April 11, 2017 - Paul Lashmet | Industry Solutions

Visual Big Data Analytics for Trade Surveillance

In order to fully demonstrate that their policies and procedures comply with regulatory requirements, financial services organizations must use information that goes beyond traditional data sources. Trade surveillance exemplifies this because alternative data sources are needed to provide a deeper understanding of the intent of trade activities. This post describes how visual analytics can provide compliance officers with actionable insight that will enable them to identify and investigate events more efficiently and establish a robust and defensible process across their enterprise.

Leverage Alternative Data for Deeper Understanding

Often the first thing that comes to mind when thinking about trade surveillance is illegal insider trading, i.e. when a security is traded based on confidential information. Illegal insider trading grabs headlines, but there are many other illegal trade practices, such as:

  • Spoofing: putting in a bid or offer on a security with the intent of withdrawing it before execution as a way to create false demand.
  • Pump and Dump: artificially inflating the price of a stock through false and misleading statements.

The instances above are not easy to find. They are buried deep within daily trading volume and communications, like a needle in a haystack. In the case of spoofing, illicit cancellations can be mixed within the many legitimate cancellations that are a normal part of trading activity, especially with low cost, high volume instruments. In the case of pump and dump, misleading statements could be hidden in public message boards that then get reported in press releases or newsletters.

To find the needle in the haystack, you must think beyond traditional data sources. For example:  

  • Data generated by machine learning routines can identify suspect trade flow patterns at a very granular level;
  • Then, specific transactions can be linked to newsletters and messages that have been analyzed for misleading statements.

These non-traditional data sources help provide a deeper understanding of both the event and the intent.

Trade Surveillance is both a Big Data and Procedural Challenge

The example above illustrates how actionable insight can be provided to a compliance officer. It is also an example of a big data problem —  consider the vast amounts of data analyzed, the variety of sources, the high frequency of trades, the complexity of cross-border transactions, and the overlapping layers of client relationships. All of these elements need to be connected for any deep data analysis to be useful.

We must also take into account the heightened expectation of financial services regulators. A financial institution must show a robust and defensible review process by responding to regulator requests in a timely, concise, and accurate manner. Preferably, a financial institution wants to show that it is proactive by identifying and mitigating issues early in the process. A review of enforcement proceedings (see links in the notes, below) shows that it is not just the misdeed that gets punished. Penalties are also levied on supervision failures, i.e. the processes and procedures that allowed the event to happen in the first place.

Empower Compliance Officers To Act Fast with Visual Big Data Analytics

Compliance officers are on the front line of responding to trading compliance requests. Providing those officers with the right tools is a necessity. Financial institutions must get ahead of potential issues by being able to act on questionable trades, navigate large volumes of data, determine the regulatory impact (if any), and respond accordingly. The tool needs to enable a creative, iterative, and collaborative approach by visually linking orders, executions, and trades to news, insider trading data, electronic communications, message boards, material nonpublic information, pump and dump scheme trackers, and surveillance alerts.

Visual big data analytics provides compliance officers with the actionable insight that they need. Visual big data analytics can:

  • Show real-time and historical data analysis side by side in one application;
  • Analyze all types of data in place, without moving it, and in an ad-hoc manner;
  • Alert designated people based on real-time data;
  • Integrate machine learning for accelerated data evaluation and modeling;
  • Visually blend data that enables cross reference and correlation across various data sources;
  • Build real-time data apps for critical metrics as well as visuals that simplify time-series analysis;
  • Securely share models with internal and external constituents; and
  • Allow compliance officers to do this in a self-service fashion.

Arcadia Data does all of this and more. Arcadia unifies data discovery, business intelligence, and real-time visualization in a single, integrated analytics platform. It enables users to directly leverage big data capabilities (running natively in Apache Hadoop clusters, the cloud, or other modern data platforms) through an intuitive and self-service interface. Referring to the examples above, a team can explore billions of rows of highly varied data, join views into interactive packages, and share that package with relevant colleagues to solve problems in a timely, secure, and collaborative way.

Notes: See links below for real world examples of illegal trading activities.


Related Posts