walk-forward-validation

What problem does it solve?

This Skill provides a robust framework for backtesting trading strategies and machine learning models on financial time-series data, preventing common pitfalls like lookahead bias and overfitting.

Core Features & Use Cases

• Time-Series Aware Splits: Implements rolling and expanding window validation suitable for financial data.
• Overfit Detection: Includes methods like Deflated Sharpe Ratio (DSR) and Probability of Backtest Overfitting (PBO) to assess model reliability.
• Use Case: Evaluate a new algorithmic trading strategy by simulating its performance on historical market data, ensuring that the validation process accurately reflects real-world trading conditions and guards against false positives.

Quick Start

Use the walk-forward-validation skill to perform a rolling window backtest with a 90-day training size and a 14-day test size on the provided price data.