Paper 2022/1424

DeFi That Defies: Imported Off-Chain Metrics and Pseudonymous On-Chain Activity

Abstract

Traditional finance quantifies risk by collecting and vetting reputation information for an individual, such as credit scores or payment history. While decentralized finance (DeFi) is an exceptionally well-suited application of permissionless blockchains, it is severely constrained in its ability to reconcile identities and quantify associated transaction risk directly on-chain. Opening the ecosystem to a broad range of use cases requires consistent pseudonymity and quantifiable reputation. This paper defies the status quo: exploring methods of assessing risk on-chain by efficiently integrating off-chain identity- and attribute- verification and on-chain transaction activity. We achieve this while preserving individual privacy within a competitive and fair environment and retaining compatibility with existing platforms such as Ethereum. Even though blockchains are inherently public, our solution gives users control over release of information that pertains to them. Consequently, our contribution focuses on customized methods that balance the degree of disclosure of provably-sourced user information against the likelihood of the user successfully gaining access to a desired resource, such as a loan under suitable terms. Our solution is consistent with the zero-trust model in that it imports explicit trust from recognized sources through relevant metrics that are subject to continuous update.