In a financial milestone, HSBC and IBM used quantum computing to execute algorithmic bond trades with real market data, marking the first demonstration of quantum advantage in active financial markets. Announced in September 2025, the trial showed that quantum systems can improve trading performance.
By combining IBM’s quantum technology with classical computing models, the joint project boosted the accuracy of predicting successful customer trades in Europe’s corporate bond market by up to 34% compared with standard methods. The results point to quantum computing’s emerging role in addressing complex financial problems and improving decision-making efficiency.
Hybrid approach and validation
The experiment used a hybrid framework that combined quantum and classical computing to address the intricate dynamics of bond trading. Quantum algorithms were applied to analyze large and noisy datasets, refining predictive accuracy beyond what traditional models typically achieve.
The system was validated using production-scale trading data across several IBM quantum computers. Central to the effort was IBM’s Heron processor, the company’s most advanced quantum system, which enhanced the detection of subtle market signals often lost in standard workflows. The findings demonstrated the potential of quantum–classical collaboration to strengthen predictive modeling in financial operations.
What is algorithmic trading?
Algorithmic trading, also known as automated trading, utilizes computer programs that execute orders based on predefined criteria such as timing, price, or volume, enabling transactions at speeds unattainable by human traders. The practice is now a mainstay of global financial markets, improving liquidity and promoting more systematic trading by reducing the influence of human emotion.
In corporate bond markets, algorithmic systems generate rapid price quotes for client inquiries, allowing banks to manage high transaction volumes efficiently while reserving human oversight for complex deals. However, fluctuating market conditions and large volumes of unstructured data continue to challenge the accuracy of classical trading models. The HSBC–IBM study indicates that quantum computing could help overcome these limitations by offering more precise forecasting and execution.
What is quantum computing?
Quantum computing applies the principles of quantum mechanics to process information in ways that surpass the limits of conventional systems. Using qubits that can represent multiple states simultaneously, quantum machines can analyze vast combinations of variables at once, making them suitable for highly complex computations.
Though still developing, quantum systems are beginning to show practical value. IBM’s superconducting quantum processors operate at extremely low temperatures to maintain stability and often work alongside classical computers in hybrid workflows.
IBM’s Heron processor, used in the HSBC trial, analyzed live trading data to test how quantum and classical computing can complement each other in financial modeling. Beyond finance, experts see potential applications in materials research, chemistry, and machine learning as the technology continues to mature.
HSBC and IBM: Innovation and intellectual property landscape
As quantum technologies transition from research to practical use, IBM and HSBC represent early adopters exploring their potential in financial applications. Their collaboration highlights the intersection of technological development and applied finance, offering insight into how each organization’s patent activity reflects its evolving priorities in quantum computing, AI, and data security.
IBM’s quantum edge
Over the years, IBM has heavily pursued hybrid cloud computing, artificial intelligence, and quantum technologies. As of October 2025, the company has a market capitalization of about US$269.6 billion, highlighting its continued influence in enterprise technology and scientific innovation.
IBM’s global patent filings rose steadily from 2015 to 2020, coinciding with the company’s strategic pivot toward hybrid cloud and artificial intelligence. The surge in 2020 reflects heightened investment in research and development during a challenging year marked by the pandemic.
Despite a decline in revenue that year, IBM expanded its hybrid cloud platform, improved profit margins, and strengthened its position through the integration of Red Hat and other AI-focused acquisitions.
IBM patents on quantum computing
IBM’s patent activity in quantum computing has grown steadily over the past decade. This trajectory aligns with IBM’s Quantum Roadmap, which charts the company’s path toward quantum-centric supercomputing. After introducing quantum computation in 2023 and improving circuit performance in 2024, IBM plans to demonstrate a fully integrated quantum-centric supercomputer in 2025. The shift in patent activity mirrors this strategic focus, emphasizing system integration and scalability as IBM prepares to translate years of research into the next era of high-performance computing.
