A lot can happen in 72 hours: Just think how blissfully endless a three-day weekend can seem the afternoon before you take off.
But when your surveillance team is rushing to meet the 72-hour requirement from regulators to reconstruct a suspicious set of trades with all related structured and unstructured data, three days never seems long enough.
In 72 hours, 958.8 billion emails and 18 billion texts are sent, 2.4 billion phone calls are made, and $15 trillion is traded—and if any of those communications are related to the trade in question, they all have to be documented and analyzed. Compiling and organizing documentation can put a strain on even the best teams.
Trade Reconstruction 101
In the financial sector, regulations regarding interactions and communications among traders have increased exponentially over the past decade—especially those related to orders and trades. If you’ve seen The Big Short, you may recall that before 2008, big firms were able to overvalue stocks and make a lot of money off of them. They would sell the stocks on the open market for higher than they should have been valued, and turn around and buy them back later at a lower price, pocketing the difference as a profit. Investment banks were manipulating LIBOR, to their benefit and the demise of the general public. The regulations in place didn’t stop this corruption.
After the 2008 financial crisis, regulations such as Dodd-Frank Act and MiFID II were set up to ensure adherence to ethical practices and keep the market fair.
Over-the-counter (OTC) traded products were heavily scrutinized for being difficult to regulate, because too much of the trading process lacked visibility. In response to these factors, the swap market is now overseen by the Securities and Exchange Commission (SEC) and the Commodity Futures Trading Commission (CFTC). To ensure that market manipulation of any kind doesn’t take place, trades are monitored and audited by regulators when found to be suspicious.
The CFTC’s specific rules require swap dealers and major participants to adhere to guidelines ensuring fair and equal practices. For instance, prior to entering into a swap, the parties have to agree in writing to all of the terms governing their trade relationship—including terms related to credit relationships.
They also have to maintain documentation of the parties agreeing on the process for determining the value of each swap, at any time, from execution to the termination, maturity, or expiration of the swap. The regulations are set up to make sure that swap dealers maintain timely and accurate confirmation of swap transactions, as well as reconciliation and compression of swap portfolios.
However, when trade reconstruction became a requirement, the data that had to be gathered was largely made up of in-person conversations and landline phone calls on the trading floor. Now, trades take place across trading desks, mobile phones, Bloomberg chat, and even video calls. Firms must find efficient ways to dive deep into the trade process in order to understand every step and be prepared to meet a broad range of record-keeping requirements.
Considering the wide range of events and communications that can relate to a single trade, including emails, calls, and meetings, not to mention the actual trade details like the time and date stamps, asset classes traded, or counterparties, it is no surprise that trade reconstructions are both complex and arduous.
The Most Common Obstacles
Today, we’re communicating across a variety of platforms both personally and professionally, and it’s natural that trading encompasses just as many channels. The data requirements for trade reconstruction involve communication data, structured data, and other unstructured data.
Communication data includes voice communications, instant messages, email, and social media. Structured data focuses on trading systems, OMS, post-trade information, and collateral systems. Unstructured data can be found in sales and marketing documentation, auditing and compliance data, financial records, and organization charts. All of these types of data must be brought together in a cohesive manner to meet regulation requirements on identified trades.
As it stands, however, most financial institutions still store data and investigations projects in silos—which requires manual, time-consuming labor, that may be less defensible, to gather. While these different platforms contribute the information needed for trade reconstructions, and may help meet record keeping obligations, they traditionally do not communicate. Add to that the fact that many of these operations are often completed by multiple people—and bringing together all the relevant data for a trade reconstruction is prohibitively difficult.
In short, disparate systems and teams make the manual process of pulling data and then assembling it coherently take hours and days for even the best analysts.
But given the 72-hour rule mandated by the CFTC, pulling and linking trade-related data across organizational silos, and then analyzing and organizing it, as quickly as possible is vital to avoid penalties.
The good news? A lot of the effort that goes into trade reconstructions comes from repetitive, tedious tasks, like identifying and analyzing unstructured data in the form of communications. If we can align on what needs to be done, in what order, how, and why—we can begin automating the process with technology.
Automating Trade Reconstruction
Often the biggest challenge to meeting regulatory requirements is understanding the unstructured data. Plus, identifying related communications is difficult and cumbersome with multiple platforms involved. The ability to see a more complete view of communications—from eComms to aComms to trades—related to risky behavior is key for effective surveillance and improves efficiency meeting trade reconstruction requirements. Once disparate data sources are brought together, trade reconstruction can become part of an organization’s proactive surveillance strategy to identify risky trades before regulators come calling.
With centralized data storage and an automated system that uses artificial intelligence to pull related communications to the trade in question, compliance teams can spend more time reviewing the data and presenting an organized, holistic view of the trade. These compliance teams will never be replaced by technology, but they can use this technology to enhance their efforts. This ensures they’re able to focus their valuable time and attention where it really matters—and automate the rest.
When a trade reconstruction request comes in, usually, an analyst would have to reach out to IT or another team to gain access to the needed structured and unstructured data. Then, they will likely have to manually log on to each system, search for related communications and trade data (usually by date, time, trade, and trader name), and then link all of this information together in a separate log. But an AI-powered system like Relativity Trace can automatically ingest and link this information from an OMS system and over 50 different kinds of communication channels, including Slack, Microsoft Teams, and Bloomberg Chat.
Relativity Trace leverages advanced lexicon searching, metadata filtering, and AI to connect trades and communications together, which cuts down the volume of false positives and duplicate material. When you can review the communications and trade side by side, analysts have a clearer picture of how an event unfolded and the corresponding action taken by the trader.
Maintaining a fair marketplace requires both regulators and users to adopt and honor financial laws and regulations. Compliance teams are put in place to do just that, and with an innovative, AI-powered surveillance solution helping to break down data silos and automate the heavy lifting of sifting through all that data, teams can better protect their firms from misconduct.
Cassandra Morrison is a senior specialist in content marketing at Relativity, with a special focus on our communication surveillance solution: Relativity Trace.