Key takeaways
- The break itself is not a signal. The opening range broke on 99.8% of 93,105 session days. Something that happens almost every day cannot tell you which days matter.
- Follow-through is a coin flip. 50.3% of breaks extended one further range height before returning through the opposite side.
- No instrument escapes it. Across 26 markets the follow-through rate runs from 47.6% to 52.5% — a spread of 4.9 points around 50%, over sixteen years.
- Upside breaks are more common, and no more reliable. 52%-odd of first breaks go up, but they hold no better than downside ones.
- After costs it is negative. A 1:1 structure needs better than 50% to survive the spread. Nothing here clears it.
First, in plain English
Two numbers on this page look like they contradict each other — 99.8% and 50.3%. They do not, and the gap between them is the entire finding. Here is the setup, step by step.
Step 1: draw a box. London's first hour is 07:00 to 08:00 UTC. Take the highest and lowest price in that hour and draw a rectangle around them. Say EUR/USD traded between 1.1030 and 1.1050 — that is your box, 20 pips tall. This is the "opening range".
Step 2: wait for price to leave the box. At some point during the rest of the London session, price will trade above 1.1050 or below 1.1030. That is a "break". We counted 92,947 of them — one per instrument per day, across 26 instruments and sixteen years. They are not trades and nobody placed them; each one is simply a day on which price left its box.
Step 3: this is where it falls apart. Price left the box on 99.8% of days. Of course it did — the box is one hour tall and the session runs another eight. Expecting price to sit inside a one-hour rectangle all day is like expecting the tide not to move. So "the range broke" is not news, and it cannot be a signal: a thing that happens every single day tells you nothing about which day you are in.
Step 4: the question that actually matters. Once price is out of the box, does it run — or does it turn around and come straight back? We set a simple, fixed test. Going up from 1.1050, the break "kept going" if price reached 1.1070 (one more box-height up) before it dropped back to 1.1030 (the far side of the box). Run to the target first and it held. Fall back through the bottom first and it failed.
The answer: 50.3% kept going. 49.7% turned around. That is a coin flip. So yes — to answer the obvious question directly — roughly half the time price breaks out and then reverses all the way back through the box. You are not being unlucky when that happens. It is what the setup does.
Why this matters more than it sounds. A trade that wins half the time and pays the same as it risks makes exactly nothing before costs. Then you pay the spread on every one of them. That is the difference between a strategy and a habit, and it is why the raw opening-range break needs a filter on top — which is what the last section of this page is about.
How this was measured
- Data. Dukascopy bid-side hourly OHLC for 26 instruments with meaningful London participation — FX majors and crosses, gold, silver, oil, indices and crypto — from 2010-08-31 to 2026-09-11.
- The opening range. The high and low of the 07:00 UTC hourly bar, the first hour of the London session in summer. Hours are in UTC throughout so the definition never moves; see our session hours guide for how the local clocks map onto it.
- The break. The first subsequent bar up to 15:00 UTC whose high exceeds the range high or whose low breaks the range low. Whichever comes first sets the direction.
- Held or failed. A break counts as held if price then travels a further one full range height beyond the break level before trading back through the opposite side of the opening range. That is a clean 1:1 test with no discretion in it.
- Ties go against you. When one bar contains both the extension and the invalidation, it is scored a failure.
The result
The first column is the reason most opening-range rules disappoint: there is almost always a break, so the break alone selects nothing.
