Stop Loss Hunting and Liquidity Sweeps: Why Your Stop Gets Hit Before the Move
The stop sat in a reasonable spot and it still got hit. Price spiked a few pips past the low, took the position out, then went straight to the target without you. That sequence has a mechanical explanation. Your stop was never hidden. It sat in a crowd of identical orders, and that crowd is exactly what a large buyer wants to fill from. Seeing your stop as somebody else's entry changes how you place it, and it explains why the fix is almost never "hide the stop somewhere smarter."
Key Takeaways
- Stop clusters are inventory. The sell stops below an obvious low are the orders a large buyer fills into, because filling there is cheaper than lifting every offer above the market, which makes your protection someone else's entry.
- Sweeps are small and specific. On EUR/USD the tag is often just 5 to 10 pips past an equal low. On gold, against a daily range near $96, a $12 poke past yesterday's low is roughly an eighth of a normal day.
- The timing rhymes. London open sweeps the Asian range, New York sweeps the London morning extremes, and red-folder news tags both directions inside the first minute.
- Placement: past the zone, then shrink size. Structure plus a volatility allowance for the stop distance, with dollars risked held constant. A 25-pip stop at 0.6 lots risks about the same $150 as 15 pips at 1.0 lot.
- The broker conspiracy is mostly wasted energy. Some dealing desks do trade against you, most retail flow passes straight through, and the cluster math runs your stop either way. Placement is the part you actually control.
How a Stop Hunt Actually Works
Every fill has two sides. A market buy needs a seller at the other end of it. When a fund wants to build serious long size in EUR/USD, it has two options. It can lift every offer in the book, paying the spread and pushing its own entry higher with each clip. Or it can let price drop into the pile of resting sell stops below an obvious low and buy from those stop-market orders as they trigger.
The second option is cheaper, and the piles are visible. Liquidity providers see where resting orders sit. Equal lows form because thousands of traders look at the same double bottom, sell the same breakdown, and park the same stops a few pips below it. The cluster prints itself on the chart in plain sight.
So price dips into the cluster, the stops trigger as market sells, and the buyer absorbs all of it in seconds. Then the real move starts. Textbooks drew this pattern decades ago and called it a spring or a shakeout. The mechanic never needed a fancier name than this: someone needed your exit to be their entry.
The conspiracy version, where a shadowy desk personally targets your 2-lot, wastes energy that belongs on placement. I have never seen the order flow that ran me, and I stopped pretending I could. What you can see is where the crowd parks its stops, because the crowd draws the same obvious levels you were tempted to use, and you already know where you would hide your own stop.
Where the Stops Actually Sit
Clusters form wherever a level is so obvious that everyone protects the same trade the same way:
| Level | Who Parks Stops There | What the Sweep Looks Like |
|---|---|---|
| Equal lows (double bottoms) | Buyers protecting longs, sellers riding the breakdown | a few pips below, then reversal |
| Equal highs (double tops) | Sellers protecting shorts, breakout buyers above | spike above, then rollover |
| Prior day high and low | Range traders fading the extremes | pierce, run a little, mean revert |
| Asian session range | Anyone trading the London breakout | London open tags one side, sometimes both |
| Big figures and round numbers | Everyone, because the level announces itself | overshoot by pips, then back inside |
The common thread is that the level is legible. Any trader with six months of charts draws the same lines you do. The stops are the most predictable thing on the screen, and predictability at scale gets used.
When the Sweeps Run
Session opens, because that is when the volume arrives to push price around. The Asian session in EUR/USD often holds a 20 to 35 pip range, against a 24-hour average near 55 pips recently, so the overnight box is roughly half the day's total travel. London only needs to run 10 pips past the Asian high to collect every trader who parked a "safe" 5-pip cushion above it.
New York does the same thing to London's morning extremes. Add an 8:30 ET data print and both directions get tagged inside a minute. Gold shows this clearly around US data: the pre-print drift toward the obvious low, the wick through it, then the actual directional move once the number lands. If you trade the NY open in gold, assume the first poke is a liquidity check before it is information.
