Combining RSI, ADX, and Linear Regression for Trend Signals

Summary

This strategy combines RSI thresholds, an ADX strength filter, and a rolling linear regression estimate to form directional signals. A long setup uses RSI crossing above the oversold level with ADX above its threshold and the predicted close above the current close; a short setup uses RSI crossing below the overbought level, strong ADX, and a prediction below the current close. The regression uses a configurable lookback to estimate a line from recent price data and project a price for the current bar.

The document lists adjustable RSI, ADX, and regression lookback settings and a published backtest configuration for SOL/USDT on Binance over about a month, with hourly bars. It supplies no performance statistics, so it does not establish predictive accuracy or profitability. The description calls the regression component machine learning, but it is a simple linear fit, and the text's five-minute framing differs from the hourly backtest configuration. The stated limitations include indicator lag, parameter sensitivity, overfitting, and weaker suitability in ranging markets.

Key ideas

• RSI threshold crossings provide candidate long and short signals.
• ADX must exceed a set level before the strategy permits trades.
• A rolling linear regression estimate is compared with the current close to set directional bias.
• The backtest configuration concerns SOL/USDT on hourly bars and includes no reported results.
• The regression is a simple statistical fit, and the document flags lag, overfitting, and range-bound conditions as risks.

Tags

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