Traders have lived with market surveillance for some time now. In fact many current equity traders probably don’t know of a time before MiFID I where market structure and hence surveillance capabilities were very different. The rest of the market will be brought under the same dominion as their equities-focused colleagues with the introduction of MiFID II and the extension of market surveillance regulations.
The directive stipulates that all trade-related communication is recorded across asset classes, be it email, instant messenger or voice. It also has to be tagged and timestamped and then stored in such a way that it’s readily accessible when requested by the regulator. This is enforced with the threat of increased fines for non-compliance.
This may sound daunting but the reality is that solutions that work for equities will work for other asset classes; potential solutions may well already exist within the institution and need only be extended to cover other trading activity. So for many, compliance with monitoring requirements will not be overly difficult. However regulatory overlap does present additional challenges.
MiFID II is not the only regulation that demands market surveillance, MAR and Dodd Frank both require oversight. Establishing what level of surveillance is common between these regulations is critical to providing implementation teams a base from which to extend changes to ensure compliance with the nuances of each one.
Compliance in three parts
In the case of MiFID II, voice and data both need to be monitored, timestamped and recorded. The compliance process can be split into three segments – storage, surveillance capabilities and market data needs.
The first place to look is the increased storage requirements that expanded record keeping necessitates. This is a relatively simple decision between hosted, on premise and off-premise. Various factors will need to be considered including the physical space needed, the storage capacity and the potential increase over time.
The main focus of the compliance effort should be the capability of the monitoring systems themselves. The solutions used to comply with MiFID I are coming up to 10 years old and that’s a long time in technology terms – the first iPhone was launched in the same year as MiFID I. And yet market surveillance systems have not been materially upgraded. Indeed in much of Europe, due to privacy laws, market surveillance is rudimentary.
Voice and electronic communications need to be handled differently and the stipulations for both have their own intricacies. Both need to be timestamped but to differing tolerances and while it’s relatively simple to scan electronic communication for signs of irregular behaviour, doing the same for voice is not so straightforward. Speech recognition technologies are still inadequate and nuances of language or dialects are all but indecipherable to software. What about private wires? These are still not IP-based and monitoring may well prove difficult.
Electronic communications by comparison are easier to process. The issue is the plethora of communication tools available, both approved by the institution and not. Clearly monitoring approved channels doesn’t present problems but what about social tools? Facebook Messenger? WhatsApp? Snapchat? How can these be monitored? While there are solutions which enable monitoring and recording, engendering a change in behaviour of traders will be key to ensuring compliance. It’s an oft-reported maxim that surveillance rules are in place to protect traders as well as the wider market.
The third piece in the jigsaw is market data. Pricing data is an essential part of market surveillance and this needs to be incorporated in any record keeping effort. Trade reconstruction is nothing without context and without a picture of the wider market it’s impossible to tell how a trade affected the market – whether the intention was fraudulent or not. Indeed, combining market data feeds with analytics technologies has the potential to reduce false positives and to spot trading behaviours before they become a problem.
Regulation, Reputation and Revenue
Since MiFID II extends coverage to multiple asset classes, all the above monitoring has to take place across trading desks. This may well mean that internal budgeting structures need to change, especially when ongoing servicing agreements and data feed costs need to be factored in to planning. Indeed the effort to comply across asset classes may uncover opportunities for cost or technology arbitrage where systems are common to many different trading desks. The technology will have the same functionality whether trading equities, fixed income or FX and the same market data feeds will be needed regardless of physical location.
And while it might seem that compliance involves a lot of effort with little compensation for the trading floor, the cost of non-compliance could be huge. Several institutions have been in the news after being fined for trading irregularities, and the size of these fines has not been insignificant. A particularly high-profile case has even been referred to the Serious Fraud Office in the UK.
Complying with monitoring requirements will not be the most difficult part of MiFID II and it may have the biggest consequences. It would be tempting to focus on the more interesting parts of the regulation to the detriment of market surveillance but, as recent fines have proved, it’s not something to be taken lightly.