A few weeks ago, I moderated a panel discussion on quantum computing use cases in finance at the Inside Quantum Technology conference in San Diego. It’s always insightful to hear from customers, and it was striking to see how closely the panelists’ perspectives echoed the patterns we’ve seen across our customer base. So, for those that couldn’t attend, I wanted to share what they had to say.
Quantum computing may be gaining attention as a major cybersecurity threat, but it also has many promising opportunities and use cases for finance. These include portfolio optimization, fraud detection, and financial modeling, to name a few. We’ve done some significant work ourselves in contributing research to advance these financial use cases for our customers.
In March, we published a paper with our customer, the global bank BBVA, in the New Journal of Physics that presented a new quantum algorithm and circuit design that significantly reduced the qubits required to speed up Monte Carlo simulations, which are commonly used for credit valuation adjustment and complex derivative pricing.
But as we heard from the panelists at IQT, the promise alone isn’t enough to adopt quantum computing. Investing in any new technology comes with risk. So, the first thing I wanted to know is where is the friction in adopting quantum, and what does the end user need to demystify the technology and justify the risk in investing?
One of the biggest friction points in quantum adoption is the immaturity of the technology. Without demonstrable results that deliver a competitive advantage, it can be difficult for executives to justify an investment. This view was shared by both representatives from financial enterprises on the panel: Peter Bordow of Wells Fargo and Roland Fejfar of Morgan Stanley.
It could be several years before quantum methods can outperform the state-of-the-art classical methods, and I can understand why some finance executives would want to wait to see results before investing. But the truth is, those results will only be achieved by the enterprises that didn’t wait. The first quantum competitive advantage will only be available to the companies that upgraded their infrastructure, upskilled or hired a quantum team, and built quantum-ready applications well before the hardware could deliver an advantage. Those that wait for results will spend years catching up with their more forward-thinking peers.
So how do we convince executives to invest in quantum technology in the absence of proven results? In Peter’s view, the answer is cybersecurity:
From my 30 years of experience in the tech industry, I’ve seen leadership take notice when you articulate a risk or threat to their thriving business. If you expound on the advantages of a new computing platform, they want to see results. They want demonstrable data and empirical evidence. But once you raise the red flag on the threats, you can get their attention and open the conversation to other potential advantages down the road. Discussing the threat gives you the opportunity to explain the capabilities of quantum computing that created that threat, which can open the executives’ eyes to the potential advantages that quantum can bring.
Peter Bordow, SVP & Principal Systems Architect for Advanced Technologies at Wells Fargo
For more information on the cybersecurity threat posed by quantum computers, check out our recent blog post answering the most frequently asked questions we get from customers.
Another point of friction that was brought up on the panel was the lack of clarity and depth in use cases. Roland in particular mentioned that it wasn’t clear which use cases were near-term, medium-, or long-term.
A good place to start when looking for viable use cases would be seeing what others are working on. “A lot can be garnered from looking externally at what use cases are gaining momentum,” said Peter. I would agree with him here, and direct curious readers in finance back to our own research in quantum algorithms for credit valuation adjustments that I mentioned earlier. Beyond that, we believe the earliest use cases in finance will be in optimizing the diversification and overall performance of portfolios to ensure the lowest possible risk for a desired investment gain. In this area, we recently published research that demonstrated how quantum-inspired generative models could outperform classical solvers for portfolio optimization problems.
At the same time, valuable use cases can also be surfaced outside of R&D by talking to internal business partners across departments. As Peter noted, “If you listen, they will tell you what their bottlenecks are.” We agree with this perspective, but as Roland pointed out, “There’s a difference between what’s desirable, which is defined by the business, and what’s feasible, which is defined by the technology providers — the two need to overlap.”
Use cases emerge from conversations, they emerge from listening.
Esperanza Cuenca Gomez, Head of Outreach and Strategy at Multiverse Computing
Roland is right to some degree, although the constraints on what’s feasible often come from within the business itself. Time and time again, we’ve seen that the rubber hits the road when we add time constraints and use real data from within the business units. Peter agreed, noting that “R&D teams have a lot to say about the practical application of use cases, given the resources and timing available. They set the guardrails for what’s possible.”
Once you’ve settled on a quantum use case, the next step is to build it into a practical application. The panelists I spoke to had different thresholds for defining what it means to go from use case to application. For Roland, “Once you start writing code, it goes from a theoretical concept to a practical solution.” Esperanza Cuenca Gomez, the Head of Outreach and Strategy at Multiverse Computing, had a different perspective, suggesting a use case becomes an application once it can provide value.
Peter had a more nuanced take, comparing the progression from use case to application to the progression of child to adult. “If your use case is a child, every child has a different maturity curve. The expectation of the value that that person provides to the world varies from person to person, use case to use case. While there’s structure to the innovation lifecycle, you have to be sensitive to the fact that some use cases may fizzle out and others will go on to win Nobel Prizes.”
There’s also the question of how “quantum” a quantum solution needs to be. At Zapata, we see quantum-inspired classical methods as part of one long interconnected trajectory with full-on quantum hardware. Many of our customers are thrilled to have quantum-inspired classical solutions that are forward-compatible with quantum hardware, while others have a quantum mandate and a longer time horizon — and although classical will be part of that, quantum hardware is essential within their purview. We believe building workflows with quantum-inspired methods today will allow enterprises to swap in more powerful quantum hardware in the future to reap an advantage.
With the quantum-inspired method, you can build the infrastructure you need to implement the quantum application and make it forward compatible with a real quantum computer later.
Roland Fejfar, Head Technology Business Development International at Morgan Stanley
Roland agreed that quantum-inspired was a necessary step in the evolution to quantum readiness. From his perspective, the quantum-inspired method is a byproduct of exploring what’s possible on an actual quantum computer. He added that “With the quantum-inspired method, you can build the infrastructure you need to implement the quantum application and make it forward compatible with a real quantum computer later.” We couldn’t agree more.
We ended the panel with a question that may be on many of your minds as you consider the merit of the current wave of hype around quantum computing: is it a bubble?
On one hand, you could think of a bubble in terms of the explosion in funding for quantum startups. When you look at the history of other innovative technologies, for instance the dot-com bubble or the early days of the automobile industry, there’s initially a massive boom in investment and new entrants, then the group winnows down to a small handful of winners. In Roland’s view, we’re still in the inflation stage of the prospective “quantum bubble,” and it’s unclear when or if it’s going to pop.
On the other hand, the rapid pace of technological advances can also give the impression of a bubble. In our view, this isn’t a bubble. Sure, there’s a lot of hype out there, but there’s certainly value to be created now in preparing organizations for the quantum future, and we just need to continue to deliver as an ecosystem. Roland agreed, noting that “The key with these long-term technologies is to maintain your focus and not dip in and out with the tide, because that actually does slow down the momentum.”
At Zapata, we believe that this momentum won’t slow down, but we would agree with Roland that enterprise adopters should maintain a long-term focus. Everything about quantum computing will take time: from building your quantum team, to prototyping and testing quantum algorithms, to adapting your IT stack to integrate with quantum technology. It takes patience and discipline to achieve a quantum advantage, but in the end, it will be worth it.
For our part, we’re focused on delivering value in the face of hype, and as Esperanza pointed out, that means providing clarity for the market: “We need to be very clear about what the technology is, what it can do, what it can’t do, and when it will do things.” By tamping down on hype, we can be clear-eyed on the hard realities of what it takes to achieve a quantum advantage and help our enterprise customers along their journey.
Sharing our understanding of the current state of quantum.