Thank you to everyone who has contributed to this thread. We sympathise with everyone affected by the Stream, Elixir and Stables Labs insolvencies.
We welcome a governance process that gives EUL holders a direct say on what should happen next. Euler Labs will not participate in the vote to avoid any conflicts of interest. We will continue to support the community’s decision-making and share factual updates as they emerge.
On the substance, we want to offer a few points for consideration.
It is important to stress that the protocol operated as designed throughout. Euler is a permissionless protocol and, by design, does not pre-screen economic risk or make judgements about the creditworthiness of assets at the protocol level. Those assessments are made by independent risk curators and market creators who choose which assets to support and how to configure them. Smart contracts can mitigate technical and on-chain risks, but they cannot prevent or remedy off-chain issuer misconduct; no lending protocol design can eliminate that category of risk. Despite claims circulating on social media, there were no protocol-level issues with liquidations or oracles, and Euler DAO had no ability to pause or modify vaults that were risk-curated and governed by others.
The pause mechanism some people have referenced applies only at the protocol level and can be triggered by security partners in response to a potential Euler smart contract vulnerability. It cannot be used to pause or intervene in individual vaults or markets, nor is it intended to mitigate economic risks arising from specific collateral assets; that responsibility sits with vault risk curators and governors. See docs here.
There have also been some misconceptions about vault verification. This process is intended to catch configuration errors that would make a vault incompatible within the Euler system and pose security risks. It is not part of the risk-curation process, and none of the recent issues stemmed from misconfiguration. In order to help protect new users from inadvertently depositing into vaults affected by insolvency, some vaults were temporarily unverified after the fact to remove them from the default UI display. Affected users could still see these vaults in their own interface, and anyone can view them at any time by toggling the verified filter in user settings. See docs here.
Curators on Euler are free to select assets, parameters, oracle sources and collateral models. See docs here. This is standard across permissionless lending protocols including Euler, Morpho, Silo and others. Those choices can and should be debated, but it is important to recognise that a major factor in recent events appears to have been the conduct of the asset issuers. These were not obscure anonymous projects; they were well-known teams backed by major industry investors. In several cases, curators themselves suffered losses in their own markets, which indicates that they believed the collateral assets they listed were legitimate and appear to have at the very least acted in good faith. The collapse in the issuers’ asset values, coupled with their sudden disappearance and refusal to engage, removed any possibility of remediation at source. Curators have been attempting to contact the issuers to understand the status of any bankruptcy proceedings and how affected users might eventually be able to seek redemptions. Euler Labs stands ready to support curators in that process, help relay updates to the community and assist wherever we can.
The proposal under discussion would require the DAO to underwrite user losses arising from third-party actions. Doing so would set a precedent for all current and future markets and would not be compatible with permissionless credit infrastructure. DeFi needs secure, neutral platforms on which to build credit markets, but those markets only work when lenders are compensated for taking risk. If losses were implicitly guaranteed by the underlying infrastructure, incentives would break down and both curators and capital would gravitate towards the riskiest assets, creating dynamics that no permissionless credit system can sustainably support.
We hope these points are helpful as the community continues the discussion and prepares for a formal vote, and we look forward to supporting whatever outcome the community ultimately decides is in the best interests of the protocol.