This design proposal presents a framework for enhancing the Relative Strength Index (RSI) by incorporating macroeconomic liquidity conditions and market sentiment. The traditional RSI, while widely used, suffers from well-documented limitations including false signals during trending markets and fixed thresholds that fail to adapt to changing market conditions. The core design principle is to preserve the canonical RSI scale [0,100] while adapting interpretation through dynamic thresholds and conditional normalization, maintaining interpretability for practitioners familiar with traditional RSI while systematically accounting for the external factors --- global liquidity and news sentiment --- that influence asset prices.

This paper is implemented live on the site as the Enhanced RSI tool, which turns the paper's context-aware momentum framework into an interactive page.How it's implemented. The standard 14-period Wilder RSI is recomputed as a regime-conditional signal. Instead of fixed 70/30 bands, the tool conditions both the mean and the spread of RSI on three live market regimes — M2 money-supply growth (pulled from FRED), a momentum-based sentiment proxy, and a realised-volatility percentile — then reports a conditional z-score mapped back onto the familiar 0–100 scale. Concretely it uses \( \mu_{RSI|R} = 50 + \beta_{M2}(M2\% - 50) + \beta_{sent}(sent\% - 50) + \beta_{vol}(vol\% - 50) \) and \( \sigma_{RSI|R} = 20\,(1 + 0.2\,(vol\% - 50)/100) \), so a z-score near 0 means "typical for the current regime" while the dynamic overbought/oversold bands widen or tighten as conditions change.Why. A raw RSI of 70 means very different things in a low-volatility, tightening-liquidity tape versus a high-volatility, easy-money melt-up. Conditioning on macro and risk regimes makes the indicator comparable across environments and cuts the false signals a static threshold produces.The feature has three parts: a live dashboard, a backtest that runs causal, commission-aware walk-forward tests over four strategy variants (Standard, M2-enhanced, Sentiment-enhanced, and Full context-aware), and a comparison view benchmarking all four on total return, CAGR, Sharpe ratio, max drawdown and win rate.