HSBC & Haiqu Solve Quantum Finance’s Data Block with IBM Quantum computing has long promised to calculate market risks at lightning speed, but this vision has been blocked by the data bottleneck. Getting complex, real-world financial data into a quantum computer is notoriously difficult. Now, new joint research from HSBC and quantum software startup Haiqu suggests a breakthrough. Their findings prove that financial risk modelling applications are much closer to practical reality than previously thought. Solving the loading problem The research shows that financial institutions can provide financial data to a quantum computer through a process called Quantum State Preparation. Normally, encoding “heavy-tailed” distributions – mathematical models used to predict extreme market crashes – requires complex circuits that today’s quantum hardware simply can’t handle. These circuits become overwhelmed, causing the quantum computer to crash before it finishes the calculation. HSBC and Haiqu solved this by using a method called Matrix Product States. This allowed them to create shallow circuits, which are essentially a more streamlined, efficient way to pack data. Instead of trying to store every single piece of data in the computer’s memory at once, they used a sampling-based workflow that “avoids storing the full discretised dataset in classical memory, enabling larger encoding circuits to be generated,” reads the firms’ press release. Real-world testing on IBM hardware IBM provides access to quantum computing processors like the Eagle and Osprey, which are designed to handle increasingly complex workloads. HSBC and Haiqu ran their tests on this hardware. The study demonstrated the method’s efficacy across increasingly complex scales. At the 25-qubit level, a threshold where physical quantum processors begin to handle complex data, the team utilised IBM hardware to successfully reproduce probability distributions that satisfied all standard statistical benchmarks To test the system’s resilience against the errors common in larger processors,