Curve Finance Soft Liquidation Model Keeps Hundreds of Loans Alive for Weeks

Hundreds of DeFi loans on Curve Finance have survived weeks inside liquidation status without being forcibly closed, validating a risk management approach that could reshape how institutional capital thinks about on-chain borrowing.

The mechanism at the center of this is LLAMMA (Lending-Liquidating AMM Algorithm), Curve's proprietary system that replaces the binary, all-or-nothing liquidation model common across legacy DeFi lending protocols. Instead of triggering an immediate forced sale when a borrower's collateral drops below a threshold, LLAMMA gradually converts collateral into stablecoins as prices fall, and back into the original collateral as prices recover. The borrower stays in the protocol. The position survives. The liquidation becomes a process, not an event.

That distinction matters enormously in practice. In traditional DeFi lending, a sharp drawdown hits a collateral ratio threshold and the protocol liquidates the position in a single transaction, often at the worst possible price. Borrowers lose everything above the liquidation penalty in seconds. LLAMMA instead creates a buffer zone where partial liquidation occurs incrementally, giving borrowers time to add collateral, reduce their loan, or simply wait for the market to recover. On-chain data from Curve shows this is not theoretical: loans have been sitting in partial liquidation states for weeks, surviving conditions that would have wiped them out on Aave or Compound.

The counterarguments are real and worth taking seriously. Extended liquidation windows increase the protocol's exposure to bad debt if collateral continues falling after the buffer is exhausted. A borrower who knows their position won't be immediately liquidated has less incentive to top up collateral when ratios get tight, a classic moral hazard problem. There is also a risk modeling headache for institutions: treasury desks and risk committees generally want clean liquidation timelines, not open-ended exposure windows that are harder to stress-test. Prolonged liquidation states also complicate protocol accounting, since partially liquidated positions sit in an ambiguous state that traditional credit frameworks were not built to handle.

That said, the institutional appeal argument has merit on the other side. One of the most frequently cited barriers to institutional DeFi participation is the risk of sudden, total collateral loss during volatile sessions. A single bad hour in a thin market can wipe a position that was healthy at the open. Curve's soft liquidation model does not eliminate that risk, but it compresses it significantly. The ability to demonstrate that loans can weather multi-week drawdown periods without forced closure is a tangible data point for any risk officer evaluating on-chain credit exposure.

Curve is not alone in exploring this direction. The broader DeFi lending space has been moving toward more nuanced liquidation mechanics for the past two years, driven partly by the brutal liquidation cascades that accompanied the 2022 bear market and the March 2024 volatility spike. But LLAMMA is among the most fully realized implementations of the gradual liquidation concept deployed at meaningful scale. The fact that hundreds of loans have now lived through the proof-of-concept phase in live market conditions gives the model empirical weight that whitepapers cannot.

What Curve is building here is infrastructure. Whether it becomes a standard that other lending protocols adopt, or remains a competitive differentiator for Curve specifically, depends on how the risk tradeoffs play out over a full market cycle. The weeks-long liquidation survival data is promising. The real test comes when collateral does not recover.