UnaVista Rolls Out Reporting Analytics for Performance and Market Surveillance

The London Stock Exchanges Group’s (LSEG’s) regulatory arm, UnaVista, is rolling out new reporting analytics and market surveillance tools on its cloud-based platform. The UnaVista Analytics services will be released in three phases, starting with performance data, followed by data accuracy and peer-to-peer assessment, and finally, market abuse detection.

The performance analytics were made available to UnaVista’s 1,000-plus clients free of charge in the second week of May. As LSEG captures vast amounts of client transaction and reporting data, UnaVista can use that data, together with proprietary algorithms, to identify potential reporting errors.

The entire suite of analytics will be delivered using a combination of dashboards, graphics, reports, and generated alerts provided daily, intraday, weekly, and monthly.

The second release will focus on enhanced analytics, data accuracy, and peer group evaluations. This offering will provide 20 to 25 types of alerts, indicating to firms potential errors within their reporting. UnaVista is using machine learning to develop analytics that will offer a more detailed look at the firms’ reporting and enable them to detect outliers in patterns, conduct data quality checks based on alerts, drill down into individual transactions, review systematic issues and analyze their reporting quality compared to their competitors.

“Most regulators would expect both counterparties, within a degree of tolerance, to report the same trade economics, with the buyer and seller fields the same on both reports. But we’re now able to look at using machine learning and a lot of complex algorithms to try and understand, not just the quality of your data versus your peers, but where your report may differ from your counterparty’s report, allowing you to analyze and confirm the details you have reported,” says Mark Husler, CEO of UnaVista.

UnaVista currently does 350 tests on each transaction report, and within each transaction, there are 65 data elements. The second phased service will provide more granular data by displaying individual data elements within each transaction and creating various combinations of the data to show where the client misreported or where there are likely to be inaccuracies.

For example, UnaVista can profile clients’ reporting times and flag potential problematic patterns. This is done by capturing a client’s reporting data, which will include data points such as execution times, and then comparing those data points with the client’s peer group trade information to see if it matches up, such as in dual-sided reporting requirements under Mifid II. Within that counterparty peer group, UnaVista works with 15 of the top 29 global banks.

Additionally, UnaVista will anonymize the personal and trade data of the peer group, to enable firms to benchmark their reporting capabilities against others and see where they might rank in the eyes of the regulator.

“It’s now a legal obligation to test and demonstrate that you are testing. It is also best practice that under the Senior Managers Regime (SMR), senior managers responsible for reporting should have confidence in the compliance of reporting from control teams. Any firm’s internal control framework should be supported by regular and comprehensive testing,” says Matthew Vincent, head of regulatory reporting strategy at UnaVista.

In April 2019, the UK’s Financial Conduct Authority (FCA) published a Markets Watch newsletter urging market participants to have the “appropriate arrangements in place to ensure and assess the completeness and accuracy of their transaction reports and instrument reference data submissions.

The second round of enhanced analytics is currently part of a beta program involving a select number of clients, who will provide feedback and suggestions on how to develop the service further. It will be made available on the UnaVista platform by the third or fourth quarter of 2020.

The new services aim to function as a second line of defense for compliance reporting. Rather than replace banks’ and investment firms’ existing reporting processes, Vincent says the analytics should be embedded into firms’ current control frameworks.

Husler says firms typically have manually laden processes for evaluating the accuracy of their data and creating analytics around that to better understand where and why the error was made.

“Not only is it manual, time-consuming, and costly, but currently, clients typically process this type of analysis at a later point to the initial trades to proactively address issues, whereas these analytics are automated daily and fully integrated into the UnaVista platform. So it’s just trying to lift out a lot of that duplicate cost and effort in the industry and bring best practice into the situation to allow firms to effectively have an automated daily capability around very detailed sophisticated analytics on reporting accuracy for the first time, to give them that external benchmark,” he adds.

The first and second suite of data analytics is designed to help firms comply with regulations such as Mifid II and Mifir. UnaVista also looks to extend the new suite of data analytics to cover other rules including the European Market Infrastructure Regulation (EMIR) and Securities Financing Transactions Regulation (SFTR) in later iterations.

Market Abuse Analytics

For the third release, UnaVista will roll out its market abuse module on the UnaVista platform. This will offer transaction-focused, multi-asset class market abuse analytics, and alerts. The service will overlay the firm’s proprietary typologies and methodologies across UnaVista’s vast proxy of instrument and transaction coverage, to identify signs of market manipulation.

Liam Smith, head of market supervision at LSEG, runs a team of technologists and surveillance specialists, with experience in trading and machine learning at the PhD level, or from previously working at a regulator.

The team developed its methodology based on the FCA’s approach to detecting market abuse combined with internal knowledge of transaction reporting.

The module will include a range of patterns for detecting classic forms of misconduct such as insider dealing, wash trading, and marking the close. Additionally, the service will offer capabilities to detect financial crime including anti-money laundering and terrorist financing.

The technology looks at trader behavior, identifying the hallmarks of how they trade and how aggressive their strategy is, and creates statistical outliers that indicate whether they acted normally or abnormally against their profile.

“Given the granularity of the dataset, we’re able to compute reasonably sensibly what is an acceptable tolerance, or a potential profit and loss, and what we would regard as standard in ordinary trading versus potential market abusive trading,” Smith says.

“We are able to provide an independent detailed perspective to end-users formed from their transaction reporting data that is used by the client’s regulators to monitor surveillance and market abuse. This includes enabling our client’s compliance departments to access behavioral profiling analysis across their organization at an individual trader level across asset classes.”

The market abuse module is still in the research and development stage and is also part of a beta program. The service is scheduled to be rolled out to UnaVista members in the third or fourth quarter of 2020.

For 2020, the three data services will be available free of charge to existing UnaVista clients. In 2021, the firm will look to develop a commercial strategy around the second and third analytics offerings.