AI and digital twins: improving operational resiliency in financial markets

Paul Brennan, Chief Strategy Officer, Imandra
First published onTabb Forum, 15 February 2024.

Operational resilience is a key KPI and a major regulatory risk factor for exchanges and trading venues, where the impacts of outages are wide-reaching, with revenues and reputation at stake. Paul Brennan, Chief Strategy Officer atImandra, looks at how digital twins and reasoning AI offer new solutions in the fight for operational resilience and better business intelligence.

Global financial markets are under pressure not only to unlock innovation but crucially, to maintain and improve operational stability in the process, as highlighted in a recentIOSCO consultationon market outages. While great strides have been made to improve system resiliency, there is always room for improvement. And there are lessons to be learnt from innovations in other industries to address this problem.

The digital twin

According to a recent McKinsey article, digital twins, born in the production lines of the manufacturing industry to drive efficiency, have revolutionised how decisions are made and performance is analysed. The concept is certainly taking hold in other industries, including finance. A digital twin - in the context of digital twins for complex systems - is a fully functional and logically precise virtual replica of any complex financial software system, which accurately simulates its operation and verifies its behaviour in real-world scenarios. When combined with automated reasoning AI, a digital twin approach can increase operational resiliency while unlocking better business intelligence and innovation within financial markets.

Addressing a key challenge: operational resilience

Delivering innovation while maintaining operational resiliency is a challenge. New products and features must be designed, built and tested by different internal groups and communicated to external stakeholders such as trading firms and regulators. The coordination required to align all interested parties is enormous, and the industry demands getting it right the first time.

In our industry, we know that operational resiliency is paramount for the well-functioning of our markets. When exchanges suffer outages, the impact is wide-reaching, with regulatory and commercial consequences, as well as reputational damage. Regulators and market participants alike strive for greater operational resiliency.

A recent IOSCO report (December 2023) highlighted to what extent software-related issues caused most market outages (see below).

All market participants feel the impact of exchange software upgrades, many of whom must alter their trading systems to accommodate them. The exchange has to change its public and private documentation, adjacent systems (post-trade, surveillance, regulatory reporting, etc.), coordinate data providers, have trading firms' conformance tested, the list goes on. The entire ecosystem must understand and implement the change precisely.

It prompts key questions for the exchange surrounding how much due diligence is required to predict the impact throughout the entire platform, how to ensure accurate understanding across all stakeholders, testing, and how defects can be identified before launch. There are also non-functional exchange upgrades, which are mandatory for numerous reasons, including information security, hardware replacements, firmware upgrades, and system performance. Each poses challenges that are hard to address and while comparable issues occur in complex software systems throughout the industry, scrutiny on exchange resilience is much more public than for other organisations.

Throughout these processes, one perceived conflict remains present: What is the trade-off between risk management and cost?

Roots of the problem: the knowledge gap

Traditional product and software development practices struggle to keep pace with algorithm complexity in modern trading and exchange environments.

The knowledge and understanding of the exchange system are distributed across multiple teams and specialists. Documentation such as business requirements, rulebooks, and technical user guides are written manually using prose, tables, and worked examples, often under-specified and quickly outdated. This leads to significant challenges with the system build as there is no way of ensuring that requirements are logically consistent across the ecosystem. Moreover, test programs lack formal measures they can be assessed by and accountable stakeholders lack transparency, exposing them to unbounded risks.

Early identification of defects

We know from real-world experience that fixing bugs in production is exponentially more expensive than during the design phase. Yet, there is a lack of scientific techniques used within financial markets to identify defects and design flaws early.

In civil engineering, for example, structural engineers can apply the laws of physics to a design before any building occurs, finding flaws and ensuring structural integrity and efficacy in the finished build. All stakeholders then unite around a single agreed blueprint. This up-front analysis at the design stage saves time, money and lives, and this model-based design verification approach is commonplace in safety-critical industries and microprocessor design.

A solution - digital twins for financial markets

With so many stakeholders, the degree of complexity and the demand for rapid innovation, financial systems need to transition from age-old, prose-based specifications to a precise design.

Breakthroughs in AI and mathematics allow us to model exchange rules and regulations precisely and apply rigorous logical AI to automate regulatory analysis and testing, all while providing logical audit trails.

Digital twins that can be interrogated, allow exchanges to:

1. Verify properties about the exchange system behaviour
2. Generate test cases to ensure the correctness of the exchange system
3. Run an audit against production trading data
4. Generate precise English-prose documentation

And in a more advanced state, machine learning and generative AI can be harnessed to create new revenue streams with innovative data products and integrate large language models for ease of access.

Based on the digital design of the exchange or venue, highly automated logical reasoning AI now exists to help ensure its compliance and correct implementation. There are two critical pillars to this:

1. System Verification - A Verified Design
   AI-powered logical reasoning can verify system behaviour, identifying exchange design defects upfront, before the software development process begins, which isn't possible in the analogue world of prose-based requirements and specifications.
2. Systematic Validation - Automated Test Plan
   Reasoning AI can then systematically analyse the digital twin to identify all possible behaviours and edge cases. This analysis, “symbolic reasoning”, is a form of generative AI underpinned by logical reasoning, which yields accurate results. This leads to high-coverage automated test generation, used to test the exchange system and identify functional and non-functional issues.

The result is that exchanges benefit from a design and system defect identification step-change. The perceived trade-off between risk management and cost is unlocked.

This approach unlocks the pace of software development, gives financial institutions confidence in their operational resilience, and proves that their system design will behave as intended.

Digitising system design is the way forward

By taking inspiration from safety-critical industries, we can move away from traditional analogue processes and use digital twins to take a leap forward in exchange system resiliency, accelerated system development, new avenues for business intelligence and new revenue streams.

What's more, automated reasoning AI and the use of digital twins bring unique business intelligence to the fore, that currently rests within an exchange system. Venue operators can analyse their customers' interaction with the trading system by running many what-if scenarios and create actionable intelligence to help customers optimise the whole potential of the exchange system and its features.

Exchanges have already begun using this approach. Platforms in Europe collectively responsible for nearly 25% of the equity market volume actively use this method to supervise exchange technology.

With industry-wide focus and mandates to explore the use of AI, Automated Reasoning is under the spotlight, and when combined with the digital twin approach, it's a game-changer.