Front office systems are an essential tool in detecting suspicious activity and churn out huge amounts of data. But it’s quality not quantity that risk and control teams really need.
Ask any senior front office controls professional what will be the dominating theme of the next few years in their corner of the financial services industry and the answer is likely to be “technology”. Indeed, it has become something of a mantra among chief executives of big banks that these days they run ‘technology firms with financial services’ and it seems unlikely that this will change any time soon.
However, while it is doubtless true that the latest generation of digital machinery has made the task of supervisors in the 1st line of defence more possible, it is also true that this machinery has thrown up particular headaches of its own as well. Managing the front office systems and the data they spit out has become a major conundrum, and one that banks are currently a long way from solving.
Let’s get digital
Clearly, the great attraction of new generation technology to front office controls systems is that, when working well, it allows headcount to be reduced. The initial outlay on systems might be hefty, but, long-term, there are significant savings in wages to be made. “There are a lot of cost savings that can be realised and absolutely must be realised,” says Russell Dinnage, head of the capital markets intelligence research practice at consultancy GreySpark Partners in London.
Without new systems, supervisors in the front office would be unable to perform their functions effectively. As Konstantinos Rizakos, global head of compliance and regulatory systems at Citigroup in New York, says: “Mortgage servicers, for example, have a prescribed process with a set number of documents to review, so they’re not suffering an information overload. But the supervisor on a trading desk has an open spectrum of things to look at, so there’s no way he or she can manually pick out the problematic transaction or communication. They absolutely need technology or they’re looking for a needle in a haystack.”
Of course, banks are operating under considerable financial constraints these days, so every area of control that clamours for extra financial investment in systems doesn’t get it. Bankers and consultants agree that those areas that have to be in shape to comply with regulations or, even more pressingly, those areas that have to be in shape to comply with a regulatory enforcement order of some kind, receive priority. What regulators want continues to be the main driver.
Too much of a good thing
Most supervisors at tier one banks have a dashboard, to which suspicious trades or communications are uploaded. In addition to user interfaces he or she will also rely, to a degree, on artificial intelligence systems to find anomalies in any pattern. But these systems are only useful to the extent that the data that is fed into them is reliable and manageable, and also to the extent that data they produce for consumption by the supervisor can be assimilated. And these are two areas that, at the moment, are often deficient.
“There’s a limit to how much information you can process. Humans can only assimilate so much. There’s a huge problem with ranking,” says Rizakos. Russell Dinnage echoes this: “I get clients coming to me all the time who are not only drowning in data that their banks generate internally but also in what clients are asking from them for regulatory reporting or compliance reporting. It’s the big challenge in investment banking and capital markets today.”
Machines have gotten better and better at producing lots of and lots of information; but banks are struggling to make sense of it all. They’re paddling hard to keep on top of the sea of information that is returned to them by ever-more comprehensive machinery.
“Talk about the poor quality of tools is generally a leading indicator you’ve got a problem with the data.” Jonathon Frieder, Accenture
Garbage in? Garbage out
Moreover, the quality of data produced is often traduced by the data digested. As Jonathan Frieder, compliance technology lead for the North American regulatory group at consultancy Accenture in New York, says: “If you put garbage in, you get garbage out. There are lots of hygiene issues banks are struggling with. There are big overarching governance and process considerations to be implemented before jumping into technology.”
The data needs to be accurately sourced and the mapping needs to be correct. Data from all areas of the bank needs to be selected with the same criteria and processes, otherwise the results will be useless. All has to be subject to rigorous data analysis.
“There are always complaints from the 1st line of defence about the effectiveness of their tools, but this is a symptom of the problem,” adds Frieder. “Talk about the poor quality of tools is generally a leading indicator you’ve got a problem with the data.”
Scott Levine, a director at PwC in New York and part of the financial services digital capital markets consulting team, agrees. “It doesn’t matter what advanced technology a bank wants to use, it’s no good unless proper data management and governance are in place. There has to be a standardised taxonomy across the bank so the data can be complete and accurate for advanced technology to be applied.” This, even more pressingly than making decisions about the types of technology on offer, leads the list of important questions to be answered at leading banks.
Still early days
It might be assumed that most banks in the tier one space are at broadly the same stage of evolution with their systems and their management of data; this would be a false assumption. Technology consultants – who see the whole gamut of banks – agree that each institution is at a different stage of development than even its closest peers.
Sam Regan, global lead for Accenture’s regulatory remediation and compliance transformation group in London, notes: “Even tier one firms vary. They’re at different stages of the journey. Some are very advanced and some have barely started yet because it hasn’t been their number one priority.”
According to one senior risk officer at a major tier one firm, IT specialists are still not always justly respected within banks. In part, this is because technology has moved beyond the java era, during which a lot of people could understand the processes and principles, into one of fiendish complexity. Now you need to be specialist to get to the bottom of how the technology works.
Unfortunately, the increasing sophistication of the role and the greater level of expertise required has not led to a commensurate rise in professional acknowledgement afforded to the specialists, he says. Bankers continue to treat them like utility suppliers.
“They are happy when it works and not happy when it doesn’t, and don’t want to know why. Technology has become like medicine in that you need to be a specialist to understand it. But doctors still get treated with respect and people don’t try to micromanage their prescriptions.
This lack of professional acknowledgement creates friction. Perhaps more worryingly, bankers still think they can understand the front office technology. “A lot of people in the business function do not appreciate the complexity. This is sometime dangerous because they make decisions they don’t understand,” the risk officer concludes.
So, if CEOs of major banks are right and they do now run technology firms with some financial services thrown in, it seems they are still struggling to make sense of this new business model.
All banks face big decisions about whether to build technology in-house or buy from a vendor. Both routes have something to be said for them, but, according to consultants, it generally makes more sense to build in-house rather than buy off the shelf because every bank is different and one size does not fit all. Even those that do buy from a vendor generally then have to spend time customizing the product to suit the exigencies of their institution.
“We were dealing with eight banks that wanted dashboards,” says Dinnage. “Six built them in-house and two bought from vendors. In time, those who built seemed to be further along the line than those who bought. Buying from a vendor ramps you up quickly, but given the complexity of each bank, you can’t just buy it, plug in and play. Nothing we have seen does this.”
So, banks have bigger and better computers than were dreamed of ten years ago. Without them, the task of a 1st line of defence supervisor would be nigh-on impossible, and, at some stage in the future, they will allow banks to make significant cost savings. But at the moment there are a great many growing pains.
“The amount of data the market has access to far exceeds its ability to do anything useful with it. It’s the challenge of the era and one that is not resolvable any time soon,” concludes Dinnage.