V4 Spoke Liquidation Parameters: Impact on Revenue

[TokenLogic]

Executive Summary

This post explores how Aave V4’s liquidation parameters can be tuned to achieve different spoke objectives. Using V3 Ethereum Core Market data and V4 simulations, we examine how parameter choices affect repayment sizing, liquidator rewards, execution reliability, and protocol revenue, including SVR.

We present two configurations as case studies illustrating opposite ends of the design space:

• Prioritizes minimal intervention when positions become undercollateralized
• Reduces losses for users being liquidated
• Compressed liquidator margins may reduce liquidator participation, particularly during congestion
• Results in lower liquidation surplus and SVR revenue compared to V3

Liquidation Efficiency Configuration

• Prioritizes reliable execution during market stress
• Enables larger liquidations with stronger liquidator rewards
• Higher post-liquidation health factors provide larger safety buffers
• Generates higher surplus and SVR revenue

These configurations are boundary cases within the design space, not expected operating regimes. The borrower UX-friendly approach yields a lower protocol surplus than V3, whereas the liquidation-efficiency approach generates substantially higher surplus, especially during market stress.

Overview

In this publication, we evaluate two configurations of Aave v4’s liquidation engine at the spoke level to demonstrate the impact of tailoring configurations to different objectives. The two configurations are:

• Borrower UX: Smaller liquidations, lower liquidation bonuses
• Liquidation Efficiency: A highly competitive and rewarding liquidation environment

By isolating individual parameters and conducting a sensitivity analysis, we can assess their impact on revenue and overall liquidation efficiency.

V4’s configurable liquidation engine lets the DAO tune how aggressively positions are liquidated. This controls the trade-off between borrower protection and execution reliability, and determines how the surplus is split among borrowers, liquidators, and the protocol.

Borrower UX Spoke

Spoke Parametrization

Parameter	Monte Carlo Prior	Economic Interpretation
Target Health Factor	U[1.02,1.06]	Encourages minimal intervention by restoring solvency with limited repayment sizing.
Stress activation threshold	U[0.82,0.90]	Defers maximum bonus activation to deeper stress states.
Minimum incentive level	U[0.20,0.55]	Implies low liquidation surplus close to the liquidation boundary.
Bonus adjustment	U[-0.05,-0.01]	Systematically compresses the maximum attainable liquidation bonus relative to the baseline.
Liquidation fee	U[0.08,0.12]	Protocol share of gross liquidation surplus. Low fee preserves minimal liquidator margin, ensuring timely execution before bonus escalates
Maximum bonus cap	1.18	Limits tail-risk penalties borne by borrowers.
Debt dust threshold	1000	Operational lower bound, held constant across spokes.

This design reduces the amount of collateral that borrowers lose per liquidation event. However, it compresses the surplus available for liquidators and SVR. In stressed markets, lower rewards may mean that some liquidations are not executed as quickly, or at all, if liquidators don’t find them profitable enough.

Bottom line: Better experience for collateral holders, but less protocol revenue and potentially less reliable execution under stress.

Liquidation Efficiency Spoke

This configuration answers the question: What if we calibrated liquidation parameters to ensure reliable liquidation execution, even during market chaos?

The goal is to make liquidations attractive enough that liquidators compete aggressively to execute them, even when gas is expensive and markets are volatile.

Spoke Parametrization

Parameter	Monte Carlo Prior	Economic Interpretation
Target Health Factor	U[1.07, 1.14]	Larger target repayment size increases liquidation opportunity value, supporting reliable execution under stress.
Stress activation threshold	U[0.90, 0.95]	Maximum bonus activates earlier, even under moderate under-collateralization.
Minimum incentive level	U[0.60, 0.95]	Ensures sufficient baseline surplus near the liquidation boundary to sustain participation.
Bonus adjustment	U[-0.01, +0.04]	Allows upward adjustment of the maximum liquidation bonus when execution capacity is scarce.
Liquidation fee	U[0.12, 0.18]	Protocol share of gross liquidation surplus. A higher fee is sustainable given larger liquidator margins, and the protocol captures a larger share of the abundant surplus.
Maximum bonus cap	1.28	Preserves execution incentives in tail-risk scenarios with elevated volatility or congestion.
Debt dust threshold	1000	Operational lower bound applied uniformly across all spoke configurations.

This design ensures liquidations execute reliably and generates substantial surplus for the protocol (including SVR revenue). However, borrowers face larger liquidations and potentially higher penalties when their positions become undercollateralized.

Bottom line: More reliable execution and higher protocol revenue, but more intervention for affected borrowers.

