Tested strategies Verdict: No edge

Do RSI strategies actually work?

By Otto Research 2 min read Tested across 17 years, with real costs

The relative strength index reads from 0 to 100, comparing recent gains to recent losses. The classic rule buys when it falls below 30, oversold, and sells when it rises above 70, overbought. Variants use divergence between price and the indicator, or trade the crossing of the 50 line as a trend signal.

The claim is that extreme readings mark exhausted moves that are about to reverse.

We tested RSI rules, overbought and oversold levels across a range of thresholds and periods, and divergence, across 17 years of data with real trading costs, and found no edge. Oversold markets kept falling as often as they bounced; overbought ones kept rising.

The reason is that an oscillator measures how far price has moved recently, and a market that has moved far is as likely to be trending as to be stretched. The indicator cannot tell the two apart, and the rule pays the cost of guessing.

If you trade it anyway, test it in a defined market state rather than everywhere, and count the results in trends separately from ranges.

The ones that survived

See what Otto trades instead.

Most strategies do not survive an honest test. Otto is built on the few that did, and every trade is recorded live on Karnek.