Liquidation in the cryptocurrency and finanza decentralizzata ecosystem refers to the forced closure of a leveraged trading position or a collateralized lending position when the value of the underlying collateral falls below a protocol-defined maintenance margin or minimum collateralization ratio. The process is designed to protect lenders, liquidity providers, and exchanges from absorbing losses caused by a borrower’s or trader’s underwater position.
In centralized exchange (CEX) trading, liquidation occurs when a trader using margin or futures leverage cannot maintain the required margin to keep a position open. When the mark price of the asset hits the trader’s liquidation price, the exchange’s matching engine automatically closes the position at market price. If the position’s remaining collateral is insufficient to cover the loss, the exchange’s insurance fund absorbs the deficit. On platforms like Binance, Bybit, and OKX, liquidation engines process thousands of forced closures during volatile market swings, sometimes triggering cascading liquidations that amplify price movements.
In decentralized finance (DeFi), liquidation operates through smart contract logic and third-party liquidator bots known as keepers. Lending protocols such as Aave, Compound, and MakerDAO define a health factor or collateralization ratio for each loan. When the ratio drops below a critical threshold, typically due to a decline in collateral value or an increase in borrowed asset value, the smart contract opens the position to liquidation by external actors. Liquidators repay a portion or all of the borrower’s debt and receive the borrower’s collateral at a discount (the liquidation bonus), creating an economic incentive for rapid position resolution.
Liquidation mechanisms are essential to the solvency of both centralized and decentralized lending markets. Without them, a cascade of bad debt could drain insurance funds, deplete lending pools, and ultimately threaten the stability of entire platforms. The design of liquidation parameters, including the liquidation threshold, liquidation penalty, close factor, and oracle price feed methodology, is one of the most critical governance decisions in any DeFi protocol.
Origine e storia
2013: Early Bitcoin exchanges like BitMEX’s predecessor platforms began offering leveraged trading products, introducing centralized liquidation engines to the crypto market for the first time. Margin calls and forced closures were modeled after traditional futures exchanges like the CME.
2014: BitMEX launched as one of the first major crypto derivatives platforms, founded by Arthur Hayes, Ben Delo, and Samuel Reed. Its socialized-loss insurance fund mechanism became the template for most subsequent crypto derivatives exchanges. Note: BitMEX’s defining product, the perpetual swap (XBTUSD), was invented and launched in May 2016, not at founding, and grew to become the most traded crypto derivative in history.
2017-2018: MakerDAO deployed the first version of its CDP system. Separately, ETHLend (founded by Stani Kulechov in 2017) conducted its ICO, eventually rebranding to Aave in September 2018 to pivot from peer-to-peer lending to the more efficient liquidity pool model.
2018: Compound Finance launched its lending protocol with algorithmic interest rates and a novel liquidation model where any Ethereum address could act as a liquidator, democratizing the process beyond centralized entities.
Gennaio 2020: Aave deployed its V1 mainnet on Ethereum on January 8, 2020, shifting from the P2P ETHLend model to pooled liquidity markets and introducing flash loans as a new DeFi primitive. In December 2020, Aave V2 launched, introducing significant improvements to capital efficiency, risk management, and the user experience of flash loans.
2020, Black Thursday (March 12): The most consequential liquidation event in DeFi history. ETH crashed approximately 43% in a single day. On MakerDAO, congestion on Ethereum caused oracle price feeds to delay and keeper bots to fail. Some liquidators won collateral auctions with bids of 0 DAI, extracting $8.3 million in ETH for free. Over $1 billion in leveraged positions were liquidated across all platforms within 24 hours.
Maggio 2021: Over $8 billion in crypto derivatives were liquidated in 24 hours, one of the largest single-day liquidation events in centralized exchange history, with Binance alone processing over $2 billion in forced closures.
March 2022: Aave V3 launched across six networks including Polygon, Arbitrum, Avalanche, Fantom, Harmony, and Optimism, introducing Portal cross-chain functionality, Efficiency Mode (E-Mode) for correlated assets, and isolation mode with granular liquidation parameters to further mitigate systemic risk.
