Future proofing your system
Buying and installing a computer system to support the operation of a pension fund is usually a major project and a major expense. Whether it is to support administration, or trading and portfolio management for those pension funds that manage their own investments, or other purpose, computer systems are now essential to the operation of most funds. They are inevitably complex and take time to implement, and are likely to be with the fund for a number of years, so making the right choice of system is important. To help ensure an appropriate choice and a successful implementation, there are a number of things a fund should look for.
It may seem obvious, but the first thing to check is that the system supports the particular type of scheme the pension fund is running - defined benefit, defined contribution, stakeholder, etc. “Suppliers tailor their systems to different parts to market,” says Adrian Howells, group marketing manager at pensions administration systems supplier aquila, based in the UK. Although a supplier may claim it supports a range of schemes, it may cover some better than others. Suppliers tend to cover all the legal requirements in a particular area, but beyond that they can take different approaches to various optional aspects of the system, says Howells.
Next, the fund should preferably look for a system that is tried and tested, has a good user base, and has also a strong user group, recommends Jocelyn Blackwell, managing director of London-based pension fund consultancy Dunnett Shaw. Howells agrees. Start-up companies may have something new to offer, particularly in terms of contemporary technology, but the fund should check that they have a thorough knowledge of the complexities of the pensions world, he says.
Is the system user friendly, and is it easy to set up interfaces to other relevant internal systems or to external information systems? These are among the first questions that Penny Green, chief executive of Superannuation Arrangements of the University of London (SAUL) Trustee Company, suggests a fund asks of a potential new system. The general aim of computer systems is to automate what were once manual processes, and one of the major spin-off benefits is that they can pass information in an electronic form to other systems. This means that information need only be entered once although it is required by several systems, thus reducing input errors, and can pass more quickly between systems, whether they are internal (say, between a document imaging system and a pension calculation module) or external (say, to a broker or investment manager’s system). Also, many organisations where pensions are incorporated into human resources now want to link their pension administration and HR systems, says Howells. This linking of systems for the electronic transfer of data is known in the computer world as straight-through processing (STP) and is something of a holy grail in the trading, risk and portfolio management and settlement area. The underlying issue is whether the system eliminates as much human error as possible, says Green.
“Is the system future proof?” is another key question to ask, says Blackwell. Are the system architecture and the languages and tools it was built with based on modern standards and able to evolve with current technological trends? “Open” is the term the computer industry uses to describe systems that allow users choice in things like the database they use, or hardware the software runs on, or links to other systems. Blackwell recommends that a fund ask suppliers to describe their IT strategy and have it checked over by internal IT staff, if they are available. “This applies to all the component parts of a system - the hardware, database, report generation tools, document management production, workflow, and so on,” she says.
One of the most important future proof aspects to look for is support for using the web. “Organisations are being driven to look at their costs and how they can reduce those, and part of the solution is the web-enabling of some of processes so that employers and members can run at least parts_of the system on their own behalf,” says Howells.
Pension funds such as the Swedish government’s PPM and the Danish government’s ATP schemes allow members to access their accounts via the web, and in some cases make investment and other decisions online. Giving members access to their records via the web is something that Lloyd’s Superannuation Fund, which manages the pensions for the associate members of Lloyd’s of London, plans to do in the next year. The fund recently implemented a system from aquila, and web-enablement was one of the things the fund looked for in a system, says Robert Clark, pensions manager at Lloyd’s.
Another critical factor in the fund’s choice of system was aquila’s ability to supply its software as an application services provider (ASP). In this approach, instead of the organisation installing the system on their site, the software supplier hosts it at its site, with the fund sending the data to the supplier over the internet (or dedicated line), and the supplier sending back the desired output - calculations, reports and so on.
“We’re only a small company so we don’t have the resources to provide things like maintenance of hardware, backup of data and, probably more important, disaster recovery,” says Clark.
The ASP approach is becoming an increasingly popular way of delivering software, and is really a return to an earlier model of computing where applications ran on mainframes as bureau services, says Blackwell. She suggests that an ASP could be particularly attractive to other funds like Lloyd’s that have little or no IT expertise in-house.
But the most important step in selecting a computer system is understanding and clarifying just what the fund wants from the system. The better the fund is able to do this, the better chance a supplier will have to deliver what is required, says Blackwell. Consultants can be helpful in this process. SAUL used Dunnett Shaw, who were very useful in terms of identifying suppliers and drawing up an invitation to tender, and in project management, says Green. The fund chose the Pensions Office system from Bristol-based pensions management practice Edis Partnerships.
Blackwell says that funds should also look beyond the technology and investigate the type of company the supplier is. “Funds should be looking for a good cultural match,” she says. The fund will most likely have to work with the company for some time - system lifespans were typically five to eight years, but now, with pressure on costs, funds are looking for systems to last them even longer periods.
In discussions about new systems for pension funds, cost usually comes fairly late into the conversation. This is not because the systems are cheap, or that cost does not matter to funds; rather its the case that it is difficult to assess the value for money of systems until it is clear what the fund wants, and how closely potential systems fit to their requirements. Once these factors are established, then a fund can assess whether it is worth paying a premium for a top-of-the-range system that has, say 10% more functionality than a rival system but costs 25% more.
Generally, pensions is a market where it is not in the interests of a supplier to underprice their software to win business, says Blackwell. “It does the supplier no good at all if they get in with a low price but then have to increase their prices because it makes the client unhappy,” she says. Usually, suppliers attempt to give an accurate estimate of the final price, but sometimes they are hampered by the fact that the fund does not give full information of what they want and expect.
Green says that the final cost will also be influenced by how much internal IT resources are available. Small, independent funds with little or no IT resources will most likely have to pay the supplier or a consultant for the implementation of the system, whereas larger funds, or those attached to corporates with large IT resources, will be able to undertake the implementation themselves.