Price makes a higher high, RSI makes a lower one, and you short it. Then price keeps running and takes out your stop. The AFM Research Lab put RSI divergence through 2,478 trades over ten years on 28 pairs. The textbook rules failed, and so did the usual fixes like extra confirmation or new RSI settings. This video covers the one version that held up. It starts with the higher timeframe trend, which has to allow the trade. The divergence also has to fire at a daily or weekly level the market has respected before. Once it does, price needs to confirm within three candles. The stop goes 2 ATR beyond the swing, with at least 1.5R to the next level. In testing, location alone took the win rate from 33 to 41 percent. Ezekiel finishes with a EURUSD setup that qualifies and one you should skip.
