RSI divergence on M1: does it actually work?
Posted: Tue Oct 27, 2026 5:27 am
Divergence on a fast chart looks convincing. Price makes a higher high, RSI makes a lower high, and a reversal seems obvious. Then price keeps going.
One problem is subjectivity. Which swing counts? How far apart can the two highs be? Two traders looking at the same chart will mark different divergences. That makes it hard to test, and untestable ideas are hard to trust.
Another issue is that divergence can persist through strong trends, giving repeated false signals while momentum continues.
To test it fairly, define the rule exactly: swing detection method, minimum distance, RSI period, entry trigger, stop, target. Then run it on a large sample.
The same logic extends to many oscillator patterns. If the rule can't be written precisely enough for someone else to code it and reproduce the same trades, it's more of an opinion than a strategy. Opinions are fine for context, but they're hard to trade consistently.
Has divergence ever survived a proper test with fixed rules, or did it collapse once the subjectivity was removed?
One problem is subjectivity. Which swing counts? How far apart can the two highs be? Two traders looking at the same chart will mark different divergences. That makes it hard to test, and untestable ideas are hard to trust.
Another issue is that divergence can persist through strong trends, giving repeated false signals while momentum continues.
To test it fairly, define the rule exactly: swing detection method, minimum distance, RSI period, entry trigger, stop, target. Then run it on a large sample.
The same logic extends to many oscillator patterns. If the rule can't be written precisely enough for someone else to code it and reproduce the same trades, it's more of an opinion than a strategy. Opinions are fine for context, but they're hard to trade consistently.
Has divergence ever survived a proper test with fixed rules, or did it collapse once the subjectivity was removed?