Summary
LlamaRisk supports the proposed parameter changes to align with the Aave v3.6 upgrade. We endorse the cleanup of legacy configurations and the enforcement of stricter E-Mode boundaries for volatile and correlated assets.
To our knowledge, USDC is not present on Plasma, we recommend removing it from the proposed XAUt0 Stablecoins Plasma E-Mode.
Volatile Asset Collateral
Establishing E-Modes with specified borrowed assets, while setting isCollateral false in the base reserve configuration, would reduce the protocol’s exposure to assets that can be borrowed against the listed volatile assets. This approach limits borrowing to select stablecoins as defined in the respective E-Mode configurations.
For the assets where no E-Mode has been established, the table below summarizes the corresponding liquidity depth and the aggregate demand impacted (as of December 19, 2025).
| Asset | Instance | Liquidity within LB (DeFiLlama) | Aggregate Demand Impacted (Chaos Labs) | Estimated Annual Revenue Impacted |
|---|---|---|---|---|
| AAVE.e | Avalanche | $ 21,000 | $ 259,809 | $750 |
| LINK | Arbitrum | $ 890,000 | $ 6,604,596 | $22,450 |
| LINK.e | Avalanche | $ 95,000 | $ 666,507 | $1,620 |
| LINK | Optimism | $ 25,000 | $ 349,896 | $970 |
| AAVE | Optimism | $ 40,000 | $ 239,258 | $560 |
| OP | Optimism | $ 250,000 | $ 752,129 | $2,280 |
| LINK | Polygon | $ 887,500 | $ 1,462,799 | $13,280 |
| MaticX | Polygon | $ 42,000 | $ 44,564 | $380 |
For each of these assets, USDC and USDT together account for roughly 80% of the aggregate demand. Based on the impact on liquidity and aggregate demand, LINK on Arbitrum stands out as an outlier. If a stablecoin E-mode (USDC, USDT) were to be set up, it would capture the majority of both the aggregate demand and the associated revenue. The annual protocol revenue is calculated using the following formula, which annualizes the current debt, where i indexes borrowed assets in the set A.
Low-demand Collateral (outside E-Mode)
For low-demand collateral outside of E-Mode, restricting LSTs and LRTs to be used as collateral only within their existing E-Modes ensures that the borrowed and collateral asset prices remain correlated under the Aave Oracle configuration. This significantly reduces the likelihood of bad debt.
We note that the following assets had their LTV and LT parameters increased under the new stablecoin E-Modes compared to the base reserve configuration, resulting in a less conservative risk calibration:
| Asset | Instance | LTV Change | LT Change | LB Change |
|---|---|---|---|---|
| rsETH | Ethereum Core | 72% to 75% | 75% to 78% | - |
| LBTC | Base | 68% to 75% | 73% to 78% | 8.5% to 5% |
Low-demand Stablecoins
Deprecating stablecoins with low demand as both collateral and borrowing assets would reduce the protocol overhead associated with maintaining these assets. While the aggregate demand for Bridged USDC (USDC.e) across Optimism, Arbitrum, and Polygon appears relatively high, at nearly $7M, it is essential to note that native USDC has already been onboarded on all these instances. Disabling both collateral and borrowing would therefore facilitate user migration without affecting the health of existing user positions.
LST/wrapping Borrowing
Disabling the borrowing of LSTs and wrappers outside their respective E-Modes would restrict borrowing activity to similarly price-correlated assets, reducing the likelihood of bad debt. Disabling borrowing would not affect the health of existing user positions, as users would be required to migrate their positions. Low aggregate demand has been observed for these assets, with the exception of wstETH on Ethereum Core, which has a borrowing demand of $26.71M outside E-Modes (primarily composed of WETH collateral).
Other Configuration Cleanup
Removing collateral status for assets currently relying on the 0.1% LT workaround, and disabling borrowing for assets with a borrow cap of 1, would not affect the health of existing user positions. Additionally, moving XAUt from isolation mode to stablecoin E-Modes with limited assets would reduce protocol exposure to broader collateral without introducing additional risks, as the LT and LTV parameters remain unchanged.
As a necessary configuration correction, USDC should be removed from the XAUt0 Stablecoins Plasma E-Mode, as the asset is not available on the chain.
Disclaimer
This review was independently prepared by LlamaRisk, a DeFi risk service provider funded in part by the Aave DAO. LlamaRisk is not directly affiliated with the protocol(s) reviewed in this assessment and did not receive any compensation from the protocol(s) or their affiliated entities for this work.
The information provided should not be construed as legal, financial, tax, or professional advice.