Financial Firms Seek Advantage in Quantum Computing-Inspired Algorithms

Quantum computing, which promises to dramatically increase processing speeds, is still years away from large-scale commercial deployment, but some financial services firms are turning to quantum-inspired technology for interim benefits.

Traditional computers store information as zeros or ones. Quantum computers use quantum bits, or qubits, which represent and store information in a quantum state which is a complex mixture of zero and one. Machines capable of supporting this quantum state have the potential to sort through a large number of possibilities in near real time, potentially allowing them to solve problems beyond the reach of today’s most advanced computers.

Quantum-inspired technology is a general term that relates to the use of certain algorithms that typically run on quantum computers on fast-processing classical computers. According to Carl Dukatz, quantum program manager at Accenture PLC, these types of algorithms are well suited for solving optimization problems, which are common in the financial services industry and include things like risk analysis and the pricing of derivatives.

Interest in quantum-inspired technology has grown as executives hear about developments in true quantum computing, analysts said. Companies such as HSBC Holdings PLC, Ally Financial Inc.

and Spanish multinational bank BBVA are turning to quantum-inspired technology for short-term benefits.

According to Troels Steenstrup, CTO of KPMG’s Global Quantum Hub, applications of quantum-inspired optimization problems can produce solutions 1-10% more accurate than existing approaches and can yield solutions two to three times faster.

According to Will Zeng, head of quantum research at Goldman Sachs Group Inc.,

it makes sense that the demand for financial services is so high because the industry has well-specified mathematical problems, in which incremental improvements in computing can make huge differences to the bottom line.

“There are certainly big, valuable problems that have a theoretical quantum advantage,” Dr. Zeng said.

Yet quantum computers today are far from ready for large-scale commercial exploitation. Qubits, the quantum version of a computer bit, are delicate, easily disturbed by changes in temperature, noise or frequency. The number of qubits running in quantum machines today remains relatively small, which means that experiments are currently limited to a narrow band of information.

When Google last year announced plans to create a commercial-grade quantum computer by 2029, it said it was aiming for a million-qubit machine, when its systems at the time had fewer than 100 qubits.

“There are certainly valuable big problems that have a theoretical quantum advantage.”

— Will Zeng, head of quantum research at Goldman Sachs

“The main focus in the near term is really on quantum-inspired type activity,” said Steve Suarez, HSBC’s global head of innovation, global functions. The London-based bank launched an official quantum program in August 2021.

Currently, the applications the bank is exploring involve portfolio optimization and pricing, said Philip Intallura, HSBC’s global business manager for quantum computing.

“Inspired technologies lend themselves quite well to this,” said Dr. Intallura.

Turning these processes into quantum-inspired solutions first involves rewriting the traditional algorithm into one that typically runs on quantum machines, according to KPMG’s Dr. Steenstrup.

Companies often do optimization problems using linear equations. In a quantum-inspired fashion, these linear equations are rewritten as quadratic equations, in which multiple variables can be multiplied by each other. This is one of the few differences between the initial equation and the quantum inspiration process, Dr Steenstrup said.

Once the algorithm is rewritten, it is a question of testing it on certain types of classic machines to determine if the new algorithm works faster and more efficiently than the traditional ones.

Ally Financial began work in this area in late 2021. Chief Information, Data and Digital Officer Sathish Muthukrishnan said the company has built quantum-inspired algorithms and is testing them. He said that currently the work is focused on areas such as pricing, portfolio optimization and other business use cases.

Muthukrishnan said he has yet to find a quantum-inspired solution that works so much better than a traditional equation that he has to rush to implement it.

“I also want to make sure I go to market with the most impactful quantum-inspired algorithm possible,” he added.

Testing these solutions is also a priority for Escolástico Sánchez, head of quantum discipline at BBVA. Dr. Sánchez said he was working on proof-of-concept for several solutions in this area, including an algorithm designed to help with portfolio management.

Running the algorithm on past datasets has shown promising results, Dr. Sánchez said, although the bank needs to do more work to ensure the solution will work on current and future datasets. before creating a plan to implement.

In terms of true quantum, Dr. Sánchez said: “Sooner or later, I think this material will be valuable and robust enough to solve some problems better than the classic. [computers].”

Write to Isabelle Bousquette at [email protected]

Copyright ©2022 Dow Jones & Company, Inc. All rights reserved. 87990cbe856818d5eddac44c7b1cdeb8

Sharon D. Cole