Stop Hunting in Forex: Why Your Stop Loss Gets Taken

Written by Dominic Walsh · Published 23 Jul 2026 · Last updated 20 Aug 2026

Stop hunting in forex is the pattern every retail trader knows too well: price dips a few pips past your stop under a clean swing low, fills you, then rallies exactly where you expected. It is not bad luck. After this article, you will know why it happens, where it strikes most, and how to place stops that survive it.

Better still, you will see how the same raids can become entries. The move that takes your stop is information, and the traders on the right side of it are usually reading liquidity, not predicting candles. So treat every raid as a lesson in where orders live.

What Is Stop Hunting in Forex?

Stop hunting is a sharp push through an obvious level where stop-loss orders cluster, followed, in many cases, by a reversal once those orders fill. The push exists to consume the orders. Your stop under a swing low is a pending market sell; a thousand stops at the same level form a wall of sell volume that someone with size wants to buy.

Notice what that definition does not say. It does not claim your broker singled you out, and it does not require anyone to see your individual order. Instead, the cluster is the target. You just happened to be standing inside it.

The Hero Chart: A Gold Stop Hunt in One Candle

Now watch one raid on a real chart. The gold (XAUUSD) 1-hour chart below comes from July 17, 2026. A support shelf near 3,970 held four separate touches through the session, and every touch added fresh longs with sell stops tucked beneath the level.

Then the hunt fires. A single bearish candle wicks down to 3,959.8, roughly ten points through the shelf, and volume jumps on that exact bar. The volume spike is the tell: clustered stops fired as market sells, and a large buyer absorbed them. That same hour closes back at 3,997, far above the broken level. Price then extends to 4,021 over the following bars. Wick, volume burst, immediate reclaim: that is the complete signature of a stop hunt. Volume matters here because it independently confirms that orders actually changed hands at the lows, rather than price drifting through an empty book.

The Mechanics: Liquidity, Not Conspiracy

Institutions face a fill problem. A desk buying 400 million EURUSD at quiet mid-range prices would push the market against its own order before a fraction filled. Instead, it waits for moments when the market hands it sellers in bulk. Triggered sell stops below equal lows are exactly that: a burst of market sell orders any large buyer can absorb.

Thus the sequence looks engineered even when it is simply mechanical. Price gets drawn below the lows, the stops fire, the size fills, and the pressure that drove price down disappears. Then the reversal begins, often violently, because the sellers are exhausted and the buyer is done accumulating. Hence no villain is needed; the fill itself explains the snap-back.

Do Brokers Hunt Your Stops?

Mostly, no. The forex market turns over trillions of dollars daily, and a regulated broker earns more from your trading volume over time than from one forced stop-out. Also, slippage on stops during fast markets is normal order mechanics, not proof of targeting. Still, dealing-desk conflicts have existed at shady offshore shops. So pick a well-regulated broker, then retire the conspiracy theory; the chart explains the raids better than any villain does. Indeed, execution quality varies far more than intent does.

Whose Stops Fund the Move

Look back at the gold chart and count the participants. Four touches of 3,970 means four rounds of dip buyers, each round parking sell stops beneath a level that looked stronger every time it held. Add breakdown sellers with sell stop entries under the shelf, waiting for the break. Both groups supplied market sells at 3,960, and both watched the close at 3,997 from the wrong side. Their combined orders were the fuel; the raid was just the match. The flow below strips that event to its five moving parts.

Where Stop Hunts Happen Most

Raids concentrate where stops concentrate. Watch four zones:

• Equal highs and equal lows. The densest clusters on any chart. Indeed, these pools are the core of buy side and sell side liquidity.
• Previous day and session extremes. The previous day low, the Asian range, the London high. Plainly, these are default stop anchors for millions of accounts.
• Round numbers. 1.1000, 1.2500, 150.00. Stops pile a few pips beyond them.
• Long-standing trendlines. Every touch adds fresh stops just beyond the line.