Key Assumptions and Limitations

Key Assumptions for V4 Counterfactual

1. Liquidation outcomes (debt repaid, collateral seized, bonuses) are accurately priced using oracle prices at the time of liquidation.
2. Only the pre-liquidation health factor is observed. Post-liquidation health is inferred indirectly through the application of a target health factor (THF) rule in the simulation, rather than reconstructed from historical balances.
3. The liquidation threshold observed at event time is applied deterministically in the counterfactual V4 sizing logic.
4. For each simulated liquidation, the repaid debt amount is determined by a target post-liquidation health factor (THF).
5. If remaining debt falls below $1,000 after liquidation, the position is fully liquidated.
6. Liquidation bonuses are modelled as a monotonic function of liquidation stress, proxied by the distance of the pre-liquidation health factor from solvency, subject to a hard upper cap.

What the Model Does Not Capture

1. All V4 counterfactual simulations are limited to the AAVE v3 Ethereum Core Market, the only venue with active SVR integration at the time of analysis. Results may differ in other markets.
2. In the V4 counterfactual, liquidators are assumed to repay borrower debt directly to the spoke-specific target health factor, bounded by total outstanding debt.
3. Liquidator and SVR participation are modelled using simplified, data-driven probability functions based on historical behaviour. The model does not capture individual strategies, capital constraints, or adversarial execution; therefore, it reflects average rather than worst-case outcomes.
4. SVR revenue conditional projections assume that liquidation volume, asset mix, and opportunity distribution remain broadly similar to the calibration period. Significant changes in borrower behaviour, collateral usage, or protocol parameters could lead to materially different outcomes.
5. The framework focuses on liquidation execution and surplus allocation conditional on liquidation eligibility. It does not model complete execution failures, residual bad debt, or cascading insolvencies under extreme systemic stress.
6. The analysis assumes that SVR participation and recapture efficiency scale smoothly with liquidation opportunity size. Possible changes in auction performance under different configurations are not explicitly modelled.
7. The participation model should be interpreted as a local approximation around observed V3 regimes, rather than as a structural equilibrium prediction under a regime change in liquidation mechanics.

Methodology

Details

Design Space of the Aave v4 Liquidation Mechanism

Counterfactual SVR Value Estimation Under Continuous Participation Models

Counterfactual SVR value is estimated using continuous participation and efficiency models that map liquidation state variables directly into execution likelihood and captured surplus, without relying on discrete bins or frozen participation assumptions.

Logistic participation model

Participation probability is modelled as a continuous function of liquidation state variables, using a logistic specification that captures how opportunity size, position stress, and fees jointly affect the likelihood of execution.

This specification captures how liquidation opportunity size, position stress, and protocol fees jointly shape execution incentives in a smooth and interpretable way, allowing participation likelihood to adjust continuously across states rather than through discrete bins or fixed assumptions.

Kernel-based efficiency model

Conditional on participation, capture efficiency is estimated as a smooth function of opportunity size using a kernel-weighted local average.

By avoiding hard bin boundaries, this approach yields a locally adaptive estimate of capture efficiency that reflects how realized SVR performance scales smoothly with opportunity size.

Results

SVR Performance on Aave V3 Ethereum Core Market

Aave Revenue from Liquidation Fees and SVR recapture in 2025

Quarter	Protocol Fee ($M)	SVR Recapture ($M)	Total Aave Revenue ($M)
2025Q2	0.57	0.07	0.64
2025Q3	0.27	1.09	1.36
2025Q4	2.23	4.93	7.16
Total	3.07	6.09	9.16

SVR accounted for 66% of total liquidation revenue in the Core Market. Q4 generated ~80% of total Aave SVR recapture.

Quarterly Recapture Efficiency (SVR Liquidations Only)

Quarter	SVR Liquidations	Total OEV ($K)	Aave Recapture ($K)	Chainlink Payout ($K)	Aave Recapture Rate
2025Q2	98	202.8	71.8	38.6	35.4%
2025Q3	437	2104.2	1088.3	586.0	51.7%
2025Q4	1,551	13863.0	4934.9	2657.3	35.6%

Recapture rate peaked at 51.7% in Q3, declining to 35.6% in Q4 as the liquidation mix shifted toward events with lower per-event capture efficiency.

OEV Value Flow by Collateral Asset - SVR Liquidations (Top 5)

Asset	SVR Liquidations	Total Bonus from SVR Liquidations ($K)	Aave Recapture ($K)	Chainlink Payout ($K)	Aave Recapture %
wstETH	72	5412.1	1730.0	931.5	32.0%
WBTC	399	5033.8	1691.5	910.8	33.6%
WETH	1,034	3816.0	1844.0	992.9	48.3%
cbBTC	59	650.7	238.8	128.6	36.7%
LINK	187	422.4	221.8	119.4	52.5%

SVR participation rate increased steadily through the year, reaching ~68% by December. OEV and recapture volumes concentrated in Q4. Liquidation fee revenue spiked in October-November.

