Summary
Chaos Labs fully supports the integration of Umbrella within Aave’s core protocol architecture. Embedding Umbrella as a native mechanism will facilitate the prompt resolution of insolvent positions through a dynamically risk-weighted junior tranche structure, enabling efficient absorption and socialization of bad debt under adverse market conditions.
We intend to collaborate closely with the Aave Finance Committee to deliver ongoing, data-driven recommendations for fine-tuning Umbrella’s parameters. This includes adaptive calibration of tranche thresholds, emission schedules, and risk mitigation strategies to ensure resilient performance across a range of market states and stress scenarios. Below, we offer a preliminary breakdown and key insights ahead of the release of our full quantitative methodology in the coming days.
Umbrella Cost Dynamics
Currently, Aave distributes substantial incentives through stkAAVE, which represents a significant ongoing expense. Under Umbrella, incentive and cost dynamics have various implications that can be considered when optimizing markets and quantifying parameters and expenses.
One key dynamic arises from Umbrella’s design, which prevents staked aTokens from being used as collateral. As a result, staking into Umbrella is inherently more attractive for assets with low collateral demand—such as stablecoins—where the opportunity cost of forfeiting rehypothecation is minimal. This typically aligns with assets that have high borrow demand, reinforcing the need for robust Umbrella coverage.
However, this relationship does not always hold. Assets like WETH, for example, exhibit both strong borrow demand and significant collateral usage. In such cases, Umbrella introduces a higher opportunity cost for participants, as staking requires foregoing valuable collateral utility. To attract sufficient liquidity under these conditions, the yield offered through Umbrella staking must be correspondingly higher to offset this trade-off.
WETH Borrow and Collateral Distribution
The financial sustainability of Umbrella is also partially correlated with the borrowing intensity of supported assets and, as such, with their necessity for Umbrella protection. This dynamic is caused by the direct correlation between borrowing demand and revenue derived by SVR and the Reserve Factor.
However, some assets, combining low interest rates and revenue with high market utilization, such as wstETH being the primary borrow option against LRTs, amplify costs disproportionately, creating a significant opportunity cost for the DAO to incentivize such markets while accruing limited revenue from them. This highlights the importance of onboarding assets and strategies with a balanced or positive risk-reward for the DAO in order to contain Umbrella expenses.
While initially, Umbrella is planned to be funded purely through AAVE incentives, the emission curve is likely to shift to a structure whereby underlying revenues from SVR or Reserve Factor of the interested pool are reinvested into the market for incentives. This reallocation incentivizes sustainable growth and improved system resilience and allows for reduced spending from the current security budget, amounting to over $55M annually.
If necessary, a long-term strategy to balance incentives can stem from increasing the Reserve Factor to distribute to Umbrella Stakers. However, this strategy will reduce revenue directed at vanilla suppliers and thus is highly reserve-dependent, taking into account the aforementioned components below.
Parameterization of maxEmissionPerYear and Target Liquidity
The derivation of maxEmissionPerYear—understood as the optimal incremental yield offered to users staking into Umbrella—is informed by a multifaceted opportunity cost analysis. This analysis captures both user-level trade-offs and systemic risk dimensions inherent to the composable architecture of the protocol. The following considerations form the foundation of its calibration:
Collateral Utility Opportunity Cost
Staking tokens in Umbrella precludes their use as collateral, in contrast to traditional aToken holdings that can be rehypothecated. The opportunity cost is therefore linked to the share of aTokens typically used as collateral within a reserve. This relative collateral utilization reflects the user trade-off between earning additional Umbrella rewards and preserving the capital efficiency of leveraging their assets elsewhere.
Native Reserve Contribution to External Reserve Risk
This component accounts for systemic risk arising from the reserve’s exposure to overcollateralized debt positions, quantified through the nominal health factors of users across the reserve. When aggregate health factors are low and a large portion of the reserve is actively used as collateral, the protocol faces elevated liquidation risk.
