Historical Testing Limitations

Key limitations

Historical tests depend on past data, model assumptions, and simplified execution rules.

Not predictive

A strong historical report can still fail outside the test. Market regimes, liquidity, broker behavior, and user configuration can change.

Review these limits

• Data gaps and adjustments.
• Overfitting and parameter mining.
• Costs and slippage underestimation.
• Latency and partial fills.
• Broker API availability.
• User runtime configuration.

Safer next step

Treat every report as an experiment summary. Review, adjust assumptions, and use dry-run-first behavior in any self-hosted runtime.