How your scheme can make the most of big data

‘Big data’ is one of the highest-profile concepts to emerge from the technology industry in the internet era and, as would be expected of an industry as beloved of buzzwords as the pensions world, it is suitably nebulous.

One person spoken to for this article shared a joke from a colleague: “If you ask three people to define big data you’ll get five answers."

However, key objectives of big data, namely amassing and studying vast datasets in the hope of gaining insights that would not be apparent in smaller samples, have begun to permeate aspects of the pensions and investment industries.

The emerging field has multiple potential applications for pension schemes, from choosing and managing investments to segmenting membership in order to predict member behaviour and target communications accordingly.

Investment channels

JR Lowry, head of EMEA for trading company State Street Global Exchange, said investment applications depended on whether investments were managed internally or outsourced.

“You’ve got a wide variation in pension schemes between insourcing and outsourcing,” he said. “If you’re a pure outsourcer the key thing is that you get good, secure data. But also that you understand the lineage of the data and where it’s coming from.”

Some of the larger providers are looking to understand the decisions that people are making and, because they have much larger datasets, they are looking to draw conclusions

Ian McQuade, Muse Advisory

He added increased transparency was in the spirit of forthcoming regulation on financial institutions.

“Solvency II is a regulatory push to get insurers to understand what’s going on in their portfolio,” he said.

Ray Pygott, head of trustee services at consultancy KPMG, said the growing profile of data science provided a good opportunity for schemes.

“We need to give members online access to what their benefits are,” he said. “Technology has got to be the way forward.”

He added: “There’s so much data in a pension plan. Plans tend to go back decades and there are thousands of points of data you have to hold for each person.”

Pygott said the pensions industry may be better placed than others to capitalise on big data, arguing that some parts have used similar approaches for decades.

“Actuaries have always had to do that. Part of a funding valuation is looking forward, so you look at the data you have and try to use it to judge what it might mean for the future. The tools for doing that are much better than they were before.”

Boundaries and uses

However, others were more sceptical, citing the typical size of pension funds' datasets as an obstacle.

Ian McQuade, client director at governance consultancy Muse Advisory, said: “Even the trustees of the largest schemes have limited datasets to work with. It may be tens of thousands of members.”

McQuade said major providers were, however, using their data to help prepare for pension flexibility and beyond.

“Some of the larger providers are looking to understand the decisions that people are making and, because they have much larger datasets, they are looking to draw conclusions,” he said. 

Beyond analysing a scheme's membership, McQuade said data are better enabling trustees to assess the value of the experts they employ.

“The reason trustees use communication experts and talk to their administrator is because those guys have been through this before,” he said. 

He gave the example of communication consultants breaking down the results of previous campaigns, such as increased engagement or contributions.

“That’s where big data will become interesting, whereas a lot of it is best guessing,” he said.