Directional Close Bias with a Volatility Filter and ATR Exits

Summary

This price-action model measures the share of up-closing candles in a lookback window and compares the high-low range with a minimum threshold. It takes a long position when at least 60% of the closes are upward and the range is large enough; it takes a short position when the corresponding share is low and the range condition also holds. ATR sets stop and target distances, with a stated two-to-one reward-to-risk ratio. The supplied settings describe a BTC/USDT futures backtest on five-hour bars over roughly one year, but no returns or other results are reported.

The approach assumes that directional consistency during an active market may persist, though the document warns of reversals, parameter sensitivity, and differing behavior across market regimes. It proposes regime filters, adaptive thresholds, and risk-based position sizing as areas to explore. The source code enables its exit orders only once the maximum holding duration is reached, so its implementation may not behave like continuously active stop and target orders. The claims of adaptability and reduced overfitting are not supported by comparative test evidence here.

Key ideas

• The model counts upward-closing candles in a lookback window to estimate directional bias.
• It requires the recent price range to exceed a minimum threshold before entering.
• ATR defines stop and target distances, with a stated two-to-one reward-to-risk setting.
• The method assumes directional consistency may persist, but can fail during abrupt reversals.
• The described backtest settings include no performance results, and the source exit condition merits review.

Tags

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