A Natural-Language Stock Screen with Position Limits and Exit Rules

Summary

This article outlines a daily stock backtest framework in which a natural-language screening query selects entries. The example framework also allows a maximum number of holdings, a holding-period exit, and optional take-profit and stop-loss thresholds. Users can change the query and selected parameters, including commissions and slippage, to approximate trading costs.

The document gives example query types, ranging from valuation or quality filters to technical signals, and specifies a one-million-unit starting capital assumption. It does not present backtest results or explain execution timing, portfolio allocation, or how the framework handles overlapping signals. The stated thresholds and cost settings are examples rather than recommendations, and any results would depend on the query, parameter choices, data, and assumptions used.

Key ideas

• A natural-language stock query supplies the backtest's entry candidates.
• The framework includes configurable limits on holdings and holding duration.
• Exits may be based on elapsed time or optional profit and loss thresholds.
• Commission and slippage settings can be adjusted to represent trading costs.
• The article describes a framework but supplies no performance results or detailed execution methodology.

Tags

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