Home Bond screener Top picks Pricing Readings About

Readings

Author
Vladimir Tarantaev, CFA, PMP
Vladimir Tarantaev, a CFA expert in fixed income, has a strong track record in credit analysis at CIS banks and a diverse background in math-physics and astronomy.
Vladimir Tarantaev
Back
29.09.2025
HSBC and IBM Test Quantum Computing in Bond Trading
HSBC and IBM Test Quantum Computing in Bond Trading
52

The idea of quantum computing in markets once seemed like science fiction, but HSBC and IBM’s new results are remarkably concrete. By applying IBM’s quantum processors to more than a million real bond trades, the two firms showed that hybrid quantum-classical models can improve prediction accuracy by as much as 34%. This is too large to ignore.

Last week, HSBC and IBM announced what they describe as a breakthrough: the first successful trial of quantum computing applied to bond trading. Their experiment looked at a very practical question in the market — when a client asks for a price on a bond, can the bank predict whether the client will actually accept it and trade? This task, known as “fill prediction,” is central to algorithmic bond trading, where speed and accuracy directly affect profitability.

HSBC used more than one million trading requests (RFQs) for over 5,000 different European corporate bonds, covering the period from September 2023 to October 2024. Each trade request was broken down into about 216 features, such as the bond’s details, the time of day, and recent market activity. These were then fed into IBM’s Heron quantum processors, which transformed them into 327 new “quantum features.” The enriched dataset was processed by regular machine learning models like logistic regression and neural networks. 

The results were promising: models that included quantum features achieved up to 34% better accuracy in predicting whether a quote would be filled compared to traditional methods. For bond trading desks, this kind of improvement could have real benefits. If banks can better forecast which quotes are likely to be accepted, they can save time, cut costs, and capture more business.

Reactions in the media were upbeat, and IBM’s stock even gained on the news. Analysts highlighted this as a proof point for IBM’s quantum roadmap, while acknowledging that independent replication is still needed. Some academic voices were more skeptical, warning against hype and stressing that broader testing is essential before declaring a true “quantum advantage.”

For everyday investors, the bottom line is straightforward. This does not mean quantum computers will take over bond markets tomorrow. But it does show that big banks and tech firms are investing heavily in this field, and that the first concrete applications are emerging. If results like these hold up, quantum computing could eventually become part of the backbone of algorithmic trading, giving early movers like IBM and its partners a valuable edge.

Author
Vladimir Tarantaev, CFA, PMP
Vladimir Tarantaev, a CFA expert in fixed income, has a strong track record in credit analysis at CIS banks and a diverse background in math-physics and astronomy.
Vladimir Tarantaev
This article does not constitute investment advice or personal recommendation. Past performance is not a reliable indicator of future results. Bondfish does not recommend using the data and information provided as the only basis for making any investment decision. You should not make any investment decisions without first conducting your own research and considering your own financial situation.
Translate
Warning! The translation is automatic and may contain errors.