Sharp moves in crypto frequently accelerate at particular levels and then stop abruptly. A large part of that pattern comes from forced closures of leveraged positions bunching together.

Liquidation is automatic and price-triggered

A leveraged position is backed by collateral, and the venue closes it when the collateral no longer covers the potential loss by a required margin.

The trigger level is determined arithmetically from entry price, position size and collateral, so it is known in advance for every position.

Closing is executed as a market order against the book, which means it consumes liquidity and moves price in the same direction as the loss.

Common choices produce common levels

Traders select from a small menu of leverage settings, and platforms offer a limited set of round multiples.

Entries also concentrate, since many positions are opened around the same events, prices and technical levels.

Identical inputs produce identical trigger prices, so positions opened by different people at similar times share a liquidation level without any coordination.

Each wave supplies the push for the next

When price reaches a cluster, those positions are closed at once, and the resulting orders push price further into the next cluster below.

Depth is usually thinnest during exactly these moments, since market makers widen quotes when volatility spikes, so each wave moves price more than its size alone would suggest.

The cascade ends when it runs out of clustered positions or reaches a level where resting bids are deep enough to absorb the flow.

Insurance funds absorb the shortfall

If a position is closed at a worse price than its bankruptcy level, the loss exceeds the collateral posted.

Venues maintain an insurance fund from liquidation surpluses to cover these cases, and its size is a reasonable indicator of how well the venue has handled past stress.

When the fund is exhausted, the loss is spread across profitable traders on the other side, an outcome that is unpopular and explicitly documented by platforms that use it.

Heatmaps estimate rather than observe

Charts showing where liquidations sit are constructed from open interest and assumed leverage, not from a published record of positions.

Actual collateral, cross-margin arrangements and added funds are invisible to those models, so the estimates are approximate.

They still capture the structural point, which is that leverage concentrates at predictable levels and creates moves out of proportion to the flow that started them.