Building and Validating Rule-Based Stock Selection Strategies
Summary
This article outlines a workflow for creating custom equity-selection rules from platform factors. It illustrates price-location features based on recent highs and lows, then combines return, volume, and market-relative conditions into a screening rule. The proposed process is to backtest a rule, inspect daily holdings and the best and worst outcomes, and adjust filters based on recurring characteristics. Multiple rules can be combined, ranked by historical performance, and paired with holding periods and trade-management settings.
The article also contrasts hand-coded rules with platform-generated AI strategies: custom rules are easier to inspect and adjust, while model-generated decisions may be less transparent. It recommends testing across additional years after tuning, acknowledging that optimizing against one year can bias results. Reported live and backtest performance claims are not independently substantiated in the text, and the described manual selection and repeated tuning leave substantial overfitting risk.
Key ideas
- Custom factors can describe a stock’s price relative to recent highs and lows.
- Selection rules can combine price movement, volume expansion, and performance relative to the market.
- The proposed tuning process reviews strong and weak holdings and changes filters based on recurring patterns.
- Combining and ranking several rules can help organize selections for different market conditions.
- Testing on additional years is recommended because tuning on one year can overfit its results.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.