Compound debates Institutional Comet governance after proposal cancellations

The Institutional Comet opened to deposits on Sept. 8. Since then, a dispute over cancelled governance proposals has started complicating the transition plan meant to bring the market under COMP-holder control. The transition-plan proposal on comp.xyz describes the USDC market on Ethereum as a V4 R&D deployment and calls for it to come under the ultimate authority of Compound governance.

The stakes for current depositors are real. As of Oct. 8, roughly $20.6 million USDC was supplied against $36.2 million in collateral, with $16.3 million borrowed — a utilisation rate of about 79%, which matches the proposal’s own figure after rounding.

The Compound Foundation, Woof and the Compound Governance Working Group are developing the plan, with the technical design for an optimistic governance council being finalised alongside ScopeLift. A full specification and timeline are due before the end of October.

The proposed council would draw its initial membership from the entities that already sign for the market’s proxy owner: the Foundation, Woof, the CGWG, Gauntlet, ZeroShadow and ChainSecurity. COMP holders would keep final authority — able to veto any council proposal and to change council membership or settings through Compound governance.

The case for this structure is straightforward: routine administration shouldn’t have to wait through a seven-day governance cycle when something breaks. The proposal points to a Kelp DAO bridge exploit and an rsETH collateral issue on April 18 as evidence of what delay costs. The Pause Guardian paused affected mainnet Comets that day, freezing withdrawals for seven days on the USDC Comet and 16 to 20 days on the WETH and wstETH Comets, until governance could set rsETH parameters to zero.

The Institutional Comet is currently administered by a 3-of-4 governor multisig, a 4-of-7 proxy-owner multisig and the 5-of-9 Compound Pause Guardian. The proposed structure would change that arrangement while preserving the COMP-holder veto — but the council’s voting period, quorum, veto threshold, veto period and timelock delay are not yet specified in the available material.

Then came the cancellations.

Proposal 612 was queued on Oct. 7 and cancelled by the Proposal Guardian less than nine hours later, according to a comp.xyz post challenging the cancellation. That proposal would have raised the Treasury Escrow cooldown from two days to 10, scheduled the same change for the Treasury Timelock, and granted the Governor both EXECUTOR and CANCELLER roles.

“The Proposal Guardian cancelled Proposal 612 less than nine hours after it was queued. I think that was unjust, for four reasons,” Gm everyone wrote on comp.xyz. The same post asked: “The Governor owns these contracts and the Guardian loses nothing. What was the cancellation protecting?”

Community Multisig signers laid out the Guardian’s rationale in a separate comp.xyz account of the cancellations. Their position: Proposals 612 and 613 each met two of the cancellation criteria that delegates established in Proposal 304 and extended through Proposal 538 — a single entity controlling the vote, and the proposals passing over substantial opposition.

The numbers they cited were specific. Humpy and directly tied delegates supplied the significant “For” votes, while the dCOMP delegate’s 1.75 million COMP was the deciding vote. More than 1 million COMP from 19 voters opposed both proposals, clearing the 400,000 “NO” threshold by a wide margin. “The Guardian’s mandate requires only one of its criteria to be met; both proposals meet two,” the signers wrote.

Those signers — ChainSecurity, Certora, Woof, PGov, cylon, Gauntlet, 0age, arr00 and zeroShadow — overlap significantly with the proposed optimistic council’s initial membership. ChainSecurity, Woof, Gauntlet and ZeroShadow appear on both lists.

That overlap is what gives the dispute its edge. The transition specification — council structure, quorum rules, veto mechanics — remains unfinished while the question of when cancellation is appropriate is still openly contested.