JUMPING THROUGH HOOPS.
Best Execution spoke to Lars-Ivar Sellberg, Executive Chairman of market surveillance solutions provider, Scila about the challenges implementing the Market Abuse Regulation [MAR].
What was the buyside’s level of preparedness for implementing MAR?
In my view, many buyside participants were taken by surprise by MAR, especially the obligation to implement an automated trade surveillance procedure. In their defence, it would seem a bit extreme to deploy a full-scale surveillance system for an investment firm that only trades a few times a day. The problem is that the term “proportionate” used in MAR is too vaguely defined. While the regulator may have had a good reason not to be too specific when formulating regulatory text, in this case details would not have gone amiss.
Many of the firms we have spoken to over the last year do not even execute trades themselves, which means they don’t have any experience in deploying IT solutions such as surveillance. Based on our conversations with the buyside, both pre and post MAR, surveillance vendors need to develop a scalable solution that works for all market participants, regardless of size, technical preparedness and financial strength.
Did firms have systems in place or did they have to start from scratch?
It differs greatly depending on the type of firm, asset class or market segment. MTFs are generally well prepared, since surveillance is a core part of their trading infrastructure and a necessity for market confidence. Large investment banks and brokers also have a history of using surveillance, although it is usually deployed T+1 (i.e. not real-time) and uses multiple systems spread across asset classes. For this category of firms, the challenge is to consolidate the various systems and conduct cross-product/asset class monitoring as per the requirements in MAR.
The markets that traditionally had the most underdeveloped trade surveillance are energy, foreign exchange and fixed income. While some participants have taken this very seriously in the past, few have deployed automated trade surveillance systems to monitor their markets prior to MAR. This is partly due to the fact that many technology vendors have focused on equities and exchange-traded products.
In fact many firms that we have met over the last couple of years are tired of being shown surveillance systems monitoring on-exchange or order book trading. Consequently we have spent a lot of time and effort to develop functionality around these areas, including request-for-quote/auction based support and specialised energy markets monitoring.
What are the key elements for implementing a surveillance system?
Implementing a trade surveillance system typically involves resources from IT, compliance and trading. It often involves a mix of people who have not worked together before, so it’s important to have a project manager who fully understands all of these areas. Having a grasp of your own data is a key factor to a successful implementation, especially if you are trading multiple asset classes and the data is spread out across different source systems.
Integrating this data into one centralised trade surveillance solution can be beneficial for both compliance and surveillance. We also encourage our clients to use the trade surveillance system for other purposes, such as business analysis, studying market behaviours, etc. There is a wealth of consolidated data generated that can be effectively leveraged.
And finally – compliance, legal and trading need to adopt a risk-based approach to trade surveillance and apply suitable monitoring algos – or alerts – to comply with both regulation and market rules. If you have direct market access or sponsored access, it’s also important to remember that exchanges and marketplaces can issue fines for market abuse, so it’s not only a matter of complying with MAR.
What were some of the biggest technical challenges of implementation?
From my experience working with both trading and surveillance systems, key challenges include the process of outputting all relevant data from separate source systems and implementing effective testing procedures. Since regulation now requires firms to monitor all orders irrespective of asset class, they need to ensure all data is processed in a centralised trade surveillance solution. That makes testing a crucial part of the implementation process. In the end you need to be confident that your source systems are pushing all data through the trade surveillance platform. While many tech-savvy firms in the financial industry have experience in testing, others do not. We recently partnered with test and change management specialists Certeco to provide an alternative to using internal resources for implementing a trade surveillance system.
How are firms meeting the new rules for monitoring the different asset classes?
Generally I think firms are doing very well. It gives them the opportunity to have a comprehensive view of their trading operations, which traditionally has been quite challenging for firms such as inter-dealer brokers. We are working with two of the largest firms, and it has been a rewarding challenge for both sides to consolidate data from five to ten source systems into one centralised solution for cross-asset class monitoring.
What is the most effective way to monitor instruments from different source systems?
In a perfect world, different source systems are perfectly in sync. In reality, however, this is not the case. The most effective way to monitor instruments that are linked together but reside in different source systems, is to base synchronisation on source timestamps and allow for non-precise and less granular time stamps.
As a vendor, we need to adapt to the client’s infrastructure. Many vendors rely on industry standard connectivity such as FIX, but we know from countless examples in the past that clients can rarely provide all the data from different source systems in the same format. Therefore, our surveillance system needs to be flexible enough to cater with a wide range of formats. To this date we have built adaptors for over 100 different protocols and file formats.
Against this backdrop, what lessons can US markets learn from MAR?
When introducing new regulations for trade surveillance and prevention of market abuse, the regulator needs to be precise enough in order to avoid putting an unnecessary burden on market participants interpreting the rules. This represents a real cost for firms, and we now see this being a problem for the buyside, in Europe especially.
However, the regulation should be flexible enough to cover new abusive behaviours and changes in market structure. Take spoofing or layering, if you look at some of the actual cases from regulators and marketplaces in Europe and the US, the definitions differ.