S&P sets out how it will grade the $10 billion on-chain lending vault market

Ten billion over one and a half is a multiple of about 6.7 ^[16], or roughly 158% a year compounded over the two years to September 2026 ^[18]. In dollars, about $8.5 billion of new deposits ^[17] went into these vaults before any S&P assessment existed ^[1].

The grade is S&P's opinion on the relative chance that an investor's position in a vault becomes impaired ^[6]. Liquidity is part of that, including whether a vault could struggle to meet withdrawals when its assets cannot be converted or recovered fast enough ^[15]. The size of a vault's return is not ^[10]. James Wiemken, head of Global Ratings Services, cited the complexity of vaults and their "varying disclosure standards" in explaining the framework ^[14].

The two published accounts of the method differ. Crypto.news lists six risk areas ^[3]. Crypto Briefing describes four main factors: portfolio quality, liquidity pressures, protocol risks and curator oversight ^[4]. The gap is blockchain risk and vault security and governance. For a vault whose weak point is its chain or the control of its contracts, that gap decides whether the weakness gets its own line in the score.

The clearest effect is on how vaults get built. S&P treats allocation caps written into smart contracts as strong risk limits, according to Crypto Briefing, because a coded cap limits how much capital can reach a given market whatever anyone decides in the moment ^[11]. Some vaults run fully through smart contracts. Others give human managers discretion over part of the strategy ^[8]. The discretionary curator has the most to gain on the score from coding caps, and the most flexibility to give up.

The first grades could land three ways. They could bunch near AAA(v), the low-risk end of the scale ^[9], and tell depositors little they could not work out from a vault's collateral. They could spread across the letters and give money a reason to move between vaults. Or allocators whose mandates are written around credit ratings could set the scale aside, since S&P says it is not one ^[2].

I'd expect the method to change how vaults are built before it changes where deposits go, because curators can read the criteria now and allocators have no grade to act on. Crypto Briefing goes further, writing that the evaluation principles are likely to become benchmarks for other financial instruments using DeFi assets ^[19]. That outcome depends on grades existing. Neither report says who pays for an assessment, the vault or its depositors. The view is wrong if the first grades arrive and deposits move toward graded vaults before curators have rewritten their contracts.