A financial risk evaluation service for integrating private portfolios securely

Abstract

Value at Risk (VaR) is a widely accepted measure of counting the risk of loss on a specific portfolio of a collection of financial assets such as stocks, bonds and cash, held by a financial institution or individual. It is quite useful to compute "Integrated VaR", a VaR for a portfolio which combines all of the portfolios in the group. However, the downside of the integration is that every financial institution needs to disclose its own portfolio to others to obtain integrated portfolio. We propose a novel approach to realize an efficient and secure protocol for N participants who hold each portfolio s1,⋯,sN to compute Integrated VaR for integrated portfolio s=s1+⋯+sN collaboratively without disclosing the value of each portfolio. This allows the financial institutions or individuals to compute cross-institutional risk more accurately to make the financial risk more transparently. Our protocol achieves high efficiency by specializing message structure and exchange to compute an integrated VaR in a privacy preserving manner.