Glossary

Auto-Deleveraging (ADL)

Auto-deleveraging forcibly reduces profitable traders' positions to cover losses from failed liquidations on derivatives exchanges.

Key Takeaways

  • Auto-deleveraging (ADL) is a last-resort mechanism on crypto derivatives exchanges that forcibly closes profitable traders' positions to cover losses when a liquidation fails and the insurance fund runs out.
  • Exchanges rank traders for ADL by combining unrealized profit percentage with effective leverage: the most profitable, highest-leveraged positions are closed first. A five-bar indicator shows each trader's position in the queue.
  • ADL is controversial because it closes winning positions without consent, can break hedging strategies, and concentrates losses on a few traders rather than distributing them across the platform.

What Is Auto-Deleveraging?

Auto-deleveraging (ADL) is a risk management mechanism used by cryptocurrency derivatives exchanges to maintain platform solvency. When a losing trader's position is liquidated but the exchange cannot close it at or better than the bankruptcy price, the resulting shortfall becomes "bad debt." If the exchange's insurance fund cannot absorb this bad debt, ADL automatically closes the most profitable opposing positions to cover the gap.

The concept originated in traditional futures markets but became standard practice in crypto after BitMEX adopted it for perpetual futures in 2016, replacing the older socialized loss model. Today, virtually every major derivatives exchange implements ADL: Binance, Bybit, OKX, BitMEX, dYdX, and decentralized platforms like Hyperliquid all use some variant.

For the affected trader, ADL means having a profitable position forcibly reduced or closed at the bankrupt trader's bankruptcy price rather than the current market price. The trader receives no warning beyond their ADL queue indicator and has no ability to opt out.

How It Works

ADL sits at the end of a three-tier loss absorption waterfall. Understanding this sequence is essential:

  1. Market liquidation: the exchange's liquidation engine attempts to close the losing position via the order book. If the position closes at a price better than the bankruptcy price, the surplus flows into the insurance fund.
  2. Insurance fund: if the order book cannot absorb the position at the bankruptcy price, the insurance fund covers the shortfall. This fund accumulates from liquidation surpluses, exchange contributions, and portions of trading fees.
  3. Auto-deleveraging: only triggered when the insurance fund is depleted or drops below critical thresholds. The exchange force-closes profitable counterparty positions to eliminate the remaining bad debt.

The Bankruptcy Price Gap

The gap between the liquidation price and bankruptcy price is central to understanding when ADL triggers. The liquidation price is where the exchange begins closing a position (when margin falls below the maintenance requirement). The bankruptcy price is where the trader's losses exactly equal their deposited collateral:

Bankruptcy price (long)  = Entry Price × (1 - 1/Leverage)
Bankruptcy price (short) = Entry Price × (1 + 1/Leverage)

Example: Long BTC at $67,500, 10x leverage
  Liquidation price ≈ $61,088  (with 0.5% maintenance margin)
  Bankruptcy price  = $60,750  (margin fully consumed)
  Gap               = $338     (maintenance margin cushion)

If the market crashes through both the liquidation price and the bankruptcy price before the exchange can close the position, the resulting loss exceeds the trader's collateral. That excess is bad debt, and if the insurance fund cannot cover it, ADL activates.

ADL Ranking Formula

When ADL triggers, the exchange must decide which opposing positions to close. Every major exchange ranks traders by a formula that combines profit and leverage. Binance's canonical formula:

PNL% = Unrealized Profit / abs(Position Notional)
Effective Leverage = abs(Position Notional) / (Wallet Balance + Unrealized Profit)

If PNL% >= 0:  ranking = PNL% × Effective Leverage
If PNL% < 0:   ranking = PNL% / Effective Leverage

Positions with both higher profit percentages and higher effective leverage rank first. This means a trader using 20x leverage with a 50% gain is deleveraged before a trader using 2x leverage with the same 50% gain. The rationale: highly leveraged profitable positions contribute more to systemic risk and benefit most from the exchange's risk infrastructure.

Variations exist across exchanges. OKX ranks hedged positions lower in the ADL queue, recognizing that closing one leg of a hedge creates unintended exposure. Bybit uses slightly different formulas for isolated margin versus cross-margin accounts.

