Fair Value Gap vs Liquidity Sweep
A fair value gap marks a band price moved through without trading evenly, and it can be drawn as soon as three bars complete. A liquidity sweep is an event — price runs past an obvious level and reverses — which is only identifiable after the reversal.
Two pieces of the same vocabulary that get offered as competing setups. One is somewhere on the chart; the other is something that happens. A method usually needs both.
What each one is
A fair value gap is a band with exact boundaries. Three bars, the outer two not overlapping, marking where price moved through without trading evenly. Fair value gap covers the rule.
A liquidity sweep is an event. Price runs past an obvious level — a prior high or low — and then reverses, on the argument that stops behind it were triggered. Liquidity sweep covers it.
One can exist before price arrives; the other cannot. That asymmetry governs everything else about the pair.
Where they differ
When it can be identified. The gap the moment three bars complete. The sweep only once price has come back, which is after the entry most people want to take.
What each depends on. The gap depends on bar geometry alone. The sweep depends on an obvious level existing, because that is where the stops are assumed to sit.
How each fails. The gap fails by never being revisited. The sweep fails by not reversing, in which case it was a breakout — and nothing distinguishes the two while it is happening.
How checkable each is. The three-bar test either holds or it does not. A sweep needs a written definition of the level, the distance and the time limit before it can be tested at all.
Where they agree
Both need a written rule. Which bars, which level, how far past, how quickly back — without those numbers both are claimed after the outcome.
Both come from the same vocabulary and both make an argument about orders you cannot observe: an untraded band, or clustered stops behind a level.
Both are frequent. On this site’s shared series direction runs average 2.01 bars with a longest of 11, so levels are exceeded and bands are left constantly.
And neither supplies a stop. The ninetieth percentile bar range here is 1.101 and the largest single bar was 2.338, which is what any nearby invalidation has to survive.
Which one to use
Use the gap as the location. It is the half you can mark before price arrives, which is what makes a reviewable record possible at all.
Use the sweep as the trigger. Something has to happen when price reaches your band, and a run past an obvious level followed by a reversal is a defined thing to require.
Use them together rather than choosing. A place plus a trigger is a method; either alone leaves the other half of the decision unspecified.
And never enter on a sweep before the reversal. Until price comes back it is a breakout, and on this site’s shared series 85% of 39 twenty-bar breakouts held.
Why the timing asymmetry matters
Because one of them is a prediction and the other is a measurement. Marking a gap requires no view about the future. Calling a sweep before the reversal requires exactly one.
And because the review is distorted by it. Sweeps that reversed are memorable; the identical moves that kept going are just breakouts, and nobody counts them.
What the rules have to specify
For the gap: whether wicks count, and on which timeframe. Two answers, and they change how many you find by a large factor.
For the sweep: which level. A prior swing high, a session high, a round number — obvious to whom, and how far back you are willing to look.
How far past counts. A wick beyond and a close beyond are different events, and the ninetieth percentile bar range here is 1.101, so a small excursion proves very little.
And how fast the reversal must be. Within how many bars, and back through what. Without a limit, every eventual return qualifies and the pattern is always available in hindsight.
The original data
Of the 24,971 unique videos in research/search-study-corpus.jsonl, no title compares these two
directly — this pair is constructed from two subjects the corpus covers separately. Separately, fair
value gaps appear in 150 titles at a median of 28,170 across 108 channels, and liquidity sweeps in 69 at
a median of 17,537 across 49. The counts come from site/corpus_count.py.
150 videos on the location at 28,170 and 69 on the event at 17,537. Twice the coverage and a comparable audience per video — both are well established, and the checkable one is the better covered of the two, which is unusual in this vocabulary.
The answer to the question on that chart is that nobody can tell yet. The reversal is what makes it a sweep, and it has not happened — so any name given to it now is a forecast rather than a reading.
When it fails
The failure is anticipating a sweep, and it puts you against a move that has just shown strength. Price exceeds an obvious high, which is what the first half of a sweep looks like, so a short is taken before the reversal. It is also what a breakout looks like — on this site’s shared series 85% of 39 twenty-bar breakouts held and 100% of the 11 fifty-five-bar ones did. The position is entered against the more likely outcome on the strength of a pattern that had not yet occurred.
The second failure is treating them as alternatives. One is a place, one an event.
A third is loosening the three-bar rule. Every fast move then qualifies.
A fourth is no definition of the level swept. Any high counts.
A fifth is no time limit on the reversal. Every return qualifies eventually.
And a sixth is assuming the stops were there. That is inferred, not observed.
Related
Fair value gap covers the markable band. Liquidity sweep covers the event. And liquidity covers what the sweep is premised on.
The asymmetry is the point. One of these you can draw on the chart today and check next week. The other cannot be named until the move has already reversed, which means any live claim about it is a prediction dressed as an observation.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.