The Evolution of Electronic Communications Surveillance


Financial services firms currently face increased risk due to collusive and unethical behaviors associated with unauthorized and rogue trading. To identify and mitigate these risks, companies must demonstrate to both regulators and their stakeholder communities that they have taken appropriate actions utilizing electronic communications surveillance solutions. Risks include: collusive insider trading, front running of trades, washed trades, and other forms of market manipulation such as access and use of material non-public information (MNPI).

Today, surveillance techniques deployed include the monitoring of employee trader(s) and trade activities, internal controls, transactions, and the communications of all participants in the financial services’ daily operations. This paper focuses on electronic communications (eComms) surveillance, what institutions do today, and what they will evolve to doing tomorrow.

New eComms solutions must mature in order to effectively monitor the exponentially increasing data volumes generated from emails, SMS messages, chats, and voice communications.
New eComms solutions must mature in order to effectively monitor the exponentially increasing data volumes generated from emails, SMS messages, chats, and voice communications.

Electronic Communications Surveillance (eComms)

eComms surveillance is the monitoring and supervision of all electronically produced communications executed by employees of a firm. Communications includes email, chat, text, and/or voice (phone & squawk-box) which can be transformed into digital form and subsequently analyzed by an automatic monitoring/surveillance system.

The Problem

Presently, firms use a variety of in-house and vendor supplied solutions to perform their eComms surveillance program. However, such solutions are rudimentary in nature and focus on lexicon-based searches, sample populations for full email review, specific watch list groups selected for heightened monitoring (to safeguard against insider trading), and other ways that an employee, counterparty, or customer may gain market advantage illegally. The problem that arises is that the monitoring and subsequent analysis is based on a predefined group of individuals, business lines, and events that have already occurred. In addition, these solutions cannot effectively monitor the exponentially increasing data volumes generated from emails, SMS messages, chats, and voice communications. Existing methods are not designed to uncover patterns, people networks, or capture sentiment and determine suspicious behaviors which can quickly lead individuals and institutions into trouble. Future surveillance capabilities must seek to prevent these inappropriate behaviors.

The Solution

As regulators identify and punish continued wrongdoings (e.g. LIBOR, SEC/Bilbao insider trading, precious metals price fixing dark pools), current eComms surveillance programs must evolve and use predictive analytics. Such evolutions aim to take advantage of a combination of statistical social networks, machine learning, and natural language process (NLP), with the support of “Big Data” analytics. To achieve this, firms must understand the behaviors of their customers, counterparties, and employees (traders, operations, and customer representatives). Companies must articulate the risks that these “actors” pose to the firm and outline standard ethical behaviors in order to allow systematic monitoring and subsequent analysis to properly differentiate between the two. Furthermore, organizations must tie its risk-based tolerances into the alerting logic to demonstrate that it has a balanced, sensible, and matured methodological approach to eComms surveillance. The operational risk here is that many behaviors could be deemed as risky, but in actuality, very few combinations of such behaviors lead to actual trouble. These are the most critical to identify.

To strengthen surveillance capabilities, firms should implement the next generation of eComms surveillance solutions that include statistical analysis tools, pattern identification (inference rules), machine learning, visualization tools, scenario generation, and network analysis functions. Risk behavior specialists and business line operations are the resources necessary to form the behaviors of interest needed for an evolved solution.

The trick here is not to overwhelm the business with unimportant suspicious alerts, but to target behavioral combinations that pose the greatest risk. In addition, as employees become aware of the extent and the sophistication of these new predictive analytics surveillance systems, their behaviors may automatically change for the better (the Hawthorne effect).

Obstacles to Solution Implementation Success

In implementing newer surveillance solutions, this process requires having management both willing to invest in technologies and in encouraging cultural change. Leaders must also spend the time to define the common sets of behaviors to observe. Vendor selection, licensing, maintenance, infrastructure, and professional services costs may be high depending on the size of the firm and the scope of the solution. The firm should ensure that their infrastructure can support and handle the exponentially increasing large data volumes and associated required searches. Data privacy regulations may impact implementation as well. Resource management targeting individuals who have the skills to interpret the results and update the solution will become critical. Lastly, firms must ensure that expectations of any new eComms solutions are realistic.

Conclusion

In summary, financial firms currently execute their eComms program with rudimentary solutions in which the monitoring and analysis carried out are based on events that have already occurred. Such solutions cannot effectively monitor the increasing volumes of data generated and must evolve to seek prevention of inappropriate behaviors. Increasing capabilities to understand behaviors of customers, counterparties, and employees is crucial. Firms should start investing time in defining common sets of behaviors for observation, and target behavioral combinations that pose the greatest risk. An evolved solution will provide firms with a holistic insight into the behaviors of the firm, and action firms to identify and mitigate these risks in order to prevent collusive and unethical behaviors.

GFT has experience in building customized solutions to monitor trading behaviors and the reports to management to aid in risk mitigation. We also have extensive experience with the leading surveillance software vendors such as: Catelas, Activio, CA Technologies, Digital Reasoning, Bloomberg, Palantir, Nice Actimize, SyntelliRead, SAS, and Behavox. In addition, we have experience with QlikView, Spotfire, MicroStrategy, and SAP BusinessObjects to facilitate effective data visualization and reporting.

Through working with our clients, vendors, and our IT and service operations team, GFT builds collaborative solutions that challenges the best of breed target operating models, technical roadmaps, data collection, and correlation, and designs for the end user team. GFT works with our clients to deliver the required solution and expected results.