Dual Linear Regression Trend Signals with an EMA Filter

Summary

This long-only trend-following approach subtracts a slower linear regression of closing prices from a faster one. A crossover of the difference above a threshold opens a long position when price is also above a 200-period exponential moving average. The strategy closes the long when the difference crosses back below the threshold; that exit rule does not require the EMA condition.

The document frames the regression spread as a way to identify shifts in trend and the EMA as a filter for weaker conditions. It describes risks including noisy signals from poorly chosen regression periods, missed trades when the EMA filter excludes strong moves, and whipsaws in ranging markets. The supplied configuration gives example periods and a BTC-USDT futures backtest window, but no results are reported. There is no stop-loss rule in the source; dynamic stops and parameter tuning are proposed as possible improvements, not tested findings.

Key ideas

• The signal is the difference between fast and slow linear regressions of closing prices.
• A cross above the threshold opens a long only when price is above the 200-period EMA.
• A cross below the threshold closes the long position.
• Poor parameter choices and sideways markets can produce noise and whipsaws.
• The document gives backtest settings but no performance metrics, and the source has no stop-loss rule.

Tags

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