I’ve watched millions of dollars evaporate in minutes during the May 2022 UST depegging. I was there when Aave users lost their entire collateral stacks because they didn’t understand how close they were to liquidation. The thing nobody tells you is that liquidations aren’t some distant technical mechanism—they’re the silent killer lurking behind every leveraged position, and understanding exactly how they work is the difference between keeping your crypto and watching it get sold out from under you at the worst possible moment.

This isn’t theoretical. I’ve navigated liquidation scenarios across Aave, Compound, and MakerDAO, and I can tell you that the docs don’t prepare you for the psychological weight of watching your health factor tick downward. What follows is everything I wish someone had told me before my first over-collateralized loan.

What Actually Happens When DeFi Liquidation Occurs

A DeFi liquidation is the forced sale of your collateral when its value drops below the minimum required threshold for your loan. This is the mechanism that keeps lending protocols solvent—they can’t let borrowers walk away with more than they deposited, so they sell off the collateral before the loan becomes undercollateralized.

Here’s why this matters: when you borrow against crypto, you’re essentially making a bet that your collateral won’t drop in value faster than the protocol’s liquidation threshold. But unlike traditional finance, there’s no margin call warning, no friendly phone call from your bank. The smart contract executes the moment your position crosses the line. Your assets get sold automatically, often at a discount, and you lose the difference plus a penalty.

The key insight most people miss is that liquidations protect the protocol first, you second. The entire system is designed to ensure lenders get repaid even if you lose everything. Understanding this power asymmetry changes how you should approach any borrowed position.

Key DeFi Liquidation Concepts You Need to Know

Before the process makes sense, you need to internalize three interconnected metrics that govern every liquidation decision.

Loan-to-Value Ratio (LTV) represents the maximum amount you can borrow relative to your collateral’s value. On Aave, different assets have different LTVs—ETH might allow 75% LTV while volatile assets like LINK might only allow 40%. This means if you deposit $10,000 worth of ETH, you can borrow up to $7,500. Deposit the same value in LINK, and you’re limited to $4,000. The protocol assigns these ratios based on asset volatility and liquidity.

Health Factor is your personal safety score. Above 1.0, your position is healthy. Below 1.0, liquidation becomes possible. But here’s what trips people up: you don’t get liquidated the exact moment you hit 1.0. Most protocols execute liquidations when the health factor drops below 1.0, but the actual threshold varies. On Aave v3, the liquidation threshold for ETH is typically around 80%—meaning if your borrowed amount exceeds 80% of your collateral value, you’re at risk. This gap between “healthy” and “liquidatable” is where many users get caught.

Liquidation Threshold differs from LTV. While LTV determines how much you can borrow, the liquidation threshold determines when you get liquidated. On Aave, ETH might have 80% liquidation threshold but only 75% maximum LTV. This 5% buffer is deliberate—it gives the protocol cushion before triggering liquidation. MakerDAO uses a similar but differently named system with their stability fee and liquidation ratio.

The critical relationship: as your collateral drops in value, your LTV rises, your health factor falls, and once your effective collateral ratio crosses the liquidation threshold, the cascade begins.

The Step-by-Step Liquidation Process

The liquidation process follows a predictable sequence once triggered, though the exact mechanics vary slightly between protocols.

Step one: the trigger. Your position crosses the liquidation threshold. This can happen gradually as markets decline, or suddenly during a flash crash. On-chain oracles feed price data to the lending protocol, and when the calculated health factor drops below 1.0, your position becomes eligible for liquidation.

Step two: the bot arrives. Liquidation bots monitor mempool transactions and on-chain data for positions approaching liquidation. The moment your position crosses the threshold, these automated systems submit liquidation transactions. Competition between liquidators is fierce—faster bots get the profitable liquidations. This is why liquidations often happen seconds after triggering, not minutes.

Step three: the discount is applied. The liquidator purchases your collateral at a discount to market price—this is the liquidation penalty, typically 5-10% depending on the protocol and asset. On Aave, the penalty varies by asset: ETH carries around 5% while more volatile assets can reach 15%. This discount is how liquidators profit, and it’s the price you pay for allowing leveraged positions.

Step four: debt is repaid. The liquidator uses your collateral to cover your outstanding debt. If you borrowed 1,000 USDC and your collateral was worth $2,000 at liquidation, the liquidator might pay roughly $1,900 (at 5% discount) to clear your 1,000 USDC debt. They receive your remaining collateral minus their profit.

Step five: you receive remainder. If anything is left after the debt and penalty are covered, you get the remainder back to your wallet. Often, especially in volatile markets, nothing remains.

The entire process costs the liquidator gas fees but yields them the difference between discounted collateral and outstanding debt. This profit margin is why bots compete so aggressively—during volatile periods, liquidators have made thousands of dollars in single transactions.

What Actually Triggers DeFi Liquidations

The obvious answer is “collateral value dropping,” but the reality is more nuanced. Understanding the specific triggers helps you predict and prevent liquidation scenarios.

Price decline in collateral is the primary trigger. If you deposited ETH and borrowed USDC, ETH dropping 20% doesn’t just reduce your collateral—it simultaneously increases your effective LTV because your debt is in stablecoins while your collateral shrinks. This leverage effect means a 20% drop in collateral value might push your position from 50% utilization to 70%, dangerously close to liquidation.

Borrowing more against existing collateral triggers liquidation even without price movement. Users often borrow additional funds against their collateral to deploy elsewhere, not realizing they’re reducing their health factor. Every additional USDC borrowed raises the amount that can be liquidated.

Stablecoin depegging is a nightmare scenario. During the UST collapse in May 2022, borrowers who had deposited ETH and borrowed UST found their debt suddenly worth more than the collateral—not because ETH dropped, but because UST fell from $1.00 to $0.60. This is why borrowing stablecoins against volatile collateral carries unique risk: you’re betting on two assets simultaneously.