Given the composable and rehypothecative structure of the system—where assets often serve dual roles as both collateral and supplied liquidity—a risky distribution of such assets can trigger cascading liquidations. If collateralized positions are forcefully unwound, corresponding aTokens are removed from the reserve, driving up utilization rates. In extreme scenarios, this dynamic can impair the protocol’s ability to effectively liquidate risk and maintain solvency.
By increasing incentives for users to stake into Umbrella under these conditions, the protocol discourages excessive relative collateral usage, reinforces reserve stability, and reduces the likelihood of systemic stress propagation.
Market Demand Reflected in Interest Rate Regimes
Since maxEmissionPerYear is expressed in raw token terms, it must be responsive to the prevailing interest rate environment, which serves as a proxy for market-wide demand for leverage. As rates rise, signaling intensified competition for capital and heightened returns on leveraged positions, the opportunity cost of supplying to Umbrella significantly increases.
In such high-rate regimes, capital agility becomes more valuable, and the trade-off of forfeiting collateral utility grows steeper. To remain attractive under these conditions, Umbrella’s reward emissions must scale accordingly, compensating users for the elevated marginal cost of locking capital in a less flexible position. This ensures that staking into Umbrella remains a rational choice even when market incentives favor rehypothecation and aggressive leverage.
Duration and Cooldown Risk
Although aTokens continue to accrue yield and remain slashable during the cooldown period, the temporal delay before assets become accessible again introduces duration risk. This immobility hampers a user’s ability to respond promptly to market shifts, reinforcing the need for a compensatory yield uplift. The longer the effective lockup, the higher the premium required to offset the strategic disadvantage of delayed capital deployment.
Target Liquidity
Target Liquidity aims to be defined by leveraging a VaR framework, applied to the distribution of underlying collateral that supports debt within the reserve. This measure seeks to estimate the required coverage to safeguard the protocol in the face of adverse market scenarios. Given the inherent composability and interdependencies within Aave, determining this value necessitates a simulation-driven, quantitative approach that accounts for the system’s complex dynamics.
This process ultimately frames the derivation of the Target Liquidity parameter as a minimization problem—where the goal is to identify the optimal nominal stkToken supply required to mitigate tail risks while respecting capital efficiency. In an ideal state, the Aave Finance Committee would have full flexibility to allocate as many resources as needed toward Umbrella coverage. However, as the reserve grows, the optimal share of Umbrella coverage is not expected to scale linearly. Instead, it is likely to follow a logarithmic growth pattern, reflecting diminishing marginal risk exposure relative to reserve size.
In our forthcoming post outlining the complete umbrella reserve methodology, we will detail the full mathematical derivation and supporting simulations.
Cooldown Period Parameterization
A 20-day cooldown and 2-day withdrawal window align well with existing Safety Module practices. However, this introduces some risks tied to oracle responsiveness—delays or inaccuracies in fundamental oracles can allow stakers early exits before slashing events are adequately reflected in the oracles and bad debt is accrued. An example of such a situation is with Lido bunker mode activation and correlation penalty, which only impacts the oracle price 18 days following the initial slashing. A further example lies in assets that require manual oracle updates through the governance process to reflect their underlying value following a backing loss.
Careful selection of cooldown and withdrawal window for these specific assets may be required in order to ensure Umbrella Liquidity at the time of oracle update.
Application of Automation
Automation can play a pivotal role in Umbrella by enabling dynamic parameter adjustments across multiple components of the system. In addition to the previously discussed elements—such as Target Liquidity and maxEmissionPerYear, which are influenced by Value-at-Risk (VaR) assessments and reserve opportunity cost—automation can also govern the initial denomination of rewards in AAVE tokens, given by the mismatch between reward token denomination and the underlying reserve denomination.
Given that the underlying objective function is expressed in raw token terms, and that market-specific nominal risk scales with reserve growth, it can be beneficial to implement a system that can continuously recalibrate these values in real time. This ensures reward emissions remain aligned with both market dynamics and protocol-level risk exposure.
Disclaimer
Chaos Labs has not been compensated by any third party for publishing this recommendation.
Copyright
Copyright and related rights waived via CC0