Key takeaways
- The stop-to-target ratio carries no information. Across six ratios and 92,585 trades, the largest gap between the measured win rate and the mathematical breakeven rate was 0.38 percentage points.
- A 1:1 stop and target is a coin flip. 49.93% of 92,585 trades reached the target first. Breakeven is 50.00%.
- Wide stops raise your win rate and change nothing else. A 2×ATR stop against a 1×ATR target won 66.97% of the time — and still returned 0.009R per trade.
- A 1:3 target is not "better risk management". It won 25.38% against a 25.00% breakeven, which is the same nothing, arrived at less often.
- This is before costs. Every ratio sits within a rounding error of zero before the spread. After it, all six are negative.
First, in plain English
Three terms appear in every table below. If they are already familiar, skip to the results — but they are the whole point, so here they are in order.
1. "Breakeven win rate" is the win rate you need just to finish level. It is not a target, it is a toll. If your winners are the same size as your losers, you need to win half your trades to end up where you started, so breakeven is 50%. If your winners are twice the size of your losers, you can afford to lose more often — win one $200 trade for every two $100 losses and you are level — so breakeven drops to 33.3%. And if your winners are half the size of your losers, you have to win far more often to survive: two $100 wins only just cover one $200 loss, so breakeven climbs to 66.7%.
So a 66.7% breakeven does not mean anything is wrong. It is simply what a small-target, wide-stop setup costs you. The arithmetic is always the same: breakeven = stop ÷ (stop + target).
2. "Measured" means what actually happened in our data. Take the 2×ATR stop and 1×ATR target row. We opened 92,577 positions in the archive, each with the stop twice as far away as the target, then followed every one of them hour by hour until one level was touched. 66.97% of those 92,577 positions touched the target first. That is the 66.97% in the table — a real count of real outcomes, not a model.
3. "Difference" is the gap between those two columns, and it is the only column that matters. You needed 66.67% to break even. You got 66.97%. The difference is 0.31 percentage points — which is nothing. Had that column read +8 points, the setup would print money. It never does.
The same idea in dollars. Risk $100 per trade with a 2×ATR stop and a 1×ATR target, so each win pays $50 and each loss costs $100. Over 1,000 trades at the measured 66.97% you win about 670 times for $33,500 and lose about 330 times for $33,000. You end up roughly $500 from where you started, before you have paid a single spread. Now run the same exercise with a 1:2 setup and you land in the same place. That is the finding.
How this was measured
The question is deliberately narrow: if you know nothing about the market and simply open a position, how often does a target at one distance get touched before a stop at another? Nothing here is a strategy. It is the background probability that any strategy has to beat.
- Data. Dukascopy bid-side hourly OHLC for 29 instruments — major and cross FX pairs, gold, silver, oil, stock indices, crypto and single stocks — from 2010-08-31 to 2026-09-11.
- The entry. A long position at the close of every 24th hourly bar. No signal, no filter, no pattern. Sampling every 24th bar keeps overlapping trades from double-counting the same move.
- The exits. Stop at S×ATR(14) below entry, target at T×ATR(14) above it, using Wilder's ATR on the same hourly series. Sizing both levels in ATR rather than pips makes a quiet pair and a violent one directly comparable.
- The walk. Forward bar by bar for up to 240 hours, checking each bar's high and low, until one level is touched. Only 2 trades out of 92,585 failed to resolve inside that window.
- Ties go against you. When one bar contains both the stop and the target, we cannot know which came first, so it is scored a loss. That happened on 1.57% of trades and pushes every win rate here slightly down.
The result: every ratio is its own breakeven
Breakeven for a stop of S and a target of T is S ÷ (S + T). If the ratio told you anything, the measured column would sit above it. It does not — at any setting.
| Stop / target | Trades tested | Target hit first (measured) | Needed just to break even | Difference | Profit per trade |
|---|---|---|---|---|---|
| 2×ATR stop / 1×ATR target | 92,577 | 66.97% | 66.67% | +0.31 pts | +0.009R |
| 1.5×ATR stop / 1.5×ATR target | 92,582 | 50.32% | 50.00% | +0.32 pts | +0.009R |
| 1×ATR stop / 1×ATR target | 92,585 | 49.93% | 50.00% | -0.07 pts | -0.001R |
| 1×ATR stop / 2×ATR target | 92,584 | 33.52% | 33.33% | +0.18 pts | +0.005R |
| 1×ATR stop / 3×ATR target | 92,577 | 25.38% | 25.00% | +0.38 pts | +0.015R |
| 0.5×ATR stop / 1×ATR target | 92,587 | 32.95% | 33.33% | -0.38 pts | -0.006R |
Six ratios, 92,585 trades each, sixteen years, twenty-nine markets. The largest deviation from breakeven in either direction is 0.38 points. That is what a market with no exploitable drift in the entry looks like when you measure it properly, and it is the cleanest confirmation we have run that our price archive and walk-forward logic are not quietly leaking bias.
Why the win rate moves but the edge does not
The win-rate column swings enormously — from 25.4% to 67.0% — which is exactly why win rate is such a poor way to judge a system. Widening the stop from 1×ATR to 2×ATR lifts the win rate by seventeen points. It does that by giving ordinary noise more room before it reaches your stop, and it charges you for the privilege at precisely the rate that cancels the benefit.
This is the same effect we isolated in the candlestick pattern study, where a hammer appeared to win 48.8% and a hanging man 15.1% despite being the same candle. The gap was stop distance, not prediction. Here there is no pattern at all, so the effect appears in its pure form.
