The governance proposal sets out initial stablecoin credit lines from Aave’s Core Hub liquidity pool of $8 million each for USDT and USDC, $1 million for GHO, and $2 million each for USDG and RLUSD.
A separate liquidity pool, the Babylon Hub, would let liquidity providers choose exposure to native Bitcoin collateral, with its own limits.
Risk adviser LlamaRisk recommends a 78% collateral factor for native $BTC. The collateral record itself would remain non-borrowable, representing $BTC securing a position.
Babylon targets mainnet for late 2026, subject to change and confirmation. It says deployment will start with restrictions on participants and volume, with a schedule and staged rollout plan still to follow. Its current FAQ directs users to a public testnet and gives a fourth-quarter mainnet target.
Keeping the underlying $BTC on Bitcoin does not remove liquidation or settlement risk. The Ethereum collateral record would be a transfer-restricted ERC-20 accounting unit, and a borrower could still lose vault collateral through liquidation.
Wrapped Bitcoin remains as part of the structure
The design also retains a role for Wrapped Bitcoin (WBTC). Under the permissionless liquidation route, BTCVaultSwap would borrow WBTC from an Aave Hub to settle immediately with liquidators.
Permissioned arbitrageurs would later repay that debt and redeem the underlying $BTC over several days, as LlamaRisk’s assessment describes.
The proposed Core Hub draw cap for that settlement venue is 100 WBTC, separate from the stablecoin lending lines. Permissioned actors can also liquidate and redeem vaults directly without drawing Hub liquidity, although fees or borrower compensation can still require WBTC.
Liquidation seizes whole vaults, even when clearing the debt requires less than a vault’s full $BTC value, and the borrower receives discounted compensation for that remainder in wrapped $BTC.
Prospective borrowers still need final AIP approval and a confirmed rollout schedule before this market moves from governance backing to mainnet access.