This article is for educational purposes only. Leverage.Trading is an independent educational and analytics publisher and not a broker, exchange, or investment advisor. Trading with leverage, margin, futures, or derivatives carries a high risk of rapid or total loss. This content is not financial advice and should not be used as a substitute for independent research or professional advice.
Look at the steepest red wicks on any Bitcoin futures chart, and most of the displacement on the way down came from forced closes, not from anyone deciding to sell. The exchange’s risk engine fires market orders into a thin book, knocks one cluster of leveraged positions into the next, and the loop runs by itself until something absorbs the flow. That loop has a name: a liquidation cascade.
Liquidation cascades can move crypto prices 10% or more in under 30 minutes. Leveraged positions caught in a cascade close automatically. Once the mark price hits the liquidation level, there is no manual window to exit. The position is gone before the exchange interface updates.
A liquidation cascade (also called a cascading liquidation) is a chain reaction. The exchange’s risk engine closes one position through forced liquidation, that close moves the price, the new price hits another position’s liquidation level, and the chain repeats. In a market with few leveraged positions, the chain dies fast or never starts.
In crypto perpetual futures, major exchanges offer extreme leverage on common contracts. Bybit allows up to 100x leverage on its BTCUSDT perpetual, and Binance offers up to 125x. At those levels, leveraged positions stack thickly at common price levels.
When the first wave of liquidations hits a dense cluster, the next wave is sitting right behind it.
Cascades usually start with a move that has nothing to do with cascades. Bitcoin drops 2% on a macro headline. An ETF outflow shows up in the order book.
Funding flips negative as a crowded long position starts unwinding. Any of these can push price into a zone where a cluster of traders has positioned at similar levels. That cluster is the first wave.
Once price reaches the maintenance margin line for the most aggressive positions in the group, the exchange’s risk engine begins closing them. The closes hit the order book as market orders, pulling the bid side down for longs or pushing the ask side up for shorts. Price moves further in the same direction, and the next layer of positions, sitting just past the first at slightly lower leverage, now has its liquidation price hit too.
A liquidation is the forced closure of a leveraged position by the exchange. That close happens through the order book as a market order.
In an order book, standing limit orders create the visible price levels. Market orders consume those limit orders, and the price shifts with each fill.
When the new price reaches the next set of liquidation levels, those positions get closed too. Their own market orders feed the same loop. Nothing in this requires anyone to be selling or buying based on conviction.
What most traders think: Liquidation cascades happen because panicked traders are selling to exit.
What actually happens: The exchange’s risk engine generated every sell order automatically. The position closes before any trader can react.
The cascade is the exchange clearing risk by hitting the book. The book moves each time it gets hit.
The price visible on the chart is the last traded price. Liquidation does not fire on this number.
Exchanges use mark price, a composite calculated from spot exchange prices across multiple venues, to determine when a position crosses its liquidation threshold. Binance’s official liquidation documentation states that “Liquidation occurs when the Mark Price hits the liquidation price of a position.” The same principle applies across major perpetual exchanges.
During a fast cascade, mark price and last traded price can diverge. A position can be liquidated even if the visible chart candle appears never to have touched the liquidation level. The mark price is designed to be resistant to manipulation from a single exchange’s order book.
Two things decide how fast a cascade moves once it starts: leverage level and position density.
Higher leverage means each position has less margin to absorb a price move before reaching its liquidation level. When the average position in a cluster runs at 25x or 50x leverage, the gap between entry and liquidation is small. The chain moves through it quickly.
The second factor is how many positions share the same entry price range. If traders entered the same setup at similar prices, their liquidation levels also sit at similar prices. Hitting one cluster means hitting most of them in close succession.
A 2% adverse price move affects a long position differently at each leverage level. The numbers show why cascade pressure accelerates sharply above 25x.
| Leverage | Margin Required | 2% Adverse Move | Margin Remaining | Cascade Pressure |
|---|---|---|---|---|
| 25x | 4% of position | Wipes 50% of margin | 50% buffer remains | Low — cascade slows here |
| 50x | 2% of position | Wipes 100% of margin | Gone — position liquidated | High — cascade accelerates |
| 100x | 1% of position | Exceeds full margin | Gone before 2% completes | Extreme — chain fires in seconds |
These figures use a simplified illustration of initial margin and price movement. Actual liquidation price depends on entry, fee structure, and whether cross or isolated margin is used. The exact liquidation level for any entry and leverage combination can be calculated using the Liquidation Price Calculator.
At 50x leverage, a 2% price move against a long position eliminates the entire margin. At 100x, the position closes before the price completes a standard daily move. Cascades are not slow events. They clear multiple leverage tiers in seconds, with no recovery window between closures.
Above any individual position, three market conditions decide whether a cluster of leveraged positions can become a cascade.
