A Weighted Rating for One-Minute Backtest Data Quality

Summary

The document gives a weighted formula for rating the quality of historical data used in one-minute strategy testing. It divides the available history into stages: bars before modeling begins, bars modeled on the nearest timeframe, bars modeled from minute data, and the remaining history. Each stage receives a different weight, with later minute-based data assigned the highest contribution. The resulting percentage is intended as a compact indicator of how much of a test’s history has higher-resolution modeling.

The article defines the bar indexes used in the calculation, including the test’s starting bar and the points where nearest-timeframe and minute modeling begin. It also notes that modeling starts no earlier than the 101st bar or the bar corresponding to the test’s initial date. The document supplies the formula and definitions but no worked example, empirical validation, or guidance on interpreting thresholds. Treat the rating as a platform-specific data-quality summary rather than a measure of strategy profitability or robustness.

Key ideas

• The rating weights historical segments according to their modeling resolution.
• Minute-modeled bars receive a greater weight than bars modeled on the nearest timeframe.
• The calculation uses the total history and the starting points of each modeling stage.
• The document defines the measure but provides no validation or decision thresholds.

Tags

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