Why your stop loss keeps getting hunted — the order book mechanics retail doesn't see

You enter long. Price drifts down to your stop. Stop fires, you're out at the worst price of the move. Then price reverses and runs in your intended direction. This happens often enough that retail traders develop a folk belief that "they" are targeting their stops personally. There's a kernel of truth — but the mechanism is mechanical, not personal, and once you understand it you can place stops where they won't be the easy meal.

What a stop loss actually does

A stop loss isn't a limit order sitting on the book. It's a trigger that converts into a market order when price touches the trigger level. The exchange sees the trigger; the market doesn't see it as a resting order. But the exchange's matching engine, and especially the data feeds that flow to high-frequency market makers, give very strong hints about where stops cluster.

Stop clustering is the whole game. Individual stops are invisible. Clusters of stops at the same price level are not — they show up indirectly in order book skew, in funding pressure, in open-interest distribution at price levels. A market maker can see clustering even though they can't see your specific stop.

The 5 places retail stops cluster

1. Round numbers

BTC at $60,000. BTC at $65,000. SOL at $200. Round-number stops are everywhere — people think in round numbers when they place stops, even though there is no special significance to those levels. Round numbers within 0.5% of spot accumulate the most stops.

2. Yesterday's low (for longs)

"Below yesterday's low" is a default stop placement taught in most technical analysis books. Everyone uses it. The level becomes a magnet for stops, especially on coins with active retail communities discussing the same chart.

3. The 200-day moving average

For swing traders, the 200d MA is the de facto trend filter. Stops placed "below the 200d MA" cluster at that level. When price approaches it during a pullback, market makers can predict the stop wave that will fire if it breaks.

4. The session high/low

For day traders, the prior session's high or low is the canonical stop zone. Stops accumulate just below the previous low for longs, just above the previous high for shorts. Around session boundaries (00:00 UTC for crypto), these levels become very visible.

5. Liquidation prices

Leveraged traders implicitly cluster at exchange-computed liquidation prices. A 25× long at entry has liquidation at roughly −3.5%; a 50× long at roughly −1.8%. These prices show up in the exchange's risk system and in third-party tools like Coinglass, which publish "liquidation heatmaps." Market makers can see exactly where the leveraged stop wall is.

How market makers exploit clustering

The exploitation is not personal. It's just opportunistic.

Say the order book shows a cluster of stops sitting around $59,800 on BTC. Spot is at $60,200. A market maker with inventory to sell can:

1. Sell aggressively into the book down to $59,810.
2. This triggers the stop cluster, which converts to market sell orders.
3. The forced selling drives price further down to $59,500.
4. Market maker buys back at $59,500, closing their initial short at a profit AND accumulating inventory at a discount.
5. Price re-equilibrates back to $60,000+ once the stop wave is exhausted.

The retail trader's stop was the gun. The market maker provided the trigger. The "manipulation" was just a market maker noticing that breaking $59,800 would unlock a wave of forced selling worth more than the cost of breaking that level. This is rational market-making behaviour, not malice.

Why this happens more on low-liquidity coins

On BTC, the order book has billions of dollars of depth. Triggering a stop cluster requires significant capital and doesn't usually move price much beyond the level itself.

On mid-cap altcoins, the book is thin. A relatively small sell order can sweep through multiple price levels and trigger cascading stops. The cost-to-payoff ratio for market makers to hunt stops is much better on thin books — which is why these moves happen more on altcoins than on BTC.

This also explains why most retail "stop hunting" complaints involve mid-cap altcoins, not BTC. The mechanic is much more profitable on lower-liquidity venues.

5 ways to place stops where they aren't the easy meal

1. Stop placement off round numbers

If everyone places stops at $60,000, place yours at $59,800 or $60,300. The slight giveaway in stop distance is worth the difference between being inside a cluster (gets swept) versus outside one (survives the sweep).

2. Stop above/below liquidation cluster price

Use Coinglass or Coinalyze liquidation heatmaps to see where the leveraged stop wall is. If the cluster is at $59,500, place your stop at $59,200 or $59,400 — below the wall, not at it. Surviving the sweep is what matters; being slightly further from entry costs you a fraction of the saved stop-hunt loss.

3. Time-based stops instead of price-based stops

For some strategies, the cleanest exit is "close after 4 hours regardless of price." This is a time stop. It cannot be hunted because there's no level to trigger it. You sacrifice some directional precision for immunity to stop hunts.

4. Wider stops + smaller position size

Per-trade risk = position size × stop distance. If you double the stop distance, halve the position size to keep dollar risk constant. The wider stop is harder to sweep; the smaller position has the same dollar exposure. Trade-off: you need higher win-rate or better R/R because you're risking the same dollars on a less-precise entry.

5. Mental stops + slow manual exits

For experienced traders only. Don't place any stop on the exchange. Watch the position manually and exit at-market when your mental stop is reached. The exchange has no visible level to defend; market makers can't hunt what isn't visible. Cost: requires discipline, presence, and emotional control that most traders don't have.

What this means for systematic strategies

Systematic strategies that use fixed-percentage stops (e.g., "always stop at −5%") are uniquely vulnerable to stop hunting because the stop level is predictable. If a market maker sees you opened a long at $100 with a known fixed stop pattern, they know stops cluster at $95.

Better systematic approaches:

• ATR-based stops (volatility-adjusted, varies per coin)
• Structural stops (below recent swing low, but with offset)
• Time stops combined with looser price stops
• Trailing stops that move with the trade, never at predictable round levels

Our internal strategies use a mix of these — none of them uses pure fixed-percentage stops on majors specifically because the predictability cost outweighs the simplicity benefit. We will not publish the exact stop logic (firm IP), but the framework is exactly what's described above: stops that don't form predictable clusters at obvious levels.

The honest takeaway

Stop hunting is real. It is also not personal. It is a mechanical consequence of clustered orders meeting profit-seeking market makers in thin books. Once you understand the mechanism, you can place stops in positions where the mechanism doesn't catch you.

None of the techniques in this article are secret. They're just rarely combined into a consistent practice. Most retail traders read about ATR stops in one article, swing-low stops in another, and then default back to round-number stops when they actually place trades because that's the path of least mental effort.

The traders who survive long enough to compound capital are the ones who consistently apply non-obvious stop placement, even when it means slightly worse entries or slightly larger losses on the trades that do hit. They take the small known cost of "better stop placement" to avoid the large random cost of "stop got hunted."