Crédit Agricole CIB, Multiverse, and Pasqal Partner to Explore Potential QC Applications in Finance

Crédit Agricole CIB is the corporate and investment banking arm of Credit Agricole Group based in Paris, France with over 8900 employees worldwide. They have been investigating potential finance use cases for quantum computing and have been working over the past year and a half with Multiverse Computing and Pasqal on two different ones. The goal was to develop Proof-of-Concept implementations using the currently available quantum processors with about 50 qubits and then scale the implementations up for production use when larger processors with roughly 300 qubits are available in the next one to two years.

The first application is a credit rating problem to estimate the probability of debtors reimbursing their loans and identify borrowers with a higher credit risk to predict a credit rating downgrade. A large amount of data can be collected relevant to this decision and it is a widely studied problem for the use of classical machine learning. Although the classical techniques can provide good results, they may not be optimum and the bank believes they might be able to find more optimal solutions using quantum-enhanced machine learning. The results have been published in a paper posted on arXiv titled Financial Risk Management on a Neutral Atom Quantum Processor and show the potential for a future quantum approach to achieve better interpretability and comparable training times versus a state-of-the-art classical algorithm called Random Forest.

A second application involves the valuation of derivatives in capital markets using neural networks. Key issues with standard classical techniques include the long times and large memories needed to accomplish the training function. In this application, Multiverse used a quantum-inspired Tensor Neural Network technique running on a classical computer to optimize the speed and memory requirements. In the future, this technique may be capable of migrating to a real quantum computer for further improvements and use in a daily production environment. Technical papers describing this research is available in two pre-print papers posted on arXiv here and here.

More information about the research the three companies have performed regarding these finance applications is available in a news release provided by Crédit Agricole CIB that can be found here.

January 27, 2023