Data has always been at the heart of the capital markets industry. Financial institutions need to be able to filter and make sense of the huge volume of data available, not just for the competitive advantage but also to better manage risk and maintain compliance.

“Five years ago the C-suite did not value data and there was a lack of commitment to experimentation, but that has changed now,” says Steve Livermore, an independent management and compliance consultant and former global head of UBS’s monitoring, surveillance, and controls function.

“There is a realization that this approach to technology is being used in many other industries and proving successful. It hasn’t really become the norm yet in banking. Most have been running projects and have moved beyond that now and have levels of program across different parts of their organization.”

Given the sheer volume of data that financial institutions generate, organizations need a strategic plan around how they intend to use it, and this needs to be led by the C-suite.

“Five years ago the C-suite did not value data and there was a lack of commitment to experimentation, but that has changed now.”

Steve Livermore, independent management and compliance consultant

“People have now realized you have to have that data strategy first – the value and relevance of good, clean data,” says Livermore. “Not only do you need the strategy and governance framework, you need to ask first what you want to do with that data.

“You need a top-of-the-house view of how you will build your data model. It needs to be something that comes out of the CEO’s office and runs at that level. A bottom-up approach doesn’t work as everyone fights for their right to get the data to work explicitly for them. But if you pull all your data together and have a clear vision of what it will ultimately do once it is in a central portal, you are in a position to start applying any advanced analytics technologies or approaches.”

Need for more than speed

The need for real-time analytics to provide fast, actionable insights has multiple use cases across banks and other financial institutions. Traders’ profits have always largely hinged on having access to information faster than anyone else, but speed on its own is no longer enough.

“Because real time is more available than it’s ever been, the informational advantages have balanced out,” says Tim Morris, director of product for S&P Global Market Intelligence’s Capital IQ platform. “At the very extreme you have Flash Boys-type cases where actually getting that extra microsecond of real time could be critical, but it’s at the point now where that’s so expensive to invest in that the returns are almost negligible.”

That means the focus has shifted to how firms can eke out analytical advantages that others may lack even if they have access to the same information at the same time.

“Behavioral insights is a blossoming area of real-time analytics, as is real-time analysis of unstructured data – things that just weren’t possible a few years ago,” says Morris. “Textual data analysis is one example, so real-time analysis of an earnings call as it happens.”

“SEC and FINRA rules have forced banks, broker-dealers and other financial institutions to provide evidence that they are making best efforts to monitor employee e-communication content to check that they are not overselling products, or laundering money, or committing wire fraud or ‘me too’-type abuses.”

Matt Barrow, Chief Data Scientist, Global Relay

Another area where real-time analytics is becoming increasingly important is risk management and helping financial institutions to anticipate issues before they happen. Take Evershed Sutherland’s legal and compliance services business Konexo. It has partnered with contract analysis platform LIKEZERO to improve the accuracy of its derivatives data to help financial institutions to monitor their derivatives risk better.

“Real-time analytics are being used in two key ways,” says Trystan Richard, Head of Legal Project Operations at Konexo.

“The first is to provide transaction-level insights and data that support front office and risk staff when making time-critical risk and pricing decisions. These contribute to whether a transaction is executed and on what basis. The second is to provide live management information on a portfolio level that helps risk staff understand the impacts of taking on or disposing of transactions and, to some extent, explore risk and pricing impacts of market, credit, and liquidity sensitivities.”  

Real-time analytics is also playing a vital role within compliance departments, such as making it easier to spot wrongdoing by monitoring electronic communication channels and quickly flagging potential issues.

“SEC and FINRA rules have forced banks, broker-dealers and other financial institutions to provide evidence that they are making best efforts to monitor employee e-communication content to check that they are not overselling products, or laundering money, or committing wire fraud or ‘me too’-type abuses,” says Matt Barrow, Chief Data Scientist at Global Relay.

“From a real-time analytics perspective, there is a requirement to do this monitoring as quickly as possible. If somebody is laundering money, you don’t want to wait five days to receive a notification saying this person sent an email and it looks like they might be laundering money; you want to know as quickly as possible whether somebody is behaving in a non-compliant way.”

Other use cases for that e-comms data are also emerging, including behavioral analytics both for internal staff members and customer trends.

“If somebody predominantly sends messages and chat items that are not business related, and the sentiment of those messages is particularly low, you might infer that that employee is not terribly engaged in their job and you might want to address that sooner rather than later,” says Barrow.

“That data can also provide intelligence around sales information, such as who a bank is selling to and what kind of price they can charge.”

Getting tech and infrastructure right

Gathering all of this data is not easy. The proliferation of digital communication channels means a single employee might use dozens of different chat tools or email accounts. Without the right technology to store all of that data in one place in a common format, financial institutions will struggle to find it when required for regulatory or legal discovery purposes.

“Take a bank like HSBC that may have a quarter of a million employees, all communicating across 150 different platforms every day. You’re not just storing big data, that’s massive amounts of data and regulators insist it needs to be stored in a safe way and can be searched accurately and quickly,” says Barrow.

Therefore, having the right infrastructure to allow for data capture, storage, and real-time analytics is critical. “If you want real-time analytics in banking, you need something that is a database and an analytics engine and ideally that utilises existing technologies like SQL,” says Livermore.

It is also important that financial institutions focus on the analytical outcomes they need rather than getting bogged down with advanced data science.

“There is no real need for bleeding-edge AI with what we are trying to achieve,” says Livermore. “Low-level AI can be very relevant and worthwhile, and is arguably more than most are doing right now. It is all about finding the right level for your organization. Data lakes, deep learning, robotic champions? All you need to do is make sense of the data you’ve got and be able to apply common sense to it so you can start to pull information or metrics out of it that’s usable and you can do stuff with – you can get too funky sometimes.”