Go jump in the lake: the trouble with data

The buy side and its service providers grapple with how to make big data useful.

120
https://commons.wikimedia.org/wiki/File:Rara_lake_2.JPG

A data lake is a repository of raw data in enormous quantities, there to be plucked intelligently and put into action. Fund management companies and their service providers appreciate the value that accessing such data enables – but legacy systems, overwhelming demand and business pressures mean firms are treading water at best – and risk drowning.

The good news is that business heads at fund houses have come around to the benefits of managing data in a centralized way. Some are outsourcing to vendors; others build their vaults in-house.

The bad news is that demand to amalgamate, access and report on data has exceeded the capability of managers and their service providers to deliver.

Rob Scott, head of operations and technology for Asia ex-Japan at Nikko Asset Management, says many companies have been slow to build a data warehouse because regulators in some markets frown on moving customer data outside the country, or make it difficult to commingle it.

Trendy and troublesome
But now that heads of business have woken up to the need for client insights, operations and I.T. departments are under pressure to hurry up projects.

“The various teams are now focused on A.I., and they’re reaching out to us [operations],” he said.

He made these remarks speaking at a recent conference in Hong Kong organized by TSAM.

“People want data but they lack the infrastructure to use it”

Fund houses, in turn, are pressuring their custodians to help them make sense of the data they have, says Caroline Higgins, head of global fund services for Asia Pacific at Northern Trust.

“People want data, but they lack the infrastructure to get it, to understand it, and to use it,” she said. Both regulators and investment firms are leaning on service providers to help them make sense of things, given custodians’ access to a fund client’s data, not to mention masses of aggregate information.

But banks in this space have to juggle every client’s demand for a bespoke solution versus the high costs of delivering on this. They would rather sell a generic product.

They are also unnerved by the tendency of individual executives among fund houses with a passion for using data getting banks to customize something for them – only to see that person change positions or leave the firm, and the successor have little appreciation for putting the reports to work.

Data as a service
Vendors see an opportunity amid such confusion. Marc Rubenfeld, Europe and Asia head of Eagle Solutions, a data warehousing provider (and a unit of BNY Mellon), says vendor solutions are moving to provide data as a service – much as fintechs sell software as a service.

Today, vendors such as Eagle sell their data lake solutions as an expensive product, but they see opportunity in hosting data services that buy sides subscribe to.

The idea of data warehousing – to organize streams of data from different feeds – has been around for more than a decade. Buy sides often choose to build these in house, as Nikko Asset Management has done, but they must continually invest to keep these up to speed with new products and new regulation.

Nikko’s Scott says the firm built its own solution because it has taken time for business heads to realize the benefit of using customer and market data to augment product design and sales. He acknowledges it’s messy to aggregate data when it comes from multiple feeds, from different departments with their own priorities.

But there are benefits to such an approach, too: “If departments own their data, it’s more likely to be good quality,” he said.