From use cases to a big data benchmarking framework in clearing houses and exchanges

Olga Lewandowska and Edgar Mai

Abstract

Although big data technology (BDT) is revolutionizing a variety of industries, including the financial sector, its architecture is heterogeneous: different companies apply different hardware, software and approaches. With advances in artificial intelligence and improved hardware computing power, a more holistic approach to benchmarking is needed. Comparing the extent of big data applications for a given use case within and between companies from a given sector of the economy can help us to identify the existing gaps in implementing this new technology. But how can we compare the use of BDT across companies, and what constitutes better usage or a greater extent of implementation? How can technology influence return on investment and other economic benchmarks? In this paper, we propose a conceptual framework that links the technical and business benchmarks in the domain of clearing houses and securities exchanges. The approach is as follows. First, we identify plausible use cases in the value chain of clearing houses and exchanges that are eligible for the application of BDT. We focus mostly on machine learning applications. Second, we present the eligible technical benchmarks that can be applied to those use cases. Finally, we map the economic benchmarks to the identified use cases and associated technical benchmarks in an integrated framework. We describe methodologies and tools to help assess and maximize the business benefits of BDT adoption for clearing houses and exchange operators, providing criteria for the selection of the most appropriate BDT solutions. Moreover, we present use cases in which BDT has already been implemented in leading clearing houses.