Linear Regression RSI Crossovers for Mean-Reversion Trading

Summary

This strategy applies RSI to a linear-regression-transformed price series, then compares the result with an EMA. The stated defaults use a 200-period regression, a 21-period RSI, and a 50-period EMA. One entry option buys when RSI crosses above its EMA; another buys when RSI is already above both the EMA and an overbought threshold. A cross below the EMA closes the long position. The document presents this as a way to seek reversal opportunities, especially in ranges, while allowing different entry logic for different market conditions.

The supplied configuration describes a BTC_USDT futures test from January 2023 to January 2024, with hourly base data and daily strategy periods, but no performance results are reported. The notes warn that indicator lag can delay entries and exits, and that changes in the RSI–EMA relationship can produce poor signals. They suggest parameter tuning, additional filters, volatility-aware sizing, or automated optimization, while emphasizing that position size should limit losses. The source implements long entries and exits only; it does not define a short-entry rule.

Key ideas

• RSI is calculated from a linear-regression-transformed price series and smoothed with an EMA.
• The primary entry option buys on an RSI crossover above the EMA, with an alternate threshold-based option.
• A downward RSI–EMA crossover closes the long position.
• The published BTC futures test settings report no performance outcomes.
• Indicator lag and parameter sensitivity can impair timing and signal quality.

Tags

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