Scenario Analysis of Secure Multi-party Computation implementation in EU-based multinational banks

Abstract

EU banks have a dual obligation to generate and protect assets as a due diligence for their customers, their business longevity, and in accordance with regulatory bodies. The effective use of data presents opportunities for banks to meet their obligations in improving financial risk and also ensuring business growth and continuity in their offerings. For these reasons, there is value in acquiring and using mixed or aggregated data sets from external sources as a means to extrapolate and gain information and knowledge. However, the balance between data privacy and data use remains a challenge for banks which impedes on their ability to proficiently and resourcefully generate and protect their assets. Privacy-preserving data sharing methods such as Secure Multiparty Computation or MPC may be a viable solution for enterprises facing this issue. Through the use of multiple cryptography protocols and computational algorithms, MPC is a technology that enables parties to anonymously compute functions on shared data without demanding a trusted third party. This study aimed to explore the potential business impact of MPC whereby semi-structured interviews were conducted with expert stakeholders from three major EU multi-national banks. The possible future outcomes of MPC implementation is determined via a scenario analysis over time horizon of five years (2020-2025). Based on value creation theories from inter-organizational systems and business model disciplines, a conceptual model for MPC implementation outcomes was developed to guide the scenario analysis process. This study found that there are generally four archetypes of scenarios for MPC implementation in banks leading to improved internal efficiency of banks and/or (in)direct MPC enabled business models. Besides known MPC use-cases in money laundering initiatives and security risk frameworks, new MPC business cases were found for data-driven "as a service" models and (cross)industry data platforms. This study also identified new business model components based on an existing taxonomy of data-driven business models.