Using Features to Filter Stocks in BigQuant

Summary

This brief BigQuant forum exchange addresses how to build a simple stock screen and output the resulting stock codes as a list. The replies explain that feature extraction, an input feature list, and a code list should be used together, and that filtering requires features to be available to the filter step. The discussion also raises an example condition: identifying stocks whose current close is the lowest over roughly three years, expressed as a rolling minimum over 750 observations.

The post is mainly platform troubleshooting rather than a trading strategy. It offers no backtest, evidence that the example filter works as intended, or explanation of how the platform’s close fields are adjusted. The questioner notes that the output list seemed incorrect and that close fields did not match expected prices, so users would need to check feature definitions, data conventions, and module wiring in their own workflow. Its practical lesson is that a screening condition depends on both the chosen feature and the platform’s data and output configuration.

Key ideas

• BigQuant filtering requires input features to be available to the data filter module.
• Feature extraction and the code list should be connected as part of a working pipeline.
• A rolling minimum over 750 observations is offered as an example for screening current lows over about three years.
• The discussion does not verify the example or clarify the platform’s price field conventions.

Tags

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