Morpho's Frambot proposes two vault categories, Aave's Kulechov calls the split self-serving

Morpho founder proposes two vault categories on X

Morpho co-founder and CEO Paul Frambot published an article on X on Sept. 24 proposing that onchain vaults be split into two regulatory categories. An onchain vault is a product that holds user deposits and puts them to work through crypto lending or trading strategies.

Frambot's post, titled "The Next Phase of Vaults," argues that "very different custody models, permissioning, and depositor guarantees sit under the same word: 'vault.'" He proposed separating "noncustodial vaults" from "discretionary vaults."

Within the hour, Aave founder Stani Kulechov, attorney Gabriel Shapiro and several vault infrastructure founders objected, according to The Defiant.

Key points

• Frambot defined noncustodial vaults as ones where curators cannot increase risk or take control of user funds without the user's explicit or implicit approval. In discretionary vaults, managers have wider latitude over allocation and strategy.
• He placed Morpho Vaults in the noncustodial group, citing timelocks on risk-increasing changes, a guardian role that can cancel pending changes, role-based access controls, in-kind redemptions and immutable contracts. A timelock is a delay before a change takes effect.
• Kulechov replied 36 minutes later that the categorisation "doesn't make sense and is pretty much self-serving."
• Vault builders argued that the safeguards Frambot listed are security features, not proof that a vault is noncustodial.
• Frambot had not responded publicly to Kulechov as of publication.

The test Frambot proposes for classifying a vault

Frambot wrote that "a practical way to classify a vault is to ask what happens if the operator is hacked, disappears, or acts in bad faith. If users can opt out within a reasonable timeframe without relying on the curator, then the vault is noncustodial."

For Morpho Vaults, he pointed to timelocks on risk-increasing changes, a guardian role that can cancel pending changes, role-based access controls, in-kind redemptions and immutable contracts. In-kind redemption means users receive the underlying assets rather than a cash equivalent.

Kulechov calls the split self-serving

Kulechov responded that "this categorisation doesn't make sense and is pretty much self-serving." He argued that calling a vault noncustodial because it has a timelock, when the curator can still expand into markets beyond a depositor's initial mandate, "is about as strong an argument as a wet European paper straw."

On role-based controls, he wrote that relying on roles "simply creates a blame game and relocates potentially liability but doesn't solve the actual problem."

He set a narrower boundary: "Vaults that could reasonably be considered non-custodial are those without a manager. For example, vaults that simply wrap deposits into a lending protocol, or the original Yearn vaults." He added that he saw nothing wrong with discretionary vaults "as long as the regulatory path is figured out," and said: "I'm all for developing industry standards, but let's at least do it in a way that serves the broader industry rather than your own interests."

Gabriel Shapiro, a corporate attorney at MetaLeX Labs and former general counsel of Delphi Labs, said he agreed, asking "why did projects like Yearn bend over backwards to make everything transparent, auditable, and non-discretionary if a literal unregulated hedge fund can get the same treatment just by having the deposit be onchain?"

Reid Yager pushed the other way, writing "maybe we acknowledge non-custodial is a spectrum too; which is how I read the piece," and that dismissing curators on a Yearn purity test was the wrong starting point. Kulechov answered that Yearn need not be the test but that "the proposed setup is too weak."

Vault builders question the safeguards

Sunand Raghupathi, co-founder and CEO of Veda Labs, wrote that the safeguards Frambot listed are security mechanisms rather than proof of noncustodial status.

On timelocks, he wrote that most depositors arrive through fintech front-ends: "A user on Robinhood is not studying the collateral composition." On in-kind redemptions, he wrote that the queue-based design creates an adverse selection problem, in which "the least sophisticated users are the ones left holding that position while the more sophisticated users escape with only exposure to the safe markets." On immutability, he wrote that modular contracts using external adapters "can still change their security assumptions / functionality," and that a bug in core logic becomes unfixable.

An account posting as cryptographic, who lists work at Avon, a credit protocol, wrote that Morpho's own documentation describes the curator configuring "liquidity allocation rules" and "abstracting risk curation decisions away from depositors." The conclusion: "So even if the code constrains the manager, it doesn't remove the manager."

Brian, co-founder of Glider, questioned the premise that institutions are already allocating, citing an unnamed founder whose clients include Fidelity and BlackRock: "Institutions are all excited about vaults, but when you ask them to put money in... whispers."

What Morpho's documentation says

Morpho's Vaults V2 documentation states that the curator sets risk caps, appoints allocators, enables adapters and adjusts timelock durations. It says the sentinel can revoke pending timelocked actions and reduce risk caps but "cannot introduce new risk."

The timelock documentation describes per-function timelocks set by the curator rather than a single vault-wide minimum, while Frambot's post describes "a minimum delay."

The SEC statement both sides cite

Both sides pointed to the same document. SEC Commissioner Hester Peirce's July 22 statement, "Headstands and Summervaults," said vaults "fall along a spectrum from programmatic allocations determined solely by immutable smart contracts, to allocations at the sole discretion of another person or group of persons," and that the term lacks a widely understood definition.

Peirce also pointed to selecting yield-generating activities and reallocating assets as examples of managing a vault, and said doing so can raise investment adviser questions. The Defiant reported at the time that she framed the statement as an invitation to dialogue rather than an enforcement threat.

Frambot first argued the vault thesis in an October 2025 Morpho blog post. He closed Thursday's article by pointing to Vault Summit Singapore, a conference Morpho is co-hosting with DigiFT at the Andaz Singapore on Oct. 9, the day after TOKEN2049.

What is confirmed

• Frambot published the proposal on Sept. 24, and Kulechov, Shapiro and several vault builders responded publicly the same day.
• Kulechov, Shapiro, Raghupathi, the account posting as cryptographic, Reid Yager and Brian are all quoted in the source material making the arguments described above.
• Frambot had not replied publicly to Kulechov by the time the article was published.

What is still unclear

No regulator has said whether Frambot's two-category idea would be accepted, and the source material does not show any formal rulemaking on the proposal. The dispute remains a public argument among industry participants. The claim about what institutions are doing with vaults also rests on an unnamed founder quoted by Brian, not on a named company statement.

Where Aave and Morpho stand

Aave holds $19.1 billion in total value locked (TVL) against $13 billion in active loans, according to DefiLlama. TVL is the total value of assets deposited in a protocol.

Morpho holds $10.8 billion in TVL and $5.28 billion in active loans, up 14% over 30 days against Aave's 6.4%, the same source says.

According to CoinGecko, MORPHO traded up 11% over 24 hours and 27% over seven days, while AAVE was up 3.2% and 15% over the same periods.

Sources

• The Defiant

Read also: Aave Opens Base Market for USDC Loans Against Seven Coinbase Stock Tokens