HSBC’s Innovation Momentum
HSBC stands as one of the world’s largest banking and financial services organizations, with operations spanning 62 countries and territories. As of June 2025, the bank managed assets totaling US$3.2 trillion, emphasizing its scale and global reach. This vast footprint and financial strength provide the foundation for its expanding innovation agenda, reflected in a growing portfolio of patents and trademarks across emerging technologies and digital finance.
HSBC’s 2022 patent and trademark activity reflected a clear move into the metaverse, highlighted by its partnership with The Sandbox. The same year, HSBC filed trademarks covering digital media, NFTs, virtual currency exchange, and virtual credit card services, alongside traditional offerings like mutual funds and bill payments. The filings signaled its bid to secure a brand presence in virtual worlds and capture early ground in the fast-growing Web3 economy.
HSBC innovations in quantum computing
In 2023, HSBC expanded into advanced computing and security-focused innovations. Patents included systems for processing network security alerts, a “data mirror” for data management, and quantum graph optimization for portfolio management.
Momentum accelerated in 2024, the bank’s most active year for intellectual property filings. Many applications centered on AI and machine learning, such as adversarial learning, data field validation, code transpilation, and process-to-API mapping. Others addressed data routing, transaction banking, and secure data sharing.
The graph highlights HSBC’s growing commitment to quantum computing within its wider patent portfolio. While overall filings have risen steadily in recent years, quantum-related patents only began appearing in 2021, signaling a strategic pivot toward advanced computational technologies. Activity reached a sharp peak in 2023, coinciding with HSBC’s entry into the UK’s quantum-secure metro network and its trial of quantum protection for AI-driven foreign exchange trading.
HSBC: Top Jurisdictions
The graph shows HSBC concentrating its patent strategy in China and the United States, highlighting the importance of protecting innovation in two of the world’s most competitive markets.
At the same time, HSBC is spreading its reach through filings with WIPO and in markets like Hong Kong, the UK, and Taiwan, ensuring global coverage. More selective activity in Europe and Singapore points to a focused approach, targeting hubs where intellectual property protection can deliver the greatest strategic value.
HSBC: Top Law Firms
HSBC’s patent portfolio is supported by a strong network of legal representatives, with Shanghai Patent & Trademark emerging as the leading partner. Jiaquan IP and Shenzhen Weiqi follow, highlighting HSBC’s reliance on China-based firms for much of its intellectual property protection. Hangzhou Zhijian and U.S.-based Nixon Peabody also play significant roles, reflecting a mix of domestic and international expertise.
Smaller but notable contributors include Kangxin Partners, Beijing Qichuangzhiheng, and Chengdu Jiuding Tianyuan. These firms provide additional regional and specialized support, ensuring HSBC’s innovation efforts are safeguarded across diverse markets and jurisdictions.
HSBC: Top Technology Areas
HSBC’s patent filings are heavily concentrated in electrical digital data processing (G06F), which accounts for more than a quarter of its portfolio. Information and communication technologies adapted for administrative, commercial, financial, managerial, or supervisory use (G06Q) and transmission of digital information (H04L) follow, highlighting HSBC’s emphasis on financial technologies, secure communications, and digital infrastructure.
Beyond these dominant categories, HSBC’s portfolio extends into computing arrangements based on specific computational models (G06N) and reduction of greenhouse gases (Y02E), reflecting strategic interests in AI-driven systems and sustainability. Additional areas include generation of electric power using photovoltaic technology (H02S), preparation for medical, dental, or toiletry purposes (A61K), and solar heat collectors (F24S). Smaller but notable shares in chemical or physical processes (B01J), treatment of water and wastewater (C02F), and climate change mitigation technologies for buildings (Y02B) further highlight diversification into environmental and industrial innovation.
