Quantum computing promises the possibility of performing an infinite amount of computations almost instantaneously. But the technology is still in its early stages. Joy Macknight investigates how banks can cut through the hype to what is real. 

Quantum fit

The buzz around quantum technology’s potential to revolutionise computing has increased exponentially over the past few years, to the point where it would be unwise to ignore the disruptive role it could play in financial services.

Driving the hype is the interest in the field signalled by major technology firms such as Alibaba, IBM, Google and Microsoft. As Madhav Thattai, chief operating officer at Rigetti Computing, a full-stack quantum computing start-up launched in 2013, says: “When leading companies make significant investments and hire great people, then the market takes notice because those decisions aren’t taken lightly.”

Making the leap

The ‘big tech’ brands have established research partnerships with select universities in attempts to stay ahead of the field in quantum technology. For example, Google has teamed up with the University of California, Santa Barbara (UCSB); Microsoft is working with the University of Sydney and Stanford University as well as UCSB, where its Station Q lab is located; Alibaba Cloud has partnered with the Chinese Academy of Sciences, which includes five universities and 104 research institutes; and IBM is collaborating with Oxford University, Keio University and North Carolina State University, to name but a few.

In addition, governments are pumping vast investment into quantum research. The Canadian government has committed C$76m ($57m) to the University of Waterloo for the Transformative Quantum Technologies initiative. The Australian government, together with telecommunications company Telstra, Commonwealth Bank of Australia and the New South Wales government and New South Wales University, has invested A$83m ($61m), while the European Commission will launch a €1bn quantum flagship initiative later in 2018. China is building a $10bn quantum computing research complex, and the US is proposing to earmark $1.25bn over the next five years for its National Quantum Initiative.

Several new investment funds focused on quantum start-ups have cropped up, including Quantum Wave Fund and Quantum Valley Investments, while many start-ups are also receiving funding from established players. Berkeley-based Rigetti Computing, for example, raised $64m in March 2017 from the likes of Andreessen Horowitz and Vy Capital. Likewise, in November that year Vancouver-based 1QBit, which builds quantum and ‘quantum-inspired’ software, raised C$45m from Royal Bank of Scotland, Accenture, CME Group and Fujitsu.

Past the hype

Much of the excitement can be attributed to the fact that universal quantum computers are now live. Fundamentally different from D-Wave Systems’ annealing machines, commercially available since 2011 but designed for specific mathematical questions, these gate model machines can run arbitrary algorithms, which makes it possible to programme a quantum computer to solve many different problems.

In March, Google announced the largest universal quantum computer to date, a 72-quantum bit (qubit) chip named Bristlecone. IBM is not far behind and has inspired wide engagement by putting its machines online for experimentation purposes.

For example, in late 2015 IBM built its first five-qubit quantum computer, which could not do any calculation that could not be done on a laptop, “but it showed that the technology is beginning to truly exist and becoming rapidly more practical,” says Robert Wisnieff, chief technology officer, quantum computing, at IBM. In May 2016, IBM put the machine online and today the IBM ‘Q Experience’ has more than 85,000 users.

In November 2017, the same month it unveiled a 50-qubit computer, IBM put its 20-qubit machine online and began working with key customers to get them quantum-ready. The IBM Q Network currently has 12 members, including Barclays, JPMorgan, Mizuho Financial Group and MUFG Bank.

As it achieved a 20-qubit machine, IBM began exploring how workflows will evolve in future. “We can now anticipate how to re-architect workflows in many different optimisation areas – such as chemistry and machine learning – to take advantage of the near-term machines that begin to give us the power to solve problems that we literally can’t solve today,” says Mr Wisnieff.

For Bob Stolte, managing director and head of post-trade technology at JPMorgan Corporate & Investment Bank, quantum computing has reached a point where the material science has caught up with the theoretical research. “This is where it starts to get interesting,” he says. “We can run our problems on a quantum computer and determine whether we can benefit from this technology in the future.”

