expected credoras counterparty scoring to be a simple Risk number someone calculates once and updates occasionally.
read through how it actually works and the picture is more Complicated then that in ways that matter for how much You can rely on it inside a newton Policy
credora builds counterparty scores using a combination of financial data credit history and onChain activity.
the score is meant to represent the risk profile of an entity on the other side of a transaction essentially a real time credit assessment applied to defi counterparties rather then traditional borrowers
what surprised me was.
how much of the input data is still offchain. to produce a meaningful credit score for an institutional counterparty credora needs access to financial statements collateral positions and business information that those counterparties choose to share.
the score is Only as good as the data the counterparty decides to disclose
that means theres a selection effect built into the scoring System. counterparties with strong financials have an incentive to share data and get a favorable score.
Counterparties with weaker profiles or less Transparency have an incentive to share less. the entities most likely to represent real counterparty risk are also the entities with the least incentive to Submit to the scoring process
i was actualy expecting this to be a cleaner data problem then it turned out to be. a score that depends on voluntary disclosure from the entities being scored is a diferent kind of tool then one derived from purely observable 0nchain behavior. both have uses
but treating them the same way inside a policy without Understanding that distinction seems like it could produce false confidence about what the score is actualy measuring
probaly still better then no counterparty check at all. the problem is that better then nothing and reliable enough to gate billions in vault capital are not the Same standard