Liquidation Heatmaps: How to Spot Trapped Traders and Avoid Getting Hunted

What a liquidation heatmap actually shows

A liquidation heatmap plots estimated liquidation density across price and time, overlaid directly on candlestick price action. The vertical axis is price. The horizontal axis is time. At each price level, the tool models how much leveraged notional would face forced closure if price traded there, and represents that concentration through color. Dark bands indicate sparse liquidation potential. Bright yellow bands indicate high concentration.

The data comes from public derivatives feeds: open interest, long/short ratios, and candle data from major exchanges. Because no exchange publishes the exact liquidation price of every open position, the heatmap reconstructs the picture using a leverage prior - a calibrated model that distributes open interest across likely leverage brackets based on exchange margin rules. The output is a probability model built from public market data.

Two things follow from that which matter for how you read it.

The clusters you see reflect a snapshot of where leverage was concentrated at the last data pull. By the time price arrives at a given level, some of those positions may have already been closed voluntarily, hedged, or rolled. The band still carries weight (enough open interest typically remains to produce real flow), but the heatmap is a probability distribution, not a live headcount of traders currently stuck at that level.

Band intensity is relative - a brighter band means more notional is concentrated at that level compared to nearby ones. The color reflects relative density, so a thick bright cluster near current price carries more weight than a thin bright band sitting far out in either direction.

Bands that have not yet been reached by price persist forward as new data refreshes. That is what makes them actionable: you can see where the market's leverage walls are sitting before price has tested them.

Why it's not the order book

The order book heatmap and the liquidation heatmap look similar on screen. They are two different tools reading two different things, and confusing them leads to bad reads.

An order book heatmap shows resting limit orders: the buy and sell orders currently sitting in the exchange's book at each price level. A dense area on the order book means traders have voluntarily placed orders there and are waiting for price to arrive. Those orders can be canceled at any moment. Large walls appear and disappear within seconds, either filled as price trades through them, or pulled before price gets there. Spoofing, where large orders are placed to signal intent and then canceled, is a known behavior on visible order books.

A liquidation heatmap shows estimated price levels where leveraged perpetual futures positions would be force-closed by the exchange. Those closures are mechanical. When a position's margin falls below the maintenance threshold, the exchange liquidates it automatically, and the trader has no say. There is no canceling a liquidation order, the event happens whether the trader wants it to or not.

The reading logic is different for each tool. On an order book heatmap, a large cluster of buy orders below price can represent support, provided those orders stay. On a liquidation heatmap, a large cluster below price represents concentrated long exposure that would generate forced selling if price reaches it. One reflects where traders want to buy. The other reflects where traders would be forced to sell.

Both tools have their uses. Reading a liquidation cluster as if it were an order book wall (treating it as a price level where buyers are waiting to step in) is where most traders go wrong.

Why liquidation clusters are not support and resistance

This is the mistake that costs traders the most. A bright cluster on the heatmap sits below current price. It looks like a floor. It looks like the level where buyers are defending. Traders treat it as support and enter long above it.

Then price sweeps through it and keeps going.

The cluster did exactly what it was supposed to do, it just wasn't doing what the trader thought.

A liquidation cluster below price represents concentrated long exposure. Traders opened leveraged long positions, and their forced-closure prices are stacked in that zone. When price enters the cluster, the exchange force-closes those longs. Each forced closure is a market sell order. Those sell orders add to the downward pressure. This forced selling can accelerate the move. The cluster that looked like a floor became the fuel that drove price lower.

The same mechanic runs in reverse above price. A bright cluster above current price represents leveraged shorts with liquidation levels sitting there. Price pushes into it, those shorts get force-closed as market buy orders, and the upward move extends.

In both cases, the cluster is where the trapped fuel sits – the level price is drawn toward, not the level where it stops.

Where confusion comes from: sometimes price does reverse at or near a liquidation cluster. When that happens, the reversal is coming from something else: a genuine structural level with passive buyers absorbing the sell flow, a key higher-timeframe zone, a coinciding CVD divergence. The liquidation cluster contributed to the volatility, the structure provided the stopping point, and traders remember the reversal and start reading clusters as support. That reading fails consistently on the setups where no real structure sits underneath.

CoinGlass, who built the primary public heatmap tool, describes these zones as "magnetic", price tends to gravitate toward high-density clusters. That framing is accurate. Magnets pull. They do not stop.

How forced liquidations create one-way flow

Every leveraged position carries a predetermined liquidation price, calculated at entry based on position size, leverage, and the exchange's maintenance margin requirement. Higher leverage compresses the gap between entry and liquidation, which is why the heatmap tends to show denser clusters in the zones where highly leveraged retail positioning is concentrated.

When price reaches a liquidation level, the exchange closes the position using a market order at whatever the book currently offers, with no slippage cap and no input from the trader. The trader loses the posted margin and the position is closed.

A cluster of long liquidations sitting below price is a concentration of pending market sell orders waiting to be triggered. When price enters the zone, those forced sells hit the book together. The additional selling pressure pushes price further into the cluster, which triggers the next cohort of positions, which generates more forced sells, which pushes price further still. Each step in the sequence creates the conditions for the next one.

That sequence is what a liquidation cascade looks like. The flow it produces is one-way because every exit is mechanical. Those sellers were removed from the market by their exchange's risk engine, and their removal added pressure to the move that removed them. A dense liquidation cluster in price's path functions as an accelerant: the larger the concentration of notional at a level, the more directional flow gets released when price crosses it.

Two mechanisms exist to handle cases where liquidations exceed a trader's posted margin. Exchanges maintain an insurance fund, built from accumulated liquidation fees, to cover residual losses when a position closes at a worse price than its liquidation trigger. 

