Overview of Aave V4’s Liquidation Engine

Greetings Aave Community,

This post provides an overview of the new liquidation mechanism in Aave V4. The changes are intended to create a more efficient and fair liquidation process to the benefit of users and Aave as whole.

The new liquidation engine is a critical component of Aave V4, and we are pleased to share the details with the community.

Aave V3 Liquidations

In V3, when a borrower’s health factor drops below 1, a liquidator can repay a portion of the borrower’s debt and claim a portion of their collateral as a reward. The amount a liquidator can repay is determined by a fixed close factor.

This means that liquidators always repay a fixed close factor percentage of the borrower’s debt and seize a corresponding amount of collateral, receiving a static liquidation bonus in the process. The default close factor is 50%, but this extends to 100% under certain conditions, such as when the health factor falls below 0.95 or when dust collateral or debt is involved.

This system is straightforward and effective at protecting the protocol from bad debt. But it has limitations. The fixed close factor means that borrowers are often over-liquidated. Even if a small repayment would be enough to restore a healthy position, the liquidator still has to repay a fixed amount of the debt. The static bonus also does not reflect the true risk of a position. Liquidators receive the same bonus percentage regardless of how far underwater a position is, which means there is no extra incentive to act quickly on the riskiest positions.

The Aave V4 Liquidation Engine

In V4, liquidations work differently. Instead of repaying a fixed percentage of the debt, liquidators can repay at most enough to return the borrower’s health factor to a specific level called the Target Health Factor. This target is set by governance for each Spoke and gives the protocol more control over how liquidations are handled.

When a liquidation occurs, the protocol calculates how much debt needs to be repaid to bring the borrower back to the Target Health Factor. In typical cases, the liquidator repays only up to that amount, not a fixed close factor. This means borrowers are not over-liquidated, and liquidators can act more efficiently.

V4 also introduces a variable liquidation bonus. The bonus paid to liquidators changes based on the borrower’s health factor. The lower the health factor, the higher the bonus. This creates a Dutch-auction-style system where the relative reward for liquidation increases as the risk to the protocol increases. Liquidators are incentivized to act quickly on the riskiest positions, which improves the overall health of the protocol.

Behind the scenes, the protocol uses several parameters to manage this process. The maxLiquidationBonus sets the maximum bonus a liquidator can receive for a given collateral type. The healthFactorForMaxBonus defines the health factor below which the maximum bonus applies. And the liquidationBonusFactor determines the minimum bonus at the liquidation threshold. These parameters define the relationship between health factor and bonus, allowing governance to fine-tune the system.

Handling Dust Debt

One of the challenges in any liquidation system is dealing with dust debt, or small leftover debt positions that are not economically viable to liquidate. In V3, there is complex logic governing the handling of dust, dependent on thresholds both for the amounts before and after a liquidation. This prevents dust, but it also creates situations where positions cannot be fully liquidated in one go.

V4 takes a different approach. If the debt or collateral remaining after a standard liquidation would be below the DUST_LIQUIDATION_THRESHOLD (e.g. $1,000), and the liquidator indicates they want to repay the entire position, the protocol dynamically adjusts the maximum amount of debt that can be liquidated. This requires the liquidator to fully clear the position, preventing dust from accumulating.

Parameters and Configuration

Aave V4 exposes several configurable parameters that influence liquidation:

Parameter Description Constraints
TargetHealthFactor A spoke‑wide value set by governance representing the HF to which a borrower should be restored after liquidation. Liquidators repay only enough debt to reach this HF under normal circumstances that do not result in dust collateral or debt remaining. Must be ≥ the HEALTH_FACTOR_LIQUIDATION_THRESHOLD constant.
DUST_LIQUIDATION_THRESHOLD Hard‑coded threshold used to prevent extremely small leftover debt. The maximum debt that can be liquidated is increased to ensure that debt or collateral dust less than this threshold does not remain unless the corresponding respective collateral or debt reserve is fully liquidated. Hard‑coded constant set to 1000 USD in base units.
maxLiquidationBonus Per reserve defined maximum liquidation bonus for a collateral, expressed in basis points (BPS). A value of 105 00 (105%) means there is 5% extra seized collateral over the amount of debt repaid in base currency. Must be ≥ 100 00.
healthFactorForMaxBonus Spoke‑wide value expressed in WAD units defining the HF below which the max bonus applies. It must be less than or equal to HEALTH_FACTOR_LIQUIDATION_THRESHOLD to avoid division‑by‑zero. healthFactorForMaxBonus < HEALTH_FACTOR_LIQUIDATION_THRESHOLD.
liquidationBonusFactor Spoke‑wide percentage (expressed in BPS) specifying the fraction of the max bonus earned at the threshold HEALTH_FACTOR_LIQUIDATION_THRESHOLD. It defines the minimum bonus; e.g., a factor of 80 00 yields a bonus equal to 80 % of the max bonus when HF equals the liquidation threshold. liquidationBonusFactor must be <= 100 00

A Side by Side Overview

Ultimately, the changes in V4 create a more refined and efficient liquidation process. In V3, the repayment amount is fixed at 50% of the debt in most cases. In V4, the repayment amount is variable and calculated to bring the borrower back to the Target Health Factor set by governance. This means only the necessary debt is repaid, preventing over-liquidation.

