Does RSI Actually Work? We Backtested It on 741 Assets

The honest answer

The RSI is the first indicator almost everyone learns, and the most over-trusted. So we ran RSI through the same out-of-sample, cost-included engine as every other indicator on 741 assets. The verdict: it's a niche tool, not a holy grail.

Across every asset we tested it on, RSI beat a simple buy-and-hold on just 11% of them, with a median Sharpe of 0.3.

What RSI actually is

The Relative Strength Index measures the speed of recent gains versus losses on a 0–100 scale; traders treat >70 as overbought and <30 as oversold. We test it as a trend filter (long while RSI holds above 50).

We test the standard, textbook version — no curve-fitting, no tuned settings per asset. That's deliberately a fair fight, and it's why these numbers are lower (and more honest) than the backtests you'll see in a course sales page.

Where it does work

It was the single best indicator we found for 1 of our 741 assets, and its strongest result was Virtuals Protocol (VIRTUAL) (+215.2% vs buy-and-hold).

Check the RSI page for the exact assets and timeframes it wins on.

The bottom line

RSI is a real tool, but it's not magic — like almost every single indicator, it loses to buy-and-hold on most assets. The edge, when it exists, is asset-specific. Don't trade it blind; check whether it actually works on your asset first.