The Five-Bar Indicator

Exchanges display a five-level indicator showing each trader's position in the ADL queue:

Bars LitQueue PositionADL Risk
5 barsTop 20%Highest priority: will be deleveraged first
4 bars20-40%High risk
3 bars40-60%Moderate risk
2 bars60-80%Low risk
1 barBottom 80%Minimal risk

Traders can monitor this indicator in their positions tab. When all five bars are lit, a large liquidation cascade could trigger ADL on that position. No trading fees are charged on ADL closures.

ADL vs. Socialized Loss

Before ADL became the industry standard, exchanges used socialized loss systems that spread shortfalls proportionally across all profitable traders. ADL replaced this approach because socialized loss created perverse incentives: traders taking extreme leverage knew their potential losses would be distributed across everyone, encouraging reckless behavior.

AspectSocialized LossAuto-Deleveraging
Loss distributionSpread across all profitable tradersTargeted at top-ranked traders by profit and leverage
Traders affectedEveryone on the winning sideA few traders at the top of the queue
Impact per traderSmall proportional haircutFull position closure for those selected
Moral hazardHigher: riskiest traders externalize costsLower: high-leverage positions rank first
PredictabilityProportional and predictableDepends on queue position

Why It Matters

ADL exists because crypto exchanges operate as both the exchange and the clearinghouse, unlike traditional markets where separate clearinghouses manage counterparty risk. Without ADL, a single failed liquidation during a market crash could leave an exchange insolvent, putting all users' funds at risk.

For traders using margin trading or perpetual futures, understanding ADL is critical. A winning trade can be forcibly closed at an unfavorable price, locking in less profit than expected or disrupting a hedging strategy. Traders with high leverage and large unrealized profits are most exposed.

The mechanism also highlights a fundamental difference between centralized derivatives platforms and self-custodial systems. In decentralized spot markets or Layer 2 networks like Spark, users hold their own assets and are not subject to exchange-level risk socialization. ADL is a byproduct of counterparty credit risk in leveraged trading, a risk that does not exist in simple asset transfers or self-custodial payments.

Risks and Considerations

The most fundamental criticism of ADL: profitable traders have their positions closed without agreement. Unlike a stop-loss or take-profit order that a trader sets voluntarily, ADL is imposed by the exchange. The affected trader's position is closed at the bankrupt trader's bankruptcy price, which may be significantly worse than the current mark price.

Hedging Strategy Disruption

ADL does not recognize portfolio context. A trader running a delta-neutral strategy with a profitable long and an offsetting short may have only the profitable leg closed, leaving them with unintended directional exposure. During the October 2025 liquidation cascade, this was a major complaint from institutional traders. OKX is currently the only major exchange that accounts for hedging in its ADL ranking, placing hedged positions lower in the queue.

Cascade Risk

During extreme volatility, ADL events can cascade. A liquidation cascade depletes the insurance fund, triggering ADL. The forced closures from ADL can move the market further, triggering more liquidations and potentially more ADL events. In October 2025, over $19 billion in futures positions were liquidated within 24 hours, with ADL triggered across multiple exchanges.

The Fairness Trilemma

Academic research from Columbia University demonstrated in 2025 that no ADL policy can simultaneously guarantee exchange solvency, preserve exchange revenue, and treat traders fairly. This is a formal impossibility result. The same research showed that the queue-based ADL system used by most exchanges concentrates losses on top winners and is suboptimal compared to pro-rata approaches that distribute haircuts proportionally.

Reducing ADL Exposure

Traders can take steps to reduce their ADL risk:

  • Reduce effective leverage by adding margin to profitable positions, lowering the profit-times-leverage ranking score
  • Take partial profits to reduce unrealized PnL, which directly lowers ADL queue priority
  • Monitor the five-bar ADL indicator and adjust positions when approaching the top of the queue
  • Trade on contracts with ADL guarantees: Binance guarantees no ADL on select BTC, ETH, and BNB futures contracts when open interest stays below specific thresholds
  • Use exchanges that account for hedging in their ADL ranking if running hedged strategies

For a deeper look at how risk management differs across traditional and decentralized systems, see the research article on TradFi-DeFi convergence.

This glossary entry is for informational purposes only and does not constitute financial or investment advice. Always do your own research before using any protocol or technology.