2023-2024: Advanced liquidation protocols emerged, including Morpho and Euler Finance (which relaunched as Euler V2 in 2024 after a $197 million flash loan exploit in March 2023 from which it recovered nearly all funds), implementing more capital-efficient liquidation models with smaller penalties and partial liquidation mechanisms.
2025, Black Friday (October 9): The largest liquidation event in crypto history. Triggered by geopolitical tensions, losses mounted to $19.35 billion within 24 hours, wiping out approximately 1.7 million trading accounts.
“On March 12th, we learned that DeFi liquidation infrastructure wasn’t just a technical feature. It was the immune system of the entire ecosystem. When it failed, everything was at risk.”
In parole semplici
Imagine you borrow money from a pawnshop and leave your gold watch as collateral. If gold prices drop and your watch is suddenly worth less than what you owe, the pawnshop sells your watch immediately to recover their money. That forced sale is liquidation, the lender protecting themselves by seizing and selling your collateral before the debt becomes unrecoverable.
Think of it like a margin call from your stockbroker. You bought $10,000 worth of stock using $2,000 of your own money and $8,000 borrowed from the broker. If the stock drops enough that your $2,000 equity is nearly wiped out, the broker doesn’t wait for you to add more money, they sell the stock automatically to get their $8,000 back. In crypto, this happens algorithmically in milliseconds rather than through a phone call.
Consider a car loan where the bank requires detailed insurance. If the car’s value drops below the loan balance (you’re “underwater”) and you stop making payments, the bank repossesses the car. In DeFi, the “bank” is a smart contract, the “car” is your crypto collateral, and the “repossession” happens automatically when prices drop. No lawyers, no paperwork, no delays.
Picture a game of Jenga where each block represents a leveraged position in the market. When one block is removed (a large position gets liquidated), it destabilizes the tower, causing more blocks to fall (cascading liquidations), which can topple the entire structure (market crash). This chain reaction is what makes liquidation events so dramatic in crypto markets.
Importante: Liquidation does not always mean total loss. In many DeFi protocols, only a portion of your position is liquidated (partial liquidation), and you may retain the remaining collateral minus the liquidation penalty. However, on centralized exchanges with high leverage (50x-125x), liquidation typically results in the complete loss of your margin deposit.
Principali caratteristiche tecniche
Health Factor and Collateralization Ratio
DeFi protocols assign each position a health factor: (Collateral Value x Liquidation Threshold) / Total Debt
A health factor below 1.0 means the position is eligible for liquidation
MakerDAO uses a collateralization ratio (e.g., 150% for ETH-A vaults): if the ratio drops below the threshold, the vault is liquidated
Aave V3 uses per-asset liquidation thresholds ranging from 50% to 93% depending on the asset’s risk profile
Compound uses a similar model with a “close factor” determining what percentage of a position can be liquidated in a single transaction (typically 50%)
Liquidation Mechanisms in DeFi
Fixed-Spread Liquidation (Aave, Compound)
Oracle price feeds detect that a borrower’s health factor has dropped below 1.0
Any external address can call the liquidation function on the smart contract
The liquidator repays up to the close factor percentage of the borrower’s debt
In return, the liquidator receives the borrower’s collateral at a discount (the liquidation bonus, typically 5%-15%)
The borrower’s remaining position is recalculated with an updated health factor
If the health factor is still below 1.0, additional liquidation calls can be made
A vault’s collateralization ratio falls below the liquidation ratio
A keeper triggers the liquidation, starting a Dutch auction where the collateral price starts high and decreases over time
Liquidators bid when the price reaches a level they consider profitable
The winning bidder receives collateral in exchange for repaying the DAI debt plus a liquidation penalty (typically 13%)
Any remaining collateral surplus is returned to the original vault owner
A “circuit breaker” (known as the Ilk box) limits total simultaneous auction exposure
How CEX Liquidation Engines Work
A trader opens a leveraged position (e.g., 10x long BTC at $50,000) with margin collateral
The exchange calculates the liquidation price based on leverage, margin, and maintenance margin requirements
As the mark price (derived from multiple exchange index prices) approaches the liquidation price, the trader receives a margin call warning
When the mark price hits the liquidation price, the exchange’s liquidation engine takes over the position
The engine attempts to close the position at the best available market price
If losses exceed the trader’s margin, the exchange’s insurance fund covers the deficit
If the insurance fund is depleted, auto-deleveraging (ADL) reduces profitable counter-positions
Oracle Dependencies and Manipulation Risk
