On-chain credit scores could theoretically be used to adjust liquidation thresholds, giving users with good credit lower collateral ratios and larger buffers. However, very few projects have actually put this logic into practice. Most protocols either have not implemented it, or have not publicly disclosed their rules after implementation.
A leading global cryptocurrency platform,suitable for both beginners and experienced traders.
New user benefit: 20% off trading fees upon registration!！
How Credit Scores Affect Liquidation Thresholds: The Gap Between Ideal and Reality
The liquidation threshold is the parameter in lending protocols that decides when liquidation is triggered. If the value of collateral falls below this line, the user's position gets liquidated. Traditional over-collateralized protocols such as Aave and Compound use fixed liquidation thresholds: for ETH it is usually around 80%, and when the collateral ratio falls below that level, liquidation is triggered.
In theory, credit scores can make this parameter dynamic. The OCCR (On-Chain Credit Risk Score) model proposed in academic research aims to let protocols dynamically adjust LTV and liquidation thresholds based on a wallet's credit risk. Wallets with good credit could receive lower collateral ratio requirements and wider liquidation buffers. The score itself is a probability value representing how likely the wallet is to default, and the model calculates it using dimensions such as historical borrowing behavior, current asset conditions, and asset volatility.
But in reality, implementation of this mechanism is still very limited.
ARCx is currently the closest example to the credit-score-adjusted liquidation threshold logic. Its credit score consists of three parts: daily score rewards, survival score rewards, and liquidation score penalties. If a user has a liquidation record within 120 days, 250 points are deducted directly from a total score of 999. Once deducted, the user can no longer use the pools with the highest borrowing utilization. It is still a semi-credit product within the over-collateralization framework: the credit score affects how much you can borrow, but it does not directly change the liquidation trigger line itself.
How Credit Scores Could Change Liquidation Buffer Logic
If credit score mechanisms were fully implemented, the buffer would appear at two levels:
First: the starting collateral ratio could be lower. In protocols like Aave, the liquidation threshold is fixed. What users can do is maintain a high enough collateral ratio to leave themselves enough buffer. If credit scores could lower the starting collateral ratio requirement, the starting point would be closer to the liquidation threshold, but users with high credit scores would have more mechanisms to protect them from being liquidated. Academic frameworks clearly mention that one goal of credit scores is to encourage responsible borrowing behavior: users with good credit can get higher capital efficiency from the protocol.
Second: the liquidation process itself could be optimized. Euler's soft liquidation mechanism is the closest existing protocol approach to dynamic protection: when a position starts to become unhealthy, the protocol liquidates only part of the collateral instead of the whole position, leaving the borrower room to recover. Combined with credit scores, this could theoretically mean the higher the credit score, the gentler the soft liquidation execution - for example, reducing the single liquidation ratio or extending the recovery window. Chainflip also uses a batch liquidation mechanism, liquidating only enough to restore health rather than clearing everything at once.
Current State of Liquidation Threshold Adjustment
Most mainstream protocols (Aave, Compound, Libre) still use fixed liquidation thresholds, with differences mainly in trigger mechanisms and buffer period design:
Aave: uses Health Factor to measure position health. When Hf < 1, liquidation is triggered, and at most 50% of the position is liquidated.
Libre: after LTV reaches 80%, a 72-hour warning period begins. During this period, users can add collateral or repay debt; if the deadline passes, execution happens automatically.
Chainflip: has soft liquidation and hard liquidation thresholds. Soft liquidation executes in batches, and hard liquidation discounts can reach 5%.
The reason credit-score-based liquidation threshold adjustment has not become mainstream is mainly about data and models. On-chain credit scoring needs a large cross-protocol database of borrowing and repayment behavior to train models, and such a database does not exist yet. The OCCR model can make probabilistic assessments, but only if protocols are willing to adopt this standard, and so far no major protocol has publicly said it will do so.
Actual Impact on Borrowers
If you go to Aave or Compound to borrow now, credit scores will not change your liquidation threshold. Liquidation is triggered based on a fixed ratio between collateral value and debt, and it has nothing to do with whether you have a credit score.
But if you use a protocol with credit scoring such as ARCx, the credit score affects the borrowing utilization cap. After the Euler attack in 2023, the market's understanding of liquidation mechanisms deepened. Soft liquidation and batch liquidation gradually became standard in new protocols, but the trigger conditions for these mechanisms are still based on fixed LTV thresholds, not dynamic credit scores.
A leading global cryptocurrency platform,suitable for both beginners and experienced traders.
New user benefit: 20% off trading fees upon registration!！
How to Check in Practice
If you care about how credit scores affect liquidation thresholds, do two things:
Check whether the protocol you use supports dynamic LTV: on the protocol's borrowing page, see whether there are parameters related to Credit Score or Risk Profile. If not, the liquidation threshold is fixed.
If the protocol has a credit score feature (such as ARCx), check your score and the corresponding benefits: protocols usually publish the maximum LTV or borrowing utilization cap for different score levels, so you can see which level you are in.
Dynamic liquidation thresholds may not become widespread until on-chain credit infrastructure is built first. For now, it is still in the stage of academic discussion and early experiments.
