Key takeaways
- The bear cross is indistinguishable from random. +0.0 points over its control across more than eleven thousand daily trades.
- The bull cross has a small real edge. Around a point on daily charts, at both targets.
- The sample is large enough to trust the null. Eleven thousand trades finding nothing is itself a finding.
- A 46% win rate is not the story. A random short in the same period won about the same, which is exactly the point of the control column.
What it does
The stochastic oscillator locates the current close within the high-low range of the last 14 bars and expresses it as a percentage: 100 means the close is at the top of the range, 0 at the bottom. That raw line is %K, and a 3-period average of it is %D. Readings under 20 are called oversold, over 80 overbought.
The signals measured here are crosses inside the extremes: %K crossing above %D while both sit under 20, and %K crossing below %D while both sit over 80. Requiring the cross to happen inside the zone is the stricter, more common reading, and it is what most charting packages highlight.
How this was measured
Every figure below comes from our own tick-derived archive, not from a vendor summary. The method is identical to the one behind our indicator signal study, so the numbers here are directly comparable to every other signal we publish.
- Data. Dukascopy bid-side OHLC for 93 instruments — major and cross FX pairs, gold, silver, stock indices and crypto — from 2010 to August 2026.
- The trade. Enter at the open of the next bar. Stop at 1.5 × ATR(14), targets at 1:1 and 1:2, walking forward up to 40 bars.
- Ties go against you. A bar holding both stop and target is scored a loss every time.
- Real costs. Per-instrument median spreads from our own bid/ask measurement.
- The control. For every signal, a trade with the same direction and the same stop distance at a random bar — five draws each. The gap between the two columns is all the signal is worth.
The result
Breakeven is a 50% win rate at a 1:1 target and 33.3% at 1:2. "Matched random" is a trade in the same direction with the same stop distance, opened at a random bar.
| Signal | Target | Trades | Win rate | Matched random | Edge | Significance | Beats breakeven? |
|---|---|---|---|---|---|---|---|
| Stochastic bull cross <20 | 1:1 | 8,351 | 53.0% | 51.9% | +1.1 pts | not significant | Yes |
| Stochastic bull cross <20 | 1:1 | 43,110 | 49.9% | 49.1% | +0.8 pts | ** | No |
| Stochastic bear cross >80 | 1:1 | 11,614 | 45.9% | 45.8% | +0.0 pts | not significant | No |
| Stochastic bear cross >80 | 1:1 | 57,601 | 47.7% | 47.7% | +0.0 pts | not significant | No |
| Stochastic bull cross <20 | 1:2 | 41,586 | 33.0% | 31.9% | +1.1 pts | *** | No |
| Stochastic bull cross <20 | 1:2 | 7,978 | 35.4% | 34.4% | +1.0 pts | not significant | Yes |
| Stochastic bear cross >80 | 1:2 | 11,262 | 28.8% | 28.6% | +0.2 pts | not significant | No |
| Stochastic bear cross >80 | 1:2 | 55,887 | 30.5% | 30.5% | +0.1 pts | not significant | No |
The bear cross result deserves attention precisely because it is so flat. An edge of +0.0 points across more than eleven thousand daily trades is not an inconclusive result from a thin sample — it is a well-measured nothing. Whatever the overbought stochastic cross is doing, it is not identifying short setups better than picking a bar at random and selling it with the same stop.
The bull cross is different in kind but not enormous in size: about a point of edge, consistent across both targets. Note the win rates themselves invert the story. The bull cross wins 53% and the bear cross 46%, which looks decisive until the control column shows a random long winning 51.9% and a random short 45.8% over the same bars.
How to use it
- Use the bull side, skip the bear side. This is the clearest one-sided result in the study, and the sample behind the null is large enough to act on.
- Require the cross inside the zone. We measured %K crossing %D while both sit under 20, not a reading under 20 on its own.
- Expect a filter, not a trigger. One point of edge improves a decision you were already going to make; it does not create one.
- Pair it with the timeframe that earned it. The daily rows carry the result.
What does not work
- The overbought bear cross. +0.0 points on eleven thousand-plus daily trades. Symmetry is assumed by almost every tutorial and is not present in the data.
- Reading 80 and 20 as instructions. The zones describe where price sits in a recent range, not whether it is expensive.
- Judging either side by raw win rate. Both numbers are dominated by the drift of the underlying market, which the control column removes.
Limitations
- Default parameters only. Tuning them on this same data would be curve-fitting.
- One stop rule (1.5 × ATR) and two fixed targets. Trailing stops, time exits and scaling are untested.
- Signals are taken in isolation, with no trend filter or confluence — which is how the isolated value of each is measured, but not how most people trade them.
- Bid data with modelled ask. Spread is a measured median, not a replayed tick-level ask feed.
- Spread only — no commission or swap. Both would push results further down.
- A statistically significant edge of one point is still only one point. Large samples make small effects certain; they do not make them large.
Cite or republish this data
Free to quote, screenshot, or republish — in an article, a video, a newsletter, or a course — as long as you credit FxBacktest with a link back to this page.
Frequently asked
Does the stochastic oscillator work?
Half of it does. The bull cross from under 20 beat a matched random long by about a point on daily charts across 93 instruments. The bear cross from over 80 produced an edge of +0.0 points across more than eleven thousand daily trades, meaning it performed identically to selling a randomly chosen bar with the same stop. Treating the two sides as mirror images is the main mistake.
What is the difference between stochastic and RSI?
RSI compares the size of recent gains to recent losses; stochastic locates the close within the recent high-low range. Measured the same way on the same data, RSI(14) leaving oversold produced a larger daily edge at a 1:2 target than the stochastic bull cross. Both are small, and both are far stronger on the long side than the short side.
Is a stochastic reading over 80 a sell signal?
Not on this evidence. We measured the stricter version - %K crossing below %D while both sit above 80 - and it produced no edge at all over a random short. A high stochastic reading means price is near the top of its recent range, which in a rising market is a normal condition rather than a warning.
What stochastic settings did you test?
The default 14,3 configuration only. Tuning the periods on the same data used to evaluate them would be curve-fitting. The result that matters here is directional rather than parametric: the bull cross carried a small edge and the bear cross carried none.
How were these stochastic numbers produced?
Enter at the open of the bar after the cross, stop 1.5 x ATR(14) away, targets at 1:1 and 1:2, walking forward up to 40 bars. Bars containing both stop and target are scored as losses. Real per-instrument median spreads are applied. Every signal is compared against five random-entry draws in the same direction with the same stop distance. Data is Dukascopy bid OHLC for 93 instruments, 2010 to August 2026.