Crypto Liquidations: How Cascades Start and How to Track Them
A liquidation is the forced closure of a leveraged position whose margin can no longer cover its losses. The exchange's risk engine takes over the position and closes it into the open market — the trader has no say in the timing or the price. In crypto, where 20x to 100x leverage is routine and markets trade around the clock, liquidations are not an edge case: on volatile days, more than a billion dollars of positions can be force-closed within hours.
Why cascades feed on themselves
Liquidations are self-reinforcing. A forced sale of a long position pushes price down, which pushes the next tier of leveraged longs below their maintenance margin, triggering more forced selling. This chain reaction is a liquidation cascade, and it explains crypto's signature move: hours of calm followed by a violent wick that liquidates both directions.
The structure of the cascade is knowable in advance, at least approximately. Because most traders enter at round-number prices with standard leverage tiers, their liquidation prices cluster at predictable distances from entry. Mapping those clusters produces a liquidation heatmap: dense zones act like magnets, because market makers know that pushing price into a cluster releases a burst of forced orders to trade against.
Using the data in practice
Three views cover most use cases. The 24h liquidation totals by coin and side tell you what already happened — a large long-liquidation skew after a dump means leverage has been flushed, which historically improves the odds of stabilization. The liquidation levels map tells you what could happen next: large clusters just below spot are downside fuel, clusters above are squeeze fuel. Max-pain levels compress that map into single price points where the most notional would be destroyed.
The main trap is treating estimates as ground truth. Exchanges do not publish every position, so heatmaps are model-based reconstructions from open interest and volume profiles. Use them for ranges and asymmetry — where the bigger cluster sits — not for precise trigger prices.