Fair value gap (FVG) in ICT trading, explained, and why the fill-rate claims are unsourced

What a fair value gap means in ICT trading

In ICT trading, a fair value gap is a three-candle price imbalance where the move is so fast that the first candle's extreme does not overlap the third candle's opposite extreme, leaving a visible gap on the chart. ICT traders read that gap as an area of inefficient pricing the market may return to fill (The Inner Circle Trader, ICT methodology).

I scope this tightly on purpose, because "fair value gap" is also a generic finance term in accounting and options, where it means something completely unrelated. This page is about the ICT trading pattern taught by Michael Huddleston, the Inner Circle Trader, and nothing else.

The pattern is a label within a methodology, the same as order blocks. The candle shape is real and easy to see, but the claim that the gap represents an inefficiency the market is obliged to rebalance is ICT doctrine, not a measured property of markets.

The three-candle pattern, drawn in words

A bullish fair value gap forms across three candles. The first candle pushes up, the second candle displaces so violently that its body covers the gap, and the third candle's low sits above the first candle's high, leaving an empty price range between them (The Inner Circle Trader, ICT methodology).

That empty range, the gap between candle one's high and candle three's low, is the FVG. A bearish FVG is the mirror, where the third candle's high stays below the first candle's low after a sharp down-move.

I describe it without a chart because the rule is mechanical enough to verify on any candlestick view. The requirement is displacement, a forceful middle candle, which is what separates a genuine FVG from any random three candles that happen not to overlap.

Bullish versus bearish fair value gaps

A bullish FVG appears during an up-move and is treated as support on a retrace. Price pushed up so hard it skipped a range, and ICT traders watch for price to come back down into that range before continuing higher (The Inner Circle Trader, ICT methodology).

A bearish FVG appears during a down-move and is treated as resistance. Price fell through a range without trading it, and the methodology watches for a bounce back up into the gap before resuming lower.

The directional logic is symmetric, and I find the cleanest way to remember it is that the gap lies in the direction price already moved. A gap left by a surge up sits below price as potential support; a gap left by a plunge down sits above price as potential resistance.

Fair value gap versus order block

An order block and a fair value gap describe different things and are often used together, which is why ICT teaches them as a pair. The order block is the specific candle where the move is said to originate; the fair value gap is the price range the move skipped over (The Inner Circle Trader, ICT methodology).

In practice ICT traders look for a fair value gap inside or just above an order block, because the combination gives both a location and an entry trigger. The order block says where to look, and the FVG says where within the zone price is likely to react.

I keep them distinct because mashing them into one idea loses the method's structure. The order block is a positional claim about institutional intent; the FVG is a structural claim about price efficiency, and they happen to coincide often enough that the pairing is useful.

Fills, mitigation, and the rebalancing claim

ICT calls a return into the gap a fill or mitigation, and the methodology's explanation is that markets dislike inefficiency and move to rebalance it. Price is said to gravitate back to the gap to "fairly value" the skipped range before extending (The Inner Circle Trader, ICT methodology).

The defensible core is narrower and less mystical. Prices do retrace, and short-term dislocations often get revisited, which is consistent with the mean-reversion behaviour traders commonly observe in intraday markets.

The rebalancing language is ICT's way of narrating that observation.

I separate them deliberately. The tendency of price to revisit recent gaps is something you can watch on a chart, but the idea that the market has an obligation to fill them is a story, and trading as though fills are guaranteed is how accounts get hurt in strong trends that never look back.

The fill-rate statistics nobody can source

You will read that fair value gaps fill 80% of the time, or 90%, or some other confident figure. I have not found a peer-reviewed or tier-1 study that tests ICT fair value gaps and publishes a fill probability, and I have not seen any page quote one with a real citation.

Every specific percentage floating around forums and YouTube traces to an unsourced claim repeated until it looks like research. The honest position is that no verified backtest exists in the public literature, and a fill rate you cannot link to a study is worth nothing.

I would rather state the gap in the evidence than hand you a fabricated number. If you want a fill rate you can trust, the only honest route is to test the pattern yourself on your own instrument and timeframe, record the results, and treat your figure as a sample from one dataset, not a universal law.

FVG as confluence, not a standalone signal

The biggest gap between how ICT uses fair value gaps and how the internet teaches them is confluence. In the methodology, an FVG is a confirming factor inside a larger setup, not a trade signal on its own (The Inner Circle Trader, ICT methodology).

The full ICT entry layers a higher-timeframe bias, a liquidity sweep, an order block, and then a fair value gap as the precision trigger inside the block. Most educational pages lift the FVG out of that stack and present it as a standalone pattern, which strips away exactly the context that makes the method coherent.

I flag this because it is the SERP's central weakness and the easiest place to be more useful than the ranking pages. An FVG aligned with the higher-timeframe move and sitting inside an order block is a setup; an FVG traded in isolation is a guess decorated with jargon.

ICT FVG is not order-book imbalance

The word "imbalance" causes real confusion, because exchange microstructure has its own concept called order imbalance, and it has nothing to do with ICT. Order-book imbalance measures the ratio of buy to sell volume resting in the limit order book at a given moment.

ICT's fair value gap is a price gap on a candle chart, a visual void left by a fast move. The two share the word "imbalance" and essentially nothing else, and conflating them leads to analysis that mixes a chart pattern with a volume metric.

I draw the line firmly because the collision is easy to make and hard to detect once made. When ICT traders say a gap is an "imbalance," they mean price moved too fast; when an exchange report says "order imbalance," it means one side of the book is heavier.

Different data, same word.

How ICT traders actually use an FVG

The methodical ICT use waits for context before the gap matters. Price sweeps a liquidity level, displaces into an order block that contains a fair value gap, and the trader enters on a retracement into the gap in the direction of the higher-timeframe bias, with a stop beyond the order block (The Inner Circle Trader, ICT methodology).

The target is the next opposing liquidity pool, mirroring the order-block setup. The fair value gap's job is to refine the entry within the zone, narrowing where price is expected to react so the risk is defined tightly.

I present the sequence as a method, not a promise. The same steps can be followed perfectly and still produce a loss when the zone fails, which is why position sizing and the broader SMC swing framework matter more than any single pattern.

Where fair value gap setups fail

The most common failure is the trend that never retrace. A strong, extended move can leave multiple fair value gaps open for days or weeks, and trading every one of them on the assumption it must fill is a reliable way to sell into a runaway move.

Partial fills are the second failure mode. Price enters the gap, reacts enough to look like a fill, then continues through, and a trader who treated the first reaction as confirmation gets stopped.

The methodology handles this with the higher-timeframe bias, but only if you actually enforce it.

The third is noise. Quiet, rangebound markets produce three-candle shapes that technically qualify but carry no displacement or context, and these false FVGs clutter the chart.

The displacement requirement exists precisely to filter these out, and skipping it converts a disciplined method into pattern-matching on static.