The current quantum computers are small and ‘noisy’, or susceptible to interference, so JPMorgan is only experimenting at this stage because the technology is not in a place where it is possible to gain a competitive advantage. “That reality is still likely to be years away, but maybe not 10 years away,” says Mr Stolte.

Picking up speed

There are two important concepts in quantum computing: superposition and entanglement. “Superposition is the ability of a quantum mechanical system to be in multiple states simultaneously, so that its state is described by a complex number. Entanglement is the ability of two or more qubits to share information about their state, so they know everything the other knows simultaneously,” says Mr Wisnieff. “That is where in many quantum algorithms we get this remarkable improvement in performance. We harness the entanglement to do the calculation.”

One way to get a sense of scale is to determine what size of a conventional computer it takes to model the performance of a quantum computer. IBM, for example, ran the Vulcan supercomputer at Lawrence Livermore National Laboratory in California for a weekend to model about 100 microseconds of calculations on a 50-qubit computer, reports Mr Wisnieff. “As a quantum computer gets even a little larger and has good coherence and control, the range of calculations we can do goes up dramatically,” he says. “That is a tipping point range that we begin to transition into being able to do things not doable by other means.”

But to preserve entanglement, a coherent quantum state must be maintained across the fabric of the quantum processor. “That is where the physics, material, fabrication, and the way in which we drive the machine, all combine to continue to extend the range of calculations that we can do. Not only have we been able to build larger machines in the past few years, but we have also continued to improve their performance. This is incredibly important, for it lets the machines solve larger and larger problems,” says Mr Wisnieff.

Recently, researchers have developed algorithms that can run on noisy quantum computers, called hybrid quantum classical algorithms. “Essentially, they work by using a quantum processor with classical computing,” says Mr Thattai. “And by leveraging both sets of resources, the algorithm splits the work between what a classical computer can handle well on its own and hand off parts of the problem to a quantum computer”. Like IBM, Rigetti Computing has put its eight- and 19-qubit machines online.

Mr Thattai expects some milestones to be passed in the next few years. First, the industry will cross the ‘quantum supremacy’ threshold, which is when a quantum computer can perform tasks that classical computers cannot. The next thing to happen will be the first examples of ‘quantum advantage’ applications. “This is where a quantum computer is used – most likely in a hybrid quantum-classical arrangement – in a way that provides commercial value for a particular application,” says Mr Thattai, adding that these could be seen inside five years.

Banking applications

Quantum computing has the potential to be transformative in financial services. “Anything that promises to exponentially increase the computing power available for certain types of problems is extremely interesting, as algorithms are used in all parts of the business, whether consumer, asset and wealth management, or corporate and investment banking,” says Mr Stolte. For example, Monte Carlo simulation, as used in option pricing and portfolio optimisation, is one area that he expects will be better carried out on a quantum computer.

Another optimisation example, says Maxwell Rounds, 1QBit’s financial application lead, is facilities location analysis. “This could take the form of determining where a bank puts its ATM machines to optimise both the cost of maintenance and ensuring customer utility,” he explains.

Anti-money laundering and fraud is an area that would benefit from quantum computing, says Narendra Mulani, chief analytics officer at Accenture Applied Intelligence. He estimates that quantum analytics could reduce the number of false positives by between 50% and 60% compared with the more sophisticated approaches that use machine learning and advanced analytics today.

He also highlights credit scoring, which is currently undergoing great change. “Optimal criteria selection is a space where quantum computing could fundamentally change the practicality of being able to create credit scores with external and internal data instantaneously, in a way that everyone hoped but has now become feasible to do,” he says.

In addition, Mr Mulani believes that quantum capabilities will allow banks to identify new asset classes, products and services that were previously inconceivable, mainly because they could not identify or manage them in a commercial manner. “What we are learning from the automation evolution is that, at first, the idea is to do what is done today but infinitely better. But then you discover that you can also do different things,” he says.

