Liquity is a decentralized borrowing protocol on Ethereum whose defining choice is the near-total absence of governance. Where most DAO-adjacent protocols concentrate power in a token vote, Liquity's rules are fixed in immutable smart contracts with no admin key and no way for anyone to alter the system. That makes it the clearest real-world example of the minimal-governance pole on the DAO governance spectrum — a useful foil to the vote-escrow and treasury-heavy models catalogued elsewhere in this wiki, and a live case study in the tradeoff between adaptability and capture-resistance. Its closest sibling is Reflexer’s RAI, which lands at the same immutable, ungoverned core from the opposite direction — deliberately removing governance rather than never granting it.

Governance-free by design

Liquity V1 has no admin key, no upgradability, and no token vote over its parameters. Instead of humans voting on monetary interventions, the protocol adjusts its own borrowing and redemption fees algorithmically: a base rate rises with redemption volume and decays over time, so fees respond to market pressure without a committee. The team's own framing is blunt — the design makes “human interventions redundant”.

Crucially, the LQTY token is not a governance token: it captures protocol fee revenue and rewards Stability Pool depositors and front-end operators, but it grants no vote over the rules (Liquity docs). This separates Liquity sharply from the one-token-one-vote norm — there is simply nothing to vote on, so a large LQTY holder cannot capture the protocol the way a whale can capture a token-weighted DAO.

The mechanism

Liquity V1 issues LUSD, a USD-pegged stablecoin, against ETH collateral with an unusually low 110% minimum collateral ratio and no ongoing interest — borrowers pay a one-time fee (0.5%–5%) rather than a recurring rate. The peg is defended by three algorithmic backstops instead of a governance dial:

• Redemptions at face value. Anyone can always swap LUSD for $1 of underlying ETH, which arbitrages the price back toward the peg (docs).
• The Stability Pool. LUSD deposited into the pool absorbs liquidated collateral, with depositors compensated in ETH plus LQTY — a permissionless, always-on liquidation engine that needs no keeper vote (docs).
• Fellow borrowers as backstop. If the Stability Pool is ever exhausted, debt and collateral are redistributed across remaining borrowers, so the system stays solvent without an emergency governance action.

Liquity V2 (BOLD): the smallest possible governance surface

Liquity V2 launched in 2025 with a new stablecoin, BOLD, multiple collateral types (ETH and LSTs), and user-set interest rates. It keeps the immutability commitment — core protocol parameters remain immutable after launch — but adds one narrowly scoped governance module, and nothing more.

Through Protocol Incentivized Liquidity (PIL), LQTY stakers vote on where to direct roughly 25% of protocol revenue as liquidity incentives. Voting power is time-weighted (longer stakes weigh more), votes run in weekly epochs starting Thursdays 00:00 UTC, proposing a new initiative needs ≥0.01% of voting power plus a 100 BOLD fee, and initiatives below a 2% vote share for four consecutive epochs can be permissionlessly unregistered. This is explicitly a Curve-style incentive market — complete with bribe markets echoing the Convex/Votium vote-buying dynamic — but bolted onto an otherwise ungovernable base layer. The governance module “has no control over any of the core protocol parameters”: it can only decide where the incentive budget flows.

The immutability commitment cuts both ways. When a Stability Pool issue was found after the January 2025 launch, the team could not patch in place — an immutable protocol has no upgrade switch — so V2 had to be redeployed from scratch on 19 May 2025 after further audits. Governance-minimisation buys capture-resistance at the cost of adaptability; there is no committee to blame, and also no committee to fix a bug.

Where it sits on the governance spectrum

Liquity matters to DAO design precisely because it is the extreme case. Most of this wiki's ecosystem entries — Curve, Convex, and the wider vote-escrow cluster — answer “who decides?” by engineering ever-more-elaborate vote-weighting. Liquity answers it by removing the attack surface: an immutable contract cannot be governance-captured, socially engineered into a malicious upgrade, or drained by a hostile proposal, because there is no privileged action to seize. The cost is rigidity — the parameters that suit 2021 are the parameters you keep, and fixing anything means shipping an entirely new protocol and migrating users. It is a deliberate bet that a small, fixed, well-audited rule set beats a flexible one that can be turned against its users. For the opposite pole — a stablecoin whose borrow rate, mint caps, and peg backstop are all live governance votes — see Aave’s GHO.

How Caper approaches this

Caper sits at the opposite pole from Liquity, and the contrast is instructive. Liquity protects a fixed monetary policy by making the protocol ungovernable; Caper makes governance the entire point but ties it to something a whale cannot cheaply buy. In a caper the treasury is the governed object, and a member's vote weight is (held × votes) / (vote_supply × circulation) — a product of tokens held and a soulbound, non-transferable proof-of-participation earned by actually voting over time (verified against compute_vote_weight and the DIVISIBILITY_NONE vote token in contracts/src/caper_dao.rs). The same figure sets each member's pro-rata share of the treasury on exit, so a purely financial position can never dominate a record of showing up.

Both designs reject plutocratic one-token-one-vote — but from opposite directions. Liquity removes the decisions so they cannot be captured; Caper keeps every decision on-chain and anchors it to earned, un-rentable stake, so a fixed rule set isn't needed to stay capture-resistant. Liquity's immutability-vs-adaptability tradeoff is the price of having no governance at all; Caper accepts governance and spends its effort making that governance hard to buy.