Skip to main content

First Steps for Artificial Intelligence Integration in the Capital Markets

Artificial Intelligence (AI) has been popularised in fiction for decades, but it wasn’t until the mid-1950’s that the concept of Machine Learning (ML) was introduced to develop efficient AI algorithms. AI and ML have since evolved into real applications used in everything from sophisticated gaming platforms to DNA sequencing technology to online search engines. 

AI forms the basis from which many real technological advancements and practical solutions for intelligence-led complex data processing, such as deep learning and neutral networks, have emerged. ML acts as an essential filtered source of data which AI uses to operate; however, when implemented on its own, ML can still be of value by providing an adaptive filtering mechanism that produces more effective solutions.

As the electronification of the capital markets continues, AI technology is simultaneously evolving to provide new opportunities for intelligent digitalisation. AI can theoretically be used to analyse trade patterns to detect market abuse, to undertake sophisticated surveillance monitoring and to reduce operating risk for algorithmic trading; these capabilities could help deliver more efficient trading strategies, better data analysis and increased operating effectiveness.

However, AI must be first better understood within a capital markets framework before being fully integrated. This could be done by identifying ongoing and anticipated problems facing the capital markets industry and how AI and ML can potentially resolve them with added value. Another possibility would be to look at how AI and ML are generating value in other industries, such as within the Fintech industry, and exploring how those strategies could be adapted for capital markets firms. The overall advantage of this approach would be the possible implementation of a solution that would differentiate a firm from its competitors, but that same implementation may bring up other potential issues relating to the fact that AI and ML remain immature technologies within the capital markets industry.

GreySpark believes that financial organisations seeking to integrate AI and ML into their operations must first continue to or start to invest in their Big Data and High Performance Computing abilities. A sophisticated and successful AI and ML maturity model inherently depends on concrete and stable ‘data lakes’ designed to accept the widest possible range of input data streams. Carefully developed and widely understood data sources are necessary to ensure filtering is adapted correctly and that the breadth of data input is accurately scoped. Only then does AI have a greater chance of generating more meaningful, derived data.

AI supporting systems must have the ability to incorporate additional amounts of computing power on demand. Without ensuring proper support, the huge potential of AI could remain untapped and could possibly weaken the business case. GreySpark believes that given the unlimited headroom of AI consumption of resources, establishing a strong and calculated business strategy is necessary to determine the potential for an investment return and govern resource demands accordingly.

Although the capital markets industry is the beneficiary of many newly emerging technologies, it is not yet adequately suited to support an environment for the research and development (R&D) of AI and ML solutions. Banks are now struggling to find skilled people in this field as potential employees may have significant R&D experience, but lack awareness of the business constraints and opportunities that AI and ML have may on the capital markets business. GreySpark’s expertise and experience within the capital markets and technology industries have sparked the development of best-practices for AI and ML solutions that will help resolve these industry-wide issues.