Testing ATR Deviations and Drift for Long Mean-Reversion Entries

Summary

This strategy compares fast and slow ATR estimates using a two-sample significance test, treating a sufficiently large difference as evidence of unusual volatility. It then estimates directional drift from a moving average of log returns and enters long when the ATR deviation condition coincides with improving drift. A trailing stop adjusts with ATR and exits when price falls through the stop level.

The document presents hypothesis testing as a way to formalize volatility deviations, but its test-statistic description and implementation deserve careful validation before use. Trend estimation can be wrong, parameter choices affect signals, and a sharp price gap can bypass a stop. The published settings describe a daily BTC/USDT futures backtest over approximately one year, but provide no outcome statistics. The stated logic and code therefore outline a research approach rather than evidence of profitable performance.

Key ideas

• The entry condition combines a statistically significant difference between fast and slow ATR with improving estimated drift.
• Drift is estimated from a moving average of logarithmic returns.
• A trailing stop is updated using ATR and is intended to limit downside.
• The method remains exposed to mistaken trend estimates, sensitive parameters, and abrupt price gaps.
• Published backtest settings are provided without performance results.

Tags

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