the screen should be treated as an unvalidated idea with implementation ambiguities.

Key ideas

• The initial screen combines amplitude above 1, MACD below zero two days earlier, and more than two limit-up sessions in ten days.
• The final rule instead describes a ten-day gain above 20%, creating an inconsistency in the stated signal.
• The post warns that sentiment-heavy filters can neglect fundamentals, overfit, and produce unstable selections.
• It suggests adding fundamental filters, using predictive models, or replacing limit-up counts with return measures.
• The code examples contain implementation choices that do not align cleanly with the prose, and no performance results are supplied.

Tags

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