Octopus: Privacy-Preserving Collaborative Evaluation of Loan Stacking
With the rise of online lenders, the loan stacking problem has become a significant issue in the financial industry. One of the key steps in the fight against it is the querying of the loan history of a borrower from peer lenders. This is especially important in markets without a trusted credit bureau. To protect participants privacy and business interests, we want to hide borrower identities and lenders data from the loan originator, while simultaneously verifying that the borrower authorizes the query. In this paper, we propose Octopus, a distributed system to execute the query while meeting all the above security requirements. Theoretically, Octopus is sound. Practically, it integrates multiple optimizations to reduce communication and computation overhead. Evaluation shows that Octopus can run on 800 geographically distributed servers and can perform a query within about 0.5 seconds on average.