HSBC Patents on Quantum Computing
HSBC’s latest quantum computing patents target some of the field’s toughest hurdles, from reducing error-prone SWAP gates in quantum circuits to boosting the efficiency of variational algorithms and streamlining Monte Carlo simulations. These filings highlight its bid to secure a foothold in quantum innovation and position itself at the forefront of finance and advanced computing.
Quantum circuit compression to reduce SWAP gate overhead
Quantum computing promises major advances in computation, but today’s hardware still struggles with practical limitations such as qubit connectivity. Many quantum algorithms require “non-local” interactions between qubits that are not physically adjacent, which typically necessitates inserting SWAP gates. These gates increase circuit depth, add error risk, and reduce overall performance. A more efficient way to restructure circuits by eliminating redundant or costly SWAP operations can make quantum workloads run faster and more reliably.
U.S. Pat App. No. 2024/0311669 describes a method for compressing quantum circuits by identifying sections of a circuit that match a predefined gate arrangement containing non-local interactions and replacing them with an equivalent, but simpler, arrangement. The approach leverages equivalence relations, such as those derived from the pentagon equation, to remove SWAP gates while preserving functional correctness. By systematically detecting and replacing inefficient subcircuits, the method reduces gate count and circuit depth, thereby improving execution on quantum hardware.
The application, titled “Quantum circuit compression,” was filed on October 20, 2023. The patent lists Christos Aravanis, Georgios Korpas, Jakub Marecek, and Philip Intallura as inventors.
Optimization techniques for variational quantum algorithms
Variational quantum algorithms (VQAs) are a leading approach for solving optimization problems on noisy intermediate-scale quantum hardware, but they face challenges such as barren plateaus and stalled searches in complex parameter spaces. Traditional optimization methods can become trapped, wasting computational resources and limiting accuracy. A more adaptive method that detects when searches stall and reassigns them new starting points can help sustain progress and improve outcomes in hybrid quantum-classical systems.
U.S. Pat. App. No.. 2025/0173596 describes a method for executing VQAs using parallel optimization searches. The process combines classical and quantum computing, with each search updating circuit parameters based on cost function evaluations. When a search stalls, detected through gradient information, it is reassigned either a fresh random set of parameters or values borrowed from another ongoing search. This dynamic reassignment prevents wasted effort on stagnant paths and improves the efficiency of finding optimal solutions.
The application, titled “Optimisation techniques for variational quantum algorithms using noisy quantum processing hardware,” was filed on November 27, 2024. The patent lists Georgios Korpas, Jakub Marecek, Philip Intallura, Daniel Mastropietro, and Vyacheslav Kungurtsev as inventors.
Quantum Monte Carlo processing for advanced simulations
Quantum Monte Carlo (QMC) techniques are widely used to simulate complex systems across physics, chemistry, and finance, but their practical application on quantum hardware remains limited. Noise, inefficiency in sampling, and heavy computational demands restrict their scalability. A framework that streamlines QMC through quantum walks and arithmetic operations promises to improve accuracy and efficiency, making these simulations more practical on near-term quantum processors.
U.S. Pat. App. No. 2024/0428112 describes systems and methods for quantum Monte Carlo processing. The invention uses a quantum processor to load variables and probability distributions, then initiates a quantum walk made up of predetermined steps tied to quantum arithmetic operations. At the end of this sequence, the system estimates the state of the quantum system to determine a target variable, providing a more efficient route to QMC outcomes. The disclosure also extends to methods and non-transitory computer-readable media implementing these procedures, broadening its applicability across hardware and software platforms.
The application, titled “Systems and methods for quantum Monte Carlo processing,” was filed on January 4, 2024. The inventors listed are Bing Zhu, Ziyuan Li, Yong Xia, Mianmian Zhang, Si Yuan Jin, Kar Yan Tam, Yuhan Huang, and Qiming Shao.
Legal representation for all three patents was handled by Dentons Durham Jones Pinegar, with attorneys Brick Power, Gregory Johnson, Michael Dukes et al. listed in the patent application.
