Transforming Trade Reconstruction from Art to Science
Evolving regulatory requirements are creating more challenges for financial institutions around the globe. One operational challenge that plagues firms of all sizes is trade reconstruction. Rapid trade reconstruction is mandated by a host of regulations, including MAR, Code of Conduct and Dodd-Frank, and (soon to be) MiFID II. In some cases, firms must adhere to a demanding 72-hour turnaround time to provide reconstructions to regulators.
Trade reconstruction is becoming inherently more complex, too. That’s because traders have more ways to communicate: via voice calls (at a desk, turret, or on a mobile phone), or by email, instant message, SMS, chat, social media, Skype, etc. All of this data is typically stored in separate silos.
To reconstruct a trade the data is typically obtained in one of two ways: either a request is made to the respective data owner(s), or an analyst needs to manually log on to each system, search for related communications and trade data (usually by date, time, trade and/or trader ID), and manually piece all of this information together.
Voice communications are especially problematic because they are, by their very nature, unstructured. Analysts can speculate (based on when a call was made, and who made it) if a call might be relevant, but they can never really know for sure until they listen to it. That’s why firms employ teams of individuals whose sole role is to sit and listen to recordings of calls all day long – to find the proverbial ’needles in the haystack.’
The problem is – a reviewer might have to listen to hundreds of calls to find the one that matters. This can be especially challenging and time consuming for trade swaps because such transactions can take months to execute and may involve many communications.
This manual process of trade reconstruction can be likened to a complex dance, or art, because it relies on people who are skilled in the ’art’ of trade reconstruction, to go through lots of manual steps, which consume lots of time, with no guarantee of an accurate end result. (Art, after all, is open to interpretation.)
Additionally, when analysts (’the artists’) leave the firm, they take their experience with them, which means new people need to be trained in the art of trade reconstruction.
Which leads to the question - is there a better way?
The science of trade reconstruction
Compliance analysts will never be replaced by technology – but as regulations place greater demands on firms to address regulator requests for trade reconstructions, in faster, mandated turnaround times, technology can certainly enhance the analyst’s performance. And that’s where the science comes in.
A lot of what goes into trade reconstructions are repetitive, tedious tasks. If we understand the science behind those tasks (what needs to be done, in what order, how, and why), it makes sense to automate as much as we can through technology.
That’s exactly what automated trade reconstruction solutions do. Using advanced correlation algorithms, they can cut trade reconstruction time from days to minutes, and help firms improve their responsiveness to regulators.
How Automated Trade Reconstruction Works
Automated trade reconstruction brings all of the various silos needed for trade reconstruction together through a data management layer. The data management layer normalizes, analyzes, indexes, and correlates data from the underlying structured and unstructured data sources, to link the different silos together.
Furthermore, because automated trade reconstruction solutions incorporate sophisticated natural language processing and text analytics, they’re able to analyze data across all communication channels (including voice) for key words. Essentially, this enables the technology to ’understand’ the context of the communications. These key word results are then added to the metadata, and are searchable as well.
The above processes are totally automated and transparent to the analyst. Additionally, because everything is now linked together, the analyst only needs to do one search in one system to pull the trade reconstruction together.
By spending less time on the mechanics of producing the trade reconstruction, the analyst can also devote more time to carefully review the results, before handling them over to the regulator. Analysts can also use their time more effectively, by proactively managing risks, instead of just reacting to requests.
Regulators benefit too. Instead of just receiving separate files, new technology lets them visualize communications on a timeline as they happened, and even replay events in their proper context. This adds a whole new dimension to trade reconstruction (which regulators will appreciate, because it simplifies their job as well).
Increasing regulations and trade reconstruction complexity: catalysts for change
As the pace of regulatory change accelerates, it’s a given that financial firms will need to produce more trade reconstructions. Very soon firms in EMEA will ’feel the pain’ that U.S. firms have felt under Dodd-Frank, when requests for trade reconstructions increase under MiFID II, which will take effect in early 2018.
While this will undoubtedly place new demands on firms, the good news is – thanks to automated trade reconstruction technology they’ll be able to manage the workload efficiently.