Per-instrument results
24 instruments with at least 2,000 trades. If any market had a persistent directional drift on the hour scale, it would surface here as a 1:1 column meaningfully above 50%.
| Instrument | Trades | 1:1 (need 50%) | 1:2 (need 33.3%) | 2:1 (need 66.7%) |
|---|---|---|---|---|
| US500 | 2,947 | 51.8% | 35.3% | 67.3% |
| US30 | 2,991 | 51.4% | 35.8% | 67.6% |
| CADCHF | 4,166 | 51.4% | 33.6% | 67.5% |
| EURUSD | 4,166 | 51.2% | 34.3% | 67.2% |
| EURCHF | 4,166 | 51.2% | 33.0% | 67.7% |
| US100 | 2,956 | 51.2% | 35.7% | 67.6% |
| GBPJPY | 4,166 | 51.2% | 34.8% | 67.1% |
| EURJPY | 4,166 | 51.1% | 34.8% | 67.8% |
| BTCUSD | 3,019 | 50.5% | 34.0% | 66.7% |
| ETHUSD | 2,884 | 50.4% | 33.4% | 65.7% |
| GBPUSD | 4,166 | 50.2% | 33.5% | 67.2% |
| USDJPY | 4,166 | 50.1% | 33.8% | 66.7% |
| USDCAD | 4,166 | 50.1% | 33.9% | 68.0% |
| GER40 | 2,876 | 50.1% | 35.3% | 67.3% |
| BRENT | 3,415 | 49.8% | 33.1% | 66.4% |
| AUDUSD | 4,166 | 49.7% | 33.4% | 67.1% |
| NZDCHF | 4,166 | 49.5% | 31.9% | 66.7% |
| AUDJPY | 4,166 | 49.0% | 33.3% | 67.6% |
| NZDUSD | 2,258 | 48.9% | 33.2% | 66.8% |
| AUDCAD | 4,166 | 48.7% | 31.7% | 66.5% |
| EURGBP | 4,166 | 48.6% | 32.6% | 66.7% |
| XAUUSD | 4,166 | 48.4% | 32.3% | 65.1% |
| GBPCAD | 4,166 | 47.4% | 31.2% | 67.6% |
| XAGUSD | 3,838 | 46.7% | 30.8% | 65.6% |
The spread across instruments is narrow and centred on the arithmetic. US500 is the highest at 51.8% and XAGUSD the lowest at 46.7% — a range of 5.1 points around a 50% expectation, across sixteen years. Nothing in that column is worth trading.
What this means for where you put your stop
- Stop placement is a risk decision, not an edge decision. Choose the distance that your entry thesis actually needs — beyond the level that would prove you wrong — and accept the win rate that comes with it.
- Do not chase a higher win rate by widening the stop. You will get the higher win rate. The measurement above is what it costs.
- Do not chase 1:3 for its own sake either. A 1:3 target is not more disciplined than 1:1. It is a different distribution of the same zero, with longer losing streaks attached — see our consecutive losses guide for what a 25% win rate does to streak length.
- The edge has to come from the entry. Every number on this page is the cost of admission. A strategy is only worth running if its entry moves the win-rate column above the breakeven column by more than the spread — which is the test the candlestick and indicator studies apply.
Limitations
Stated plainly, because a study without them is marketing.
- Long trades only. The entry has no directional logic, so shorts are close to the mirror image, but they are not measured here.
- No spread, commission or swap applied. This measures the raw geometry. Real costs make every row worse, never better.
- Hourly bars. A stop and a target inside the same hour cannot be ordered, and 1.57% of trades hit that case and were scored as losses.
- ATR(14) on hourly data is one volatility definition among many. A different lookback shifts the absolute win rates slightly and does not move the gap to breakeven.
- Sampling every 24th bar reduces overlap but does not eliminate it. Trades lasting more than 24 hours can still share price action.
Download this dataset
Every figure on this page as raw CSV and JSON, regenerated whenever the archive updates. Free for any use, including commercial, with attribution to FxBacktest.
- the-result-every-ratio-is-its.csvThe result: every ratio is its own breakeven
- per-instrument-results.csvPer-instrument results
- stop-target-hit-rates.jsonAll tables plus metadata, one file
Cite or republish this data
This research is 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. Copy the attribution below.
Frequently asked
What win rate do you need for a 1:2 risk-reward ratio?
33.33% to break even before costs. Measured across 92,584 trades on 29 instruments, a 1×ATR stop with a 2×ATR target was reached 33.52% of the time - 0.18 points from breakeven. So a 1:2 setup with no entry edge lands exactly on zero, and goes negative once the spread is paid.
Does a wider stop loss improve your win rate?
Yes, and it buys you nothing. A 2×ATR stop against a 1×ATR target won 66.97% of 92,577 trades versus 49.93% for a 1:1 - seventeen points higher. Expectancy was 0.009R, indistinguishable from zero, because each win is worth half as much as each loss costs.
Is a 1:3 risk-reward ratio better than 1:1?
Not by itself. A 1:3 target was reached 25.38% of the time against a 25.00% breakeven. It is the same expectancy as 1:1, delivered through a much lower win rate and much longer losing streaks. The ratio is a choice about the shape of your equity curve, not about whether it rises.
Where should I actually place my stop loss?
At the price that proves your entry wrong, and nowhere else. Because no ratio carries an edge, the stop distance should be decided entirely by the structure you are trading - beyond the swing, the range, or the level your thesis depends on - and your position size adjusted so that distance equals the money you are willing to lose.
What data was used?
Dukascopy bid-side hourly OHLC for 29 instruments including major and cross FX pairs, gold, silver, oil, stock indices, crypto and single stocks, from 2010-08-31 to 2026-09-11. Stops and targets were sized in ATR(14) on the same hourly series, and every trade was walked forward bar by bar for up to 240 hours.