Kairos Research supports this proposal. The analysis is thorough and the core framework is sound, a dynamic cooldown tied to Tier 1 coverage is the right approach, and the data clearly shows that any static period (whether 3 or 7 days) would be inefficient on a large fraction of days.
A few points worth flagging:
The revised USDtb classification is the right call. Reclassifying the 99% BUIDL-backed portion to Tier 2 makes sense given the banking hours constraint on replenishment. The atomic swap pathway is fast but circular for USDtb already held as USDe collateral as it routes through the same stablecoin pool that sUSDe unstakers draw from. Only the ~$10M Anchorage fiat buffer is genuinely independent 24/7 liquidity.
Competing demands on Tier 1 deserve monitoring. sUSDe unstaking and USDtb atomic swaps both draw from the same liquid stablecoin pool with no replenishment mechanism outside US banking hours. The framework should account for these overlapping claims when calibrating the trigger thresholds.
We would also suggest exploring an express unstake mechanism as a complement. A fee-based option for users who need faster liquidity during extended cooldown periods would generate revenue during stress, provide a real-time signal of redemption urgency, and reduce queue pressure for standard-path users without weakening the cooldown framework itself.
Overall Kairos Research is supportive and we look forward to continued monitoring of the protocol’s liquidity. Thanks to our colleagues from Blockworks and Llama Risk on all the work here!