Skip to content
EthCC 9 Cannes 2026

Andrew O'Neill (S&P Global Ratings) - Bringing traditional risk analysis to DeFi

Andrew O'Neill

Session summary

Andrew O’Neill from S&P Global Ratings explains how his team adapts traditional credit-style risk analysis to DeFi, focusing on Sky Protocol (formerly MakerDAO) as a case study. He frames USDS holders as economic creditors of the protocol and defines “default” as a haircut on USDS when liquidity or capital are insufficient to support redemptions into USDC. O’Neill states that S&P’s B- credit rating on Sky reflects high vulnerability, emphasizing that Sky’s own internal model targets a substantially higher capital buffer than it currently holds while its “endgame” capital and governance overhaul remains incomplete.

O’Neill then broadens to systemic issues at the intersection of permissionless design, liquidity, and off-chain exposure. He describes how large, fast inflows such as Athena’s multi‑billion dollar deposit into Sky can create depositor concentration risk, and uses the Resolve hack and related Morpho vault losses to illustrate the dangers of thinly traded collateral backing large loans. He concludes that tokenized real‑world assets add layers of legal, operational, and liquidity complexity, arguing that explicit risk frameworks are essential if permissionless architectures are to scale without recurring crises.

Key takeaways

  • O’Neill says Sky’s own framework targets about 125 million USDS in capital reserves, versus roughly 42 million at the time of his analysis, and that this shortfall is a central reason S&P assigned the protocol a B- credit rating indicating material vulnerability to stress events.
  • He highlights that Sky’s new “stars/prime agent” subsidiaries are supposed to hold risk-based capital, but notes that sizing and verification of those buffers are entirely off-chain and dependent on human risk teams, with no automated mechanism to resize capital as lending volumes or risk intensities change.
  • Using the Resolve exploit and subsequent Morpho vault losses, O’Neill argues that allowing a token with only tens of thousands of dollars in DEX liquidity to serve as multi‑million‑dollar collateral creates arbitrage opportunities when oracle prices lag market prices, leaving vault depositors exposed to likely losses when such collateral collapses in value.