Nadaraya-Watson Regression and ATR Channels for Trend Trading

Summary

This strategy combines two lagged Nadaraya-Watson kernel regression curves with an ATR channel around one curve. A crossover between the curves is described as a trend-direction signal, while channel breaks provide trade triggers: the overview describes buying below the lower band and exiting above the upper band, and the detailed logic also describes short entries above the upper band. A position is closed after a configured run of closes below its average entry price. The supplied settings include regression, lag, ATR, and stop-check parameters, and the published test uses BTC/USDT futures over a limited period.

The document argues that regression and volatility bands can help identify trend and entry levels, but it provides no performance statistics to substantiate those claims. Channel breaks can occur during reversals or consolidation, and results may depend heavily on kernel and band parameters. The source also plots EMA lines that do not appear to filter entries, and its close logic and prose do not fully align on short-side exits. These implementation details should be checked before interpreting a backtest or reproducing the approach.

Key ideas

• Two lagged kernel regression estimates are compared to infer directional bias.
• ATR bands around a regression estimate define price breakout thresholds.
• The description and source differ on whether upper-band breaks initiate shorts or only close longs.
• A consecutive-close condition is used to exit a losing long position.
• The document gives parameter examples and test settings, but no performance evidence.

Tags

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