Why banks built quantum teams early
Finance runs on three computational workloads that map plausibly onto quantum algorithms: pricing (valuing derivatives by simulating many possible futures), risk (the same simulations, asking how bad the tail gets), and optimization (choosing portfolios or trade executions under constraints). All three burn enormous classical compute today, so even modest speedups would be worth real money, and banks can afford research teams as insurance.
The theoretical hook is quantum amplitude estimation: for Monte Carlo-style simulation it promises a quadratic speedup. A million classical samples versus roughly a thousand quantum ones for the same accuracy. Quadratic is real but modest, and it only pays once hardware overhead stops eating the advantage.
What is actually documented
The serious work is published research, not production trading. Bank teams (JPMorgan Chase's is the best known) publish papers with hardware partners on option pricing, risk analysis and optimization benchmarks. Goldman Sachs researchers have published on quantum Monte Carlo for derivatives, including honest resource estimates of how much better hardware must get before the method pays. In 2025 JPMorgan Chase and Quantinuum published a certified-randomness protocol in Nature. Randomness whose quantum origin can be mathematically verified, relevant to security and fair sampling.
Read bank quantum announcements with the same framework as everything else: a research paper is a milestone; "we explored" is not "we deployed"; and no public evidence yet shows a quantum computer beating classical systems on a production financial workload.
The realistic state today
Finance will likely be among the first commercial adopters IF general-purpose advantage arrives, because the workloads are already formalized, the data pipelines exist, and the buyers are sophisticated. Until then, expect a steady stream of research papers and pilot announcements. Real work, honestly valuable for readiness, and not yet advantage. Nothing on this page is investment advice; it is a map of who is doing documented research.
La mise en garde honnête: Classical solvers are excellent; credible claims must beat them end-to-end, and amplitude estimation's advantage needs substantial fault-tolerant resources.