GHO is a decentralized, overcollateralized stablecoin native to the Aave protocol: borrowers mint it against collateral supplied to Aave V3, and it holds a US-dollar peg backed by that collateral rather than by fiat reserves. What makes GHO worth cataloguing here is not the peg mechanics but the control surface: nearly every lever that governs GHO — the interest rate borrowers pay, how much any given contract may mint, the discount stakers receive, which assets can defend the peg — is a live decision of the Aave DAO. That places GHO at the maximally-governed pole of the DAO governance spectrum — the exact opposite of Liquity’s ungovernable BOLD, and a clean case study in treating governance itself as monetary policy.

The facilitator model

GHO is not minted by a single vault. Instead, Aave governance approves facilitators — contracts that can trustlessly mint and burn GHO up to a governance-set bucketCapacity. The original facilitators were the Aave V3 Ethereum Market (where users borrow GHO against supplied collateral) and a FlashMinter; later additions include the GHO Stability Module. Each facilitator’s bucket is a hard, governance-owned ceiling on how much GHO it can bring into existence, so the total supply is bounded by an explicit ledger of votes rather than by market demand alone.

This is a deliberately different answer to “who decides how much stablecoin exists?” than an algorithmic issuer gives. Where Liquity lets any borrower mint against ETH under fixed immutable rules, GHO’s supply is gated by a governance-maintained set of facilitators and caps that can be raised, lowered, or revoked by proposal.

An interest rate set by vote

The defining feature: GHO’s borrow rate is a fixed number chosen by Aave governance, not a curve that floats with utilization. Holders vote a rate based on the peg and market conditions, and the protocol charges it uniformly. To make routine adjustments without a full multi-day vote each time, governance delegates a bounded lever to GHO Stewards, who can change the borrow rate by at most 500 bps per two-day period, up to a 25% APR ceiling, and raise the borrow cap toward a 100M threshold — a guarded, rate-limited delegation rather than open discretion.

Layered on top is a staker discount: users who stake AAVE in the Safety Module (stkAAVE) receive a reduction on their GHO borrow rate, applied per staked token up to a per-token GHO cap. The discount rate is itself a governance parameter. The rate a borrower actually pays is therefore a function of three governed inputs — the base rate, their stkAAVE balance, and the per-stkAAVE cap — a striking contrast to a purely mechanical fee.

The GHO Stability Module and the governed peg

When the peg drifts, GHO does not rely solely on redemption arbitrage. The GHO Stability Module (GSM) lets users swap GHO 1:1 for governance-approved tokens (such as major fiat-backed stablecoins), with exposure caps, dynamic fee strategies, and oracle-based price-freeze protections — all governance-configured. If GHO trades above peg, minting and selling into the GSM pushes it down; below peg, buying and burning pulls it up. Because which assets back the GSM and at what caps are votes, the peg backstop is a governance instrument, not a fixed algorithm.

Aave has continued to extend this surface: 2025 brought stkGHO, anti-GHO, and a savings variant (sGHO) as governance-shaped ways to hold and stabilize GHO, each adding another parameter set the DAO maintains.

Where it sits on the governance spectrum

GHO and Liquity’s BOLD are the two poles of the same design question, and reading them together is the point. Liquity protects a stablecoin by making its rules immutable — no vote can touch the borrow rate or the peg logic, so there is nothing to capture. GHO protects a stablecoin by making its rules maximally responsive — the DAO can retune the rate, the mint caps, and the peg backstop as markets move. Each buys something and pays for it: immutability buys credible neutrality at the cost of adaptability; governance buys adaptability at the cost of a standing governance attack surface and a dependence on votes turning out. The post-vote-escrow turn elsewhere in this directory is a third answer again — retiring a governance lever once it has served its purpose.

How Caper approaches this

Caper is a treasury-and-governance protocol, not a stablecoin issuer, so the honest contrast is about who holds the parameters and whether a dissenter has an exit. On Aave, a borrower or AAVE holder who dislikes a governance-set GHO rate has no structural recourse beyond selling their position — the vote binds everyone who stays. In a caper, the treasury is the governed object, and two things differ. First, a member’s vote weight is (held × votes) / (vote_supply × circulation) — a product of tokens held and a soulbound, non-transferable record of having actually voted, so a large bag alone cannot capture the control surface the way it can accumulate stkAAVE. Second, when you lose a parameter fight you can exit with your share of the treasury rather than only sell into the market — the majority cannot trap you in a policy you fought. (Mechanics verified against compute_vote_weight and exit in contracts/src/caper_dao.rs; framed at the outcome level, no invented features.)