Invest1 publisherNot yet confirmed elsewhere2 min readPublished
S&P writes a risk scale for crypto lending vaults that now hold $10 billion
S&P Global launched a framework on Oct. 4 to grade crypto lending vaults, a market it says grew from $1.5 billion to about $10 billion in two years. The same day, an incident on Base pulled about $6 million out of a vault without exploiting Aave, the vault-level risk S&P now scores apart from the protocol.
The Investor · Invest desk
Drafted by a language model from the sources cited here and checked against its claim ledger before publication. How we use AISend a correction
What happened
- S&P said lending vaults can replicate the functions of money-market, private-credit, private-equity and hedge funds while running directly on blockchains.
- S&P's grades carry a (v) suffix, with AAA(v) the lowest relative risk, and are not credit ratings, do not assess expected yields and do not guarantee recovery of capital.
- S&P has not yet published grades for individual vaults and said those will follow in future announcements.
Compiled by The InvestorSomething wrong?How this is made
Why it matters
- exposure Curators running vaults on the same protocol can now be ranked against each other, so a manager's allocation and permission choices get a grade of their own.
- precedent A clean onchain history will not protect a vault's grade, because S&P can revise it when contract features or liquidity conditions change, before any loss occurs.
- decision Until vault-level grades appear, institutions putting money into a market compounding at roughly 158% a year still have to vet curators and permissions on their own.
Aave returned about 1,783 wstETH for about 1,783 aBaswstETH, one for one, as a lending protocol should [4]. The money had moved one step earlier. According to CertiK, the proxy it flagged borrowed those receipt tokens, worth roughly $6 million, out of an unnamed vault on Base [3][4]. A depositor whose checks stopped at Aave would have found nothing to flag, because Aave itself was not exploited [5].
S&P's framework separates the two layers. Of its six scoring areas, only one is protocol risk [17]. The others include curator risk, liquidity mismatch, and vault security and governance [6]. Vaults pool deposits and allocate them by strategy, through smart contracts or through human managers called curators, and the depositor holds a token claim on the pool [11]. A newly deployed proxy that can borrow a vault's holdings raises a permissions question, and it would most plausibly be scored under security and governance.
For scale, $6 million against about $10 billion of deposits is 0.06% [15]. S&P puts deposits at $1.5 billion two years ago and about $10 billion in September [2]. The difference is $8.5 billion of new money [16], a multiple of about 6.7 [13], or roughly 158% a year compounded [14]. Onchain records show where that money sits at any given moment. S&P says its assessment is built to examine how a vault could behave as conditions change [12].
The most interesting term is the one S&P left out. The grades rank the chance of impairment and do not assess expected yield [7]. S&P has taken on the job of ranking the risk and left the price of that risk to whoever allocates. CryptoSlate argues investors could demand higher returns from weaker-rated vaults or steer deposits toward stronger ones [18].
This can go more than one way. Allocators could treat a minimum grade as an entry filter, so that deposits concentrate in the top-graded vaults and the rest have to pay more to attract capital. The grades could instead cluster by protocol, in which case S&P will have put protocol risk under a new suffix. Or the grades could arrive after the market has moved on, since S&P has published the method but no individual vault's grade [9].
I'd expect the first grades to separate vaults built on the same protocol, because five of the six scored areas sit above the protocol [17]. If vaults on the same protocol come back within a notch of each other, that view is wrong and the vault layer is not being priced on its own.
What to watch
- S&P's first published grades for individual vaults, and whether vaults on the same protocol land far apart or within a notch of each other.
- Identification of the Base vault and its curator, and whether depositors recover the roughly $6 million.
- Whether vault offerings begin quoting S&P grades beside their yields, and whether lower-graded vaults start paying more for deposits.