At the end of the day, a derivatives exchange is just one big pot of money being shuffled around. Every long is matched by a short, every dollar of profit on one side is a dollar of loss on the other, and the total PnL of the system is always zero before fees. There is no magical external source of money: everything that winners “make” is coming from losers in that same pot. Auto-Deleveraging (ADL) is basically the mechanism that kicks in when that entire pot has accrued to the winners and the losers have all been liquidated. It’s how the exchange enforces the rule: “This pot must always balance,” even in extreme conditions.
ADL Mechanics
Normally, things work pretty cleanly. Traders take leverage, price moves, and the exchange’s risk engine liquidates people before they lose more than their margin. If liquidations slip a bit, there is typically an insurance fund that steps in and plugs the gap. As long as losers can be closed in time and the insurance fund is sufficient, winners get fully paid, the pot stays balanced, and no one talks about ADL at all.
ADL only becomes relevant when something breaks that flow: the market gaps, liquidity evaporates, a chunk of shorts or longs blow past their margin, the insurance fund is exhausted, and now there are traders showing big unrealized profits against counterparties who literally can’t pay them. The system is still flat in notional terms, but it’s no longer flat in credit terms: some of those “profits” aren’t actually collectible. That’s where ADL comes in.
Think of ADL as a targeted way to resettle the pot so that it makes sense again. If a bunch of over-leveraged shorts go bankrupt in a violent move up, then by definition there are longs on the other side who are showing profits that exceed what those shorts (plus the insurance fund) can cover.
The exchange cannot conjure additional capital from nowhere; it can only rearrange who has claims on the pot. So it does the only solvency-preserving thing available: it force-closes or haircuts some of the winning longs, at a fair reference price, and uses the “clawed back” PnL to plug the hole left by the bankrupt shorts. PnL is conserved at the system level: losers are fully tapped out, the insurance fund is fully tapped out, and the remaining shortfall is absorbed by reducing the profits of the traders who were directly benefiting from those failed accounts.
From a pure solvency perspective, ADL is a good thing. Without it, the exchange is faced with either going insolvent (not paying out winners in full) or implementing a blunt, retroactive socialized loss across a much wider set of users. ADL is the mechanism that lets the exchange say: “We will always be solvent; we will always make the pot add up; and if that means some of your profits get reduced because they were coming from people who can’t pay, we will do that in a systematic way.” It’s the last line of defense that ensures there is never a state where the exchange owes more than the combined capital in margin, insurance, and the positions of other traders.
ADL Fairness
Under reasonable design, ADL is also mostly fair in a specific sense: if you don’t have exposure in the product that blew up, your PnL shouldn’t be touched. That’s the key invariance principle: no PnL adjustments for accounts that are flat in the affected instrument. If a BTC perp book explodes because a bunch of shorts go bankrupt, the exchange shouldn’t haircut users who are only trading ETH, or who closed their BTC positions hours ago, or who are just sitting in cash.
ADL is about localizing the damage to the part of the pot where the credit failure occurred: you look at the instrument and the side where people are showing profits against bankrupt counterparties, and you adjust there. That’s already much more targeted than a global “everyone who made money today loses 20%” kind of approach.
Where ADL sometimes earns its reputation as “unfair” is at the portfolio level, especially for hedged or cross-venue setups. Imagine you’re long BTC-PERP on one exchange and short BTC spot or another perp elsewhere as a hedge. In your mental model, you’re roughly market-neutral: if BTC goes up, your long makes money and your short loses; if BTC goes down, vice versa. Now imagine a huge squeeze wipes out a wall of over-leveraged shorts on your long exchange. You’re one of the most profitable longs there, with high leverage and big unrealized gains.
The system, trying to restore solvency, ADL’s you: it forcibly closes your long at or around the bankruptcy/mark price of the losing shorts. That may even be a better exit price than you’d get via the order book in a stressed moment. But your hedge leg on the other venue is still open. Suddenly you’re just short BTC somewhere else, right as the market is ripping, and you may have to chase a painful re-hedge. From the exchange’s perspective, this was fair and necessary, as your long was directly benefiting from people who couldn’t pay. From your perspective, your carefully constructed hedge has been broken by an event you didn’t control.
The same pattern shows up in cross-margin. Suppose you’re on a venue where all positions share one collateral pool: you’re long BTC perp, short ETH perp, and holding spot BTC as collateral. The BTC short side blows up, and you’re a profitable BTC long. ADL kicks in and closes all or part of your BTC perp. Your ETH perp and spot BTC remain untouched, because they weren’t directly involved in the credit failure. On paper, this is consistent with the “only touch the affected instrument” rule.
In practice, your overall portfolio risk has changed drastically. Maybe BTC and ETH were structured as a spread; now one leg is gone, the other is still there, and your margin usage and hedge ratio are all different. Again, this feels unfair at the level of your intent and strategy, even though the exchange was acting fairly within its own accounting framework.
Nonetheless, if you like trading on a venue that always pays what it promises, even in insane conditions, you implicitly like something like ADL. You may not like it when it hits you, but you very much like the fact that when other people blow up, your balances are on a platform that remains solvent. That’s the trade: one shared pot of money, strict conservation of PnL, and an explicit, mechanical way of rebalancing claims on that pot when the losers can’t make good.
Disclaimer
The information provided on TheLogbook (the “Substack”) is strictly for informational and educational purposes only and should not be considered as investment or financial advice. The author is not a licensed financial advisor or tax professional and is not offering any professional services through this Substack. Investing in financial markets involves substantial risk, including possible loss of principal. Past performance is not indicative of future results. The author makes no representations or warranties about the completeness, accuracy, reliability, suitability, or availability of the information provided.
This Substack may contain links to external websites not affiliated with the author, and the accuracy of information on these sites is not guaranteed. Nothing contained in this Substack constitutes a solicitation, recommendation, endorsement, or offer to buy or sell any securities or other financial instruments. Always seek the advice of a qualified financial advisor before making any investment decisions.