The liquidation bonus also works differently between the two versions. In V3, the bonus is static and does not change based on the borrower’s health factor and instead requires a governance decision. In V4, the bonus is variable and increases as the health factor decreases, creating better incentives for liquidators to act on riskier positions.

Why This Matters

The new liquidation engine in Aave V4 makes the protocol more competitive and user-friendly. The adaptive repayment amount means borrowers are not over-liquidated. If only a small amount of debt needs to be repaid to restore a healthy position, that is all the liquidator is allowed to repay. This is fairer to borrowers and more capital-efficient.

These changes result in a more precise and capital-efficient liquidation process that better protects the protocol from bad debt while improving the experience for both borrowers and liquidators. The system is more flexible, more fair, and better aligned with the true risk of each position. Given the scale of liquidations on Aave, with up to hundreds of millions of dollars liquidated each month, these improvements will have a meaningful impact on the protocol’s health and user experience.

Aave Labs

8 Likes

Hello @AaveLabs,

thanks for explaining the liquidation process for v4.
When can the DAO expect to understand the revenue mechanism and implications with V4?
Could you explain where revenue is being captured for the DAO?
Are there new ways of value capture for the DAO?

Would be great if you could cover this in your next overview episode, considering that release isn’t that far away anymore and the DAO needs to decide on the activation.

Thank you, wishing you happy holidays.

8 Likes

Hello @EzR3aL thanks for the question.

First, on the V4 liqudation engine: the V4 engine will indeed work with SVR in the same exact way as the V3 engine, with the additional benefits given by decreased gas costs. In terms of expected SVR revenue, this will ultimately be defined by how the risk managers will configure the parameters that determine liquidation bonuses. If the risk managers believe that is beneficial to be less agressive on bonuses (to be more competitive for the users towards other protocols) then the expected revenue could be lower (unless offset by the additional users the protocol may gain due to better competitivity). Otherwise it might be on par or even above the V3 engine, if the decision is to configure it more conservatively (higher bonuses, no dutch auctions, lower competitivity). In general the goal of the new engine is to be more efficient and provide risk managers with more options to address changing market conditions and pressure from competitors.

In terms of additional possible revenue sources coming from V4 developments, while at the beginning the initial configuration will resemble Aave V3 (with only ERC20 collaterals but more diversified venues for yield seekers, see the current Aave V4 developers preview) in the future we expect a variety of new venues to open up due to the modular architecture of V4. Aave labs has in the works, or planning to implement, a variety of new etherogeneous spokes:

  1. Vaults spoke, that allow to keep the collateral segregated from the pools while still allow borrowing; this will satisfy long term needs especially of institutional actors or cefi services that currently face regulatory challenges in using Aave due to the pooled nature of the collaterals
  2. Debt trading spoke: similar to what Fluid does today, it will be possible to build spokes that leverage the deep Aave liquidity to let users trade with borrowers debt, in order to generate new revenue stream from swap fees directly at the protocol level and reduce borrow rates, increasing competitivity
  3. CDP spokes: V4 can be integrated with dexes that support liquidity ranges like uniswap V4 to allow LPs to unlock liquidity by borrowing from Aave V4 while keeping liquidity positions open.

These are just examples - since V4 moves away from the ERC20-as-collateral model, any asset or position that can be moved onchain can become source of new borrowing demand,.

8 Likes

@Emilio When should we expect a proposal “template” for V4 spoke proposals (including items like credit limits, risk parameters, and asset configuration), so builders can align early designs with DAO expectations and help standardize governance Spoke Proposal submissions. Or will it follow the formal Temp Check → ARFC → AIP stages?

2 Likes

Hi @JosueMpia the goal at the moment is to proceed with the initial Aave V4 deployment. The initial deployment will execute according to the standard process concluding with an AIP for the onchain activation. After that, there are a variety of topics to be discussed with the DAO around innovation and possibly permissionless developments on top of Aave V4. Aave labs can provide the technology but cant unilaterally decide how to use it, and there are security considerations around that as well (eg spokes, due to their nature, cannot be permissionless by definition). Therefore definining safe guardrails for the usage of V4 and innovations that can be built on top will require a series of governance decisions that will also help determining the direction that Aave V4 takes as a protocol. This is great, to be clear: innovation on top of Aave technologies has been stagnant for a while now, and having a chance to actually open up new venues for borrowing demand and protocol level renue is exciting.

2 Likes