This is the one popular smart-money claim that survives being measured. Gaps really do get filled, and they get filled fast. Here is the concept in plain language, the numbers behind it, and the reason a high fill rate still is not a strategy.
A fair value gap is a band of price that three consecutive candles skipped over, left behind when one candle moved so fast that buyers and sellers never met in the middle.
Take any three candles in a row. Normally the middle one overlaps both its neighbours — price moves smoothly enough that every level gets traded on the way. Occasionally it does not. Occasionally the middle candle is so large that the third candle's low finishes above the first candle's high, leaving a band of price that got skipped.
That band is a fair value gap. Other names for the same thing are imbalance and inefficiency; they all describe one situation, which is that price moved through a range so fast that buying and selling never properly met there.
The measurement is mechanical and there is no judgement in it, which is unusual for this family of ideas. Either the third candle's low is above the first candle's high or it is not. That is why the numbers below are the most reliable in this whole series — there is no room for the definition to flatter the result.
Imagine an auction where bidding is going normally, then someone shouts a number far above the current bid and everyone stops. The auctioneer skips straight there. Nobody got to bid at the prices in between — those prices were never tested, and nobody knows whether there were buyers at them.
Later, when the excitement dies down, the auction often drifts back through that skipped range to find out. Not because of a rule, but because unfinished business tends to get finished. A fair value gap is that skipped range on a price chart.
Every three-candle gap of at least one pip in five years of hourly candles, followed forward for two trading days.
| Instrument | Gaps found | Median size | Touched in 6h | Touched in 24h | Touched in 48h | Fully closed |
|---|---|---|---|---|---|---|
| XAUUSD | 6,009 | 17.7 pips | 69.4% | 82.2% | 85.6% | 76.7% |
| GBPUSD | 5,636 | 5.5 pips | 70% | 82.4% | 86.3% | 76.7% |
| EURUSD | 5,426 | 4.4 pips | 68.9% | 82% | 85.5% | 76.4% |
| US30 | 4,756 | 23.4 points | 69.2% | 80.4% | 82.6% | 74% |
| NAS100 | 4,957 | 14.3 points | 67.9% | 80% | 82.5% | 74.4% |
What was counted: Three consecutive H1 candles where the third candle's low is above the first candle's high (bullish) or its high is below the first candle's low (bearish), leaving a band of at least one pip. Measured on H1 candles across five instruments, 2021-01-03 to 2026-08-30.
"Touched" means price re-entered the band at all. "Fully closed" means it traded all the way through to the far side.
This is the one case in these lessons where the popular claim holds up. Gaps really do get filled, and they get filled quickly — most within the first six hours, the large majority within two days, and around three quarters get closed completely rather than just tagged.
The consistency across instruments is what makes it convincing. Gold, two currency pairs and two American indices all land within a few points of each other, on samples of several thousand each. That is not a pattern someone found by looking; it is a property of how price moves.
"Eighty-six per cent get filled" sounds like an eighty-six per cent strategy. It is not, and the difference is the whole lesson.
You do not know when a gap will fill, only that it usually does within two days. You do not know how far past it price will go first — plenty of gaps fill only after price runs a long way in the other direction, taking out any stop you placed nearby. And the ones that do not fill are disproportionately the ones that occur in strong trends, which is exactly when a trader fading the move is most exposed. A high fill rate is a fact about price. It is not, on its own, a trade.
Daniel trades the Dow and is waiting for the New York open.
What happens. At 14:30 a data release lands and the index runs several hundred points in one hourly candle. The candle before it topped at 44,900. The candle after it bottoms at 45,150. Between those two prices, nothing traded.
What he marks. A box from 44,900 to 45,150 — a 250-point gap, comfortably above the median for the Dow, which tells him this was a genuinely violent move rather than ordinary noise.
What he does not do. He does not short into it. Price is running, the gap will still be there in an hour, and standing in front of displacement is how accounts get hurt.
What he waits for. Price drifts back over the next few hours and enters the top of the box. Now he has a defined area and a defined invalidation: through the bottom of the gap, the idea is finished.
The part most people skip. Before taking anything, he checks his stop distance. A 250-point Dow gap means a stop of at least that order if he is trading the far edge — and the Stop Loss Reality Checker tells him what share of four-hour windows travel that far against a random entry. If the answer makes the position too large for his account, the honest response is a smaller position, not a tighter stop.
