A fair value gap is a three-candle sequence in which the middle candle moves so quickly that the wicks of the candles on either side of it never overlap, leaving a band of price that was traded through in one direction without meaningful two-sided trading. On a bullish sequence it is the space between the high of the first candle and the low of the third. On a bearish sequence it is the space between the low of the first candle and the high of the third.
What that empty band represents is an imbalance. In normal trading, orders rest on both sides of the book and price grinds through them, filling buyers and sellers at every level. When a large one-sided flow arrives, it consumes the resting liquidity faster than it is replenished and price skips across several levels with almost nobody trading against it. The gap on your chart is the footprint of that skip: prices at which one side of the market was effectively absent.
This page covers where the term comes from, the exact conditions a sequence must meet, how the zone is used for entries and stops, and what makes a fair value gap fail. Gaps get filled often and ignored often; treating one as a signal on its own is the fastest way to lose money with the concept.
Where the term comes from
"Fair value gap" is not a term from academic market microstructure or from exchange documentation. It comes from the ICT (Inner Circle Trader) framework, a body of retail trading material developed by one educator, along with related vocabulary like order blocks, liquidity sweeps and displacement. You will not find the phrase in a textbook on limit order books, and no exchange publishes fair value gap data.
That matters for two reasons. First, when definitions conflict online there is no authority to appeal to โ different people draw the boundaries differently and both can be internally consistent. Second, the idea is old even though the label is recent: traders have described the same thing as an imbalance, a single-print area, a low-volume node or a liquidity void. Volume profile traders look for thin areas in the profile for the reason ICT traders look for fair value gaps โ price moved through without building acceptance, so it may move back through easily too.
Use the term as shorthand, not as though the market recognises it as a special object. You are marking a stretch of price where trade was one-sided, and the question is whether that is still true by the time price returns.
How to identify a fair value gap precisely
Work with three consecutive candles on a single timeframe. Label them candle 1, candle 2 and candle 3, left to right.
For a bullish fair value gap:
- Candle 2 is a strong up candle, visibly larger than the recent average range.
- The low of candle 3 is above the high of candle 1.
- The gap is the range between those two prices.
For a bearish gap, mirror it: candle 2 is a strong down candle and the high of candle 3 sits below the low of candle 1.
Two conditions people routinely drop, and shouldn't:
- The middle candle has to be a genuine expansion. Three small candles stepping upward technically produce non-overlapping wicks, and marking every one of them makes your chart unreadable. A workable rule is that candle 2's range should be at least twice the average range of the preceding ten candles โ pick a threshold, write it down and apply it consistently rather than deciding candle by candle.
- The gap needs an origin. A gap that forms mid-range, with no clear structure behind it, has far less to say than one that forms as price breaks a structural high.
| Criterion | Bullish FVG | Bearish FVG | Why it matters |
|---|---|---|---|
| Direction of candle 2 | Strong up close | Strong down close | The imbalance has to have a direction |
| Non-overlap test | Low of candle 3 > high of candle 1 | High of candle 3 < low of candle 1 | This is the gap itself |
| Gap boundaries | C1 high to C3 low | C1 low to C3 high | Defines the zone you will trade |
| Size filter | C2 range well above recent average | Same | Filters out noise gaps |
| Context | Formed on a break of structure | Same | Separates imbalance from drift |
Timeframe changes the picture more than most traders expect. A gap that is obvious on a 5-minute chart may not exist on the hourly, because the hourly candle containing it has continuous range. Neither chart is wrong; they describe different resolutions of the same order flow. Decide in advance which timeframe you take gaps from rather than switching until you find one you like.
A worked example
Take a hypothetical EUR/USD 15-minute sequence. Candle 1 trades between 1.0828 and 1.0840. Candle 2 opens near 1.0838 and closes at 1.0893, far larger than the preceding candles. Candle 3 pulls back and puts in a low of 1.0862.
The low of candle 3 is above the high of candle 1, so a bullish fair value gap exists between 1.0840 and 1.0862 โ 22 pips wide, midpoint 1.0851.
Three levels come out of that, and they are the ones traders use:
- Gap high (1.0862) โ first point of contact on a retrace. Shallowest entry, widest stop relative to the gap.
- Gap midpoint (1.0851) โ a compromise entry, on the reasoning that a retrace stalling halfway suggests demand has not disappeared.
- Gap low (1.0840) โ a full fill. The best price if the idea works, and the point at which the imbalance argument has largely been spent.
A stop does not belong inside the gap. If the thesis is that buyers step back in within the imbalance, price trading cleanly below the gap low says they did not. A stop under the low of candle 2, or under the swing low that preceded the expansion, respects that. Sizing is then arithmetic rather than judgement: distance to stop, account risk, and the EUR/USD lot size calculator gives you the volume. Run the same numbers on gold before assuming a 22-point gap there is a comparable trade โ the XAUUSD stop loss calculator makes the difference obvious.
