December 2014 marked a significant point in the timeline of quantum computing, with the Commonwealth Bank of Australia (CBA) becoming one of the earliest commercial organisations to invest significantly in its emergent capabilities. In April 2017, CBA increased its $5m investment by a further $14m. November 2017 saw Allianz and Royal Bank of Scotland join a $45m investment group for quantum computing, which also included Fujitsu, CME Group and Accenture.
Since 2020, we’ve seen uptake by Goldman Sachs, JPMorgan, HSBC, BNP Paribas, Crédit Agricole, Japan Post Bank, Citigroup, Wells Fargo, Barclays, Royal Bank of Canada, BBVA, ABN Amro and ING. Not only did these banks make investments, but they have also achieved results using quantum systems to calculate stock and mortgage portfolio risk.
The financial industry has become one of the fastest growing markets for quantum technologies. What impact will this have on the banking sector?
Tangible benefits for banks
CBA initially wanted to replace the work done by high-performance computing (HPC) for customer analysis. This is because classical binary computers – even powerful ones – aren’t always best-suited to handle statistics, especially not for large sets of possible outcomes.
As asset portfolios become larger, the requirement for precision becomes higher
Quantum computers, working with a superposition of bits, or ‘qubits’, work statistically from their very core, and are therefore better at handling complex, interdependent and time-specific statistical calculations, such as causality and Bayes’ theorem (a mathematical formula for determining conditional probability).
These statistical laws are a knowledge gap rarely considered in human-generated algorithms. And as a problem’s complexity increases, it becomes impractical for classical supercomputers. Quantum computers can do this naturally with only a handful of qubits. Though each calculation step is slower in a quantum computer than with a classical one, they require exponentially fewer steps to arrive at the same answer, and are therefore more beneficial for large calculations.
Quantum and banking
In traditional ‘Monte Carlo’ methods to calculate capital optimisation, we use a unique small set of parameters applied billions of times; but the outcome is again just a small set of numbers. And this is what makes quantum computers so relevant now. A capital optimisation calculation has a small input, a small output and extremely large sets of intermediate unused values.
For smaller sets of assets, the classical computer is still faster and cheaper. But as asset portfolios become larger, the requirement for precision is higher and the prediction time periods become longer. As a result, quantum computers, while still early in their evolution, could become more desirable alternatives.
Of course, this presents an interesting challenge: when is it best to use classical computers and when is it best to use quantum? The solution may be hybrid computing platforms primed to a business application that can use both methods. Once these applications are tested and implemented, we will start to see quantum computing become part of day-to-day business activities.
Securing quantum computing
One key challenge around quantum – especially for the banking industry – is to ensure tight security. Financial services centre on risk management. The institution that manages to minimise risk while maximising profits dominates the market. While becoming quantum-safe as a company is technologically simple, it is a long-term challenge that cannot be sped up purely through investment; starting early is advisable and not too costly, but can easily be overlooked among other pressing challenges.
Another challenge is the usage of cryptocurrencies and distributed ledger technologies. These are often not quantum-safe. However, most banks don’t use public blockchains and they have the trust and influence to evolve to quantum-safe trust technologies independently. With exception of the public ledgers, the impact for banks is expected to be low.
Getting public networks quantum-safe is important and time-consuming. A lot of the financial trading data is run via low-latency networks. These are likely to be privately owned, and so fixing encryption in these networks is more straightforward than securing public networks.
A bright future
Over time, financial institutions will expand their quantum technology capabilities and increase the number of possible applications. In the future, hybrid quantum-HPC systems will be at the heart of their businesses. Those that don’t join in could be running serious commercial risk, and financial organisations are aware of this.
Quantum has a bright future, with the potential to make the sector more profitable and less risky. One day it might even make the global economy more stable, as fiscal risks can be better predicted with quantum computers. But quantum computing is not the only quantum technology. What would finance look like once we have a quantum internet that allows for instantaneous, faster-than-light correlations? Will we again change the statistics of algorithmic trading, as the rules of the game change? Nobody knows, but it is interesting to consider.
Frederik Kerling is head of fintech and senior quantum expert at French IT services and consulting group Atos.
