Backtesting Frameworks

What it does

This skill outlines the design and implementation patterns for robust backtesting systems. It covers point-in-time data handling, avoidance of look-ahead and survivorship bias, transaction cost modelling, event-driven simulation, and walk-forward analysis. The guidance is aimed at researchers and engineers building reproducible performance evaluation pipelines rather than giving trading advice.

When to use it

Use when developing or validating trading strategies, building data pipelines for historical simulation, running sensitivity analyses, or preparing research-grade backtests that can withstand audit and review. Do NOT use it for live trading execution or to provide financial advice.

What's included

• Scripts: none included (has_scripts=false)
• References: points to an implementation playbook in resources when present (has_references=false in this fetch)
• Instructions: stepwise checklist for hypothesis definition, data curation, cost modelling, split strategies, and simulation validation.

Compatible agents

Best used by research-assisting agents that can help with code generation and validation (Copilot-like tools, Claude Code, Gemini). The content is methodology-focused and complements toolchains for Python/R numerical work.