Pre-liquidation HF is tightly concentrated just below 1.0, with ~6000 events at 0.99-1.00. Liquidators execute immediately after positions become eligible. Post-liquidation HF clusters around 1.0-1.25.

V4 Counterfactual

We replay all 7,027 observed V3 liquidation events after SVR was introduced through two proposed V4 spoke designs:

• Borrower UX: THF = 1.05, tighter close factors → smaller liquidations, lower OEV per event
• Liquidation Efficiency: THF = 1.25, higher close factors → larger liquidations, more OEV per event

Each spoke is simulated using 500 Monte Carlo draws of parameters (uniformly distributed across plausible ranges). SVR participation and efficiency are predicted using logistic regression
and kernel-based estimators trained on observed V3 SVR data. Features are truncated to V3 bounds to avoid extrapolation.

The Liquidation Efficiency spoke diverges above V3 from the start, accelerating through Q4 as larger per-event bonuses compound. By December, the median LE path reaches approximately $11M, compared with roughly $6M for V3. The Borrower UX spoke flatlines well below V3 throughout; its tight target HF produces smaller liquidations with less surplus to recapture. The confidence bands show that even at the 10th percentile, Liquidation Efficiency outperforms V3 actual from October onward, while the 90th-percentile Borrower UX never reaches V3 levels.

Borrower UX is entirely negative: the distribution centres on ~-$4.8M, with a tight range, confirming that borrower-friendly sizing consistently reduces OEV. Liquidation Efficiency is predominantly positive: the median shift is approximately $4.7M, but the distribution is wider, reflecting greater sensitivity to parameter choices, particularly the bonus adjustment and liquidation fee. A small fraction of Liquidation Efficiency is drawn by underperforming V3.

SVR Recapture Overview

Scenario	Total (M USD)	vs V3 (Delta)	Uncertainty Range	Relative to V3
V3 Actual	6.09	-	-	1.00×
Borrower UX	1.30	−4.79	[0.50, 2.48]	0.21×
Liquidation Efficiency	10.92	+4.83	[7.66, 14.76]	1.79×

Total Revenue (liquidation fees included)

Scenario	Total (M USD)	vs V3 (Delta)	Uncertainty Range	Relative to V3
V3 Actual	9.16	-	-	1.00×
Borrower UX	3.13	−6.03	[1.47, 5.37]	0.34×
Liquidation Efficiency	27.60	+18.43	[19.30, 39.68]	3.01×

Liquidation Efficiency Spoke outperforms V3 across both views, raising SVR recapture to $10.92M (+$4.83M, 1.79×) and total liquidation revenue to $27.60M (+$18.43M, 3.01×), remaining above V3 even at the 10th percentile. Borrower UX Spoke consistently underperforms, reducing SVR recapture to $1.30M (−$4.79M, 0.21×) and total revenue to $3.13M (−$6.03M, 0.34×) due to tighter health factor targets and compressed bonuses. Narrow uncertainty under Borrower UX implies stable but lower outcomes, while Liquidation Efficiency exhibits convex upside and parameter sensitivity, highlighting the revenue–borrower protection tradeoff.

Conclusion

This analysis shows that Aave v4’s liquidation system can be tuned to exhibit markedly different behavior. By replaying real liquidation events from Aave v3 with different v4 settings, we demonstrate that simple parameter choices directly affect the amount of debt repaid, the attractiveness of liquidations to liquidators, and the protocol’s revenue, including from SVR.

Settings that favor borrowers result in smaller, less aggressive liquidations. This improves the experience for users whose positions are liquidated, but it also reduces incentives for liquidators and diminishes the surplus the protocol can capture. While this approach may perform well under stable market conditions, it can make liquidations less reliable during periods of high volatility.

In contrast, settings that prioritise liquidation efficiency make liquidations more appealing to execute, even in stressed markets. These configurations result in larger liquidations and higher protocol revenue, but they also mean more forceful action against borrowers once their positions become undercollateralized.

The main takeaway is that there is no single “right” liquidation configuration. Aave v4 enables the DAO to tailor liquidation behavior at the spoke level by adjusting parameters. This flexibility enables balancing user experience, system safety, and protocol sustainability in ways that align with the risk profile and objectives of each spoke.

Disclosure

The full methodology, simulation code, and dataset are available upon request.

TokenLogic did not receive any payment for this post.

