Overview
Chaos Labs supports the proposal to enable sUSDe liquid E-Mode, including sUSDe, USDC, sUSDS, and GHO, as well as the listing of sUSDe in the Aave’s v3 Ethereum Lido Instance. However, we recommend adjustments to the LT, LTV, and Liquidation Bonus parameters to reduce the liquidation risk posed by the new E-Mode to Aave’s protocol.
Motivation
Understanding the Risk with sUSDe Current Oracle Setup
Given the structure of the oracle chosen for sUSDe, the asset differs from others, like LSTs, in terms of how price deviations affect liquidation risk. In LSTs, deviations in market price do not typically trigger liquidations thanks to the debt being denominated in the same asset as the oracle market price. For example, in the case of wstETH, the oracle reflects the ratio of wstETH/stETH multiplied by WETH/USD. Given the predominant WETH debt, the only variance in the oracle setup that could lead to liquidation is the exchange rate of wstETH/stETH decreasing. This setup means that market price deviations don’t impact the collateral’s valuation relative to the debt, avoiding unnecessary liquidations.
In contrast, sUSDe’s oracle is structured as sUSDe/USDe multiplied by USDe/USDC. When the market price of USDe declines, it directly impacts the sUSDe oracle price, increasing the likelihood of liquidations. Additionally, sUSDe has a 7-day unlock period, during which deviations between its sUSDe/USDe market price and sUSDe/USDe exchange rate can occur. As USDe deviates from USDC and sUSDe deviates from USDe, cascading liquidations may occur and thus be triggered when the asset has already moved beyond profitable liquidation levels.
Liquidity Concerns with sUSDe Market Oracle
Furthermore, a sUSDe/USDC market oracle is not adopted because while it would lead to faster liquidations, avoiding the risk of bad debt from the previously explained oracle setup, sUSDe lacks sufficient market liquidity relative to its total supply to ensure that liquidations do not lead to liquidation cascades and eventually to the creation of bad debt. Although the current liquidity of sUSDe has grown significantly over the last month, the buy liquidity is highly volatile, leading to a sustained price impact. Over the last 4 months, the average buy liquidity has been less than $10M sDAI in the most liquid DEX pool.
Comparison with Other Protocols
Some competitors mitigate these risks by hardcoding the USDe/USD component of the sUSDe oracle to $1. While this process commonly poses significant risks, in those instances, the protocol is effectively able to internalize the liquidation process. This approach allows them to reduce the risk of bad debt and allows for higher LT and LTV ratios. However, Aave’s open-market liquidation mechanism requires more conservative parameters to ensure the protocol’s safety and solvency.
E-Mode LT/LTV
Given the previously discussed risks, we recommend setting the LB according to the maximum discount observed in recent months for sUSDe relative to its exchange rate. This discount reached 2.47% on August 5 during the broader market decline. Using this, we recommend a 3% liquidation bonus to ensure liquidations are processed even in the event of a sUSDe discount.
With this LB, bad debt would begin accruing at a theoretical LTV of 0.97. We recommend providing a buffer set to 1.5 times the maximum discount observed, leading to an LT recommendation of 92% and LTV of 90%.
Listing sUSDe in Lido Instance
Given the absence of an sUSDe listing in the Lido instance, we propose the following initial parameters.
LTV, Liquidation Threshold, and Liquidation Bonus
As the primary purpose of sUSDe is to function as collateral through E-Mode, we recommend setting highly limiting values for its LTV and LT outside of the E-Mode to minimize unintended usage. To avoid freezing the asset, we propose an LT of 0.1% and an LTV of 0.05%, ensuring minimal borrowing capability without impacting the asset’s ability to be used as collateral.
Interest Rate Curve
Since sUSDe is recommended as a non-borrowable asset, there is no need for an interest rate curve or UOptimal setting.
Supply and Borrow Cap
Following Chaos Labs’ methodology for initial supply caps, we typically set the supply cap at 2x the liquidity available under the Liquidation Penalty’s price impact. However, given the volatility of sUSDe’s available buy liquidity, we recommend basing the cap on the average buy liquidity over the past three months. Using this adjusted metric, we propose an initial supply cap of 20,000,000 sUSDe. Since the asset is non-borrowable, a borrow cap is unnecessary.
Recommendation
Below are the initial listing parameters Chaos Labs recommends for sUSDe in Aave’s V3 Ethereum Lido instance, along with an associated sUSDe Liquid E-Mode configuration in both the Ethereum Lido and Main instances.
Specification
sUSDe Market Configuration (Lido Instance)
| Parameter | Value |
|---|---|
| Network | Ethereum Lido |
| Isolation Mode | No |
| Borrowable | No |
| Collateral Enabled | Yes |
| Supply Cap | 20,000,000 |
| Borrow Cap | - |
| Debt Ceiling | - |
| LTV | 0.05% |
| LT | 0.10% |
| Liquidation Bonus | 7.50% |
| Liquidation Protocol Fee | 10.00% |
| Variable Base | - |
| Variable Slope1 | - |
| Variable Slope2 | - |
| Uoptimal | - |
| Reserve Factor | - |
| Stable Borrowing | Disabled |
| Flashloanable | Yes |
| Siloed Borrowing | No |
| Borrowable in Isolation | No |
sUSDe Liquid E-mode Configuration
| Parameter | Value | Value | Value |
|---|---|---|---|
| Asset | sUSDe | USDS | USDC |
| Collateral | Yes | No | No |
| Borrowable | No | Yes | Yes |
| Max LTV | 90% | - | - |
| Liquidation Threshold | 92% | - | - |
| Liquidation Bonus | 3.0% | - | - |