Cred Protocol, a decentralized credit scoring startup has unveiled the results of its first automated credit scoring system for users of decentralized finance (DeFi).
Cred Protocol CEO Julian Gay outlined the results in a Twitter thread, which showed how Cred successfully utilized past transaction behavior on the Aave protocol to assess the creditworthiness of future borrowers based on on-chain behavior in the DeFi space.
By using machine learning to assess time-based account attributes and analyze the user’s past transaction behavior, Cred Protocol generates a health factor score that predicts the likelihood of future liquidation for a single address, which, according to Gay, was one of the strongest baseline creditworthiness predictors.
Cred Protocol claims to make decentralized finance more accessible to the world by implementing trustworthy credit scores that would see “anyone with an internet connection” and “a good financial reputation” gain access to loans.
Where borrowers and lenders have their loan worthiness assessed by a central authority such as a credit bureau, DeFi makes it possible to run financial services with a peer-to-peer (P2P) system, eliminating the idea of an intermediary or central authority.
Prominent DeFi researcher Chris Blec raised concerns that a borrower could use multiple Ethereum addresses to skirt credit scoring — to which Gay responded that a potential solution was in Beta.
Cred Protocol is a small nine-person team based out of San Francisco with additional “hubs” in New York and London. However, Gay says that he aims to bring DeFi technology to more than one billion people.
In a Medium post, Cred outlined its plans to grow from the Aave protocol and expand its data analysis to other lending protocols like Compound and MakerDAO.
Two years ago, blockchain lending protocol Teller raised $1 million in a seed funding round to incorporate traditional credit scores into DeFi.
In November 2021, Credit DeFi Alliance (CreDA) officially launched a credit rating service that would ascertain a user’s creditworthiness with data from multiple blockchains. CreDA was developed to work using the CreDA Oracle by evaluating records of past transactions carried out by the user across several blockchains with the help of artificial intelligence (AI).
Recently, P2P lending protocol RociFi labs concluded a seed funding of $2.7 million in partnership with asset management firm GoldenTree, which is aimed toward expanding on-chain credit ratings for decentralized finance.