Historical Technical Indicators and the Risk of Look-Ahead Leakage

Summary

This exchange asks whether MACD components and a short-window swing-volatility feature used in model training and prediction contain future information. The response says these factors are calculated from historical data, implying that the indicators themselves need not use future observations. The examples include MACD, its histogram and signal line, and a volatility feature.

That answer is limited: a historical formula does not by itself guarantee a leak-free dataset. Whether a feature is valid depends on when its input prices become available, how the feature is aligned with the prediction target, and whether preprocessing or data joins introduce later information. The document gives no formulas, code, timing conventions, or checks for leakage, and it reports no empirical test. It is a brief conceptual reassurance rather than a complete audit method, so users should verify feature timestamps and the full training pipeline before relying on the claim.

Key ideas

• The response identifies MACD components and a volatility feature as indicators calculated from historical data.
• Features based on historical prices can still leak information if their timestamps are misaligned.
• A leak check should account for data availability, target timing, preprocessing, and joins.
• The excerpt gives no formulas or validation procedure for ruling out future information.

Tags

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