Regulatory reporting—A view from the SMF “hot” seat

Paul Graham talks with MarketAxess.

Accuracy and timeliness should go hand in hand. However, for Senior Managers, regulatory reporting often means dealing with problems today, caused by inaccuracies introduced in the past.

To help you guard against this eventuality, MarketAxess recently spoke with Paul Graham, former Head of IB Operations at UBS. We discussed what it’s like to be in the hot seat as changes in technology, the rising importance of high-quality data, and the growing demands of the regulatory environment increasingly impact the role of the SMF.

MarketAxess: In tests performed by SensAI^®, MarketAxess’s AI-powered regulatory reporting assistant, we found that 57% of transactions had at least one previously undetected error. What challenges do firms face in getting transaction reporting right?

Paul Graham: My experience was that transaction reporting was historically derived from applications and data sources not originally designed with a primary function of transaction reporting

Complexity in technology architecture combined with growing data requirements from regulators can make delivering accurate and timely data more challenging than those not close to this space may think.

MarketAxess: Two years ago the CSSF explicitly stated that “…a low level of data quality (e.g. incomplete or missing transaction reports) will not be considered as a negligible offence but may constitute a serious infringement of professional obligations under MiFIR.” Other regulatory bodies have expressed similar sentiments. What can senior managers do to ensure their firms adhere to the demand for completeness and accuracy?

Paul Graham: "My approach was to think about my role as being personally responsible for continuously asking questions that related to the foundations of the process and the culture. Foundations—Does the firm have the following aspects correct to enable completeness and accuracy: Technology applications; Data; Organisational construct; Governance; Operating procedures; And a control framework to positively affirm adherence and/or detect issues early?

Culture—Is there a deep and wide understanding of the importance of completeness and accuracy across the firm? Do we have the right people in the right roles? Is the culture balanced correctly: People understanding the need for high standards whilst equally feeling comfortable making mistakes and speaking up?"

I also periodically sought out external advice and perspective. Even with a highly effective team, there can, over time, be a drift toward groupthink.

MarketAxess: Given the global move to electronic trading and the resulting vast pools of data exhaust, where do you think third parties and vendors can add the most value in helping firms with data quality?

Paul Graham: They provide a really valuable service. Firstly, they can provide ready-made technology solutions that firms may not have had the time and resources to build themselves. But beyond this, I think they provide two additional benefits a firm won’t get on its own:

1. They can reconcile your data vs other users of their platform. This is an excellent way to validate your own reporting and detect problems early.

2. They have a unique perspective on how all of their member firms are operating and can give valuable insight to best practices and problems that other firms are encountering.

That said, the Senior Manager is always ultimately responsible, so they must be clear on what they are getting from the third party and make sure it is of an appropriate standard.

MarketAxess: Ever-changing regulatory requirements are best met with advanced technology. However, investing in automated AI QA processes consistently gets deprioritised. What is the surest way to make the case that technology is vital in helping deliver the above needs?

Paul Graham: It’s absolutely critical. Without a robust technology architecture and the associated controls around that, it becomes exponentially challenging.

Much of the new and emerging technology solutions represent a great opportunity to proactively interrogate the accuracy of TR data. I had previously overseen the implementation of quality assurance processes which were heavily reliant on SMEs. Whilst these are very useful controls, they are challenging to scale up.

Technology can scale up those processes now to a very effective level. Of course, you need SMEs involved but the combination of SMEs and tech is a key component of your operating model that all senior managers should be leveraging.

MarketAxess: As data now drives so much decision-making, can improving reg data accuracy have other business benefits?

Paul Graham: Of course, the vast majority of data used for TR is used for multiple other processes across any firm from finance, to risk management and trading. In my experience, I witnessed improvements in reg data accuracy having a positive impact on other operational processes I oversaw such as settlements, confirmations and collateral management.

MarketAxess: The last thing any company needs is for regulators to spot inaccuracies first. What one bit of advice would you give to senior managers to help drive accuracy?

Paul Graham: I would say you need to be the "person that asks for more", because if it's not you, who will it be?

You are the person who needs to ask your team to continuously improve. To ask Technology to deliver better solutions. And where needed, ask the executives above you for support. This is a never-ending cycle as requirements, technology and people change.

It is very valuable to speak to external parties to see how they approach their TR. Additionally, having a third party come and pressure test your process or review your data is highly valuable and a good example of an SMF taking ‘reasonable steps’ in relation to their prescribed responsibilities.

Time for change

When regulators take action, the average age of incorrectly reported transactions is 254 days. In a business based on forecasting market movement, it is incredibly disruptive and expensive to wind back the clock and deal with errors that in many cases could have been avoided. With MarketAxess’ SensAI you can fast forward to real-time reporting.

When you are the "person that asks for more", as Paul Graham suggests, shouldn't you be asking for SensAI?  Not only are you benefiting from expert outside help, but our machine learning algorithm derives insight from over 20 billion transactions.

SensAI not only monitors your reporting but also learns from your trading habits and predicts the expected number of transactions reported. It compares what you’ve reported against your counterparty’s version of the trade, and flags potential errors using market-wide data. Plus it stays up to date with ever changing regulatory rules.

To learn more about SensAI, contact our team.