A liquidation heatmap shows price levels where clusters of leveraged positions would be forcibly closed. It is one of the most-shared charts in crypto and one of the least-understood, because of how it is built.
It is modelled, not observed
Venues do not publish where each account’s liquidation sits. The map is reconstructed: take open interest, assume a distribution of leverage and entry prices, and compute where those hypothetical positions would be closed.
The assumption is the whole product. Change the assumed leverage mix and the clusters move. Two vendors can publish materially different maps of the same market, both internally consistent, because they assumed differently.
What it does capture
Something real: leverage is not evenly distributed. Round numbers, obvious highs and lows and prior ranges attract entries, and entries clustered at a level produce liquidations clustered somewhere below or above it.
So the map is a reasonable picture of where forced flow would be concentrated if price got there — and forced flow is the mechanism that turns an ordinary move into a cascade. Read as structure, it earns its place.
The two misreadings
As a magnet. “Price will go there to take the liquidity” treats a statistical estimate as an intention. Sometimes price reaches a cluster and the cascade follows; often it does not go near it for weeks. Nobody is steering.
As precision. Reading a cluster as a level, and placing a stop just beyond it, applies a decimal point to something built on an assumed leverage distribution. The honest resolution is a zone, and a wide one.
How to use it
As a companion to funding rather than alone. Funding tells you which side is crowded; the heatmap suggests where that crowd would be forced out. Together they describe the fuel and roughly where it is stacked — still with nothing to say about the spark.