DeFi liquidations rely on price oracle feeds (Chainlink, Pyth, Uniswap TWAP) to determine collateral values
Oracle manipulation can trigger artificial liquidations. The Mango Markets exploit ($114M, October 2022) demonstrated this attack vector
Time-weighted average prices (TWAPs) provide resistance to flash loan-based oracle manipulation
Multi-oracle systems with median pricing add additional security layers against manipulation
Vantaggi e svantaggi
Vantaggi
Svantaggi
Solvency Protection: Ensures lending protocols and exchanges remain solvent by preventing the accumulation of bad debt that could drain liquidity pools
Cascading Risk: Large liquidation events can trigger cascading sell-offs, amplifying market downturns beyond the initial price decline
Esecuzione automatizzata: Smart contract-based liquidation removes human delay and bias, executing instantly when thresholds are breached
MEV Extraction: Liquidation bots engage in priority gas auctions (PGAs) and sandwich attacks, extracting value from liquidated users through front-running
Incentivi economici: Liquidation bonuses create a competitive market of liquidators who ensure rapid position resolution, maintaining protocol health
Dipendenza da Oracle: Faulty or manipulated price feeds can trigger unjust liquidations or, conversely, prevent necessary liquidations from occurring
Trasparenza: On-chain liquidations are publicly auditable, allowing anyone to verify the fairness and accuracy of the liquidation process
Disproportionate Penalties: Liquidation penalties (5%-15%) can significantly exceed the actual loss, punishing borrowers more than necessary
Efficienza del capitale: Enables leveraged trading and undercollateralized borrowing, increasing capital efficiency across the crypto ecosystem
Gas Cost Barriers: During network congestion, high gas fees can make small liquidations unprofitable for keepers, leaving bad debt in the protocol
Partecipazione senza autorizzazione: Anyone can act as a liquidator in DeFi, democratizing a role traditionally reserved for institutional market makers
Complessità per gli utenti: Liquidation mechanics, health factors, and margin requirements are difficult for retail users to understand and monitor
Insurance Fund Backstop: CEX insurance funds and DeFi protocol reserves provide additional solvency protection when liquidation proceeds are insufficient
Black Swan Vulnerability: Extreme market events (Black Thursday 2020) can overwhelm liquidation infrastructure, causing systemic failures
Partial Liquidation Options: Many protocols support partial liquidation, allowing borrowers to retain a portion of their position rather than losing everything
Centralization of Liquidators: Professional liquidation bots operated by well-funded teams dominate the market, marginalizing smaller participants
Risk Management
Collateral Ratio Management
Always maintain collateral ratios well above the minimum liquidation threshold. A 200%+ ratio provides a reasonable buffer for ETH-collateralized loans on MakerDAO (minimum 150%)
Set up automated health factor monitoring using tools like DeFi Saver or Instadapp that can automatically repay debt or add collateral when ratios approach danger zones
Diversify collateral across multiple assets and vaults to reduce concentration risk
Leverage Discipline
Avoid maximum leverage on centralized exchanges; using 3x-5x instead of 50x-125x provides substantially more room for price fluctuations before liquidation
Always use stop-loss orders set above your liquidation price to exit positions voluntarily before forced closure
Calculate your liquidation price before entering any leveraged position and set price alerts at 10%-20% above that level
Protocol Risk Assessment
Evaluate the liquidation mechanism design before depositing collateral. Dutch auctions (MakerDAO) generally provide better outcomes for borrowers than fixed-spread liquidations (Compound)
Review the oracle infrastructure: protocols using Chainlink with multiple data sources are generally more resistant to manipulation than those relying on single DEX price feeds
Check the protocol’s track record during past market crashes to assess liquidation system reliability
Assicurazione e copertura
Consider purchasing DeFi insurance from providers like Nexus Mutual or InsurAce to cover potential smart contract failures in liquidation mechanisms
Hedge leveraged positions with options or opposing positions on different platforms to reduce net directional exposure
Maintain a reserve of stablecoins or liquid assets that can be quickly deployed to shore up collateral ratios during market downturns
Rilevanza culturale
Liquidation events have become defining moments in crypto market culture, shaping narratives and community behavior in profound ways. The term “rekt,” meaning completely liquidated or suffering total loss, originated in gaming culture but was adopted by the crypto community to describe the devastation of leveraged positions being forcibly closed during market crashes. Websites like Coinglass (formerly Bybt) aggregate real-time liquidation data, and large liquidation spikes are watched as cultural events, the crypto equivalent of watching a slow-motion car crash.