Two practical reads fall out of the timing. If your setup needs a level to survive the open, place for the overshoot rather than the exact level. And entries taken well inside the range have a smaller sweep to survive than entries grabbed at the extreme, because chasing the session low means owning a stop that sits inside everybody else's.
Placement That Survives the Tag
Go past the zone, then add volatility
The stop belongs beyond the cluster plus an allowance for ordinary noise. A working baseline is structure plus 1.2 to 1.5 times the ATR of your trading timeframe. On EUR/USD 1h with ATR near 12 pips, that puts you 15 to 18 pips past the swing low, and the 5 to 10 pip sweep that collects the crowd closes out at your entry price instead.
Cut size to keep dollars constant
A 25-pip stop at 0.6 lots and a 15-pip stop at 1.0 lot both lose about $150 on a standard $10-a-pip pair. The wider stop does not add dollar risk. It is the same risk spread over fewer lots, and it stops breaking on ordinary noise. Widening the level while keeping the size is the combination that bleeds accounts out quietly.
Keep the level unattractive
Skip the exact tick everyone quotes. The pile sits at the obvious distance under the obvious low, so an odd offset a few pips deeper keeps you out of the pile itself while the cluster gets bought. There is nothing clever in this. You are declining to stand in the most crowded few pips on the chart.
Fit the prop math
On a funded account this discipline doubles as breach protection. A $2,500 daily loss limit on a 50k account dies to two sweep-tagged trades at $800 each plus one $1,000 revenge entry, which adds up to $2,600. Three ordinary-looking trades and the account is gone, and only the third one was the actual mistake. The buffer arithmetic behind that sizing is in the drawdown buffer guide below.
Place the stop like it will be hunted
ChartWise sizes stops from your risk budget, moves them to break-even on your rules, and enforces daily loss limits on MT4 & MT5 mechanically. Placement discipline stops depending on the kind of morning you are having.
Get Early Access →Frequently Asked Questions
Do brokers hunt your stops?
Some do, in the sense that a dealing desk which fills your order internally profits when you lose and sees the flow it holds. Most retail brokers pass orders through to liquidity providers, where your stop is one row in an aggregate book next to thousands of identical ones. Either way the placement conclusion is the same: clusters get run because running clusters is how large size moves, and no desk needs to know your name for that.
Is stop hunting illegal?
Triggering a visible cluster of resting orders is just trading, and regulators do not treat it as manipulation. What they do prosecute in this area looks different: spoofing, where fake orders paint the book and get cancelled before they fill. A sweep through an equal low is none of that. Build placement around the reality that the sweep is legal, and that it is not going away.
Where should I put my stop to survive a sweep?
Past the cluster, plus a volatility allowance: structure plus roughly 1.2 to 1.5 times the ATR of your trading timeframe, then size down so the dollar risk stays fixed. Accept that some sweeps still reach you. A placement that gets tagged once in ten trades with honest sizing is a working placement. The pattern that kills accounts is the tight stop with big size.
Is it different on gold and the indices?
The mechanics transfer unchanged, but the magnitudes grow. Gold has been running a daily range near $96, so a $12 poke past yesterday's low is ordinary noise rather than a signal, and index futures tag the first minute after the cash open constantly. Scale the allowance to the instrument's own ATR and the same placement logic applies.
The Short Version
- Your stop is part of a cluster, and clusters are the liquidity large players fill into. The run before the move happens because refilling inventory is worth paying for.
- Sweeps stay small: 5 to 10 pips past an equal low in EUR/USD, about $12 on gold against a $96 daily range.
- Windows that repeat: London open against the Asian range, New York against London's morning extremes, the first minute after 8:30 ET data.
- Stop past the zone plus 1.2 to 1.5x ATR, then cut lots so dollars stay constant. Wider stop, smaller size, same risk.
- Skip the exact tick everybody quotes. An odd offset past the visible extreme keeps you out of the pile itself.
- On funded accounts, a sweep loss followed by a revenge re-entry is the classic breach pattern. Honest sizing turns the sweep into a cost instead of an emergency.