NLP Machine Learning for Trader Surveillance

Intrepid uses machine learning to support trader activity surveillance on instant messaging applications.

About Our Customer

A global investment bank and financial services firm founded and based in Switzerland. Headquartered in Zürich, it maintains offices in all major financial centres around the world and provides services in investment banking, private banking, asset management, and shared services.

Their Use Case

Investment banking trading uses numerous communication channels to conduct business. Many of those communication channels are not monitored for compliance with trading regulations. Bloomberg Chat is a popular instant messaging platform that investment bankers use to communicate with each other and potentially do business. Our client sought to bolster their trader activity surveillance to identify the trader’s deal intention when discussing business over instant messaging.

Our Solution

Our team in collaboration with our client prepared a Natural Language Processing (NLP) solution to monitor Bloomberg chat records. The NLP solution successfully extracts and compiles a report on a weekly basis, showcasing all the deal intents expressed through non-electronic trading channels. In order to provide a comprehensive investigation, the report includes the true intent and complete product information, which is determined through the utilization of advanced machine learning algorithms.

Here is a list of surveillance items the ML algorithms were processing and identifying in the chat data:

  • True Deal Intents

  • Financial Products Mentioned

  • Distinguishing the Unreported and Mass Data

  • Flagging the Responsible Person

To generate the weekly report, the NLP model automatically processes an enormous amount of data, specifically 10GB of Bloomberg Chat records. This ensures that a thorough analysis is conducted and accurate insights are derived. The primary objective of this project is to fulfil a regulatory requirement by reporting all communications containing deal intents to the regulator using a 'Request for Quote' (RFQ) format. This ensures transparency and compliance with the necessary regulations governing the industry.

The Results

Our client was very satisfied with the results of our solution. The client was given a solution that helped reduce their exposure to regulatory risk which meant a lower probability of financial and reputation loss. The solution added data insights that could be used in many other current and future trader surveillance and regulatory tools.