“Liquidation cascades are the crypto market’s way of reminding leveraged traders that the market can stay irrational longer than you can stay solvent.”
Black Thursday (March 12, 2020) became a watershed moment that fundamentally changed how the DeFi community thinks about risk. The $8.3 million in zero-bid MakerDAO liquidations exposed critical infrastructure weaknesses and led to sweeping protocol upgrades across the entire DeFi market. The event is referenced as a cautionary tale in almost every governance discussion about liquidation parameters.
The phenomenon of “liquidation hunting,” where large traders or “whales” deliberately push prices toward clusters of liquidation levels to trigger cascading forced sales, has become a widely discussed strategy and a source of significant controversy. On-chain analytics platforms have made liquidation level mapping possible, creating a cat-and-mouse game between leveraged traders and liquidation hunters.
In DeFi culture, the role of liquidator bots has sparked debates about fairness, MEV extraction, and the centralization of critical protocol functions. The emergence of Flashbots and MEV-protection protocols was partly driven by the desire to reduce the extractive nature of liquidation bot competition on Ethereum.
Esempi del mondo reale
Black Thursday MakerDAO Liquidation Crisis (March 2020)
Scenario: On March 12, 2020, ETH dropped 43% in a single day, pushing thousands of MakerDAO vaults below their 150% collateralization ratio simultaneously.
Implementazione MakerDAO’s liquidation system used English auctions where keepers bid DAI to purchase discounted ETH collateral. Ethereum network congestion caused gas prices to spike over 200 Gwei, and many keeper bots were priced out or failed to submit transactions. Some liquidators won auctions with bids of 0 DAI because no competing bids were submitted within the auction window.
Risultato: Approximately $8.3 million in ETH was extracted at zero cost, creating a $5.4 million deficit in the MakerDAO system that required an emergency MKR token auction to recapitalize. MakerDAO subsequently redesigned its entire liquidation system (Liquidations 2.0) with Dutch auctions, circuit breakers, and improved keeper infrastructure.
May 2021 Crypto Market Crash: CEX Cascading Liquidations
Scenario: Bitcoin dropped from $58,000 to $30,000 in a single week (May 12-19, 2021), triggered by Elon Musk’s reversal on Tesla accepting Bitcoin and China’s renewed crackdown on crypto mining.
Implementazione Across centralized exchanges, over $8 billion in leveraged positions were liquidated within 24 hours on May 19 alone. Binance’s liquidation engine processed the highest volume, with approximately $2.6 billion in forced closures. Bybit and FTX each processed over $1 billion. The cascading nature of liquidations, where forced selling pushes prices lower, triggering more liquidations, amplified the crash significantly beyond the initial selling pressure.
Risultato: The event led to widespread calls for exchange reforms, including lower maximum leverage limits. Binance subsequently reduced maximum leverage from 125x to 20x for new accounts. The crash demonstrated how interconnected CEX liquidation engines can create feedback loops that destabilize the entire market.
Aave Flash Loan Liquidation Strategy
Scenario: A DeFi developer built a liquidation bot that monitors Aave V3 positions for health factors approaching 1.0 on the Ethereum mainnet.