Open interest is the total leveraged capital sitting on a contract at any given moment. When it grows quickly in one direction, the build-up is one-sided.
The liquidation levels of those new positions, often chasing the trend with high leverage, stack on the same side of the current price. Bigger open interest means more fuel in the cluster.
Liquidity depth determines how much damage each forced close does. A book with thick limit orders near the current price soaks up forced selling, and each market order moves the price less.
A thin book amplifies every market order. The same number of liquidations on different days can produce vastly different price impacts depending on what the book looks like at the moment the cascade starts.
Higher volatility widens the range where positions get hit and shrinks the room between an entry price and a forced close. In a quiet market, a cluster might never be reached. In a volatile session, the same cluster can be reached three times in a week.
The three conditions compound each other: open interest builds the fuel, a thin book turns small liquidations into large price moves, and volatility brings the trigger closer.
A single position closing is a small event. Several hundred thousand positions closing in minutes is a market-wide move.
The bridge between the individual and the systemic is positioning. When funding rates stay positive and elevated for weeks, longs are crowded. When total open interest hits records during a one-way move, the build-up has nowhere to go but a forced unwind.
The trigger is the spark. The cluster is the fuel.
When those conditions converge, the same trigger that would move price 1% on a balanced day can move it 10% on a stretched one.
Historical cascade events have generated nine- and ten-figure liquidation volumes within minutes of the trigger. The conditions that produce them, sustained elevated funding rates, record open interest on a one-directional move, and thin order books, are all visible in real time. Visibility does not make the trigger predictable. It makes the fuel level quantifiable.
A cascade does not run forever.
The most common stopping point is exhaustion. Every cluster of liquidation levels is finite. Once the engine has cleared the densest part, the next layer of positions sits further away in price, and the forced orders no longer reach them.
Absorption brings a different kind of stop. At some point, price drops to a level where buyers, typically spot traders or counter-trend traders, see value and start absorbing the orders. The book deepens on the other side, and the same dollar volume of forced selling stops moving the price as far.
Rebalancing is the third mechanism. Once the dominant positioning has been forcibly cleared, the imbalance that fueled the cascade is gone. New positions come in more cautiously, and the next significant move requires a fresh build-up to start.
When liquidations happen faster than the order book can absorb, the exchange’s insurance fund covers the gap between the liquidation price and the actual market execution price. In extreme events where the insurance fund is depleted, auto-deleveraging (ADL) takes over, closing profitable positions on the opposite side of the market to cover the shortfall.
Leverage concentration, open interest, and order book depth determine how severe a cascade becomes. All three are observable before the trigger arrives.
The trigger itself is not predictable. The fuel level is.
For a position caught in a cascade, the only variable that mattered was the buffer set before the event started. Once the mark price reaches the liquidation level, the position closes automatically. The cascade makes no distinction between bad timing and bad sizing.
Any price move that pushes the mark price into a dense cluster of leveraged liquidation levels. Common triggers include macro headlines, ETF outflows, and large funding swings. The trigger is secondary to the setup. Pre-existing leverage density, crowded open interest, and thin order books determine how far the chain runs.
Most cascade events run from minutes to under one hour. The chain stops when the densest liquidation clusters are exhausted, when buyers absorb the flow, or when the positioning imbalance is cleared.
The fuel level is visible. The trigger timing is not. Sustained elevated funding rates, record-high open interest on a one-directional move, and recent dips that disproportionately cleared one side are all readable before a cascade fires. Coinglass publishes a real-time liquidation heatmap showing where leveraged position clusters are concentrated.
If the mark price reaches the liquidation price, the exchange closes the position automatically at market. In fast cascade conditions, execution can occur at a price worse than the listed liquidation level. No manual exit is possible after the level is hit.
Related but distinct. A short squeeze is buyer-driven. Rising prices force short positions to cover, which pushes prices further. A liquidation cascade is engine-driven. The exchange’s risk system generates all the market orders automatically. A cascade can run in either direction. A squeeze is directionally specific.
A liquidation cascade is a mechanical event. The selling that drives price down during a cascade comes from the exchange’s risk engine, not from traders deciding to exit. Understanding this changes how the event reads on a chart and how the setup reads beforehand.
Leverage concentration, open interest, and order book depth determine how severe a cascade becomes. All three are observable before the trigger arrives.
The trigger itself is not predictable. The fuel level is.
For a position caught in a cascade, the only variable that mattered was the buffer set before the event started. Once the mark price reaches the liquidation level, the position closes automatically. The cascade makes no distinction between bad timing and bad sizing.
This article is published under Leverage.Trading’s leverage trading & crypto derivatives education , an independent risk-first learning system built to help traders quantify and manage risk before trading.
