sGHO — Risk Report
Moderate risk · 5.5/10
sGHO is Aave’s ERC-4626 savings vault over GHO, at 0xE1753F2e00940cC31213dd92013cF019DFE4ca1d. You deposit GHO, you receive sGHO, and the share price accrues.
⚠️ It holds 161,905,236.04 GHO against a GHO supply of 699,000,000.00 — 23.16%. That makes it the largest single destination for the stablecoin, not a peripheral product, and it means a reader looking at GHO’s supply is looking at a number that is nearly a quarter this one vault.
| Yield | Underlying | Redemption | Assets vs shares | Chain |
|---|---|---|---|---|
| Accrues in share price | GHO | Instant, ERC-4626 — no cooldown, lockup or queue | 161.9M GHO against 159.9M sGHO | Ethereum |
⚠️ There is an older sGHO, and it is not this contract
A legacy sGHO deployment exists, and reporting indicates holders were required to migrate off it.
⚠️ Its address has not been located, so this report cannot tell you what it is — only that it is not the contract above. Ticker resolution can land on the wrong one: a search, an aggregator listing, or a wallet’s token list may resolve “sGHO” to the legacy deployment.
The contract this report covers is 0xE1753F2e00940cC31213dd92013cF019DFE4ca1d, and that address — not the ticker — is what identifies it. Check the address before depositing anything.
The wrapper is good, and that is the finding
⚠️ This report scores 5.5, the same as GHO — deliberately equal, not lower. It is worth saying why, because a wrapper scoring level with its underlying is unusual on this site.
Three things are genuinely strong here:
- Redemption is instant. Standard ERC-4626 withdrawal, no cooldown, no lockup, no queue. For a savings wrapper that is the whole question, and this one answers it well.
- Assets exceed shares — 161.9M GHO against 159.9M sGHO. The vault holds GHO, not a strategy book, so what backs a share is the underlying itself rather than positions taken with it.
- Governance is mutually locked. Aave Governance v3’s Executor (
0x5300a1a1…192a) and PayloadsController (0xdabad81a…aec5) own each other, so neither can be changed except through a passed payload. ⚠️ No key set can act unilaterally — which is a materially different posture from a multisig that merely has a threshold.
Compare sfrxUSD, scored the same day at 5.0 against frxUSD’s 5.5. Same wrapper shape — ERC-4626 over a stablecoin, holding a large share of its supply — and the opposite admin quality: there, an accessor named timelockAddress() resolved to a 3-of-6 Safe with no execution delay at all. The wrapper was what dragged that score down. Here the wrapper is not the constraint.
What you actually inherit
A savings wrapper improves none of the underlying’s problems, and GHO’s are open. Holding sGHO means holding GHO with a yield accrual on top, so its findings pass through undiluted:
- 44.3% of GHO supply is minted by a role-gated contract whose authorised callers cannot be enumerated on-chain, and whose “GSM” label describes machinery it does not implement.
- Supply has been exactly flat for sixteen days, with 65.8% sitting at hard caps.
- Both mainnet GHO Stability Modules are seized — verified 2026-08-13, unchanged 2026-08-29 — and Curve’s GHO PegKeeper debt ceiling is 0, so the peg’s defence mechanism is wound down.
That is why the score lands at 5.5 and not higher. The vault is well built; what it holds is the limiting factor.
Secondary depth — an LP finding, not a peg signal
The frxUSD/sGHO Curve pool is 68.4% sGHO against 31.6% frxUSD and holds roughly 0.29% of supply, about $2.0M.
⚠️ Read that as a warning to liquidity providers, not as a peg signal. A stableswap pool accumulates whatever is sold into it, so an LP in that pool is long the tilted leg and gets longer as the tilt worsens — an exposure that does not appear in a spot quote and is not visible to someone checking the price.
⚠️ And do not read the thin pool as an exit problem. Primary redemption is open and instant, so the front door is unlocked regardless of what secondary depth looks like. The pool matters for people supplying it, not for people leaving.
Three things not established
⚠️ Each is a gap in what could be read, not a property of the system.
- The governance execution delay. Every delay accessor reverts. The 7-day
GRACE_PERIOD()is the execution window, not the delay before execution. “The delay could not be read” is supported; “there is no delay” is not, and this report does not claim it. - The rate. The roughly 4.25% figure in circulation is press-reported. Every rate accessor reverts, so it is not presented here as measured.
- The legacy contract’s address, as above.
Who should avoid this
- Anyone who has not checked the contract address. A legacy sGHO exists and the ticker does not distinguish them.
- Anyone who would not hold GHO itself. This is GHO with accrual; none of its open findings are improved by the wrapper.
- Anyone providing liquidity to the frxUSD/sGHO pool without understanding the tilt. You are long the heavier leg and get longer as it worsens.
What to watch
- Everything on the GHO page. The underlying is the constraint here, so its re-score trigger is effectively this report’s too.
- Whether the governance delay ever becomes readable, which would close the largest open question about this vault’s own admin.
- The vault’s share of GHO supply. 23.16% today — a rising share makes the two assets one risk rather than two.
- The Curve pool’s tilt, if you are supplying it.
Revision history: 2026-08-29 — first publication, staged. sGHO holds 161,905,236.04 GHO — 23.16% of supply — making it the largest single destination for the stablecoin. ⚠️ Scored 5.5, deliberately EQUAL to GHO rather than below it, because the wrapper is not what limits it. Redemption is instant ERC-4626 with no cooldown or queue; assets exceed shares (161.9M against 159.9M), so the vault holds GHO rather than a strategy book; and Aave Governance v3’s Executor and PayloadsController own each other, so no key set can act unilaterally. The contrast that makes the equal score legible: sfrxUSD was scored 5.0 against frxUSD’s 5.5 the same day — same wrapper shape, opposite admin quality, since its timelockAddress() resolved to a 3-of-6 Safe with no delay. What a holder inherits is GHO’s open findings, none of which a savings wrapper improves. ⚠️ The Curve pool is written as an LP finding rather than a peg signal: at 68.4% sGHO it makes a provider long the tilted leg and longer as it worsens, invisible from a spot quote — but thin secondary depth is not a gate here, because primary redemption is open and instant. ⚠️ Three items are recorded as not established, each a reading limit rather than a system property: the governance execution delay (all delay accessors revert; the 7-day GRACE_PERIOD() is the execution window, so “unread” is supported and “none” is not), the ~4.25% rate (press-reported, every rate accessor reverts), and the address of a legacy sGHO contract that is not this one — carried prominently because ticker resolution can land on the wrong deployment.