Implementazione When a liquidation opportunity is detected, the bot executes a flash loan from Aave itself: borrowing the debt token, repaying the borrower’s debt through the liquidation function, receiving the discounted collateral, swapping the collateral back to the debt token on a DEX, repaying the flash loan, and pocketing the difference, all in a single atomic transaction with zero upfront capital.
Risultato: The bot processes profitable liquidations averaging $500-$5,000 per event, earning roughly 3%-8% after gas costs and flash loan fees. This model has become the standard approach for DeFi liquidators, making the process accessible to any developer willing to build and maintain the bot infrastructure.
Mango Markets Oracle Manipulation Liquidation ($114 Million, October 2022)
Scenario: Trader Avraham Eisenberg identified that Mango Markets on Solana used a single oracle source to price MNGO tokens for its perpetual futures and lending markets.
Implementazione Eisenberg opened a massive MNGO-PERP position, then used a second account to buy large quantities of MNGO on spot markets, artificially inflating the price. The manipulated oracle price increased his perpetual position’s unrealized PnL, which he used as collateral to borrow $114 million in other assets from Mango’s lending pools, effectively liquidating the protocol’s entire lending pool.
Risultato: Mango Markets was drained of $114 million. Eisenberg was subsequently arrested by the FBI and charged with commodities fraud and market manipulation, setting a legal precedent for DeFi exploit prosecution. The incident catalyzed industry-wide improvements in oracle design and liquidation parameter governance.
October 2025 “Black Friday” Liquidation Cascade
Scenario: On October 9, 2025, geopolitical instability arose as China announced the weaponization of rare earth metals, followed by US threats of 100% tariffs. This sparked a massive market shock, exacerbated by concentrated short positions on the Hyperliquid perpetuals exchange.
Implementazione The sudden price drop triggered cascading liquidations, forcing exchanges like Hyperliquid to use Auto-DeLeveraging (ADL) to maintain solvency. Binance’s ADL mechanism faced API failures that disrupted stablecoin pegs, causing Ethena (USDe) to briefly depeg to $0.68. Conversely, DeFi protocols like Aave demonstrated resilience, successfully liquidating $180 million in collateral in a single hour without downtime.
Risultato: The event recorded $19.35 billion in total losses, marking it as the largest liquidation event in the history of the asset class. It highlighted critical vulnerabilities in centralized infrastructure and the systemic risks of excessive market leverage.
Tavola di comparazione
Caratteristica
DeFi Fixed-Spread (Aave/Compound)
DeFi Dutch Auction (MakerDAO)
CEX Liquidation Engine (Binance/Bybit)
Velocità di esecuzione
Instant (single transaction)
Minutes (auction duration)
Near-instant (matching engine)
Liquidation Bonus/Penalty
5%-15% fixed
Variable (market-determined)
Varies by asset/leverage tier
Liquidator Access
Permissionless (any address)
Permissionless (keepers)
Exchange-operated (closed)
Requisiti patrimoniali
Zero (flash loan enabled)
Requires DAI capital for bids
N/A (exchange handles internally)
Partial Liquidation
Yes (close factor ~50%)
Yes (proportional to debt)
Varies by exchange and position size
Dipendenza da Oracle
Chainlink / protocol-specific
Chainlink + OSM (1-hour delay)
Internal index price + spot markets
Meccanismo assicurativo
Protocol reserve / safety module
MKR dilution auction
Insurance fund + ADL
Termini correlati
Richiesta di margine: A demand from a broker or exchange for additional collateral when a leveraged position’s equity falls below the maintenance margin, often serving as a warning before full liquidation.
Collateralization Ratio: The ratio of the value of pledged collateral to the value of borrowed assets, which must be maintained above the liquidation threshold to avoid forced closure.
Health Factor: A numerical representation of the safety of a DeFi lending position, where values below 1.0 indicate the position is eligible for liquidation.
Prestito Flash: An uncollateralized loan that must be borrowed and repaid within a single blockchain transaction, commonly used to fund DeFi liquidations without upfront capital. Introduced in Aave V1 in January 2020.
Oracle: A service that feeds real-world price data to smart contract, enabling liquidation mechanisms to determine when collateral values have fallen below required thresholds.
