Risk Stewards: Supply and Borrow Cap Changes on Aave V3 / 2026.09.22

[LlamaRisk]

Summary

LlamaRisk recommends the following parameter changes based on user behavior, on-chain liquidity, and position health observed in the latest review of Aave V3 reserves.

Aave V3 Core:

• Increase the borrow cap for wstETH from 14,000 to 28,000.
• Reduce the borrow cap for USDe from 700,000,000 to 100,000,000.

Aave V3 Plasma:

• Increase the supply cap for USDe from 550,000,000 to 750,000,000.

wstETH (Aave V3 Core)

wstETH sits at 99.8% borrow cap utilization on Aave V3 Core, with 13,977 borrowed against a cap of 14,000.

Borrow Distribution

Source: LlamaRisk, September 22, 2026

The top wstETH borrowers have health factors ranging from 1.00 to 4.63, with a median of 1.11 and most of the cohort below 1.16. Their collateral is predominantly weETH by value at $51M (80%), followed by WETH at $8.5M (14%). Given the close correlation among these assets, health factor stability is primarily sensitive to the relative performance of wstETH against the collateral assets, rather than broad market movements.

Recommendation

The proposed cap of 28,000 is approximately 2.00x the current outstanding borrow, resulting in approximately 49.9% post-change cap utilization.

USDe (Aave V3 Plasma)

USDe sits at 100.0% supply cap utilization on Aave V3 Plasma.

Supply Distribution

Source: LlamaRisk, September 22, 2026

The top eight suppliers include six carrying outstanding debt, at health factors between 1.01 and 1.05 with a median of 1.02, and two supplying on a spot basis. All six borrow USDT0 against USDe collateral, and three also hold sUSDe. Since USDe is priced at par with USDT, these positions have very low liquidation risk. Supply is concentrated, with the largest position holding 64.6% of the reserve and the top five holding 94.8%.

Recommendation

The proposed cap of 750,000,000 is approximately 1.36x the current outstanding supply, resulting in approximately 73.3% post-change cap utilization.

Cap Reductions

USDe on Aave V3 Core sits at 11.1% borrow cap utilization (77.9M / 700M). The proposed cap of 100,000,000 is approximately 1.28x current outstanding borrows, landing at approximately 77.9% post-change cap utilization.

Specification

Instance	Asset	Current Supply Cap	Recommended Supply Cap
Aave V3 Plasma	USDe	550,000,000	750,000,000

Instance	Asset	Current Borrow Cap	Recommended Borrow Cap
Aave V3 Core	wstETH	14,000	28,000
Aave V3 Core	USDe	700,000,000	100,000,000

Next Steps

We will move forward and implement these updates via the Risk Steward process. The implemented changes can be reviewed on the LlamaRisk Risk Stewards Dashboard.

Disclosure

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.

[Roch33]

Thanks for sharing the September 22 update on USDe on Aave V3 Plasma.

Looking at this recommendation together with the previous September 15 cap adjustment gives an interesting first data point.

On September 15, USDe was already at 100% of its 450M supply cap. The largest supplier represented 66.6% of the reserve and the top five represented 97.4%. The cap was then increased to 550M.

By September 22, the 550M cap was again fully utilized. The largest position represented 64.6% of the reserve and the top five 94.8%. The new recommendation is to increase the cap to 750M.

This gives us a useful observation.

The previous increase appears to have produced modest relative diversification, while the absolute exposure of the largest positions continued to grow.

Using the published percentages as an approximation, the largest position increased from around 300M USDe at the 450M cap to around 355M at the 550M cap.

That leads to a first hypothesis for the move to 750M.

The diversification outcome will depend heavily on who absorbs the additional 200M of capacity.

If the current largest position remains around 355M while new suppliers provide most of the additional liquidity, its share would fall toward 47%.

If the current distribution remains broadly unchanged, the largest position would stay around 65%.

If most of the additional capacity is absorbed by the existing largest supplier, concentration would increase materially.

These are illustrative scenarios, not forecasts. Their purpose is to identify what should be measured after implementation.

A second point relates to market stress.

The February 2025 Bybit event gives a useful historical reference. During that episode, USDe traded around 0.994 on Curve and fell to 0.96 USDT on Bybit. The Chainlink USDe/USD feed reached approximately 0.977. Around 117M USDe was redeemed. Aave recorded six sUSDe collateral liquidations totaling approximately 22M, without bad debt.

The current Plasma structure is different because USDe is priced at par with USDT. That difference is important.

It raises a different monitoring question: during a secondary-market discount, stress may appear first through withdrawal liquidity, redemption activity and changes in supplier behaviour while the collateral valuation used by Aave remains comparatively stable.

Against that background, we would suggest adding a short post-adjustment monitoring block to future cap reports.

1. Largest supplier and top-five concentration before and after each cap change.
2. Absolute exposure of the largest positions, alongside percentage concentration.
3. Share of incremental supply provided by new suppliers versus existing large suppliers.
4. Available withdrawal liquidity and material changes in redemption activity.
5. Simple reference scenarios for a 1%, 3% and 5% secondary-market discount in USDe.
6. A predefined concentration, liquidity or secondary-market price threshold that would lead to an interim cap review.

This is not an argument against the proposed increase.

The objective is to test the assumptions behind the increase after implementation.

In this case, one useful question for the next review would be:

Did the additional 200M capacity result in genuine diversification of the supplier base, or mainly increase the absolute exposure of the positions already dominating the reserve?

A second question would be:

If USDe trades at a material secondary-market discount while Aave continues to reference USDe through USDT pricing, which liquidity or market indicators would trigger an earlier reassessment of the 750M cap?

This would also provide a practical way to extend the Risk Steward reporting framework discussed previously from decision-time reporting to outcome monitoring over time.