Defined Benefit

Automation could save defined benefit (DB) pension schemes tens of thousands of pounds a year, according to research by Spence & Partners.

The advisory firm analysed running costs associated with DB schemes with between 100 and 10,000 members, including administration, actuarial and investment costs, as it said these were the areas in which most efficiencies could be realised.

The research, Spence said, showed that an average DB scheme with 1,000 members could save more than £100,000 a year by automating functions and eliminating other inefficiencies.

It argued that, if this figure was extrapolated across the universe of approximately 3,000 DB schemes with between 100 and 10,000 members, the sector could save as much as £300m a year.

Alistair Russell-Smith, head of Spence & Partners’ corporate advisory practice, said: “Many schemes are serviced by older software, augmented by a range of sticking plaster solutions to get them to provide a basic service for pension members and fulfil their regulatory requirements. This creates inefficiency within the market.

“Every pound companies spend on running their pension schemes is a pound not spent on member benefits or other company priorities. This is a challenge for policymakers as well as individual schemes.”

He also pointed out that modernising systems and processes would be “necessary for plugging DB schemes into the pensions dashboard [ecosystem] in a timely and cost-effective way”.

‘Overdue an overhaul’

Philip Dickinson, director of Cosan Consulting, said: “I don’t think anyone involved in running pensions would challenge the findings. The way schemes are run is overdue an overhaul.”

Quoted in Spence & Partners report, Dickinson said there were “some examples of good practice”, particularly in larger schemes, but “truly slick tech led operations are rare”.

“As well as the obvious implications on cost there is a big impact on risk,” he continued. “Where operations are using sub optimal models and poor IT architecture, risk will be increased, whether that is simply getting things wrong (i.e. paying the wrong benefits), reputational risk from delivering poor service or cyber risk.”

Spence – whose sister company Mantle provides DB scheme technology including automation software – said modernising systems could help mitigate these risks while also helping schemes reach buyout sooner, or potentially generate a higher surplus for those schemes planning to run on.

“One of the schemes that we administer has two million members,” said Bala Viswanathan, CEO of Aptia, quoted in Spence’s report.

“We have about 150 people that we require to service that scheme. It’s a DC master trust. A significant proportion of the work is automated. But then we’ve got a DB scheme which has got 10,000 members and you’d probably take 150 people to do that as well.”

Most inefficiencies, Spence & Partners said, occurred when scheme advisers such as actuaries and investment professionals have to interact and share data.

In addition, the firm said “typical” DB scheme technology was “fragmented”, with data moving between multiple systems, an overreliance on spreadsheets, and unnecessary manual tasks all examples of inefficiencies.

Cost savings

Martin Freeman, director at Pensio, told Pensions Expert: “Smaller DB schemes have often tolerated higher levels of manual calculations, because they compared the cost of automating them with the cost of an administrator, often with a very short payback time.

“But this judgement seems shortsighted when you factor in the higher risk of making mistakes, and of data breaches from having so many spreadsheets and files whizzing around.  

“It also means you can't tell people what their benefits are online, which erodes members' trust.”

The advent of pension dashboards were “the final nail in the coffin” for spreadsheet-based calculations, Freeman added, as the increased volumes and reduced timescales would “make automation a necessity”.

“Trustees who look after schemes with these inefficiencies will need to prioritise repaying this technical debt, or else they will struggle to comply with the dashboard duties,” he said.

Full automation of the administration of member events, additional member self-service technology and the streamlining of systems for communications, valuations and pension increase exercises could help the average scheme in Spence’s research save £35,000 a year.

Additional savings could be recognised with automation for actuarial and investment processes, Spence claimed, with similar efficiencies available for project work such as guaranteed minimum pension equalisation and dashboard preparations.