Fondo assicurativo: A reserve pool maintained by exchanges to cover losses from liquidated positions that result in negative balances exceeding the trader’s margin.
Auto-Deleveraging (ADL): A mechanism used by derivatives exchanges to automatically reduce the positions of profitable counter-traders when the insurance fund is depleted during extreme liquidation events.
Margine di mantenimento: The minimum amount of equity a trader must maintain in a leveraged position to avoid liquidation, expressed as a percentage of the position’s total value.
MEV (Valore Massimo Estraibile): The value that block producers or searchers can extract by reordering, inserting, or censoring transactions, significantly relevant to the competitive dynamics of DeFi liquidation bots.
Cascading Liquidation: A chain reaction where one liquidation event pushes prices lower, triggering additional liquidations that further depress prices, creating a self-reinforcing downward spiral.
FAQ
Q: What happens to my remaining collateral after liquidation in DeFi?
In most DeFi protocols, only a portion of your position is liquidated, known as partial liquidation. For example, on Aave, the close factor is typically 50%, meaning only half your debt can be repaid in a single liquidation event. After the liquidation penalty is deducted (5%-15%), any remaining collateral stays in your position. However, if your position’s health factor is deeply underwater, multiple successive liquidations may consume most or all of your collateral.
Q: How is DeFi liquidation different from CEX liquidation?
In DeFi, liquidation is permissionless: any external address can trigger it by calling a smart contract function, and liquidators compete for the liquidation bonus. On centralized exchanges, the exchange’s internal liquidation engine handles everything automatically. DeFi liquidations are transparent and on-chain, while CEX liquidations are processed in private order books. DeFi also allows flash loan-powered liquidations requiring zero capital, which is impossible on centralized platforms.
Q: Can I prevent my position from being liquidated?
Yes, by maintaining a healthy collateral ratio well above the liquidation threshold. You can add more collateral, repay part of your debt, or use automated tools like DeFi Saver’s automation system to programmatically adjust your position when it approaches danger zones. Setting price alerts and maintaining reserve funds for emergency collateral deposits are essential risk management practices.
Q: What caused the Black Thursday liquidation crisis in March 2020?
A combination of factors converged: ETH dropped 43% in one day, Ethereum network congestion caused gas prices to spike above 200 Gwei, MakerDAO’s keeper bots failed to submit competitive bids in collateral auctions, and the English auction mechanism allowed zero-bid victories when no competing bids were submitted. The crisis exposed fundamental weaknesses in DeFi liquidation infrastructure and led to detailed protocol upgrades across the ecosystem.
Q: What is a liquidation cascade, and why is it dangerous?
A liquidation cascade occurs when a price decline triggers forced selling from liquidated positions, which pushes prices lower, triggering more liquidations, which causes more forced selling, creating a self-reinforcing feedback loop. This mechanism can amplify a modest market correction into a severe crash. The May 2021 event, where $8 billion was liquidated in 24 hours, demonstrated how cascading liquidations can destabilize the entire crypto market.
Q: How do liquidation bots make money?
Liquidation bots monitor blockchain mempools and lending protocol positions for liquidation opportunities. When they detect an undercollateralized position, they submit a transaction to repay the borrower’s debt and receive the collateral at a discount (the liquidation bonus). Using flash loans, bots can execute this entire process: borrow, liquidate, swap, repay, all in a single transaction without any upfront capital, profiting from the spread between the liquidation bonus and transaction costs.
Q: Are high-leverage positions more likely to be liquidated?
Absolutely. Higher leverage means a smaller price movement triggers liquidation. With 100x leverage, a mere 1% adverse price movement can liquidate your entire position. With 2x leverage, the price would need to move approximately 50% against you. This is why most experienced traders recommend using conservative leverage ratios of 2x-5x and always calculating the exact liquidation price before entering a position.
La piattaforma di scambio di criptovalute Bitget sta portando avanti i piani per entrare negli Stati Uniti, indipendentemente dal fatto che il Congresso approvi o meno la proposta.
La Russia si è avvicinata alla creazione di uno dei suoi quadri normativi più completi in materia di criptovalute dopo che la Duma di Stato ha approvato un
