An inverse fair value gap, or IFVG, is a fair value gap price has traded through instead of bouncing off. The imbalance has not disappeared, it has switched sides: a bullish gap broken downwards then reads as a resistance area, exactly as broken support becomes resistance.
The gap that was traded through flips role
Contents
The idea of an inverse fair value gap rests on a simple observation: when a level many people watch does not hold, the positions taken on it end up on the wrong side. Price trading through a fair value gap therefore leaves orders that need correcting behind it, and that is what explains why a return towards that level can produce a reaction in the opposite direction.
It is the same mechanism as the role swap of broken support turning into resistance, applied to an imbalance rather than a horizontal level. It is also called an inversion fair value gap, or simply an inversion FVG: those three phrasings mean the same thing, and nothing separates them.
One point matters more than all the mechanics: an inverse fair value gap is not a new object, it is a fair value gap in another state. You do not hunt for it on a blank chart, you observe it on a gap you were already following. Our lesson on the fair value gap and the inverse FVG covers the full method; this page stays on the definition and on the difference between the two.
IFVG = an FVG price has closed on the other side of — the edges do not change, only the role does
The two edges stay those of the original gap. That is what makes an inverse FVG as precise as the FVG it comes from, and what distinguishes it from an area redrawn after the fact.
The break condition is the only setting, and there is no consensus on it. Some methods require a candle close beyond the gap, others accept price simply passing through. The stricter rule, the close, produces far fewer cases and far fewer false ones.
Whichever rule you pick, it has to be written down first. Decided afterwards, the break criterion is the exact place where analysis turns into justification.
| What price does | How it reads |
|---|---|
| It bounces inside the gap | The FVG held. Nothing flips, and the object stays an ordinary fair value gap. |
| It pierces with a wick | Ambiguous, and the most frequent case. Without a rule written beforehand, this is where analysis bends to the outcome. |
| It closes beyond | The break is clean. This is the strict version, the rarest and the most defensible. |
| It comes back to test from the other side | This is the moment ICT methods look for, and the only one where an inverse FVG is of any use. |
| It leaves and never returns | Nothing to do. An inverse FVG with no retest is an observation, not an opportunity. |
The table above shows that an inverse fair value gap is not decided on the gap, it is decided on the break rule. It is the only place where a decision has to be made, and the only place where you can lie to yourself: a wick two points beyond is a break if you were looking for a reason to sell, and a mere sweep if you were looking for a reason to buy.
The second point is rarity. A strict reading, on the close, produces few cases, and that is a quality rather than a flaw. A loose criterion produces inverse FVGs constantly, and an object you can find at will no longer distinguishes anything. The whole value of the concept rests on it staying exceptional.
The edges have not moved by a point, only their role has changed: the 1.0960-1.0980 band is now an inverse fair value gap, and it is watched from above. The test at 1.0962 is what ICT methods call the retest of the IFVG.
This is the concept's real contribution, and it is rarer than it looks in technical analysis. Most objects describe what works: a support that holds, a gap that fills. The inverse FVG describes what did not work, and uses that failure as information.
It changes how you follow a chart. A level you watch no longer disappears when it gives way, it simply moves to the other side, and your list of useful levels stops shrinking with every break.
An inverse FVG is almost never taken on its own. It confirms an existing reading, for instance a structure already broken in the same direction, and it brings what structure does not give: two precise prices at which to act.
It is also the most quoted combination in ICT methods, known as the IFVG and SMT model, where a divergence between two correlated markets says the move lacks support while the inverted gap says where price should react.
The central flaw is the break criterion left implicit. Almost no presentation of the concept says what counts as a break, and the ambiguity is convenient: it lets you declare an inverse FVG after the fact, once you know where price went. A concept that can only be validated in hindsight cannot be tested, so it cannot be improved.
The second is inflation in the number of levels. If every fair value gap that gets traded through becomes an inverse FVG, a chart carries as many as it carried gaps, that is dozens. The reading then loses what made it interesting, rarity, and turns back into a grid where a level can be found at any price.
The third is the most general: none of this is measured on your own trades. That the inverse FVG works in the videos explaining it says nothing about what it produces for you, with your break rule and your way of exiting. A tag on those trades and thirty occurrences answer the question; no amount of reading does.
The classic mistake. Deciding after the fact that a gap was inverted. Because the concept describes a failure, it is always perfectly visible once you know where price went, and that is exactly what makes it unusable in real time. Without a break rule written beforehand, an inverse FVG is not a reading, it is a commentary.
| Indicator | Formula | What it measures | When to use it |
|---|---|---|---|
| Inverse FVG | An FVG price has closed on the other side of | A level's failure, and its role swap | Confirming a reversal you already suspected |
| Fair value gap | 3 candles, the 1st and the 3rd do not overlap | A price range that barely traded | Locating a return, or aiming at a fill |
| Breaker block | An order block traded through, then retested | The same role swap, on a candle | Looking for a reaction when price returns |
| Support turned resistance | A horizontal level broken | The classic role swap | Reading a chart without ICT vocabulary |
Close beyond the gap, or price simply passing through: choose, write it down, and apply it even when it does not suit you. It is the one action that turns the inverse FVG from a justification into a criterion, and it costs nothing but a sentence.
A move that breaks several gaps does not hand you several opportunities, it hands you one. Keeping only the gap closest to the turning point avoids the inflation of levels that empties the concept of its interest.
They are two different readings and there is no reason for them to produce the same result. Two distinct tags in your journal, and after thirty or so trades you will know which of the two pays you, which no video can tell you.
It is a fair value gap price has traded through instead of bouncing off, and whose role has flipped: a bullish gap broken downwards then reads as a resistance area. The two edges stay those of the original gap, only their reading changes.
FVG vs IFVG: they are not two objects but two states of the same one. The FVG is the imbalance left by three candles; the IFVG is that same imbalance after price has traded through it. So you do not choose between them, you observe which of the two states a gap you were already following is in.
Yes. Both terms refer to the same object, and no method gives them different meanings: the difference comes from translation, not from a technical distinction.
It is the most quoted combination around the inverse FVG: an SMT divergence between two correlated markets says a move lacks support, and the inverted gap says at what price to act if price comes back. Our lesson on the inverse fair value gap covers that case.
There is no universal rule, and that is the concept's weak point. The strict version requires a candle close beyond the gap; the loose version accepts price simply passing through. The first produces far fewer cases and far fewer false ones. What matters is having chosen beforehand, not afterwards.
In Tradoshi, the question raised by the inverse fair value gap is settled with two tags rather than one: one for trades taken on an intact FVG, one for trades taken on an inverted gap. The Performance by tag block in the Journal screen then gives, for each separately, the trade count, the win rate, the average R and the net P&L, and clicking a row filters the journal down to those trades.
It is the only way to arbitrate between the two readings on your own numbers rather than somebody else's. Backtesting speeds up the accumulation of occurrences by replaying the market candle by candle, with the same statistics at the end, which saves waiting six months of live trading for an answer.
To go deeper on this: read the full guide. Every term is gathered in the glossary.