Bayesian RSI Signals from Price and Momentum Probabilities

Summary

This strategy estimates whether price movement and subsequent RSI movement support continuation, then uses a Bayesian-style probability score to drive long and short signals. It counts historical price moves over a lookback and checks RSI movement over a later range, forming separate upward and downward scores. Crossovers of smoothed scores trigger entries, with the described setup also using opposing signals to exit or reverse positions.

The document gives parameter defaults and published backtest settings for BTC/USDT futures, but reports no performance statistics. Its explanation presents probability combination as improving judgment without showing calibration, out-of-sample validation, or evidence that the probabilities predict returns. Results may depend heavily on lookback choices and market regime; limited samples, unusual events, and RSI failures are acknowledged risks. The code's signal logic and the prose description are not fully aligned, so the method should be treated as an outline requiring careful verification before evaluation.

Key ideas

• The strategy counts historical price changes and RSI changes to estimate directional probabilities.
• It combines the estimates with a Bayesian-style calculation to create smoothed signals.
• Crossovers of the upward and downward scores are used to enter long or short positions.
• The document provides BTC/USDT futures backtest settings but no performance results.
• Small samples, parameter sensitivity, and divergence between the prose and signal logic limit confidence.

Tags

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