Chaos Labs Risk Stewards - USDe Interest Rate and Borrow Cap Adjustments - 05.28.25

[ChaosLabs]

Overview

Chaos Labs recommends:

• Modifying the interest rate curve for USDe on Aave V3 to capture additional protocol revenue while keeping PT-USDe and PT-eUSDe looping strategies highly effective and more profitable than loops with USDT, USDC, and USDS.
• Increasing USDe’s borrow cap on Ethereum Core instance. The cap increase is backed by Chaos Labs’ risk simulations, which take into account user behavior, on-chain liquidity, and price impact, ensuring that higher caps do not introduce additional risk to the platform.

The proposal is motivated by the new, more optimal collateral parameters for PT-USDe-31JULY2025, PT-eUSDe-14AUG2025, and USDe E-Modes, made feasible through the Principal Token LT/LTV Risk Oracle dynamic parameterization. At present, most of the PT-token looping positions must borrow USDC, USDT or USDS because Stablecoin E-Mode is effectively the only borrowing market enabled for PT-collateral, with the exceptions being PT-USDe and PT-eUSDe tokens mentioned above. The recommended changes will be performed through the use of Risk Stewards.

Motivation

When PT-USDe and PT-eUSDe assets are leveraged with USDe rather than USDT or USDC, risks are reduced significantly, due to the aforementioned collateral assets converging to USDe at maturity. As a result, bad debt is less likely to occur, and it is possible to recommend higher final collateral parameters to which the PT assets will converge when they are leveraging their underlying asset relative to merely correlated assets. Borrowing USDe directly against PT-USDe collateral aligns the asset of debt and collateral, significantly limiting price-divergence risk in stress events, while the PT Risk Oracle’s automatic market-freeze on extreme PT price moves will make protocol bad debt from PT swings mathematically infeasible.

This is explicitly recognized in previously passed proposals(1, 2), which show that the final LT for USDe E-Mode is higher, and does not max out until 0.075 years to maturity, whereas the Stablecoin E-Mode’s LT stops increasing after 0.14 years to maturity. ******

USDe E-Mode

Stablecoin E-Mode

Higher LTVs and LTs facilitate greater leverage, which in turn reflects in increased earnings in the looping strategies in the E-Modes that leverage the underlying. With USDe E-Mode’s (3, 4) LT for PT-tokens rising from 91% to 96% (rather than a maximum of 93% in the generic Stablecoin E-Mode) over time, potential USDe leverage will increase up to 25x, higher than the 14.3x that can be achieved in the Stablecoin E-Mode, enabling USDe-leveraged positions to effectively capture a higher net APY, and increasing their elasticity to the borrowing rate of USDe.

Below, we detail the net APY of the various strategies and the risks inherent, as well as a proposal to increase the target rate of USDe to better optimize the protocol.

Competitiveness

Current on-chain data show PT-token borrowers averaging 87% LTV, with USDT, USDC, and USDS accounting for the bulk of outstanding debt. These positions are highly leveraged with a concentrated Debt-to-Collateral ratio. Introducing USDe E-Mode reduces the concentration and improves capital efficiency for participants.

Debt vs. Collateral Distribution for Top PT-USDe-31JULY2025 Suppliers

Debt vs Collateral for PT-USDe-31JULY2025 Suppliers

Thanks to the higher leverage available in USDe E-mode compared to Stablecoins E-Mode, the net APY of the strategies that can be run also increases significantly. With this in mind, we can optimize the borrow rate for USDe while keeping the looping strategy competitive. Here, net APY is calculated as (supply APY - borrow APY) × leverage, where leverage is approximated as 1 ∕ (1 − LT), borrow APY is the Slope 1 at optimal utilization, and supply APY is the Fixed APY of the underlying on Pendle.

As can be observed from the charts "Expected Net APY without / with hikes”, even with three 50 bps rate hikes, PT-tokens looping in USDe E-mode remains profitable and appealing at low maturities compared to Stablecoin E-Modes. This will incentivize users to switch between USDe and Stablecoin E-Modes, reducing concentration.

Expected Net APY without hikes

Expected Net APY with hikes

USDe (Ethereum Core)

Following the recent listing of PT-eUSDe-31JULY2025 and PT-USDe-14AUG2025 USDe has almost reached 99% of its borrow cap.

USDe borrowing has surged because traders are now using PT-tokens as collateral in the highly capital-efficient USDe E-Mode, which delivers far more leverage than the Stablecoin E-Mode. For instance, the loan-to-value (LTV) range for PT-USDe-31 JUL 2025 rises from 91% to 93% with maturity in Stablecoin E-Mode, compared to 92% to 96% in USDe E-Mode. Since USDe E-Mode went live, outstanding USDe loans have grown roughly by $150 million, roughly a 50% increase in the last 48 hours.

Aggregate Supply Side Exposure: Collateral Utilization and Rehypothecation Risk

USDe suppliers on Aave are frequently highly leveraged—depositing USDe as collateral and borrowing USDT (or other stablecoins) against it. This rehypothecation loop increases the concentration of the borrow-to-collateral ratio, which in turn increases the liquidation volume at a specific price point of USDe. This is the primary reason why the optimal‑utilization parameter for USDe is calibrated at 80% rather than the 90–92 % common for other stablecoins. A sharp contraction in on‑chain USDe liquidity or an unwind of these leveraged positions could rapidly push utilization beyond the kink, causing borrow rates to spike and elevating bad debt risk for the protocol. Such an interest rate curve configuration aims at the minimization of unnecessary rate volatility given the associated USDe market size and collateral utilization.

