Restaking Risk Management
Ebisu's LRT Parameterization & AVS Risk Underwriting Framework
For a backround on LRT risk, check out our article with Shoal Research.
In order to properly underwrite the AVS slashing risk of a given LRTs, Ebisu evaluates
1) AVS Individual Risk
A normalized risk score is calculated with a formula specific to the type of AVS based on its underlying infrastructure composition. This formula accounts for factors such as restaking modality, operator reputation, audits, total restaked capital, TVL, business model etc.
Try a simple version of our AVS risk score calculator here https://eigenavsrisk.streamlit.app/
The end score is a number between 0-100 that illustrates the risk level of the AVS incurring a fault (either malicious or non-malicious) resulting in capital delegated to its operators being slashed.
2) AVS Portfolio Risk in a Pooled Security Environment
Learn more about pooled vs attributable security here https://x.com/0xMeir/status/1737892943745216914
Given EigenLayer currently only offers pooled security, it's the role of an LRT provider to construct an AVS portfolio. To underwrite AVS portfolio slashing risk, Ebisu considers:
Number of AVSs being delegated to by an LRT.
Number of LRT AVSs in the the same category.
AVS Portfolio aggregate risk as a function of their individual risk.
Particularly at this early stage, LRTs delegating to an excessive number of AVSs, which are in the same service category and have high individual infra risks, are significantly risk exposed.
3) Recommendations on Borrowing Parameters per LRT
AVS risk is one type of risk factored into setting borrowing parameters for each of Ebisu Money's supported collateral assets. Other factors taken into consideration are liquidity, smart contract risk, withdrawal queue implementation, and more. Ebisu will support multiple LRTs as deposit collateral. The goal of collateral parametrization is to ensure that LRTs that are more likely to be devalued via slashing events or otherwise do not make up a large portion of the collateral backing the supply of ebUSD.
The risker the LRT, the lower its protocol-wide deposit cap.
The risker the LRT, the higher its MCR is ie. users will be able to borrow less debt against risky LRTs
Ebisu has collaborated with TokenSight to develop empirical ways to quantify risk for each step above in order to best parameterize borrowing.
Last updated