The numbers on this page say gaps get filled. They say nothing about entries, stops, targets or how it feels to sit through the move that happens first. Replay gold or the Dow bar by bar, take fifty gap trades with rules you wrote down in advance, and find out what the fill rate is actually worth to you.
Test gap fills in the free backtester →There are indicators that draw these automatically. They are fine and they will match the definition used here. Do it by hand for the first fifty though — the point of the exercise is learning to see displacement, and an indicator does that part for you.
Run the drill in the free backtester → Free, no sign-in to start.
A fair value gap is created by the same violent move that produces a break of structure, and the candle that starts it is usually an order block. When all three line up — a break, a block, and a gap sitting on top of each other — you have one price with three independent reasons to watch it, which is worth far more than any of them alone.
It also pairs naturally with the Key Levels Calculator. A gap that overlaps yesterday's pivot is a considerably more interesting area than a gap floating in empty space.
What is a fair value gap in simple terms?
It is a price range that got skipped. If you look at three candles in a row and the third one's low is above the first one's high, the space between them never traded. That space is the gap, and price tends to come back to it.
Do fair value gaps really get filled?
Yes, mostly, and this is unusually well supported. On hourly gold candles 69.4% were re-entered within six hours and 85.6% within two days, with 76.7% closing completely. The other four instruments land within a few points of that on samples of several thousand each.
If 86% get filled, why is that not an easy strategy?
Because you do not know when, and you do not know how far price will run first. A gap that fills two days later, after price has travelled twice the gap size in the opposite direction, counts as filled and would still have stopped you out. The fill rate is a fact about price behaviour, not a description of a trade.
What is the difference between a fair value gap and an imbalance?
Nothing meaningful — they are two names for the same three-candle pattern, along with "inefficiency". Different teachers prefer different words. The measurement is identical whichever term you use.
What timeframe should I use for fair value gaps?
Hourly is a sensible default and it is what these numbers are based on. On very low timeframes gaps appear constantly and almost none of them matter; on the daily they are rarer but more significant. Pick one timeframe and stay on it — a gap only means something relative to the chart it was found on.
Should I wait for a full fill or a partial touch?
That is a rules decision you should make before you trade, not during. Roughly ten percentage points separate "touched" from "fully closed" in the measurements above, which means the choice materially changes how often your setup triggers. Test both in the backtester and let your own numbers pick.
Eighty-six per cent is a fact about price, not a strategy. Replay gold or the Dow, take fifty gap trades with rules written down in advance, and measure what the entries, stops and targets actually produced.
Open the free backtester →Beginner exploration
Open each answer for a plain-language way to read Fair Value Gaps, And Whether They Really Fill, test it carefully and decide what to explore next.
This page focuses on “Fair Value Gaps, And Whether They Really Fill”.What a fair value gap is and how to spot one, plus the measured fill rate from 26,784 real gaps: 85.6% were revisited within two days on gold. With diagrams and a practice drill.For “Fair Value Gaps, And Whether They Really Fill”, a beginner should identify what the learning guide measures, assumes or teaches before acting on its conclusion.Treat this page's account of “Fair Value Gaps, And Whether They Really Fill” as a learning reference rather than a prediction, signal or promise of future performance.
For “Fair Value Gaps, And Whether They Really Fill”, translate the idea into a definition you could apply the same way on two different charts.While exploring “Fair Value Gaps, And Whether They Really Fill”, work through one example slowly and record which inputs or observations determined the result.Keep your “Fair Value Gaps, And Whether They Really Fill” record honest: list the limitation or counterexample before using the concept in a trading plan.Before leaving “Fair Value Gaps, And Whether They Really Fill”, practise the definition on unseen history and review consistency before judging performance.
Turn one idea from “Fair Value Gaps, And Whether They Really Fill” into a rule with explicit inputs, dates, costs and pass-or-fail conditions.Ask AI to expose missing assumptions in that “Fair Value Gaps, And Whether They Really Fill” test, not to guess the next market move.Use the FXAbsolute AI Backtesting Lab to inspect calculations connected to “Fair Value Gaps, And Whether They Really Fill” and the assumptions behind them.Reproduce any important “Fair Value Gaps, And Whether They Really Fill” result and reserve unseen data before deciding that an apparent pattern is useful.
Continue your exploration of Fair Value Gaps, And Whether They Really Fill with the beginner AI prompt guide, or inspect public calculations in the AI Backtesting Lab.