Notice what the example does not tell you: whether to take the trade. The gap defines a zone and a level at which you are wrong. Trend, session and what is on the economic calendar in the next hour sit outside the pattern.
How fair value gaps fail
This is the part that decides whether the concept costs or makes you money.
Price never comes back. In a strong trend the market can leave a run of gaps unfilled for days. Traders who insist on an entry inside the gap miss the move, then chase it. There is no rule that imbalances must be rebalanced, only a tendency that shows up often enough for people to build methods around it.
Price fills the gap and keeps going. The trader is in the position expecting a bounce and price continues straight through. This is the common losing outcome, and it usually means the flow that created the imbalance was a news reaction or a liquidation rather than sustained positioning. Nothing in the candle sequence distinguishes the two at the time.
The gap was noise. A small middle candle on a quiet pair in a thin session produces a technically valid gap that means very little. The size filter above removes most of these.
News reprices the market. A rate decision or inflation print can invalidate a structural read entirely, and a gap formed shortly before a scheduled release deserves much less confidence than one formed in a quiet hour.
Treat these as invalidation rules. If price closes decisively beyond the far side of the gap on your working timeframe, the setup is done. If price reaches the gap but the reaction is slow and overlapping rather than immediate, the imbalance is being accepted rather than rejected, and the argument weakens before your stop is hit. Waiting for a lower-timeframe reaction inside the zone instead of resting a limit at the edge is one way traders handle that; it costs some price and filters some losers, and only your own records will settle whether the trade-off is worth it.
Common mistakes
- Marking every gap on the chart. Twenty zones on a screen is not analysis. Keep the ones that formed on a break of structure.
- Assuming a gap is support or resistance. It is an area where price previously moved fast โ a reason to watch, not a reason to expect a reversal.
- Shrinking the stop into the zone. If the sensible stop is too wide for your risk, size down or skip the trade.
- Confusing it with a weekend or opening gap. Those are literal gaps in traded price between sessions; a fair value gap occurs within continuous trading.
How it fits with everything else
A fair value gap is one piece of a read, not the read itself. It asks the same question as supply and demand trading โ where did price move away from with force โ in different vocabulary. Within the wider ICT trading framework it is usually the entry refinement applied after a higher-timeframe bias is set, and it forms the displacement leg of the accumulation, manipulation and distribution sequence.
Candlestick reading works alongside it. If price retraces into a bullish gap and prints a hammer candlestick at the edge of the zone, you have two independent observations pointing the same way.
Frequently asked questions
Do all fair value gaps get filled?
No. Some fill within minutes, some days later, and some never fill because the market has repriced permanently. Any source telling you gaps must fill is describing a tendency as though it were a rule. Filling is common enough to be tradeable and uncommon enough to require a stop.
What timeframe is best for fair value gaps?
There is no single best one. Higher timeframes leave fewer, wider gaps that require larger stops and more patience; lower timeframes produce many more with more noise among them. Most traders use a higher timeframe for bias and a lower one to locate the gap they trade. Pick a pairing and test it rather than switching mid-analysis.
Is a fair value gap the same as an imbalance?
In practice, yes. "Imbalance" is the older and more neutral word for the same observation, and some platforms mark these zones automatically. Fair value gap adds a specific three-candle definition to the general idea.
Should I enter at the top, the middle or the bottom of the gap?
That is a trade-off to measure, not a question with a fixed answer. The near edge fills more often at a worse price; the far edge gets a better price on fewer trades and misses the ones that turn early. Log which edge you used on every gap trade and let your own data settle it.
Can I trade fair value gaps on stocks and indices?
Yes โ the definition applies to any candlestick chart, and index products produce them frequently around the open. Instrument volatility changes what a given gap width means, so the same number of points on an index and on a currency pair are not comparable risks.
Turning this into something you can measure
Nothing above tells you whether fair value gaps work for you, because that depends on which gaps you take, on what timeframe, in which market and with what stop placement. Those are your variables, and the only way to learn them is to record them.
Log every gap trade with the same fields: the timeframe, the width of the gap, whether it formed on a break of structure, which edge you entered at, where the stop went and the outcome. After fifty entries in a trading journal, patterns start to show โ usually that one subset of gaps carries the results and the rest are habit. Backtesting the same rules over a longer history gets you there faster, provided you fix the size filter in advance rather than eyeballing it afterwards.
Keep the sizing mechanical alongside it. Gap widths vary enormously between instruments and sessions, so a fixed lot size gives you inconsistent risk. Run the distance to your stop through the lot size calculator every time, so a wide gap and a narrow one cost the same when they fail. The pattern is probabilistic; the sizing does not have to be.