When the insurance fund runs short, exchanges fall back on auto-deleveraging (ADL): profitable positions on the opposite side are automatically reduced to keep the system solvent. For standard market conditions, the insurance fund handles the gap. ADL is reserved for extreme dislocations.

Why stops cluster at predictable levels

Put your stop below the last swing low, below the round number, and below the level that held three times. These are standard instructions in every trading course, and they are reasonable on their own terms. 

The problem is that every trader following the same logic places their stop in the same place, which clusters liquidation potential at the most obvious structural points on the chart.

This is why bright bands on the heatmap tend to sit at levels that already look significant. The clustering is a consequence of rational individual behavior: traders independently making sensible decisions that, in aggregate, create predictable concentrations of leverage at predictable locations.

Large players operating with significant size have a structural incentive to trade into these zones. Institutions and well-capitalised traders need to fill large orders without moving price too far in the process. Zones where liquidity is concentrated, where stops and liquidation triggers are stacked, give them the volume to do that. 

When a large buyer needs to accumulate a position, a zone with dense long liquidations below price can offer a pool of forced selling to buy against. The liquidations provide the liquidity. The buyer absorbs it and builds the position.

Whether this constitutes deliberate manipulation or structural market behavior is a debate without a clean resolution. 

On highly liquid instruments like Bitcoin, the cost of intentionally pushing price to sweep retail stops is high relative to the value of those stops. What you are usually observing is a structural cluster: price gravitating to where liquidity lives, for the same reason water finds low ground. The framing of "stop hunting" often captures the mechanics accurately and the intent less reliably.

What matters for reading the heatmap is the outcome: predictable levels produce predictable clusters, and those clusters have real consequences for price when they are triggered. The heatmap makes that visible before the fact.

Why combining exchanges matters

A liquidation heatmap built from one exchange's data only shows that exchange's leverage. Binance's heatmap shows Binance's open interest distributed across price levels. Bybit's shows Bybit's. 

A cluster sitting on Binance may have no corresponding concentration on Bybit or OKX, which means the notional behind it is thinner than it appears and the cascade it could trigger is limited to that venue's participants.

When a cluster appears across multiple exchanges simultaneously, the picture changes. The same price zone is, for example, carrying concentrated leverage from Binance's broad retail and institutional base, Bybit's leveraged positioning, OKX's participant mix, and Gate's flow on top of that. If price reaches that level, the resulting liquidation flow comes from all of those venues at once. The cascade is wider, the one-way market orders are larger in aggregate, and the move that follows tends to be more sustained than a single-exchange sweep.

The practical implication for reading a heatmap is that cluster weight should be assessed relative to how many venues are contributing to it. A bright band that only appears on one exchange when you filter the view is a local concentration. A bright band that persists across the aggregated view, visible regardless of which exchange you filter to, represents genuine market-wide leverage sitting at that level.

MindPillar's Intel Dashboard aggregates across Binance, Bybit, OKX, Gate, and other major venues by default. The dropdown lets you filter to a single exchange, which means you can check whether a cluster you are watching is a market-wide concentration or specific to one venue. That distinction changes how much weight you give it before price arrives.

How to read MindPillar's Liquidation Map

If you are looking for a live liquidation heatmap aggregated across Binance, Bybit, OKX, Gate, and other major venues in a single view (free, no account required) MindPillar's dedicated Liquidation Map is at mindpillar.com/liquidation-heatmap.

The map plots projected liquidation density as horizontal bands over the candlestick chart, with price on the vertical axis and time running left to right. 

Band intensity scales with concentrated notional: a thick, bright band means significant leverage is stacked at that level across the aggregated exchanges. Thinner, darker bands indicate lighter concentration. The data is reconstructed from public derivatives data (open interest, long/short ratios, and klines), projected forward against a calibrated leverage prior (an estimate of how leverage is distributed, and refreshed every five minutes.

The data panel alongside the chart shows three figures worth checking before you read the bands: total at-risk notional across the market, how much of that sits on the downside (concentrated long exposure below current price), and how much sits on the upside (concentrated short exposure above current price).

 When downside notional significantly outweighs upside notional, the market is carrying more leveraged long exposure overall, which means more potential forced selling is sitting beneath price than forced buying above it.

Four controls shape what you see:

• Exchange. The default view aggregates across all venues. Filtering to a single exchange lets you check whether a cluster you are watching is a market-wide concentration or isolated to one venue. A band that holds its intensity in the aggregated view carries significantly more weight than one that only appears when you filter to Binance or Bybit alone.
• Interval. The 24h setting shows liquidation density built from the most recent session's data. Shorter intervals give you a tighter, more current picture of where intraday positions are concentrated. Longer intervals show larger structural walls that have been building over days.
• Side. Toggling between Both, Longs, and Shorts lets you isolate the direction of the exposure at each level. A bright cluster below price in the Longs view means leveraged longs would be force-closed there, generating forced selling. A bright cluster above price in the Shorts view means leveraged shorts would be force-closed there, generating forced buying. Checking the side breakdown before placing a trade removes a significant source of misreading.
• Bands persist until crossed. Any level that price has not yet reached remains visible on the chart as new data refreshes. That persistence is what makes the tool useful before a move rather than as a post-event explanation.

For a structured session on integrating the heatmap into a live trading process: how to combine it with CVD data for order flow confirmation, how to read cascades as they develop, and how to build it into a pre-trade checklist - MindPillar's dedicated Liquidation Heatmaps Playbook session covers it in full.

Disclaimer: Trading involves substantial risk of loss. This content is for educational purposes only and is not financial advice. Individual results vary.