What Is a Liquidity Sweep? How to Spot and Confirm One

Most explanations of this stop one step too early. They show a long wick, say smart money took the stops, and move on. That leaves you with the hardest part unanswered: on a live chart the sweep and the breakout look identical for as long as the candle is still printing, and the difference between them is the difference between a level that held and a level that is gone.

So this page is about the mechanics. What the event actually consists of, which candle makes it official, what to do when the first attempt is followed by a second and a third, and what changes on your chart the moment it is done. If the levels themselves are still fuzzy, where those pools sit covers the ground underneath this one.

What is a liquidity sweep on a chart?

A liquidity sweep is one candle doing two things: reaching beyond a level that was obvious to everyone, and then closing its body back on the side it came from. The reach fills the orders parked past the level. The close says the move past it was not accepted. Both parts are required, and neither one on its own is the event.

Think about what sits past that level. Traders who sold at the high left their protective stops above it, and those stops are buy orders. Traders waiting for a breakout left their entries there too. When price reaches up, all of that gets filled at once, which is exactly what a large participant needs in order to sell size without pushing the price down against themselves.

Then the body closes back, and that is the tell. If a move through the level were genuine, there would be no reason for the candle to give the ground back inside the same bar. The methodology this site works from reads it plainly: price collected what it came for and had no business going further.

The event travels under several names. Some traders say the level got raided, others call it a stop run or a liquidity grab, and there are small differences in how those terms are used. Mechanically they describe the same candle behaviour, and this page uses "sweep" throughout.

What matters more than the name is what the sweep does not do. It does not break the level. After a sweep the level is still there, still meaningful, and still the reference point for whatever comes next. That is the whole reason to tell the two cases apart.

What has to be on the chart before you can spot one?

A sweep is defined against something, so the level comes first. You need a price that was obvious before the move, one closed candle to read, and nothing else. No indicator is required for any part of this, which is worth saying because the event is often presented as something only a tool can see.

The levels worth watching are few. The structure levels your markup already uses. The previous day's high and low. The high and low of a trading session. Equal highs or equal lows sitting side by side. The edges of zones already on your chart, such as an order block or a stretch of order flow. Every one of them was visible to every trader on the instrument, which is the only reason orders collected there in the first place.

Marked before the move is the part that gets skipped. A level you drew after seeing the wick is not a level, it is a memory of a wick. If you cannot point at a high and say it was already on your chart an hour ago, you are not reading a sweep, you are decorating one.

Timeframe deserves the same discipline. The sweep belongs to the chart where the level lives. A one-minute candle stabbing through yesterday's high is not a sweep of that high until the timeframe you marked it on produces a close, and a decisive-looking wick down there is regularly one bar inside an hourly candle that is still open. Read the event where the level was drawn.

How do you spot a liquidity sweep step by step?

The procedure is short enough to run in your head, and running it in that order is what keeps you honest: the level is chosen before the candle, and the verdict comes after the close instead of during it.

Step three is where most of the damage happens. An unclosed candle is genuinely undecided, and traders who act inside it are not reading the market faster than everyone else, they are guessing earlier. The candle that looks like a perfect sweep with four minutes left often closes as a clean break with the wick gone entirely.

Which wick counts, and which one does not?

Not every poke past a level is a sweep, and the filter has nothing to do with how long the wick is. What counts is whether the object being tested is one that demands a body close. Most of them are, and there is exactly one that is not.

The body rule applies to structure levels, order blocks, order flow, any point of interest you have marked, session boundaries, and the previous day's high and low. For all of those, a wick through with the body closing back means the level held and the liquidity there was collected. A body closing beyond means the level is broken and the reading changes completely.

Two habits follow from this. The first is that wick length is not evidence. A long wick looks dramatic and tells you only that price travelled and came back, which a small wick also tells you. The second is that you have to know what you are testing before you can judge the candle, because the same bar can sweep one object and break another sitting a few pips away.

There is also the boring case that nobody illustrates: the wick that does not reach the level at all. It is not a sweep, it is a miss, and the orders past the level are all still sitting there. Untouched pools stay on your chart. They are the most likely destination price has, and the fact that a candle got close changes nothing about that.

What confirms the move after the wick

The single-candle version above is the simplest of three cases, and the other two are where live charts spend most of their time. Once a wick has pierced your level, the level you measure against stops being the line you drew and becomes the extreme that wick reached.

