Fair value gap (FVG)

What a fair value gap is

A fair value gap is an unfilled slice of price inside a fast move: three consecutive candles where the first candle's high and the third candle's low leave a band that the middle candle raced past without the neighbours trading it. On a chart it reads as a blank corridor.

The name shortens to FVG, and the same construction is called an imbalance, an inefficiency, or a liquidity void depending on who is teaching it. The terminology is not standardised, which is worth knowing before comparing two explanations that look like they disagree.

Why the gap forms, and why it is partly about your chart

Two things are going on, and most explanations only mention the first.

The market part: price moves through resting orders. When a move is fast, it is passing prices at which little interest was waiting, so it travels far for the volume involved. A band nobody traded is therefore a record of a shortage of resting interest at those prices, at that moment.

The measurement part, which is usually left out: a candle is not a thing the market produced. It is a summary of a continuous stream of transactions, aggregated into a fixed interval that somebody chose. The gap is defined by the highs and lows of neighbouring bars — so it is a property of the sampling as much as of the market. Trades may well have occurred inside that band; they just did not land in the neighbouring bars' extremes at the resolution you are looking at.

This has a consequence you can check in about a minute. Open the same move on a lower timeframe. Gaps routinely shrink, split, shift, or vanish, because finer bars capture prints that coarser bars averaged away. A construction that changes when you change the chart's settings is not purely a market fact.

None of this makes the idea worthless. It does mean that saying there is an FVG here is a statement about a chart, and needs the timeframe attached to be meaningful at all.

How to identify an FVG

The common three-candle reading:

• Find a decisive candle — a long body relative to its neighbours, with little overlap.
• Take the previous candle's extreme on the side the move came from.
• Take the following candle's extreme on the far side.
• The band between those two extremes is the gap, if they do not overlap. If they overlap, there is no gap at this resolution.

Two judgements hide in those steps

How long is long, and whether you measure from wicks or bodies. Wick-to-wick produces fewer, larger gaps; body-to-body produces more, smaller ones. Both are in use.

Pick one, write it down, and keep it fixed. Otherwise you are not identifying gaps, you are choosing them.

FVG, imbalance, order block: what the words mean

Mostly vocabulary, with a few real distinctions.

• Fair value gap, or FVG, usually means the unfilled band between neighbouring candles' extremes — the most specific of these terms when defined as above.
• Imbalance usually means any area where one side clearly dominated; it is broader and may not require the three-candle geometry.
• Inefficiency is the same idea framed as price having skipped fair value — a rhetorical framing rather than a different measurement.
• Liquidity void describes a region with little resting interest, so it names the cause rather than the chart shape.
• An order block is a different object: the origin candle of the move, where it started, not the gap inside it. This is the distinction that matters most, because order blocks and supply or demand zones frequently sit near a gap and are routinely confused with it.

Do fair value gaps always get filled?

No, and the honest answer has three parts.

Not always. Price often returns through such a band, and often does not. Anyone quoting a fill rate should be asked which instrument, which timeframe, which gap definition, and over which period — because all four change the number, and a figure without them is not a measurement.

Eventually is not a claim. Given enough time and a wide enough range, most price bands get revisited. A statement that holds for nearly any band is not telling you anything about this one.

And it is partly a definitional question. Because the gap depends on bar resolution, so does filled. A band fully traded on a five-minute chart may still look unfilled on a four-hour one. Two readers can disagree about whether the same gap filled and both be reading their charts correctly.

Is FVG trading reliable?

Reliability is not a property the construction has on its own, for the same reasons: the definition is discretionary, the measurement depends on timeframe, and no general figure survives the four questions above.

Where the idea does earn its place is as a description. It names a fast, thin move compactly, and that is useful shorthand. The error is treating a compact name as evidence of a reliable behaviour.

Common mistakes

The first two account for most of the trouble.

• Quoting a fill rate. Without instrument, timeframe, definition, and period, the number is decoration.
• Marking gaps without recording the timeframe, which makes the mark unreproducible even for you.
• Switching between wick-to-wick and body-to-body depending on what looks better on the chart in front of you.
• Treating an FVG as an entry. It marks a band; it does not say direction, size, or invalidation.
• Confusing it with an order block and then reasoning about the wrong object.
• Hunting gaps on a lower timeframe to justify a higher-timeframe view. Drop the resolution far enough and there is always a gap somewhere.

A review checklist

Settled before the next chart, not during it.

• Wick-to-wick or body-to-body? One answer, fixed.
• What makes the middle candle decisive, as a number?
• Which timeframe governs, and do I record it with every mark?
• What counts as filled — any touch, a full trade-through, or a close beyond?
• Does a gap alone trigger anything, or does it need a second condition?
• Reviewing my own gap entries, did I apply these answers, or find the gap after the outcome?

Methodology and limitations

This guide is an original editorial explanation based on the primary sources listed below. It uses no live prices, personal data, forecasts, or automated trade signals. U.S. sources are identified as such; product rules and protections vary by provider and jurisdiction.

Primary sources and scope

Frequently asked questions

Continue learning