Swimming with FIS: Navigating the data ocean

The investment industry has a problem: it is drowning in a digital ocean of data. From a regulatory point of view, advisers need to collect a huge (and growing) amount of data; from their clients; and then, there is data on the clients’ investments, and how they are performing.

When an advice firm wants to call-up a complete view of a client’s wealth, to provide a report; or to assess how its own business is travelling, it must quickly find and interrogate the relevant data it holds, and compile the response quickly. But it’s a morass.

The industry’s data management problem stems from not only the sheer amount of data that threatens to overwhelm it, but the different kinds of data — especially the distinction between ‘structured’ and ‘unstructured’ data.

‘Structured’ data is all the data that is standardised in tables and spreadsheets that arrange it in rows and columns that clearly define data attributes — such as name, address, and phone number. Computers can easily and effectively process structured data for insights, due to its quantitative nature.

But ‘unstructured’ data is a different beast. It is information that is not ordered in a pre-defined way through data models and stored in a structured database format; it could be anything: from text files, video or audio files, images, reports, emails, images, social media posts, recorded call-centre interactions, sensor data, to customer records. It might be human-generated, or machine-generated in a textual or a non-textual format.

Unstructured data is ubiquitous and abundant: more than 90% of all enterprise data is unstructured data. And, given this vast majority, enterprises require more money, space, and resources to store it, let alone use it efficiently.

And that’s where the industry’s problem lies.

“You have all of these different participants, from wealth providers, asset managers, general partners (GPs), limited partners (LPs), and individual investors of all different types of flavours and types,” says Andres Choussy, group president, trading and asset services at global financial technology firm FIS.

“Data is coming in from many different places, and there are so many different sources of data that need to be collated, standardised, normalised, cleansed, stored, and maintained, so that it can then fuel whatever activity, whatever systems, whatever processes each of those individual players have.”

Even structured data has its limitations: investors filling-out forms can often use different arrangements of their names and initials. “That’s a huge problem for the industry, because it happens all the time: sometimes investors use their middle name, sometimes an initial, sometimes two initials. That’s where you need a cleansing tool that actually has the intelligence to realise all those different name configurations are the same person,” says Choussy.

This task was difficult enough when the financial industry was mainly trading liquid instruments; but it “becomes much more complicated when you’re talking about alternative investments,” he says.

The problem is that much of the liquid investment world is structured data; but alternative investment structures mostly sit on unstructured data sources.

“Alternative investments is a sector where essentially, each individual prospectus is different from the next prospectus, and therefore the data is not really flowing into any one system or being captured in a way that can actually be then easily distributed, says Choussy.

“It’s unstructured data that humans understand quite easily, but typically, it has been difficult for a computer to understand. But now with AI we can go through it, prospectus by prospectus, and extract the information that’s important — we can have the machine self-learn how to read the data, how to interpret what data is important, what data is not important, and to capture that data.”

This ability is at the heart of FIS’ investment data management platform, which — in partnership with data and analytics company Exafluence and cloud-based data storage company Snowflake, Inc. — is revolutionising the ability of industry participants to aggregate, store, cleanse, mine and interrogate their data. FIS’ system collects, integrates, continuously cleans and standardises data from any dataset — structured or unstructured — to provide accurate client reports. The system can reduce the time spent on data aggregation from 20 hours to minutes, significantly improving efficiency.

“Industry feedback suggests the average number of data sources that need to be aggregated is between five to 20 different data types,” says Jon Hodges (pictured), senior vice-president, head of trading and asset services, Asia Pacific, FIS. “The industry needs a data management solution, so we developed an investment data platform in collaboration with the industry. We see the platform as a market utility that all wealth and financial advice firms can use to solve their data management problems and unleash the power of AI, to produce accurate, on-demand data in minutes, not hours, or even days.” 

The FIS investment data platform brings its users not only cost and time savings, but also the highest level of cyber-security controls, and built-in compliance with the Australia Privacy Act and Personally Identifiable Information (PII) regulations, which are laws and standards that govern how organisations handle personal data that can be used to identify an individual. Not only does this enable quicker and more accurate reporting, it helps firms meet legal and regulatory standards around privacy and data retention/deletion. The platform also allows data owners to control access and visibility.

“Our observation is that more than 50 per cent of the cost of a wealth manager is spent on this task, of aggregating, cleansing and then producing information,” says Hodges. “That’s because the wealth industry has been growing exponentially, with resultant growth in the quantum, and different sources, of information, with different structured and unstructured formats.

“It has overwhelmed the average wealth manager, who finds themselves needing about 20 hours to produce a client report, because they have to aggregate the data, fix it, then compare it to the previous quarter’s report, with commentaries and recommendations attached. But with sophisticated technology, we have been able to build a solution that solves that problem,” says Hodges. “We have automated that whole process so it can be done in a matter of minutes.”