The three cases, in the order you meet them:

1. The candle settles it by itself. Price reaches past the level and the same candle closes its body back on the original side. The liquidity is taken, the level held, and there is nothing further to wait for.
2. The level was pierced by a wick, and after that a close beyond the original line no longer settles anything. Confirmation now requires a close beyond the extreme of that piercing wick, because the wick's tip is the furthest price the market has been willing to reach and reject.
3. Several candles take turns poking past, each leaving another wick. Confirmation is a close beyond the furthest wick of the whole series, not beyond the nearest one and not beyond the line you started with.

That shifting reference is missing from most explanations of the idea. It is a markup rule of the methodology rather than a measured market law, and the author's stated basis for it is his own testing, which is worth knowing when you decide how much weight to give it. What it buys you is protection from a very common false read: a body closing "beyond the level" while sitting comfortably inside a wick from four candles ago.

The practical effect is a small change in how you watch a level under pressure. After the first pierce, move your attention to the tip of that wick. That is where the market's last rejection happened, and until something closes past it, the level is still doing its job.

What happens to the level once it has been swept?

It moves. This is the step that keeps a chart honest a week later, and it is skipped almost universally: after a sweep, the structure point relocates to the new extreme the wick made, and every later event is measured against that new position, not the line you originally drew.

The logic is short. The old extreme no longer represents the furthest price the market rejected, because the wick went past it. Orders that were resting there have been filled. What is left is a fresh extreme, unfilled and untested, and that is now the price other traders can see and react to.

Leaving the old markup in place produces a quiet, repeatable error: you read the next candle against a level that no longer holds anything, and everything downstream of that reading inherits the mistake. Nothing looks wrong at the time. The chart is simply out of date, and traders who mark by hand usually find out weeks later while going through screenshots.

So the routine after a sweep is not "note it and wait". Move the level to the new extreme first, then keep reading. Whether what follows amounts to a genuine shift in structure or another trap is a separate question, and telling a real break from a trap takes it on directly.

How do you check that what you saw was a sweep?

You check it on closed candles, one at a time, against a level that was already there. Reviewing history this way is the cheapest practice available for this idea, because the answer for every candle is knowable and you get a verdict per bar instead of an opinion per chart.

Take a level from last week and walk forward bar by bar. For each candle that interacts with it, answer one question and write the answer down:

• The wick went past and the body closed back on the original side. That is a sweep. The level held.
• The body closed beyond the level. That is a break. Stop treating the level as intact.
• The wick never reached the level. Nothing happened, and the orders past it are still there.
• The candle has not closed. There is no verdict yet, and inventing one is the habit the exercise exists to break.

Two things usually surface in the first session of this. Sweeps are much more frequent than the highlight-reel examples suggest, and most of them are unremarkable: a small wick, a close back, price carrying on with whatever it was doing. And the candles that felt most obvious in the moment are frequently the ones whose body closed on the wrong side of the level for the story you were telling. A handful of annotated examples is a reasonable companion to the exercise, but doing it on your own instrument is what makes it stick.

What a liquidity sweep does not tell you

A sweep is an event, not a signal. It tells you the orders resting past a level have been filled and that the level survived. Direction and distance are outside what it establishes, and pretending otherwise is the fastest way to make the idea expensive.

Price does turn after sweeps, often enough that the pattern is worth knowing. It also takes a level and carries on in the same direction it was already going, or takes one at each end of a range within the same session and leaves you with two contradictory stories. What follows depends on which level was taken and on the context around it, which is why the same wick means different things at the previous day's high and at the edge of a zone you have already marked.

This is also where the honest limit of the whole idea sits. Retail traders on forex and CFDs cannot see resting orders. What you are doing is inferring that a level was obvious, therefore orders collected there, therefore filling them was worth a trip. The inference is reasonable and it remains an inference. Treat a sweep as information about a level, put it next to the rest of your read, and keep the risk plan doing the job that a candle pattern cannot do. Where the swept level sits relative to the range around it usually matters more than the wick itself.

Whatever trading the event looks like for you, it starts after the close, not during it.

Marking sweeps without staring at the chart

The rules on this page are simple and the work is not, because it is the same check repeated across every level on every instrument you follow, at the moment each candle closes. That is what an indicator is actually good for: applying one rule the same way every time while you are asleep, at work, or watching a different pair.

Money Hunter marks the sweep variants on the chart as they occur, alongside the levels they belong to, so the level and its verdict arrive together rather than as a judgement you make from memory. It is non-repaint by code, which matters more here than almost anywhere else: a tool that quietly re-draws its markings after the fact would be showing you sweeps that were never visible when the candle closed. You can check that claim yourself with the 100-bar replay test rather than take it on trust. What the tool cannot do is tell you the trade. It marks the event, and the reading stays yours.

FAQ