When a Leveraged Trader Gets Liquidated Underwater, Someone Still Has to Eat the Loss. That's What an Exchange's Insurance Fund Is For
6 min read
Loading...
6 min read
On 10 October 2025, more than $19 billion in leveraged crypto positions got liquidated in about a day, the largest single-day deleveraging crypto markets had ever seen. When a losing position gets liquidated at a worse price than the trader's remaining margin can cover, someone still has to absorb that shortfall. An exchange's insurance fund exists to cover exactly that gap. When even the fund isn't enough, a separate mechanism called auto-deleveraging (ADL) takes over, and it can close out a profitable trader's position without asking.
When you trade with leverage, you post margin, a fraction of the position's full value, as collateral. If the market moves against you enough, the exchange liquidates your position to stop your losses from exceeding what you put up.
In fast, thin markets, a liquidation doesn't always fill cleanly. By the time the exchange's system closes your position, the price may have already moved further against it than your remaining margin can absorb. That leftover loss doesn't just vanish. Someone on the other side of your trade was expecting to get paid.
This is what an insurance fund is for. It's a pool of money the exchange sets aside, usually funded by a cut of liquidation fees over time, specifically to cover the difference between what a liquidated trader's margin actually pays out and what their counterparty is owed. As long as the fund has enough in it, profitable traders on the other side of a liquidation get paid in full and never notice anything happened.
A big enough crash can liquidate thousands of leveraged positions within minutes, all at once, in a market where there isn't enough liquidity to fill them at a fair price. If the shortfall from those liquidations is larger than what the insurance fund holds, the fund can be drawn down fast, and in an extreme case, exhausted.
When that happens, most major derivatives exchanges fall back on auto-deleveraging, usually shortened to ADL. Instead of leaving the shortfall unpaid, the exchange forcibly closes out some of its most profitable, most highly leveraged traders' opposite positions, using their unrealized profit to cover what the bankrupt trader and the insurance fund together could not.
ADL doesn't pick traders at random. Most exchanges rank open positions by a combination of unrealized profit and effective leverage, then show each trader an indicator, often a row of lights or bars, signaling how close they sit to the front of that queue. The more segments lit up, the sooner that position gets closed if ADL has to fire. A trader running high leverage while sitting on a large paper profit is exactly who the mechanism is built to reach for first.
The difference from ordinary liquidation matters. In a normal liquidation, you lose your posted margin. Under ADL, you keep whatever profit you'd already made, but your winning position gets closed for you, at a moment you didn't choose, with no say in the matter.
Say a trader is short Bitcoin with 20x leverage and has a large unrealized profit as the price falls sharply. At the same time, a wave of long positions gets liquidated so fast that the exchange can't fill them at a price that covers what their counterparties are owed, and the insurance fund is already thin from absorbing earlier liquidations that same day. To close the gap, the exchange's ADL system reaches for the short trader's profitable position first, since it ranks near the top of the queue on both profit and leverage, and closes part or all of it at the current market price. The trader keeps the profit already earned. They just don't get to decide when the position ends.
This example is illustrative only, built to show the mechanism, not a real trade.
The mechanism above isn't theoretical. It ran, live, at scale, on 10 October 2025, when a sudden risk-off shock (triggered by a tariff announcement) sent Bitcoin down roughly 14% in hours and dragged the wider crypto market with it. More than $19 billion in leveraged positions were liquidated across exchanges in around a day, according to CoinGecko's own writeup of the event, with roughly 1.6 million trader accounts affected, corroborated by separate reporting from Forbes and FTI Consulting.
Binance's own published data shows the scale of what its insurance fund absorbed that day: the shared insurance fund balance covering BTC, ETH and BNB USDT-margined futures contracts dropped from about $1.23 billion to about $1.04 billion, with roughly $188 million deployed to cover shortfalls during the worst of the volatility, according to Binance Futures data reported by the analyst account Wu Blockchain and corroborated by separate coverage from Yellow.com and HokaNews. Auto-deleveraging activated that day not just on Binance but also on other major derivatives platforms, including OKX, Bybit and Hyperliquid, based on subsequent industry reporting on the event.
No single exchange's insurance fund, no matter how large, is sized for every trader's liquidation hitting it in the same few minutes. That is precisely the scenario ADL exists for.
It's easy to confuse an exchange's futures insurance fund with a separate, differently-purposed reserve some exchanges also keep. Binance, for example, also runs the Secure Asset Fund for Users (SAFU), a fund set aside since 2018 specifically to reimburse users after a hack or a security breach on the exchange itself, not to cover leverage-trading shortfalls. The two funds solve different problems: the futures insurance fund exists because of how leveraged trading and liquidation math work, while SAFU exists because of exchange security risk. A large SAFU balance says nothing about how well-funded an exchange's separate futures insurance fund is, and vice versa.
None of this is a reason to avoid leverage entirely, but it is a reason to understand what actually backstops your position when markets move fast:
If you hold leveraged positions or are weighing one, checking an asset's own volatility and drawdown history before you size a trade gives you a much better sense of how fast a move like this can happen than any single headline number does. The Explore screener lets you compare that volatility across assets directly, before leverage turns a normal pullback into a liquidation.