Key results
- Total return
- +16.5%
- Buy & hold: +136.4%
- Annual return (CAGR)
- +1.5%
- Buy & hold: +9%
- Max drawdown
- -14.1%
- Buy & hold: -41.9%
- Win rate
- 38.68%
- Profit factor
- 1.16
- Trades
- 106
- Longest losing streak
- 8 trades
- Period
- Oct 4, 2016 to Oct 1, 2026
- 10 years
- Market
- DAX/EUR
- 4-hour
Equity curve vs buy & hold
Returns year by year
| Year | Return | Trades |
|---|---|---|
| 2016 | +4.3% | 1 |
| 2017 | -4.8% | 12 |
| 2018 | +14.7% | 10 |
| 2019 | -3.2% | 17 |
| 2020 | +5.2% | 8 |
| 2021 | +2.4% | 7 |
| 2022 | -0.4% | 12 |
| 2023 | -3.7% | 10 |
| 2024 | +0.9% | 11 |
| 2025 | +0.5% | 10 |
| 2026 | +1% | 8 |
Calendar years, compounded from monthly results. The first and last years can be partial.
The Relative Strength Index was introduced by J. Welles Wilder in 1978, and divergence is one of its most popular readings: when the price makes a new low but the RSI does not, the fall may be running out of steam. On its own, divergence fires often and in every kind of market. This version adds a location filter borrowed from smart money concepts: bullish divergences count only in the "discount" half of the recent range, and bearish ones only in the "premium" half. The goal is to buy weakness near the bottom of the range and sell strength near the top, the classic mean-reversion logic described, for instance, by Quantified Strategies.
The rules we tested
- Market and period: DAX (DAX/EUR), 4-hour candles, October 2016 to October 2026.
- Long entry: a bullish divergence on the 14-period RSI (high sensitivity setting) appeared at least once in the last 10 candles, and the price is in the discount zone, the lower half of the range of the last 100 candles.
- Short entry: a bearish RSI divergence appeared at least once in the last 10 candles, and the price is in the premium zone, the upper half of that range.
- Stop and target: a 1.5% stop-loss and a 3% take-profit (2:1). No separate exit rule.
- Sizing and costs: 100% of equity per trade, 0.02% commission per side, $10,000 starting capital.
What the backtest shows
The bracket did what it was designed to do. The average win was 2.85% and the average loss 1.54%, close to the 3% and 1.5% levels once costs are included. With that payoff the break-even win rate is about 35%, and the strategy won 38.7%. Of the 106 trades, 38 hit the target, 65 hit the stop and 3 closed on a signal. That margin of three or four points gives a profit factor of 1.16 and +16.5% over ten years.
The yearly table is where to look before trusting it. 2018 alone returned +14.7%, with 7 winners out of 10 trades. The other years, compounded together, add up to roughly +1.5%. Several were negative: 2017 (-4.8%), 2019 (-3.2%) and 2023 (-3.7%). The longest losing streak was 8 trades.
Against the index, the gap is wide: +16.5% versus +136.5% for buy and hold. The strategy did take far less risk, with a worst drawdown of 14.1% against 41.9% for the index, but most of that comes from being in the market only around ten times a year.
Why it works, and where it falls short
Combining divergence with location makes sense: a bullish divergence near the top of a range is often just a pause in a rally, while the same signal near the bottom has room to play out. The positive profit factor suggests the filter does add something. But a 2:1 bracket on an index that trended up strongly for most of the decade is a tough setting: shorts taken in the premium zone fight the trend, and long trades are capped at 3% while the index keeps climbing. One strong year carrying the whole result is also a warning that the edge may be a matter of a few market phases rather than a stable property.
What you could test next
- Use the normal sensitivity setting to keep fewer, more pronounced divergences.
- Shorten the window from 10 candles to 5 so the entry stays close to the divergence.
- Trade longs only, in the direction of the DAX's long-term trend, and compare.
- Try a wider target (for example 4.5% for a 1.5% stop) to see whether winners have more room.
Methodology and assumptions
- Starting capital
- 10,000
- Position size
- 100% of equity
- Commission
- 0.02% per side
- Data window
- Oct 4, 2016 to Oct 1, 2026
- Run date
- October 2, 2026
- No repainting: every signal is computed on closed candles only, so the backtest never acts on a price a trader could not have seen yet.
- Past performance does not predict future results. A backtest is a historical simulation, not a forecast.
Frequently asked questions
Is RSI divergence a profitable strategy?
In this test on the DAX, slightly: +16.5% over ten years with a profit factor of 1.16. Almost all of it came from 2018 (+14.7%); the other years taken together added about 1.5%.
What is a bullish RSI divergence?
The price makes a lower low while the RSI makes a higher low. It suggests the selling pressure is fading even though the price is still falling. A bearish divergence is the mirror: a higher high in price with a lower high in the RSI.
What stop-loss and take-profit should I use with RSI divergence?
We used a 1.5% stop-loss and a 3% take-profit, a 2:1 ratio. That bracket needs about 35% winners once costs are included; the strategy won 38.7%.
What timeframe is best for RSI divergence?
We tested 4-hour candles, which gave 106 trades in ten years, about ten a year. Lower timeframes produce more divergences, but each one tends to be less significant.
Reproduce it, then change it
Every number on this page comes from the Backtrex engine. Rebuild the strategy in the app, then change the market, the timeframe or a parameter and see whether the result holds.
Ready to backtest your strategies?
Create your free account and take the guided tour: a real 10-year backtest in 3 minutes, no coding, no credit card.
Create your account in 30 seconds and run your first backtest. No credit card.
Other backtests
- Strategy
- +14.4%
- Buy & hold
- +136.7%
- Strategy
- +13.6%
- Buy & hold
- +136.4%
- Strategy
- -98.6%
- Buy & hold
- +137.5%
- Strategy
- +26.9%
- Buy & hold
- +528.3%