Crédit Agricole CIB has announced the successful completion of two quantum computing experiments focused on financial product valuation and credit risk assessment.
The bank collaborated with French quantum computing company Pasqal and Spanish algorithm specialist Multiverse Computing over the past 18 months. The first experiment examined the performance enhancements offered by quantum computing in the valuation of derivatives.
Recent studies have highlighted the advantages of using neural networks for such calculations. However, neural networks can be challenging to implement due to their high resource demands and lengthy processing times. Quantum-inspired algorithmic techniques can optimize both speed and memory usage during the training phase, resulting in quicker valuations and more accurate risk assessments, according to the bank.
The second experiment aimed to evaluate a quantum computer’s capacity to tackle a real-world problem while investigating performance variations based on the number of qubits utilized. The bank selected a practical use case: predicting a counterparty credit rating downgrade within a six to 15-month timeframe. While conventional computer technology and heuristics can yield favorable results, quantum parallelism theoretically allows for more efficient discovery of optimal solutions.
Ali El Hamidi, project sponsor at Crédit Agricole CIB, stated, “These two Proofs of Concept showcased the potential and current capabilities of quantum computing in the financial sector, even as the technology is still in its early stages. We leveraged this initiative to begin building internal expertise in preparation for a technological advancement that could significantly influence our sector’s competitiveness.”