We agree that minimizing rollover friction is important; however, sizing the initial May-2026 PT caps based on the outstanding February-2026 PT supply is not feasible to maintain a consistent risk profile. The two instruments do not share the same risk profile, and as such, should not be considered equivalent when determining initial caps. The near-expiry PTs have reached the low-duration regime where convergence to par is mechanically tighter, the remaining downside to boundary conditions is smaller, and liquidation feasibility improves as maturity approaches. A longer-dated PT sits in the higher-convexity part of its lifecycle: implied-yield repricing can generate meaningfully larger mark-to-market moves over a longer window, and the system must remain robust to larger tail moves before the PT Linear Risk Oracle naturally becomes more deterministic closer to expiry. This is why we quantify risk as a function of time-to-maturity and implied-rate volatility, and why the initial configuration must be anchored to the higher-risk regime rather than assuming “rollover size” is a proxy for “safe exposure”.
The liquidity picture of these assets reinforces this constraint. The outgoing sUSDe Feb-2026 market has materially deeper liquidity (roughly $37.6M) than the new sUSDe May-2026 market (roughly $7.5M). When the new expiry has significantly less executable depth, raising caps increases the probability that Aave accumulates an exposure that cannot be unwound efficiently under stress, precisely when duration risk is higher, and the market is more sensitive to repricing. For that reason, we set initial caps and parameters to reflect the new PTs’ longer-dated risk and current liquidation capacity, then scale caps as the PT risk profile decays over time and the liquidity is migrated from the prior expiries.
Finally, the migration constraint does not imply that users are forced to exit, idle capital, or stop earning while caps are full. Because USDe and sUSDe are part of the PT E-Mode collateral set, users who cannot immediately migrate to the new PT maturity redeem the expiring PT into their underlying and maintain an open, yield-bearing position within the same correlated collateral environment, as observed in the plots below. This provides continuity of exposure and reduces the practical risk of TVL leakage during the transition window, while allowing caps for the new maturity to be increased in a manner consistent with the quantified risk decay over time and demonstrated depth in the relevant Pendle markets.
However, given that the AIP has not yet been launched over the last week, the May-2026 PT has mechanically moved further down its lifecycle and the associated duration risk has materially decayed. On our time-to-maturity framework, this translates into an approximately ~10% reduction in duration-driven mark-to-market sensitivity versus the point at which the initial parameters were set. This incremental decay supports cap increases of $30M and $15M for PT-sUSDe-7MAY2026 and PT-USDe-7MAY2026, respectively, relative to the previously parameterized value, while remaining consistent with the principle that caps should scale with quantified risk decay (and not simply mirror rollover sizing).
Specification
| Asset | Previous Recommended Supply Cap | Current Recommended Supply Cap |
|---|---|---|
| PT-USDe-7MAY2026 | 30,000,000 | 45,000,000 |
| PT-sUSDe-7MAY2026 | 70,000,000 | 100,000,000 |
Disclaimer
Chaos Labs has not been compensated by any third party for publishing this recommendation.
Copyright
Copyright and related rights waived via CC0