At some point, you stop asking whether it’s just another rough cycle.
When enough builders leave, enough projects close, and enough communities are left defending what used to be possible, the feeling changes.
Not anger or panic. Just a quiet loss of conviction.
Darren Camas (@IPOR_io) at the NYSE Vault Summit just laid out the cleanest case we've heard for what actually
unlocks institutional capital on-chain.
verification → risk assessment → underwriting → allocation at scale.
His line says it all: "You can't ensure what you can't risk-rate, and you can't risk-rate what you can't verify."
Risk rating isn't a "nice to have". It's the crucial step between "this vault exists" and "a regulated institution can allocate into it."
That's exactly why every Fusion vault deployment gets risk-rated, and we're proud to be the partner doing it.
Segregated, per-client vaults mean each strategy's risk can be decomposed from its on-chain footprint and underwritten on its own. That's the world risk ratings were built for.
On-chain asset management isn't waiting for permission. It's waiting for infrastructure institutions and regulators trust.
Rating the risk is how that trust gets built.
Had a amazing day @proofoftalk the quality of conversations and the attendees are all top notch. Privileged to attend with my colleagues from @xerberus
We will be around again for the second day, drop by to say Hi.
5/ So we built a POC and wrote a spec. Now we're looking for a team to take it to production and run with it.
Open source, open data layer, named steward, no composite scoring. Just the feeds, side by side.
Applications close June 15. Feel free to reach out or apply directly here: https://t.co/yrkKQfcZiM
YieldNest Max USDC: 0.79/1.00
✅ Why it's strong (the code):
- Clean ERC-4626 mechanics (receipt-token integrity 1.00/1.00)
- BeaconProxy verified
- non-custodial (curator can't touch deposits)
- EVC integration intact
- first-depositor + donation-attack protection
- 90% live utilisation
- LlamaRisk-operated curator.
One of the cleanest vault implementations on @eulerfinance V2.
⚠️ What could still improve (the operating envelope):
• Same multisig governs the supply vault, both collateral vaults, the underlying tokens AND the price oracle.
• Oracle has no fallback, single adapter, governed by the same multisig that sets the 90% LTV
• No pauseGuardian, no hookTarget, no supply or borrow caps → emergency response is fully manual
• Collateral is 365-day Australian private credit with no on-chain redemption until 2026 = liquidators must clear on thin secondary markets
The risk is that one operator holds every key on the loop, and nothing on the risk path fires automatically.
🔁 We corrected our own scorecard this week:
K_oracle_fallback_existence moved YES→NO after we read fallbackOracle() == address(0) on-chain
(dispute approved, confidence 0.92).
That's the standard working.
Full article below.
Full rating → https://t.co/3cT7JX1QzR
Aave v3: 0.82/1.00
✅ Why it's strong (the code):
Governance-gated multi-chain deploys · deep audits + Certora formal verification · snapshot voting · real
supply/borrow caps.
One of the cleanest contract layers in DeFi.
⚠️ What could still improve (the operating envelope):
• Oracle reads raw Chainlink, but no on-chain staleness/deviation guard
• Emergency response is manual, as there are no automated circuit breaker
• Safety Module slashing has never once fired
• 1-token-1-vote, but~33% held by one entity
The risk isn't the code. It's that almost nothing on the risk path fires automatically.
🔁 We corrected our own scorecard this week: 3 findings moved NO→YES (0.79→0.82), a 4th challenge rejected.
That's the standard working.
Full article below.
Full rating → https://t.co/FjHMVRwCT3
Immensely privileged to be representing @xerberus at Proof of Talk next week.
Swing round our booth to chat about how we are changing the game of risk in defi !
If you are a FoF, hedge fund, curator, vault platform or LP- we would love to chat!