Comment on page
Risk models
Appropriate risk assessment frameworks ensure that the probability of a position going under-collateralized (a default event) is below a certain threshold while maximizing borrower capacity and avoiding premature liquidations.
In lending applications, two main parameters need to be optimized to fulfill the needs of both borrowers and lenders: the collateral threshold and liquidation threshold. These parameters serve as a cushion to mitigate the risk for both parties.
- The collateralization threshold (also known as LTV, loan-to-value, or haircut in TradFi) is the percentage difference between an asset's market value and the amount that can be used as collateral for a loan. It is the initial safety margin and is a function of the volatility of underlying assets and the price correlation between different assets.
- The liquidation threshold is the percentage difference between the collateral current market value and the loan amount. It is a function of the available liquidity on-chain and a parameter to optimize LGD (loss given default).
These values are not arbitrary. They depend on an asset's risk profile, market conditions, and user preferences. We are working on dynamic models based on on-chain data to estimate optimal parameters. However, until our research reaches maturity, we are following traditional best practices to determine these parameters.
Our risk models will be mainly off-chain at the beginning. However, we will bring our risk models on-chain in due time. Arcadia values open-source, and we plan on making our risk models and research public.
Last modified 7mo ago