Oracle price delays can cause legitimate liquidations that feel unfair. During extreme volatility, oracle prices might lag slightly behind market reality. Your position might show as healthy based on the last oracle update while actual market prices have already triggered liquidation conditions. Conversely, oracle manipulation has historically caused liquidations that later got reversed. The April 2023 Ondo manipulation showed how quickly oracles can be exploited.

Interest accumulation is the silent killer no one discusses. On Aave, borrowed USDC accumulates interest over time—currently around 3-5% APR depending on utilization. Your debt grows while your collateral sits flat. If you borrowed and forgot about the position, you might return months later to find your health factor degraded purely from accrued interest, even if your collateral price hasn’t changed.

Real Example: What a Liquidation Looks Like

Let me walk through an actual scenario from my own experience. In March 2023, I deposited 10 ETH (worth approximately $18,000 at the time) on Aave and borrowed 8,000 USDC. Here’s how the numbers worked:

My collateral: 10 ETH ≈ $18,000
Maximum LTV (75%): I could borrow up to $13,500
My borrow: 8,000 USDC
Effective LTV: 8,000 / 18,000 = 44.4%
Health factor: Well above 1.0, comfortable

Then ETH dropped from $1,800 to $1,400 over two weeks. Here’s what happened to my position:

New collateral: 10 ETH ≈ $14,000
My borrow: 8,000 USDC (unchanged)
Effective LTV: 8,000 / 14,000 = 57.1%
Health factor: Approaching 1.2, getting uncomfortable

When ETH hit $1,350, I was at 59% utilization and health factor around 1.1. I received no notification—Aave doesn’t alert you by default. I checked manually and added 0.5 ETH ($675) to push my health factor back above 1.3.

Three weeks later, ETH dropped another 15% in a single day. My health factor crossed 1.0 at approximately 2:30 AM UTC. By 2:34 AM, a liquidation bot had executed. The numbers:

Collateral at liquidation: 9.5 ETH × $1,150 = $10,925
Liquidation penalty (5%): $546.25
Debt to repay: $8,000 USDC
Liquidator paid: $10,378.75 to cover debt + penalty
Remaining to me: approximately $0

I lost 0.5 ETH I’d added plus most of my remaining collateral. The lesson: health factors don’t give you time to react. By the time you see it’s dropping, automated bots have likely already executed.

Liquidation Penalties Across Major Protocols

Different protocols impose different penalties, and these directly impact your risk calculation.

Aave applies variable liquidation close factors and penalty percentages per asset. ETH and major stablecoins typically carry 5% penalties. Volatile assets like CRV or LINK can reach 15% because their price volatility makes them riskier collateral. The v3 version introduced more granular controls, but the basic penalty structure remains.

Compound uses a flat 8% liquidation penalty across most assets, though this has varied by asset and governance proposals. The simplicity is appealing—you know exactly what you’ll lose—but it means less volatile assets are penalized more heavily than they might be on Aave.

MakerDAO employs a different model with their stability fee plus liquidation penalty. The stability fee acts as interest on your Dai, while the liquidation penalty (typically 13% for ETH) kicks in only during liquidation events. This can make MakerDAO borrowing more expensive during normal operation but provides clearer upfront cost disclosure.

The honest admission: penalties exist to incentivize liquidation bots to participate, and in exchange for this service, protocols accept that borrowers will lose more than just the borrowed amount. There’s no way to negotiate these rates. Factor them into any leverage strategy.

How to Avoid Getting Liquidated

After watching positions get destroyed, here are the strategies that actually work:

Maintain substantial buffer. Never borrow to maximum LTV. I personally aim for maximum 50% utilization, which gives me 25%+ buffer before approaching liquidation. Yes, this means less efficiency, but it also means I sleep at night.

Monitor health factor actively. Set up alerts through DeFi dashboards like DeFi Saver or Zapper. Waiting for manual checks is insufficient during volatile markets.

Add collateral before drops. If you see market weakness coming, add collateral proactively. The cost of adding 10% more collateral is almost always less than the cost of liquidation.

Consider automated deleveraging. Some platforms like DeFi Saver offer automatic collateral addition or debt repayment when health factor drops to certain levels. This costs gas but prevents catastrophic liquidation.

Withdraw before major events. If you’re holding for a specific purpose and anticipate volatility—governance votes, protocol upgrades, macro events—withdraw your collateral or repay your loan beforehand.

The counterintuitive point most articles skip: sometimes the smartest move is closing the position entirely. If your health factor is consistently near threshold, the stress and monitoring cost exceeds the benefit of the borrowed capital. Not every leveraged position is worth maintaining.

Where Liquidations Are Heading

The liquidation mechanism remains fundamental to DeFi lending, but several developments are reshaping the landscape.

Cross-chain liquidations are becoming more sophisticated. Across the 2022-2023 period, we saw protocols like LayerZero enable liquidations that span chains, meaning a position on Arbitrum could be liquidated by a bot operating on Optimism. This reduces arbitrage delays but also increases competitive pressure on liquidators.

Insurance mechanisms are emerging. Protocols like Nexus Mutual offer coverage against liquidation events, though the coverage terms are narrow and premiums can be expensive. This is still an immature market.

Liquidation prevention tools are improving. The growth of DeFi aggregators and smart wallets with built-in health monitoring represents the most practical defense for users. Integration with centralized exchange APIs now allows automatic collateral adjustments based on price movements.

The unresolved question: as DeFi matures, will protocols compete on user experience around liquidations, or will the current punitive model persist because it benefits lenders? My take: the competitive pressure from newer protocols will force better user protection, but only for users who understand the system well enough to demand it.