Finance
Time Series Generation
Although real-world financial time-series data is abundant (stock prices, credit card transactions, ...), there is a growing need for representative synthetic time-series data[1] because synthetic data can: a.o. be shared without confidentiality concerns, it can be provided ‘clean’ from real-world noise, it can enable rare event modeling and it has been shown to improve the accuracy of supervised ML.
Monte Carlo simulations compute time evolution of individual samples directly from the SDE, which can be computationally demanding for complex portfolios and lacks flexibility.
State-of-the-art methods rely on generative adversarial networks (GAN).
PASQAL has developed two [3][4] unique and proprietary quantum methods which enable efficient generation of synthetic time-series data, by efficiently sampling from the solution of an SDE [5] based model. Our methods run either on large GPU clusters (quantum ready solution) or on our neutral atoms quantum computers (quantum solution).

Quantum in Real Life
%201.png)
Crédit Agricole CIB
"Quantum computing is radically different from almost everything we know and use today, in terms of theory, hardware and algorithms. This project will assemble many different competencies around the table: bankers, physicists, mathematicians, computer scientists, IT architects, all cooperating to this remarkable journey. This is a huge challenge, and we are confident to make it a success, jointly with our talented partners PASQAL and Multiverse Computing.”
Ali El Hamidi
Department Head Capital Markets Funding - Global Markets Division
Learn More
Explore Other Industries
%20(1).webp)