Timing clusters too. Most raids fire at the London open, roughly 2:00 to 5:00 a.m. New York time, and again around the New York open, 7:00 to 10:00 a.m. Volatility expands in these windows while fresh institutional flow arrives. So check the exact windows for your time zone with our forex market hours tool. Meanwhile, news releases add a third window, since thin pre-news books let price leap through several levels at once.

How to Place Stops That Survive the Hunt

You cannot make the raids stop. Yet you can stop parking your risk inside them. Work through this checklist on every trade:

1. Find the level that proves you wrong. Your stop belongs where the trade idea is invalid, not at a fixed pip distance from entry.
2. Move it beyond the sweep zone. If the obvious low sits at 1.0785, the raid will likely probe a few pips under it. So place the stop under the deeper structural swing, past where a sweep would reach.
3. Add a volatility buffer. Pad the stop by a fraction of the average true range (ATR), for example half the 14-period ATR on your entry timeframe, so ordinary noise cannot tag it.
4. Size from the stop, never the reverse. A wider stop means a smaller position at the same account risk. Our position size calculator does the lot math from stop distance in seconds.
5. Refuse to tighten into a pool. Trailing a stop to just under the latest minor low hands it straight back to the next sweep.

Step two deserves emphasis, because it changes outcomes fastest. Most stopped-then-reversed trades die from a stop placed at the obvious level rather than beyond it. Truly, a few pips of structure-aware placement, funded by a slightly smaller size, removes you from the raid path entirely. Next, rehearse that placement on past charts until it feels automatic.

Stop Hunts in the SMC Workflow

Smart money concepts (SMC) traders slot the raid into a fixed sequence: higher timeframe (HTF) bias first, then a point of interest (POI), then lower timeframe (LTF) confirmation, then the entry. The daily chart names the draw on liquidity. Next, the 1-hour chart maps the stop clusters along the route. Then a 5- or 15-minute chart confirms any sweep with displacement before an order goes in. A raid only matters inside that context; the same wick means opposite things in a trend and in a range.

Run the gold example through that filter. The higher timeframe leaned bullish with a draw above the market, the four-touch shelf at 3,970 was the POI’s guardian pool, and the hourly reclaim supplied the confirmation. Every stage of the sequence was present before the rally to 4,021. Absent any one stage, the same wick would have deserved no trust.

Pairing Timeframes and Sessions

Keep the pairing fixed: H4 or H1 for the zones, M5 or M15 for the trigger, and no switching once the trade is on. Since raids cluster in the two opening windows, the LTF watch only needs a few focused hours a day. The London sweep of the Asian range is the single most repeated sequence in this niche, with the New York open running whatever London left behind. Outside those windows, a poke through a level carries far less information, because thin markets drift through prices without intent.

Turning Stop Hunts Into Entries

Experienced traders flip the event. Instead of fearing the raid, they wait for it, because a completed sweep marks the spot where large orders just filled. The classic model works like this: price runs the pool, displacement snaps it back inside the range, and entry comes on the retrace with a stop beyond the raid’s extreme. Thus the trap becomes the trigger.

This idea predates modern smart money content. Linda Raschke published the Turtle Soup setup in the 1990s, fading false breaks of 20-bar highs and lows; our turtle soup strategy guide covers her rules and the modern variants. Meanwhile, the school of Michael Huddleston, the Inner Circle Trader (ICT), formalized the same event as the liquidity sweep and built entry models around it; the full guide to liquidity sweep trading walks through them step by step.

Also, you do not have to eyeball raids in real time. Several of our ICT indicators for MT4 and MT5 mark equal highs, equal lows, and completed sweeps directly on the chart. Still, learn the manual read first.

