AI Liquidations Model
Last updated
Last updated
We have developed an arbitrary and permissionless Machine Learning model, which should help Sirio users reduce the risk of their loans being liquidated. But how? When you take out a loan on a lending protocol, there are dozens of variables to consider, such as general market conditions, ecosystem market conditions, liquidity, the tokens you use as collateral, the allocations of each token used as collateral, and so on. Our model is capable of analyzing these and many other variables to estimate the probability that the loan you are about to take out will be liquidated, with high risks for the integrity of the tokens.
The operation of the model is quite straightforward: when you borrow a specific asset by following the steps in the dedicated section, and when you click on the 'Borrow' button, before proceeding with the transaction, a popup will appear showing a percentage. This represents the probability that your loan will be liquidated. Afterwards, you can decide whether to proceed with the loan or to decrease the amount of assets borrowed/increase the collateral, in a more informed manner.