| Instrument | Days tested | Days price left the box | ... and kept going | Left via the top |
|---|---|---|---|---|
| XAUUSD | 4,317 | 99.9% | 52.5% | 52.3% |
| USDJPY | 4,162 | 99.7% | 52.5% | 56.1% |
| GBPJPY | 4,163 | 99.8% | 52.4% | 57.4% |
| AUDJPY | 4,163 | 99.7% | 52.1% | 56.8% |
| AUDUSD | 4,163 | 99.9% | 51.9% | 55.0% |
| GBPUSD | 4,162 | 99.8% | 51.9% | 55.8% |
| EURUSD | 4,162 | 99.8% | 51.5% | 55.5% |
| EURJPY | 4,162 | 99.5% | 51.4% | 56.9% |
| GER40 | 3,316 | 99.4% | 51.4% | 57.8% |
| EURGBP | 4,163 | 99.8% | 51.3% | 56.5% |
| NZDCHF | 4,163 | 99.8% | 50.8% | 56.3% |
| NZDUSD | 2,257 | 100.0% | 50.5% | 55.5% |
| CADCHF | 4,162 | 99.8% | 50.2% | 54.3% |
| EURCHF | 4,163 | 99.6% | 49.4% | 55.2% |
| XAGUSD | 4,001 | 100.0% | 49.3% | 55.7% |
| BTCUSD | 3,015 | 99.8% | 49.2% | 58.0% |
| DXY | 1,029 | 99.8% | 49.2% | 56.7% |
| US100 | 3,388 | 100.0% | 48.9% | 58.1% |
| AUDCAD | 4,162 | 100.0% | 48.9% | 56.8% |
| ETHUSD | 2,881 | 99.8% | 48.9% | 57.5% |
| GBPCAD | 4,163 | 100.0% | 48.8% | 58.2% |
| BRENT | 3,853 | 100.0% | 48.6% | 57.6% |
| US30 | 3,392 | 99.9% | 48.3% | 58.1% |
| US500 | 3,381 | 100.0% | 47.8% | 58.6% |
| USDCAD | 4,162 | 100.0% | 47.6% | 57.4% |
Pooled across all 26 instruments: 93,105 session days, 92,947 breaks (99.8% of days), 43,091 held and 42,608 failed — a follow-through rate of 50.3%.
Why a one-hour range breaks every single day
A single hour of the London session is small next to the eight that follow it. Our hour-by-hour volatility measurements show the 07:00 UTC hour is itself one of the more active of the day, but the range it prints is still a fraction of the day's eventual span. Price leaving that box is close to inevitable, which is why the break rate sits at 99.8% rather than somewhere useful.
This is the trap in the popular version of the strategy. A rule that triggers almost every day feels like it is working because it is always in the market during the session that moves. What it is actually doing is taking a coin-flip position every morning and paying the spread for the privilege.
What would make it tradeable
The measurement says the raw break has no edge. It does not say opening ranges are useless — it says the break needs a filter that this test deliberately withholds. The honest places to look:
- Range size. An unusually tight opening hour and an unusually wide one are different setups. This test treats them identically.
- Where the range sits. A break that also clears the Asian session high is a different event from one that happens mid-range.
- Time of the break. A break at 08:00 UTC and one at 14:00 UTC have very different amounts of session left to work with.
- News. A break driven by a scheduled release behaves differently — our release volatility study measures how much.
- Direction of the larger trend. Breaks aligned with the daily direction are the obvious first filter to test, and the one this study is built to be compared against.
Each of those is a hypothesis you can now test against a known baseline: 50.3%. A filter is only worth keeping if it beats that by more than the spread.
Limitations
Stated plainly, because a study without them is marketing.
- The opening range is fixed at the 07:00 UTC bar all year. London shifts to 08:00 UTC in winter, so roughly half the sample uses the hour before the true open. A DST-aware definition is the obvious next cut.
- One hour, one extension target. A thirty-minute range or a two-hour range is a different setup, and a trailing exit may behave differently from the fixed 1:1 used here.
- Hourly resolution. A break and its invalidation inside the same hour cannot be ordered and were scored as failures.
- No spread, commission or swap applied. Real costs move every row down.
- Bid-side data. Ask prices are not replayed, so a break measured on the bid is marginally early for a long and late for a short.
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.csvThe result
- london-opening-range-break.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
Does the London opening range breakout strategy work?
Not in its raw form. Measured on 92,947 breaks across 26 instruments and sixteen years, a break of the 07:00 UTC opening range extended one further range height before failing 50.3% of the time. That is a coin flip before costs and a losing proposition after the spread, unless a filter is added.
What time is the London opening range?
The London session opens at 07:00 UTC during British Summer Time and 08:00 UTC in winter when London is on GMT. This study uses the 07:00 UTC hourly bar year-round as a fixed definition, which is one of its stated limitations.
How often does the London opening range get broken?
Almost always - 99.8% of 93,105 session days across the sample. A single hour is a small box relative to the eight hours of London trading that follow it, so price leaving that box carries very little information on its own.
Which pairs have the best opening range follow-through?
XAUUSD at 52.5% and USDJPY at 52.5% were the highest of the 26 measured, and USDCAD the lowest at 47.6%. The whole range spans 4.9 points around 50%, which is too narrow to build a strategy on by instrument selection alone.
What data was used?
Dukascopy bid-side hourly OHLC for 26 instruments with meaningful London participation, from 2010-08-31 to 2026-09-11. The opening range is the 07:00 UTC bar, the break is the first bar through either side before 16:00 UTC, and follow-through is a further one range height before the opposite side of the range is traded through.