Worked Example 1: EURUSD London Raid

Suppose EURUSD closes the Asian session with equal lows at 1.0842 and 1.0843. Retail longs from New York hold stops near 1.0838. Also, the round figure 1.0840 thickens the cluster. Then, at 3:15 a.m. New York time, a fast bearish candle drives price to 1.0834, nine pips through the lows, filling the entire cluster.

Next comes the tell. Within two 5-minute candles, price reclaims 1.0845 with a wide-bodied bullish close. That displacement says a buyer absorbed the sell orders; no genuine breakdown behaves that way. So a long on the retrace at 1.0843, a stop at 1.0830 beyond the raid low, and a first target at the Asian high of 1.0871 builds a defined-risk trade directly on the hunt.

Compare the two traders in that move. Trader A kept a stop at 1.0838 and got swept. Meanwhile, trader B placed the stop at 1.0830 with a smaller size and held through the same raid. Same idea, same direction; placement decided who survived.

Worked Example 2: A Raid on Buy Stops

Now invert the hero chart. Picture EURUSD stalling twice at 1.16205 and 1.16208 during the New York afternoon, printing equal highs while the daily trend points lower. Shorts from the decline hold protective buy stops above 1.16215, and breakout buyers rest entries beside them. So a dense cluster of buy orders now caps the market, directly against the HTF bias.

Work the sequence step by step:

1. Mark the pool of buy stops above the equal highs at 1.16208.
2. Confirm the daily draw points down, toward the previous week low.
3. Wait for the London morning spike: price wicks to 1.16222 and stalls immediately.
4. Demand displacement: a wide bearish 15-minute body closes back below 1.16190.
5. Short the retrace near 1.16200 with a stop above the wick, at 1.16240.
6. Target the sell side cluster at the previous day low near 1.16050, taking partials at each swing.

The annotated chart below walks the same sequence with every stage labeled, from the marked cluster to the entry.

Why This Entry Logic Holds

The stop sits beyond the raid extreme because a second push through the cluster would mean real breakout demand, which breaks the read. Meanwhile, the target sits at the opposite pool, since a market that just rejected buy side liquidity has a mapped reason to seek sell side liquidity. Gold delivered the bullish version of this on July 17; this EURUSD sketch is the same trade wearing a bearish coat. Direction changes, mechanics never do.

Manage the position at structure once displacement extends. Trail the stop behind each new lower high, and take partials at every intraday swing low along the route. If price stalls short of the target pool and prints its own sweep of a minor high, close the remainder rather than negotiating with the map. Exits deserve the same rules as entries.

Common Mistakes and How to Fix Them

Five errors dominate the losses around stop hunts. Each has a correction:

• Judging the wick before the candle closes. Half-formed hunts look like breakdowns. Fix: wait for the close and the displacement candle before labeling anything.
• Calling every break a hunt. Trends break levels legitimately all day. Fix: no reclaim, no hunt; treat a holding break as a breakout.
• Revenge re-entry after a stop-out. Getting swept then chasing doubles the damage. Fix: require a fresh displacement signal before any re-entry, and log the first loss.
• Tightening stops after losses. Fear moves stops closer to the obvious level, straight into the raid path. Fix: keep structure-based placement and cut size instead.
• Trading raids in dead hours. A quiet-session poke through a level rarely carries intent. Fix: weight the London and New York windows, and distrust the rest.
• Blaming the broker instead of the level. The conspiracy story teaches nothing and repeats the loss. Fix: audit where the stop sat relative to the cluster, then move the next one beyond it.

Keep the placement rules visible while the habits form. The card below condenses them.

Pre-Trade Stop Hunt Checklist

Run these checks before entering near any obvious level:

1. Where is the nearest stop cluster relative to my planned entry?
2. Is my stop beyond the deeper swing, with an ATR-based buffer added?
3. Did I size the position from the stop distance, not the other way around?
4. Does the trade point with the daily draw on liquidity?
5. Is a session window active, or is this the quiet drift between opens?
6. If price sweeps the level first, what is my re-entry plan?
7. Have I logged the setup before the outcome is known?

