Trend Following with Regression Slope and Residual Dispersion

Summary

This trend-following method fits a linear regression to a rolling window of prices and measures the fitted slope relative to the standard deviation of deviations from the fit. The resulting trend-confidence measure is compared with separate entry and exit thresholds. Crossing the long-entry threshold opens a long position, while falling through the long-exit threshold closes it; corresponding negative thresholds govern short trades. A percentage stop is also included in the described configuration.

The document argues that this measure can distinguish persistent linear movement from noisier price action, but it does not provide empirical results to substantiate claims about win rate or generalization. Published settings specify BTC/USDT futures and daily bars over a limited date range; the source also includes sample parameter values and a stop-loss setting. Trend reversals can still cause losses, and poorly chosen thresholds can lead to excess trading or missed opportunities. The text recommends testing across markets and timeframes and avoiding parameter overfitting; it suggests adaptive or machine-learning extensions as possible future work.

Key ideas

• Trend confidence is calculated as regression slope divided by the dispersion of prices around the fitted line.
• Separate thresholds govern long and short entries and exits.
• The example includes a percentage stop loss, but trend reversals can still produce losses.
• The document gives settings rather than measured performance evidence, so its claims need independent testing.

Tags

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