Due to the nature of the looping strategy, USDe borrowed on Aave will be deposited into Pendle to mint PT-tokens. This can pose a potential liquidity risk for the positions looping USDe collateral to stablecoins, which, as mentioned in the previous section, are highly concentrated. During a liquidation of such users or exogenous supply contraction, caused by a significant USDe market event, the USDe market can become illiquid due to the simultaneous inability of PT token suppliers to redeem to the underlying and reduce their USDe borrows, potentially driving USDe utilization to 100% and accruing significant interest and potential bad debt under the assumption of a significant permanent deviation.

Borrow Distribution

Borrow distribution shows some concentration: the largest wallet accounts for ≈20 % of outstanding USDe debt and collateralizes it with volatile assets such as WETH and weETH. Despite that exposure, its health score signals no immediate risk. Many of the other dominant USDe borrowers use USDe-correlated collateral (PT-tokens), which further lowers liquidation risk.

Newly added PT-tokens are taking up over 145M USDe with the other 290M mostly being comprised of Stablecoins and ETH/BTC-based assets.

Recommendation

In light of passed proposals to list PT-eUSDe-14AUG2025 and PT-USDe-31JULY2025 with Stablecoin E-modes and USDe-specific E-Mode, we propose increasing the Slope 1 for USDe from 5.5% to 6%, and potential increases to 6.5% and 7% respectively later on. This change is expected to increase revenue by $600K over a 2.5-month period (~$2.8M annualized), and not significantly impact demand for looping PT-tokens with stablecoins. Apart from this, we recommend increasing the borrow caps for USDe to satisfy demand and further increase protocol revenue.

Specification

We will move forward and implement these updates via the Risk Steward process.

Parameter	Current Slope 1	Recommended Slope 1
USDe	5.5%	6.00%

Instance	Asset	Current Supply Cap	Recommended Supply Cap	Current Borrow Cap	Recommended Borrow Cap
Ethereum Core	USDe	960,000,000	-	440,000,000	660,000,000

Next Steps

Depending on the observed state of demand through implied yield dynamics, we aim to perform additional changes in the USDe Slope 1 in two distinct phases since the execution of the aforementioned recommended parameters.

Parameter	Slope 1	Recommended Slope 1
2025/06/31	6%	6.5%
2025/07/31	6.5%	7%

Disclaimer

Chaos Labs has not been compensated by any third party for publishing this AGRS recommendation.

Copyright

Copyright and related rights waived via CC0

[myriad]

If you’re intent on rugging like this, can you at least make USDe borrowable in the Stablecoin e-modes so that borrowers can switch to/from borrowing USDe vs other stables without incurring a huge gas expense?

I see no reason why USDe should not be borrowable (with the lower LT equal to USDT, etc.) in that mode

[myriad]

I should be more specific:

You correctly note that USDe-denominated debt is better for AAVE given correlation to the collateral, and that there is an incentive to borrow in USDe given the higher LT

yet at the same time you are proposing a change that will immediately put users who borrowed USDe in a materially worse position than those who borrowed other stablecoins. The higher APY more than negates any advantage from increased leverage and leaves USDe borrowing disincentivized.

Furthermore, you propose another increase to 7%, which is only 1% lower than either of the PT prices implied APY. 7% is also higher than that the PT-USDe-31JULY2025 was priced in the past.

Irrespective of the USDe borrow curve decisions, I suggest enabling USDe borrowing in the stablecoin e-modes for all ethena PTs

[ChaosLabs]

Thank you for your feedback. We acknowledge the importance of maintaining competitive net APYs for USDe looping, and we will continue monitoring the rates to implement additional increases only if the market allows for it. The rationale for the proposed rate adjustments is also rooted in stabilizing the rate after the increase in borrow cap, as the additional leverage is expected to drive utilization above USDe’s UOptimal and cause the rates to become highly volatile and unpredictable for users. Thanks to this change, we will allow the market to maintain a greater utilization buffer prior to rate volatility. Additionally, we emphasize that while the Slope 1 will be increased, the Borrow rate will adjust to the market’s demand, hence not necessarily fully reflecting the increase. Regarding the inclusion of USDe in the PT asset stablecoin E-Modes we appreciate the suggestion and will investigate its addition and implications on risk management.

[myriad]

I would just suggest sizing/timing any changes to the curve in a way that does not disincentivize borrowing USDe, in your example graphs there were periods like that after each rate hike.

Also notable that rates (as a consequence of funding rates) were at the highest they had been in months at the time you made this proposal, and both relevant PTs have traded below 8% today.

When the optimal borrow rate is close to the PT yield, leveraged looping strategies are very sensitive to rate changes.

I expect the utilization will easily reach the cap whenever the supply caps for the PTs finally get lifted.

The May PT-eUSDe and PT-sUSDe expired, and the only place to go has been morpho due to the caps

edit to add: I agree with everything you said about market forces, and thank you for taking my feedback into account.