Risk management: quantum computers roll dice at the casino
Getting your Trinity Audio player ready...

Quantum computers are often pitched as being able to perform calculations – or at least certain kinds of calculations – more efficiently than classical machines. However, for companies, the advantage of using quantum technology could relate to more than just a speedup of their operations. Monte Carlo integration engines give users the opportunity to run risk management algorithms on qubits, which could give businesses a clearer view of the road ahead.
What is Monte Carlo Simulation?
Understanding the likelihood of an event taking place is helpful across a range of industries. Stockbrokers, in particular, have a keen interest in how different assets could fare as market conditions develop. And trades are based not just on price expectations, but on whether those thresholds will be reached at a certain point in time – for example, when considering financial derivatives such as futures and options contracts.
Naturally, stock prices can move along various paths and are dependent on numerous parameters, to a greater or lesser extent. What’s more, considering portfolio management, the price of one stock may be dependent on another, which can send analysts in circles if they approach the modeling process as a conventional maths problem.
An alternative way of picturing how pathdependent behavior could play out is to use Monte Carlo methods, which rely on the fact that a random sample can – under the right conditions – shine a useful light on future events. “We care about quantities right down the tail,” Steven Herbert – Head of Algorithms at Quantinuum, a fullstack quantum computing company combining Cambridge Quantum and Honeywell Quantum Solutions, told TechHQ.
Herbert and his colleagues have built what they dub a Modular Engine for Quantum Monte Carlo Integration that gives users in financial markets and other sectors a computational tool for evaluating the interplay between parameters in complex systems.
Simulations can be run to consider large numbers of possible paths based on random conditions to reflect the likelihood of events, such as the probability of catastrophic failure. Repeating the process over and over again emphasizes strong signals in the results. And data could help companies to determine how much capital to hold to prevent default. “Risk management is going to be the killer app,” said Herbert.
Quantum computing is based on the idea that it’s possible to prepare a quantum state that encodes every answer for every input. Given sufficient coherence time, measurements indicate the most likely outcome. However, qubits are highly susceptible to noise and can lose their state, which is a challenge for developers of quantum computing technology.
Quantinuum gets around this problem in its quantum Monte Carlo engine by chunking the computation into shorter segments. Rather than consider the Monte Carlo integral as a whole, it’s possible to decompose the expression as a Fourier series, with each harmonic estimated using a shallow quantum circuit.
The strategy allows the team to run its computation as blocks that fit into the coherence time. “Fourier series computation works really nicely,” notes Herbert.
The Quantum Revolution is here! 🧠 Discover how Quantum Computing is supercharging AI in finance.
GeorgesOlivier Reymond, CEO of @pasqalio, and Christophe Michel from @CreditAgricole dive into Quantum Accelerated AI.
Watch now: https://t.co/SZE9zF8Kvt#B2BRocks #Quantum #AI
— OVHcloud (@OVHcloud) September 21, 2023
Another approach to squeezing as much performance as possible from current hardware, known as Noisy IntermediateScale Quantum (NISQ) devices, is the use of TKET – an SDK that optimizes quantum gate design.
Quantum algorithms can be assembled using wellknown building blocks, but a less intuitive gate arrangement could be better suited to the task, and TKET solves that design problem. “It’s highly optimized and means that you’re getting as much computation as you can in the box,” Herbert explains.
Clients want the higher precision that techniques such as quantum Monte Carlo integration provide to be straightforward to adopt for use in financial risk management and to benefit other use cases. And hiding the complexity of building quantum circuits in software is a step in that direction.
In fact, analysts may one day sit down and not even know that they are using a quantum computer as the user experience evolves. But, as Herbert points out, quantum computing is a whole new industry that’s being built from scratch.
Future of risk management and other applications
Given how novel the technology is, it’s not surprising that a learning curve still exists in understanding how enterprises can benefit fully from advances in the field. On TechHQ, we’ve written about how quantum computers can help to solve supply chain issues and improve logistics, but the applications don’t stop there.
Quantum computing is extremely broad in scope. Tools can help with not just risk management in the financial sector, but also modeling chemical reactions, developing better medicines, and much more besides. And while that may spread developers thin, in terms of focusing on each of the individual use cases, it is great news for investment.
Earlier this year, Deloite’s Insights team updated its analysis of the quantum computing sector based on levels of capital investment and other signals, such as patent filings for hardware technology. “Globally, the financial services industry’s spending on quantum computing capabilities is expected to grow 233x from just US$80 million in 2022 to US$19 billion in 2032, growing at a 10year CAGR of 72%,” write the researchers, highlighting how financial services firms are amongst those loading up on the opportunities to get ahead.
It could be the case that customers using quantum tools for derivative pricing, portfolio risk calculations, regulatory reporting, and other risk management activities play a role in softening the sharp edges – making quantum computing solutions easier to adopt elsewhere.