Quantum computing interests the French central bank in more ways than one. First on the risk aspect. Users of encryption solutions, such as the Banque de France, must anticipate threats.
At the end of 2022, via its innovation lab, the BdF was experimenting with a post-quantum security solution for data exchanges. But quantum can also be a source of new opportunities, for example in portfolio management.
Opportunities, but persistent challenges
The financial institution conducted a first exploration by focusing on quantum algorithms for the optimization of asset portfolios. These experiments, of which the Bank provides a report, included an algorithmic implementation on a quantum computer simulator.
From this research work, the central bank draws various lessons. First of all, it underlines the interest of hybrid algorithms to answer the problems of optimization of portfolios of assets.
The approach therefore consists in combining classical and quantum algorithms – of the Variational Quantum Eigensolver (VQE) or Quantum Approximate Optimization Algorithm type. The Banque de France has also confirmed the possibility of implementing a VQE algorithm on a quantum simulator.
The results obtained are valid. Their validity was verified “by comparison with the result of the classical approach known as the efficient portfolio, obtained by Monte Carlo (MC) simulation. »
Possible investigations on stress testing
However, it is not tomorrow that quantum computing will disrupt portfolio management. In its summary note, the institution underlines the “difficulty, in the state of technology, of programming quantum algorithms. »
Another reported problem is the lack of quantum memory (QRAM) to store data and tasks to be performed. Existing simulators also have limitations. In particular, they do not make it possible to obtain a practical estimate of the error rates and the calculation speed.
However, the Banque de France already foresees other applications of quantum, in particular for stress testing. She identifies it as a use case to be investigated for Quantum Monte Carlo or Quantum Annealing type algorithms.
Public finance players are of course not the only ones to explore the potential of quantum computing. JPMorgan is interested, for example, in financial modeling methodologies in pricing and risk analysis.
Private banks active in quantum
Goldman Sachs develops sophisticated instruments applied to pricing. Wells Fargo uses time-series machine learning to predict stock market indices and other financial assets.
In France, Crédit Agricole CIB signed a noteworthy partnership in 2021 with two quantum nuggets: Pasqal and Multiverse Computing. The objective is to achieve the improvement of algorithms in the areas of capital markets and risk management.
The applications of quantum in finance are multiple. They also reside in the areas of the fight against fraud, credit scoring and customer segmentation. Quantum computing should, in theory, speed up the functioning of these complex models.