Building and Combining Quantitative Stock Selection Rules
Summary
This Chinese-language article outlines a six-step process for developing stock selection rules: find an idea, define its logic, run an initial backtest, tune parameters, check for overfitting, and select strategies for combination. Its example setups look for stocks that pull back after consecutive limit-up sessions and close higher on the day, aiming to capture another rise. It also describes combining several models to increase returns and reduce time out of the market.
The article lists example inputs, including limit-up status, the number of rising A-shares, recent high returns, lower-shadow length on bullish candles, and large-order flows. These examples suggest ways to turn price and trading activity into selection factors. However, the article does not provide full rules, backtest results, validation procedures, or enough implementation detail to reproduce the models. It presents the material as an introduction to a training session, so the strategy claims cannot be assessed from the text alone.
Key ideas
- A proposed workflow moves from strategy inspiration through rule design, backtesting, parameter tuning, overfitting checks, and strategy selection.
- One example looks for a pullback after consecutive limit-up sessions followed by a bullish close.
- Suggested selection inputs include recent returns, market breadth, candle shape, and large-order flow.
- Combining several models is presented as a way to reduce idle capital, but supporting performance evidence is not included.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.