Curve founder Michael Egorov has proposed a new potential answer to one of DeFi’s most challenging problems: What happens after bad debt creates a vulnerability?
Yegorov outlined the proposal to recover non-performing loans by converting non-performing assets into tradable investment products, initially testing the concept on Curve’s own CRV long LlamaLend market.
This promotion comes at a time when DeFi is embroiled in public debates over recovery mathematics, bailout schemes, and who should bear responsibility for failures spreading across protocols.
The Kelp DAO vulnerability and the resulting bad debt risk to Aave triggered a flurry of rescue proposals, ranging from token allocations to emergency loans, with contributions from ecosystem participants including Mantle, Lido, and Aave itself.
Yegorov's version attempts to steer the discussion in another direction.
In his proposal, he wrote: "I have proposed a bad debt recovery mechanism that is not a donation, but an investment tool applicable to all participants. If the pilot study is successful, it can be applied to Curve or other protocols facing similar challenges."
Yegorov's promotion
The issue he is now trying to solve predates and is smaller in scale than the kelp problem, but it has gained new practical significance because of the kelp problem.
Curve's CRV long-term LlamaLend market experienced bad debt in October 2025, resulting in approximately $700,000 of undercollateralized funds in the vault and preventing lenders from fully withdrawing their funds.
Although this amount is negligible compared to some other events, the assets stolen in the KelpDAO incident were valued at $292 million; despite subsequent billions of dollars in Aave fund outflows, Egorov believes these compromised vault tokens are not dead assets.
In his description, they have unusual return characteristics.
If the Credit Risk Value (CRV) increases, the debt can be recovered as collateral is liquidated, leading to a smooth liquidation. Conversely, if the Credit Risk Value decreases, the collateral does not deteriorate like traditional bad debt.
Egorov believes this gives the vault token an "option-like" structure: if the collateral rebounds, there is upside potential; if the collateral does not rebound, there is downside risk. To make it tradable, Egorov says he has created a Curve stableswap pool with approximately 71% solvency, where distressed vault tokens can be exchanged.
Liquidity providers in the pool earn trading fees and may also receive CRV incentives if approved by the governance body, while the DAO itself can accumulate degraded tokens through management fees without requiring a direct bailout vote.
This is precisely the core appeal of the design. It seeks to replace social welfare with market mechanisms. Rather than requiring the treasury of a decentralized autonomous organization (DAO) or associated protocols to fill the funding gap, the model explores whether bad debts can be cleared if bundled with sufficient returns for sale.
Traders can purchase vault tokens at a discount. If the price is low enough, arbitrageurs can exploit the opportunity. Liquidity providers can earn fees while waiting for the price to rebound. Curve DAO can participate, but Egorov stated that it is not required to do so.
The timing is no accident.
Since seaweed utilization, DeFi has been forced into a chaotic real-time rescue. Protocols such as Lido and Mantle have pledged support. Aave has also discussed direct donations and... ETH frozen on Arbitrum has been unfrozen.
All of this highlights a larger issue: Should bad debt be socialized, isolated, refinanced, or simply left to the market to resolve?
Yegorov clearly favors the last option—or at least prefers to package it with market-based approaches.
On the other hand
The initial response shows that coming up with an idea is much easier than selling it.
One commentator pinpointed a clear primary concern, saying: "The reality is that no one will buy the affected positions because they generate no yield."
Another user countered that if CRV eventually recovers, the yield still exists, and the instrument resembles a discounted perpetual option with limited downside risk rather than a dead debt.
The same commenter explained the trade in simple terms: do nothing and hope CRV rises above the threshold needed to repay the debt; sell now at a discount; or become a liquidity provider in the new recovery pool and earn fees while waiting.
A more skeptical response questions whether sophisticated capital would bother with this. If similar returns can be synthesized elsewhere through cheaper means, this pool may struggle to attract genuine buyers without substantial subsidies.
Egorov responded that traders may prefer Curve stableswap LP itself rather than just the vault tokens, and noted that its payout structure is different and potentially more attractive.