If the borrower owes WBTC, the keeper must transfer WBTC to repay the loan even when the application fee and borrower-surplus payment are zero.
The assessment said the integration contracts had not been deployed on Ethereum mainnet. Babylon’s contract directory, checked Oct. 25, identifies its published addresses as Sepolia testnet deployments.
Two routes, three payment requirements
A borrower’s native $BTC would remain in a Bitcoin vault, and Aave would recognize that collateral through a restricted internal ERC-20 accounting record, which users cannot trade as a wrapped Bitcoin token.
Through the proposed Liquidation Liquidity Provider (LLP) route, an outside liquidator would repay the borrower’s debt and receive an immediate WBTC payout.
The settlement component, BTCVaultSwap (the Swap Spoke), would place the seized vault in escrow and draw the payout from Aave’s shared liquidity pool (the Hub). The underlying $BTC would remain on Bitcoin.
An authorized keeper acting as an arbitrageur would then be expected to buy the escrowed vault and handle redemption on Bitcoin. The process depends on a willing eligible buyer completing that purchase.
The Babylon proposal also describes a direct route for keepers with the vault’s committed Bitcoin redemption rights. Such a keeper would fund debt repayment and take the seized vault into redemption without the Swap Spoke’s new Hub-funded payout or its escrow step.
When the Hub WBTC allowance runs out, a keeper able to supply the required payment assets can avoid that funding constraint. However, the pool of available participants would narrow to operators equipped and authorized to claim the underlying Bitcoin.
Babylon’s liquidation operator guide says it collects debt repayment and two additional charges: the application liquidation fee and any fairness payment owed to the borrower. For direct liquidation, both additional charges are denominated in WBTC and funded by the caller.
The debt repayment uses the token of the borrower’s selected debt reserve. The adapter pulls the repayment input and refunds whatever Aave’s lending contract does not consume. The repayment input is separate from the caller’s limit on additional WBTC payments.
The keeper therefore needs assets for the debt, the fee, and any surplus payment. Its own inventory supplies funding outside the Hub, subject to each required token transfer working.
The assessment describes the Babylon application fee as 0.5% of liquidation bonus profit. The assessed design separately forces the Babylon lending Spoke’s own liquidation fee to zero while retaining the application fee at the adapter.
For the direct route to avoid the stated WBTC fee and surplus obligations in LlamaRisk’s transfer-failure scenario, the application liquidation fee must be zero and no fairness payment can be due. Those conditions address only those two payments.
A WBTC loan still needs WBTC repayment even in that exception. For a loan in another asset, repayment uses that asset, while the keeper still needs transferable debt tokens, redemption rights, and functioning liquidation contracts.
Setting the application fee to zero would itself require a time-locked Babylon application-parameter change. LlamaRisk places liquidationFeeBps changes behind the application timelock, separately from immediate emergency pause and freeze controls.
The fee-specific delay remains unspecified in the assessment, so an individual keeper has to use the effective fee setting, including any WBTC charge still due when it acts.
Whole vaults can create a WBTC surplus
The fairness payment follows from the way the collateral is held, while partial-position liquidation can seize some of a borrower’s vaults while leaving others in place. Each seized Bitcoin vault is indivisible.
Under the model described in the Oct. 23 assessment, the governance proposal and the operator guide, each liquidation call targets the first whole vault in the borrower’s ordered list. If the liquidator must liquidate multiple vaults, it checks the position’s health again.
If the borrower’s entire debt is cleared before the selected vault’s value is consumed, the liquidator still receives the entire vault. It must compensate the borrower for the remaining WBTC value at the liquidation-bonus discount.
Better sizing of the first vault can reduce the surplus a liquidator must fund while leaving other vaults in the position. Whether a fairness payment arises depends on the debt and collateral valuation at execution.
Partial-position liquidation describes how many vaults are taken, so a borrower-surplus payment can arise even on a direct call that charges no application fee.
Liquidation on Ethereum must also be distinguished from receiving spendable Bitcoin. Direct liquidation initiates an authorized redemption, but the Bitcoin proof-and-challenge process still takes multiple days. On the direct route, the keeper finances that multi-day wait itself.
Babylon’s integration documentation describes removing vaults subject to the remaining position staying healthy. The assessment additionally requires no remaining debt to withdraw every vault. Borrower withdrawals remain subject to those conditions even when a keeper can liquidate.
LlamaRisk recommends an initial 100 WBTC Core Hub credit line for BTCVaultSwap, separately from the proposed 300 $BTC collateral launch cap.
The assessment envisages an initial rollout with 10 allowlisted borrowers and 50 vaults. That would be a limited launch alongside permissioned Bitcoin redemption operators.
The lending line would determine how much funding is available for the LLP route. Beyond that line, the direct fallback would rely on keepers repaying the debt and meeting any WBTC fee or borrower-surplus payment.
Final fee settings, debt assets, and vault composition would determine how far that fallback extends.
