Euler (EUL)

Overview

Euler is a permissionless lending protocol on Ethereum, governed by the Euler DAO through its EUL token. Euler v1 launched in December 2021 as a reactive-interest-rate money market that let anyone list a lending market for almost any ERC-20 asset without a central listing gate – a contrast to the curated asset lists of Aave and Compound.

After a $197 million exploit in March 2023 – the largest DeFi hack of that year – and the subsequent return of the stolen funds, the team rebuilt the protocol from scratch. Euler v2 shipped in September 2024 as a modular credit layer: instead of one monolithic pool, it is a toolkit for deploying and connecting independent lending vaults. That architecture is what makes Euler structurally distinct from every other lending DAO in this directory.

The 2023 exploit and the recovery

On 13 March 2023 an attacker drained roughly $197 million from Euler v1 in a flash-loan attack that exploited a missing health-check in the protocol's donateToReserves function, letting the position be pushed into a state where the liquidation logic mispriced the collateral. Chainalysis documented it as the largest DeFi exploit of 2023.

What followed is a widely cited case study in on-chain negotiation. Over roughly three weeks the Euler team combined public appeals, a bounty offer, and back-channel messages sent as on-chain transactions; the attacker ultimately returned effectively all recoverable funds, and Euler used the reclaimed assets to make depositors whole. The team's own post-mortem recounts the recovery in detail. The episode is a useful counterpoint to the assumption that a "DAO" resolves a crisis by vote: the recovery was coordinated by the core team, with governance ratifying the redistribution afterward – the pattern documented on how DAOs fail, where decisive crisis response tends to route through a small core rather than a token vote.

Euler v2: modular vaults and the EVK

Euler v2 is not a single lending contract. It is a system of ERC-4626 vaults built with the Euler Vault Kit (EVK) – a vault-development toolkit that, per the EVK white paper, is deliberately agnostic about governance, upgradability, and oracle choice. Anyone can deploy a lending vault permissionlessly and configure its own collateral set, interest-rate model, oracle, and liquidation parameters.

Each vault is an isolated lending market: lenders deposit an asset, borrowers post collateral held in other vaults, and the vault's parameters bound the risk. Because vault creation is open, Euler v2 supports a long tail of markets – from blue-chip pairs to exotic or real-world-asset collateral – that a monolithic, DAO-curated pool would never list. Two products sit on top of the kit: EulerEarn, ERC-4626 yield aggregators that route deposits across vaults, and EulerSwap, an AMM that borrows against vault liquidity.

The Ethereum Vault Connector

Isolated vaults would fragment liquidity if they could not share collateral. The Ethereum Vault Connector (EVC) is the interoperability primitive that solves this: it lets a vault recognise a deposit sitting in a different vault as collateral for a loan, and it batches multi-step operations (deposit, enable-collateral, borrow) into a single atomic call while deferring the account health check to the end.

The EVC is an open-source standard, not Euler-specific plumbing – other protocols can build on it. In practice it turns Euler v2 into a network of composable credit markets: vault creators chain vaults together so that liquidity deposited in one market can back borrowing in another, recovering the capital efficiency that pure market isolation would otherwise sacrifice. This is the same isolation-vs-composability tension explored on the Morpho page, reached by a different design.

Governed, ungoverned, and escrow vaults

The EVK draws a sharp line between vaults that a manager can change and vaults whose rules are frozen. A governed vault hands a DAO, a professional risk manager, or an individual ongoing authority over its parameters – loan-to-value ratios, interest-rate model, accepted collateral – so passive lenders are effectively trusting that governor. An ungoverned (finalised) vault has fixed parameters and no active manager; its two variants, 0x and nzx, differ only in whether the collateral it accepts itself carries any governance exposure. Escrow vaults hold deposits usable only as collateral, earn no interest, and are ungoverned by construction.

This is the heart of Euler's governance model: the protocol base layer is immutable, and risk curation is pushed to the vault level. A depositor chooses their trust assumption by choosing a vault, rather than inheriting one global risk surface set by a single DAO.

What the Euler DAO controls, and what it gave up in 2026

Because the base protocol is immutable and anyone can spin up an ungoverned vault, the Euler DAO never sat astride a single monolithic risk surface the way an Aave-style money-market DAO does. EUL-weighted votes, run through the Euler Improvement Proposal (EIP) process, directed the protocol treasury and the fee switch; which official governed vaults were curated and surfaced in the front end; grants and incentive programmes; and the parameters of the governed vaults the DAO itself operated. The permissionless long tail of ungoverned vaults was outside the DAO's reach by design.

In April 2026 that remit was cut back to the treasury. Euler Labs proposed sunsetting all directly managed markets and vaults, arguing that running markets is "an operational activity that sits outside this core competency" and that outsourcing risk curation to providers who simultaneously operate competing vaults elsewhere makes a conflict of interest unavoidable. The handover was carried out, not merely debated: on 28 April 2026 Gauntlet posted for the public record that curation of Euler Base and Euler Yield had transferred, with Euler Unichain, Linea, Arbitrum, Arbitrum Theo and Plasma in a two-day timelock pending transfer. The next day Euler Labs asked the DAO to reduce protocol fees to zero across all live and future deployments, leaving Fee Flow paused.

The voting surface tells the same story from the other side. The DAO's Snapshot space records no proposal after 26 August 2025 (EIP65, activating fees on Euler Lending), so the two decisions that redefined what the DAO does were carried through the forum rather than the ballot. Euler is a working protocol run by an active team, and the point for a directory of DAOs is not decline but scope: a governance body can be narrowed back to a treasury without ever being wound up, and the change is legible only in the record, never in the description. Compare SafeDAO, which reached a similar place by pausing its own resource allocation, and progressive decentralization read in reverse.

How Caper approaches this

Euler's control token, EUL, is freely tradeable: voting power over the governed core and the treasury is capital that can be bought on the open market. That is the standard DeFi arrangement, and it is the thing a caper deliberately does not replicate. In Caper, a member's influence is set by an earned vote weight that folds together how much of the token they hold and a soulbound, non-transferable proof-of-vote balance minted on one surface only – ranked ballots cast, with a buy, a sell or a transfer minting none – so that balance can be earned by anyone who votes, but never bought and never moved between wallets.

The same weight drives the exit right: a member redeems a share of the treasury proportional to that weight and burns their vote balance on the way out. A large bag still matters – t is a multiplier, so holdings are not irrelevant – but a bag alone cannot capture control the way accumulating EUL can, and the treasury claim is a first-class protocol primitive rather than something a governed vote has to be trusted to honour. Where Euler curates trust at the vault level, a caper builds the accountability into the weight itself.