trading-strategy-backtest

To backtest a trading strategy from natural language, this Skill converts plain-text descriptions into a runnable Python and pandas script, generating an equity curve, trades, and a summary for an HTML dashboard.

Event-driven backtesting here prevents look-ahead bias by enforcing next-day-open execution, ensuring trades only act on information available after the signal triggers without using future data.

Yes, you can run portfolio backtests across multiple symbols and date ranges with pandas, handling stock screening and portfolio construction logic directly from natural-language inputs.

The best way to generate a performance dashboard is producing standard backtest artifacts like equity curves, trades, and summary JSON, which render into a shareable HTML dashboard for insights.

Yes, this backtesting approach supports warmup handling and stock screening, accommodating event-driven signals and market-specific rules across A/HK/U.S. markets as described in the referenced rules.

Limitations of natural-language backtests include potential interpretation biases, which the Skill mitigates by providing an auditable narrative of methods and limitations for responsible analysis.