Loading
Sharpe FinanceDID YOU KNOWnext in 7s
There are more EVM L2s now than there are countries in the eurozone.
Morpho Blue · Aave V3 · 11 chains · scoring refreshed every 5 min
There are more EVM L2s now than there are countries in the eurozone.
USDG supply on Aave V3 (Ethereum). Read-only on Sharpe — deposits, withdrawals, and borrows execute on app.aave.com. We surface parameters, rates, and risk so you can compare this reserve against Morpho vaults on the same asset cohort.
Plain-English summary of this vault — what it does, who runs it, where the yield comes from, and what could break it. Generated from the same deterministic inputs shown elsewhere on this page; the numbers are the source, this is just the explanation.
You deposit USDG (a stablecoin issued and controlled by Paxos) into this Ethereum reserve and earn 1.37% APY. Your deposit is lent out to borrowers at 3.31% variable rate; you capture the spread after Aave's 20% cut. USDG cannot be used as collateral on Aave, only borrowed against other collateral. Your funds are redeemable at the available liquidity of $8.09M; deposits beyond that queue until borrowers repay.
Aave governance has marked this reserve borrow-only, signaling acceptance of USDG's counterparty model but caution against using it as collateral.
At 52% utilization, borrow demand is moderate. The 1.37% supply APY comes directly from the 3.31% borrow rate × 52% utilization × (1 − 20% reserve factor). Utilization in the mid-range means rates remain stable if demand shifts moderately.
Paxos controls USDG issuance and can freeze the asset at its discretion. That's the material risk here—not market price volatility, but counterparty action. The $8.09M liquidity buffer is thin relative to the $16.88M supplied; a freeze would prevent redemptions entirely.
Good fit for fixed-income allocators comfortable with Paxos counterparty risk seeking steady yield; avoid if you need fast exit options or cannot tolerate issuer freeze.
On Aave V3, a supplier can withdraw up to the reserve's available liquidity instantly. Above that, the withdrawal must wait for borrowers to repay or new suppliers to deposit. Markets at 95%+ utilization can keep large redemptions waiting for hours to days at current rates.
Reserve parameters are set by Aave governance (BGD Labs + Aave Chan Initiative + Risk Council). Listings, caps, IRM curves, and LLTVs change through on-chain votes.
This reserve isn't usable as collateral, so LTV / Liquidation LLTV / liquidation bonus don't apply. Borrowers can take it against other collateral; suppliers earn the variable rate driven by borrow demand.
On Aave V3, supply APY = borrow APY × utilization × (1 − reserve factor). There is no curator skim; the only intermediation cost is the reserve factor, which funds the Aave treasury. No emissions slice in v1 of this integration — the headline is fully sustainable.
Aave V3 reserves carry parameter risk (LTV/LLTV adequacy vs collateral volatility), liquidity risk (high utilization slows exit), and depeg risk on the underlying. Governance can pause / freeze / adjust caps in response to incidents. Sharpe's full risk decomposition for Aave reserves rolls in v1.1.
How the composite risk score breaks down. Every number traces to an explicit input — /methodology documents each factor's formula.
Per-collateral decomposition: Sharpe-style realized σ from 30 daily log returns, blended 50/50 with the asset-taxonomy tier fallback. Phase 2 will replace the tier slot with liquidity depth, mechanism classification, and holder concentration. Methodology in /methodology.
The honest version. Every structural failure mode this vault is exposed to, ranked by severity. If you want to know whether to invest, start here.
USDG is a centrally-issued asset (Paxos). The issuer can freeze specific addresses (including aToken holders) and may comply with regulatory demands that affect this reserve. Aave is the secondary-market exposure layer; the primary obligation sits with Paxos.
Aave V3 prices and liquidations route through Chainlink's aggregator feed for this asset. A stale, manipulated, or wrong-side update is the protocol's largest single-point-of-failure for this reserve. Chainlink has a strong track record but the failure mode is binary.