Sorry, you need to enable JavaScript to visit this website.

How AI is playing a role at Investment Banks and Capital Markets Trading Desks

How AI is playing a role at Investment Banks and Capital Markets Trading Desks
July 13, 2021

The LIBOR scandal, which made headlines in 2012, revealed that traders at major international banking institutions in the U.S. and the U.K. acted in collusion to manipulate the London Interbank Offer rate (LIBOR). This revelation, which, in financial markets, was thought to predate back to 2003, led to massive reputational damage for those investment banks. This culminated in banking regulations, lawsuits, and fines against the traders and the banks that had failed to uncover the problem.

How did this all begin within financial markets? The traders communicate with each other via chatrooms, mobile telephones, and Bloomberg terminals to keep up with the capital market trends from peers and get actionable insights.  However, the results of investigations in the U.S. for example concluded that certain Wall Street investment banks had allegedly manipulated the market for U.S. Treasury bonds. There were documented cases where these platforms had been used by some traders to coordinate clearly fraudulent transactions within financial services. The conversations held in trading chat rooms revealed that the traders rigged the LIBOR benchmark interest rate and a sum of benchmark foreign exchange prices over eight years.

These findings raise the question whether this fraudulent behavior could have been detected earlier. With improved risk controls and technology systems industry participants accept the total impact and cost of the fraud could have been significantly lower.

Further investigation into the respective platforms and communication channels by Risk and Compliance officers, looked at sources of activity particularly those offering free access which were reviewed to ensure messages were compliant with banking regulations. With technology advances, new and improved solutions involve the use of automated AI systems.

Use Case with HCL’s Partner Mathlabs - Trader Fraud Detection and Explanation

Mathlabs has developed a system that monitors trading patterns in FOREX (flow) traders to detect anomalous behavior. The framework detects anomalous patterns and offers analysis with explanation. The solution’s unique detection framework captures regressive patterns and explains the location of irregularity together with continuous auto capture and update of its libraries of dominant anomalies.

Case study

Challenge

An investment bank looked to monitor its flow traders (FOREX) for anomalous behavior found in trading patterns to continually create and update the bank’s Risk and Compliance.

Solution

Math Labs (MLR) defined a unique data space where it applied its own anomaly-analysis framework to identify suspicious traders based on dominant behaviors. It also recognized persisting old patterns vs newly formed anomalies.

Impact

The benefit of this solution is a significantly increased sensitivity to novel anomalies without human intervention. It allows the Risk and Compliance function to focus on the results and check for false positives or further investigate if any fraud has indeed been committed.

Credits

Pravesh Johri, Partner – Financial Services, Math Labs Research Limited, United Kingdom

HCL Risk & Compliance Solutions

HCL’s global Risk and Compliance team helps our financial services clients to deliver technology change from the impact of risk, regulations, financial crime, cybersecurity, and emerging technologies. Please contact our team for further assistance on this article or for any of our services.

HCL ‘s global Risk & Compliance team helps our Financial services clients to deliver technology change from the impact of risk, regulations, financial crime, cybersecurity and emerging technologies.

Email: Kinsuk.Mitra@hcl.com

HCL’s Risk & Compliance microsite