Liquidity Sweep vs Support and Resistance
Liquidity sweeps are moves that run past an obvious level and reverse, on the argument that stops behind it were triggered. Support and resistance is that obvious level, so the first concept depends on the second rather than competing with it.
These get presented as belonging to different eras of analysis. In fact one is built on top of the other: without an obvious level, the newer concept has nothing to describe.
What each one is
Support and resistance is a level price has already reacted to, usually more than once, visible to anybody looking at the chart. Support and resistance covers it.
A liquidity sweep is a move past such a level that then reverses, on the argument that stop orders clustered behind it were triggered. Liquidity sweep covers it.
The second requires the first. Stops cluster behind prices lots of people can see, so the sweep’s whole premise rests on the level being widely watched.
Where they differ
One is a place and one is an event. The level sits on the chart; the sweep is something that happens to it, and only sometimes.
When each can be identified. The level as soon as price has reacted there twice. The sweep only after the reversal, which is later than the entry most people want to take.
What each claims. The level claims price has reacted here before, which is visible. The sweep claims stops were sitting behind it, which is not.
How each fails. The level fails by being passed through. The sweep fails by not reversing, in which case what happened was simply a breakout.
Where they agree
Both need the level marked in advance. A line drawn after the reaction describes history, and both concepts look flawless when identified that way.
Both are frequently irrelevant. On this site’s shared series direction runs average 2.01 bars with a longest of 11, and price travels through marked levels constantly.
Both cost a round trip when acted on — about 2% of the median bar range of 0.493 here — which a method taking every excursion past a level pays very often.
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 an invalidation just beyond a level has to survive.
Which one to use
Draw the levels first. Nothing in the sweep concept works without one, and the more obvious the level the better the premise holds.
Use the sweep as a trigger at those levels. A run past followed by a return is a defined thing to wait for, and it is more specific than simply trading the level.
Prefer the most obvious levels available. Prior session highs, prior swing highs, round numbers — places where a lot of people would have placed a stop.
And when the level is one only you can see, drop the sweep framing. There is nothing behind it to take, so the mechanism the concept relies on does not apply.
Why obviousness is the mechanism
Because stops cluster where people can see. That is the entire argument for the sweep, and it applies in proportion to how many people are watching the level.
And because a private level has nobody behind it. A line derived from your own rule may still work as a level; it does not support the stop-hunting story at all.
What to write down before using either
How many touches make a level. Two, three, or a wick each side. Without a number, any area qualifies once price has bounced somewhere near it.
How far past counts as an excursion. A wick beyond and a close beyond are different events, and the ninetieth percentile bar range here is 1.101.
How quickly the reversal must come. Within how many bars, and back through what. Without a limit, every eventual return qualifies.
And an expiry on the level. An untouched level from months ago is either live or it is not, and deciding once price approaches is deciding by outcome.
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,
liquidity sweeps appear in 69 titles at a median of 17,537 across 49 channels, and support and
resistance in 145 at a median of 30,434 across 112. The counts come from site/corpus_count.py.
69 videos on the event at 17,537 against 145 on the level at 30,434. Half the coverage and half the audience per video — the older concept leads on both counts, which is unusual and reflects that it is the one the newer idea is built on.
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 — and on this site’s shared series 85% of 39 twenty-bar breakouts held, so the base rate favours the other reading.
When it fails
The failure is anticipating the sweep and shorting into a breakout, and the base rate is against you. Price exceeds an obvious high, which is the first half of a sweep, so a position is taken before the return. It is also the first half of a breakout — on this site’s shared series 85% of 39 twenty-bar breakouts held and 100% of the 11 fifty-five-bar ones did. The trade is entered against the more likely outcome, on the strength of an event that had not yet occurred.
The second failure is drawing the level after the move. Everything works backwards.
A third is using a level only you can see. Nothing is behind it.
A fourth is no time limit on the reversal. Every return counts eventually.
A fifth is no expiry on the level. They accumulate until the chart is unusable.
And a sixth is assuming the stops were there. That is inferred, not observed.
Related
Liquidity sweep covers the event. Support and resistance covers the level it depends on. And liquidity covers the premise underneath the whole idea.
The sweep idea only works because the level is obvious. Stops cluster behind prices lots of people can see, so a privately derived level has nothing behind it to take. That makes the older, plainer concept the load-bearing half of the pair.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.