When It Is Not a Hunt: The Genuine Breakdown

Not every poke through a low is a hunt. Trends legitimately break structure, and a sweep only earns the label after displacement confirms the reversal. So trade the confirmation, never the wick alone. Plainly, patience filters most false reads.

The chart below shows the failure case. Price breaks a marked support shelf, prints one weak bounce, then keeps falling with full-bodied candles. No reclaim ever arrives, and anyone who bought the “hunt” is trapped in a real breakdown.

Invalidation and the Journal

The invalidation rule is binary. A hunt read stays alive only while price closes back above the broken level within a few candles; the gold example reclaimed its shelf inside the same hour. Once a candle closes beyond the level and holds, treat the move as a breakout, stand aside, or flip the analysis to continuation. Then log it: instrument, session, the touches that built the shelf, and how the failure unfolded. Reviewing those logs weekly shows which sessions and pairs betray your reads most often.

Batch review beats trade-by-trade tinkering. After twenty logged raids, patterns surface: perhaps your failed reads cluster around news days, or on one pair, or in the drift after the New York lunch. Cut the weakest context first, and the surviving reads improve without a single new rule. Keep the failed charts filed next to the clean ones, since the contrast trains faster than either set alone.

Honest Limitations

Wider stops are not immunity. In high-volatility regimes, even buffered stops get hit, and the smaller size they force means slower equity growth in exchange for fewer shakeouts. That trade-off is real and permanent. Meanwhile, hindsight flatters this concept: swept levels look obvious after the reversal prints. Judge your read at the hard right edge of the chart, and distrust anyone quoting precise reversal statistics, because no public data set counts the stops behind a level. Survivors of every raid write the folklore, which inflates the pattern’s reputation; your own journal is the only antidote to that bias.

Related Concepts to Study Next

Two neighbors deepen the picture. First, our guide to the liquidity pool in forex explains how the clusters that attract raids build in the first place, rejection by rejection. Second, inducement in trading covers the engineered version: minor levels created to bait entries so their stops can fuel delivery into a deeper zone. Then reread the gold chart above with both ideas in mind; the four touches built the pool, and the wick to 3,959.8 collected it. Every raid you study from now on should answer two questions: whose orders funded it, and who was waiting on the other side.

FAQ

Why does price reverse right after my stop loss gets hit?

Your stop sat inside a cluster of similar orders at an obvious level. The push that took it out fed a large player the volume it needed. Then, once those orders were consumed, the pressure vanished and price snapped back. The reversal is the natural aftermath of that fill.

Is stop hunting illegal?

Trading into levels where stops rest is legal and routine; it is ordinary order-flow behavior in every liquid market. Deliberate price manipulation by a dealer against its own clients is a different matter and is prohibited under most regulatory regimes. So the concept itself is not evidence of wrongdoing.

Should I trade without a stop loss to avoid hunts?

No. Removing the stop swaps a small controlled loss for an uncapped one, and a single news spike can do severe damage. Instead, keep the stop, place it beyond structure, and cut position size to match the wider distance.

How do large players know where stop losses sit?

They do not see individual orders in spot forex. Instead, they infer clusters from structure, because retail habits are uniform: stops go under swing lows, above swing highs, and past round numbers. The chart itself reveals the placement, which is why obvious levels attract raids.

What does a stop hunt look like on a candlestick chart?

The classic print is a long wick through an obvious level, often on a volume spike, followed by a fast close back inside the range. The gold example above shows the shape: a probe to 3,959.8 and a close at 3,997 within the same hour. Without the reclaim, the same wick may simply be a breakout starting.

How far beyond a swing low should my stop go?

Past the level a sweep would reach: beyond the deeper structural swing, plus a volatility buffer such as a fraction of the ATR. There is no universal pip number, because the right distance scales with the pair and the timeframe. Results are not guaranteed; past performance is not indicative of future results.