Liquidation Auctions: Why DeFi Positions Vanish in a Single Block
When a leveraged position gets liquidated in DeFi, it can look like a switch got flipped. In reality, it's a sale, and every sale needs a buyer.
When collateral value falls below a protocol's required health factor, the position becomes eligible for liquidation. Bots monitor these thresholds constantly. Once a position crosses the line, liquidators race to repay part of the debt in exchange for the collateral, plus a bonus, often 5-15 percent depending on the asset. Whoever confirms first captures the discount, which is why liquidation events cluster during network congestion.
Some protocols use explicit Dutch auctions instead, where collateral price starts high and drops over time until a buyer steps in. This spreads the sale out rather than resolving it in one block, softening some price impact.
A practical example: ETH used as collateral drops 12 percent in an hour. The position crosses its liquidation threshold, and within the same block a bot repays part of the debt and receives discounted ETH. That ETH usually gets sold immediately on the open market to lock in profit, often funded by a flash loan. When many positions liquidate at once, this creates a wave of forced selling stacked on top of the original move, which is part of why sharp crypto drops often look like waterfalls rather than smooth declines.
The key insight for traders is that a health factor isn't just a personal risk metric, it's public on-chain data. Anyone can see which large positions sit close to their liquidation threshold. Clusters of at-risk collateral represent latent selling pressure that can activate mechanically, independent of any new information hitting the market.
This also explains why leverage behaves differently on-chain than on centralized exchanges. There's no internal risk desk managing the unwind, just an open market of competing liquidators, where speed, gas costs, and bot density determine how much price impact spills into the broader market.
#DeFi #Crypto #Trading #RiskManagement #MarketAnalysis
When a leveraged position gets liquidated in DeFi, it can look like a switch got flipped. In reality, it's a sale, and every sale needs a buyer.
When collateral value falls below a protocol's required health factor, the position becomes eligible for liquidation. Bots monitor these thresholds constantly. Once a position crosses the line, liquidators race to repay part of the debt in exchange for the collateral, plus a bonus, often 5-15 percent depending on the asset. Whoever confirms first captures the discount, which is why liquidation events cluster during network congestion.
Some protocols use explicit Dutch auctions instead, where collateral price starts high and drops over time until a buyer steps in. This spreads the sale out rather than resolving it in one block, softening some price impact.
A practical example: ETH used as collateral drops 12 percent in an hour. The position crosses its liquidation threshold, and within the same block a bot repays part of the debt and receives discounted ETH. That ETH usually gets sold immediately on the open market to lock in profit, often funded by a flash loan. When many positions liquidate at once, this creates a wave of forced selling stacked on top of the original move, which is part of why sharp crypto drops often look like waterfalls rather than smooth declines.
The key insight for traders is that a health factor isn't just a personal risk metric, it's public on-chain data. Anyone can see which large positions sit close to their liquidation threshold. Clusters of at-risk collateral represent latent selling pressure that can activate mechanically, independent of any new information hitting the market.
This also explains why leverage behaves differently on-chain than on centralized exchanges. There's no internal risk desk managing the unwind, just an open market of competing liquidators, where speed, gas costs, and bot density determine how much price impact spills into the broader market.
#DeFi #Crypto #Trading #RiskManagement #MarketAnalysis