Copyright

Copyright and related rights waived via CC0.

[JosueMpia]

@TokenLogic

The main takeaway is that there is no single “right” liquidation configuration. Aave v4 enables the DAO to tailor liquidation behavior at the spoke level by adjusting parameters. This flexibility enables balancing user experience, system safety, and protocol sustainability in ways that align with the risk profile and objectives of each spoke.

I think this is a great summary of this analysis. There is literally no single right liquidation configuration.

What v4 really unlocks here isn’t just higher liquidation revenue in aggregate, it’s the fact that liquidation behavior is now a design choice at the spoke level. That’s a big shift from v3, where everyone was effectively forced into the same trade-offs regardless of collateral quality or market structure.

The Borrower UX vs Liquidation Efficiency framing makes sense @TokenLogic , great job on this, so, let me brainstorm which spokes works well with which design.

Core Spoke

For Core, general-purpose markets, I don’t see much room for debate. This spoke will have the highest TVL, the most diverse collateral, and the most liquidation volume. Here, execution reliability matters more than being gentle. If liquidations fail during congestion or stress, the downside is bad debt that completely overwhelms any UX benefit from smaller liquidations. The numbers in the post make this pretty clear, Liquidation Efficiency is better in terms of SVR capture and total revenue.

My takeaway is that Core spoke should default to Liquidation Efficiency. This isn’t about extracting more from borrowers, it’s about making sure liquidations actually clear when they need to.

Frontier Spoke

Frontier spoke feel different to me. ETH, BTC, and similar assets have deep liquidity, lower volatility, and tend to attract more sophisticated borrowers who care a lot about capital efficiency. In that context, I actually think a Borrower UX leaning configuration makes sense. Yes, you give up liquidation revenue but these markets aren’t really competing on liquidation yield anyway. For high-quality collateral, smaller liquidations and lower bonuses feel like a reasonable trade-off if it keeps large borrowers happy and sticky.

Prime Spoke

Prime Spoke look safe most of the time, but taking at looks at the current Aave v4 testnet prime tokens, when something breaks, it will breaks fast. In those moments, you don’t want delayed incentives or hesitant liquidators. Because of that, I think stable spokes should still bias toward Liquidation Efficiency, just tuned carefully. Early stress activation and reliable execution matter more here than minimizing liquidation size, especially during depeg scenarios.

Yield Seeking Spoke

For yields with maybe some LRTs and other complex or newer assets, I’m firmly in the “be aggressive first” camp. These assets have layered risks and the highest chance of nonlinear failure. A single failed liquidation can easily turn into meaningful bad debt. In that world, maximizing execution probability is the priority, even if it’s harsher on borrowers.

Anyway, here is my final take:

Great work as usual for this. One thing I really like about v4 is that you can start these spokes with very conservative parameters and relax them over time as markets mature, instead of doing the opposite during a crisis. Aave v4 finally lets each spoke choose the liquidation behavior that actually matches its risk and goals.

The same protocol can now run:

• Highly efficient, revenue-optimized Core market.

• Borrower-friendly Prime spoke.

• Fast-acting, defensive Frontier spoke.

• Aggressively protected high-risk Yield seeking spoke.

All at the same time.

[Kene_Anode]

First off, I think this is an excellent quantitative framework for understanding V4’s liquidation engine.

There is an observation that prompts a question worth discussing, in my opinion.

The 3× revenue multiplier under the Liquidation Efficiency configuration is compelling, but I am pretty curious about the second-order effects on borrower behavior. If users anticipate more aggressive liquidations, we might see structural shifts toward more conservative LTV usage or migration to competing protocols, which could compress realized upside over time. Has the model considered any behavioral elasticity in borrowing demand as a function of perceived liquidation severity?

[stani]

Thanks for the detailed analysis, excellent work @TokenLogic.

This research demonstrates the flexibility of the Aave V4 liquidation engine and how it can be tuned to achieve outcomes favored by the community.

My main comment is that while capturing the maximum amount of liquidation spread may look great from a balance sheet perspective, it can actually hurt user acquisition if liquidations are not perceived as fair by borrowers.

As someone who reads every customer support ticket, I often notice that liquidations cause significant frustration, as pricing is seen as an efficiency play rather than fair risk reward compensation. We need to find the right balance between capturing rewards for collateral risk and ensuring that liquidations feel fair to borrowers, remain competitive in nature, and do not increase churn and reduce borrows. At the same time, ensuring sufficient incentives is especially important during the volatile market periods we have seen recently.

After the initial hub and spoke configuration is finalized, the next focus for risk managers should be assessing asset parameters, including liquidation fees and risk premiums.