Artificial intelligence Part 5: specific industry impacts - finance and banking

ChatGPT image

AI is particularly suited to back office operations in banks and financial institutions analysing large amounts of financial data. Typically banks and financial institutions employ large numbers of people to undertake functions such as compliance documentation, fraud review, transaction monitoring, credit analysis and financial reporting. AI systems designed by Palantir Technologies and SAS Institute for example, can review financial data and identify anomalies much faster than manual teams. The impact of AI on specific industry segments is summarised as follows -

Investment banking and capital markets

• Analyst roles are vulnerable to AI systems. The business folklore of junior bankers working 100 hour weeks using Excel models, pitch books and risk management due diligence is already under severe pressure as these tasks are highly structured and can be executed by AI. Examples already known include Goldman Sachs and JPMorgan deploying AI for financial modelling, earnings analysis and report generation.
• Equities market research has been transformed as AI can monitor thousands of stocks, synthesise earnings and generate initial research notes faster than any human team. 

Asset management

• Quantative analysis and factor modelling can be easily augmented by AI and is increasingly occuring already. This situation is leading to a changing and evolving role for quantitative analysts.
• Portfolio reporting and client communication is increasingly being automated with AI at the commodity end.
• Active investment funds management comes under further pressure as passive funds are now better guided by AI-driven strategies.
• Compliance reporting which is a very large cost centre in financial markets is being substantially automated with AI. The use of automation was an existing trend for many years but AI enables a faster rate of uptake. 

Retail and commercial banking

• Loans underwriting is already largely algorithmic and automated for retail consumers and the SME business level already. AI does not alter the trend but merely further reduces the remaining human review layer.
• Customer service and branch banking continues a long decline with face-to-face service reduction. This situation however is subject to fluctuations due to community pressure and increasing consumer preferences for personal interaction for specific services. AI's influence is limited in this line of business activity.
• Fraud detection and ani-money laundering (AML) monitoring is already within the AI-dominated sphere. Human reviewers have been shifting to exception handling only.
• Financial advice at the mass market level  already has limited use of robo-advisers. This segment is however subject to regulation and government oversight and the requirement for financial advice licenses, accountability and legal liability. The use of robo-advisers beyond limited information provision and recommendations for the mass retail market has not yet occured. High-net individuals particularly prefer human advisers and personal banking managers rather than an automated service. Various financial advice scandals in the sector may also limit the use of AI for the time being.

As with all industries, the use of AI in finance and banking is most easily implemented in large data analysis, administrative and reporting tasks. It is not well suited to client relationships and regulatory, legal and compliance responsibilities.