The quantum threat

Michele Mosca, co-founder of the Institute for Quantum Computing at the University of Waterloo, points to one of the dangers of quantum computing: a fully fault-tolerant quantum computer could potentially crack the current foundations of cyber security, the public key infrastructure. “This poses an unprecedented systemic risk, affecting not just banks but the whole internet,” he says. “The bright side is that maybe this will motivate us to fix cyber security. Security by design is not a broadly adopted principle. It is always reactionary: we wait for things to break and then we fix it. However, quantum is a more well-defined threat and will hopefully motivate us to adopt a better discipline at the foundation.”

While agreeing that quantum decryption is a serious concern, and should be addressed in a security audit, Mr Mulani believes it is not achievable in the near term. “We aren’t waking up worrying about whether a bank will be assaulted by a quantum-generated decryptor – that is not a feasible issue in the next decade,” he says. Mr Thattai agrees: “Banks should be more concerned about how their business applications will be disrupted in the next five or so years than about quantum security concerns that will happen further in the future.”

Mr Rounds points out that quantum computing could also play a part in advancing the field of cryptography. “The idea of quantum-powered encryption, which utilises the entanglement concept, is something worth highlighting. Because quantum entanglement can be susceptible to noise, the signal would fall apart if someone tried to tamper with the data,” he says, adding that quantum encryption is still a long way off.

Developing a strategy

For a top investment bank such as JPMorgan, having a research and development approach to quantum is vital. “We consider ourselves as an emerging technology company as well as a bank – it is a big part of our DNA. As such, we are serious about the level of investment in, and constant evaluation of, emerging technologies,” says Mr Stolte. “That mindset is the driver behind why we are involved in quantum and other emerging technologies.”

However, a bank’s quantum strategy depends on its ambition level, says the chief digital officer at one European bank. “Our level of ambition is to understand the technology and follow the markets. If there is something commercially usable, for example in portfolio or risk management, we would always be one of the first European banks to deploy it,” he says. “Our focus is to be early adopters, not leading edge researchers.”

Milos Dunjic, associate vice-president of payments innovation technology solutions at TD Bank, speaking to The Banker in a personal capacity, says that banks need to remain abreast of quantum developments, be engaged and stay close to the field. While TD Bank is not currently investing large sums in quantum, it understands the importance of the technology and is building in-house knowledge. As such, the Canadian bank has allocated several people to keep eye on it and stay close to the research facilities. “We are raising our knowledge level so that we can have intelligent conversations with the ultimate experts,” says Mr Dunjic.

The University of Waterloo’s Mr Mosca argues that, at a bare minimum, top financial players need to have part of their innovation team exploring quantum and developing in-house expertise. “It is getting risky now to not have any skin in the game because it will be hard to catch up when the breakthrough happens, say, in five years’ time,” he warns. “At that point an organisation will pay a premium to get into the game, whereas some competitors will have critical pieces of intellectual property and talent. It is also a hard time to build relationships with industry leaders.”

While noting that it is unlikely quantum computing will replace classical computing anytime soon, Accenture’s Mr Mulani agrees it is important for banks to understand the value for when quantum becomes commercialised. “We recommend that our clients identify use cases that will scale with quantum computing, do some testing and, in the longer run, create a roadmap for how quantum will transform the business,” he says. “Clients need to become aware, because three to five years comes on very quickly, particularly given everything else that is happening with artificial intelligence, blockchain and so on.”

PLEASE ENTER YOUR DETAILS TO WATCH THIS VIDEO

All fields are mandatory

The Banker is a service from the Financial Times. The Financial Times Ltd takes your privacy seriously.

Choose how you want us to contact you.

Invites and Offers from The Banker

Receive exclusive personalised event invitations, carefully curated offers and promotions from The Banker



For more information about how we use your data, please refer to our privacy and cookie policies.

Terms and conditions

Join our community

The Banker on Twitter