The search for a way to share risk between employers and employees continues in the UK with the recent publication of a consultation document. Gail Moss explores its proposals
The consultation period for the UK government’s proposals to open defined benefit (DB) schemes to risk sharing between employers and members finished on 28 August.
Published after intense lobbying by the industry, the proposals are designed to make DB schemes less costly for employers, and therefore more viable. But if the proposals do become law will they have any effect in slowing down the rate of closure of DB schemes?
Risk sharing originated in the Netherlands, although the way in which it works is different from the plans mooted for the UK. When introduced in the Netherlands, retirees assumed the risk of waiving indexation in case of a low funding ratio in the scheme. When many pension schemes switched from final salary to career average, active members also saw their wealth accumulation subject to conditional indexation. However, in most cases, there is an agreement that waived inflation can be made up for when things go well again in the long run.
“This may seem as taking rights away from beneficiaries, albeit with the full agreement of all parties involved, but has prevented the system from collapsing,” says Theo Kocken, chief executive, Cardano Risk Management.
“From the fund’s point of view, conditional indexation ensured that solvency requirements became easier to comply with,” he says. “Solvency can be measured in nominal, instead of real, terms, which - depending on the maturity of the fund - can easily be 30-40% higher. On the other hand, the pension funds’ stakeholders, including the employers, still have the ambition to pay full indexation.”
Kocken says the alternative would have been to close the schemes by moving the accrued benefits to an insurance company.
“However, this mostly involves a downgrade towards a nominal annuity promise, without inflation-linking,” he says. “Future benefit accruals would have to be managed in a DC framework, with all the costs and behavioural finance disadvantages and no access to inflation-linked annuities in the European market. In Holland, only a very few pension funds followed this route.”
A step further in shifting the risk towards employees is the collective DC approach, with employers paying only fixed contributions.
“This approach also works well in Dutch practice and has positive welfare effects, compared with a shift towards DC, though it is only applied by a handful of pension funds,” says Kocken. “However, many pension funds may eventually follow this direction, as ageing will leverage pension liabilities more and more compared with company wage costs, putting more pressure on the corporate balance sheet.”
According to John Smolenaers, senior pension consultant with Watson Wyatt in Eindhoven, the Dutch system uses conditional indexation, as 80% of Dutch pension schemes are based on career average salaries.
Smolenaers says: “Indexation is more a communication issue than an actuarial issue, because the legislation says funds have to inform scheme participants as to what they can expect from target indexation in future. It is important to clarify for participants how indexation is to be funded - whether by higher premiums or extra returns on investment.”
He continues: “In the Netherlands, the threat of a move to DC is coming more from IFRS rather than legislation on pensions. And there is still a threat - new IFRS standards are being developed, especially for industry funds or multi-employer funds. Also, the development of the Solvency II guidelines by the EC may be unaffordable for pension schemes.”
But would risk sharing work in the UK? The proposals set out three options:
Conditional indexation for career average schemes would allow schemes to halt indexation and salary increases in rocky times but would oblige schemes to start redressing this when scheme funding has improved. Conditional indexation for all DB schemes, including final salary, would work on the same basis, but with no provision for changing the normal retirement age. This would give a lower retirement income than final salary schemes, but more certainty than DC schemes. The third option is the collective DC scheme, a new structure where employers pay a fixed contribution rate and the members share all risks.
Peter Routledge, (pictured right) head of pensions at Towers Perrin, gives a guarded “yes” when asked if this is a good idea. “Conditional indexation is a useful safety valve for employers who are still committed to DB schemes but want to find a way of managing and sharing some of the risk with members.”
He says: “It only relates to a small band of employers - over 80% of DB schemes are now closed to new members. It will slow down the trend of DB closures a little, but not reverse it.”
Routledge says that risk sharing might not be as successful in the UK as in the Netherlands because of fundamental differences between the two systems.
“The Dutch have always been pretty conservatively invested, predominantly in bonds, whereas UK pensions funds are generally much more heavily invested in equities, and are less well funded,” he says. “So when the performance does not come through, there is a much bigger impact. Also, UK plans are generally bigger relative to the business, and more generous, so there is more risk to begin with.”
Furthermore, he says: “The Netherlands could be seen as a bit of timebomb. Risk sharing has not really had time to bed in yet. The test will come when the employer tries to reduce benefits - they could end up with employee unrest and industrial action.”
As for the three options, Routledge says that conditional indexation for career average schemes would be helpful for a company, because UK pension schemes have high guarantees.
He warns: “Whether companies will be able to hold back benefits in practice remains to be seen. The situation is similar to with-profits policy payouts, where insurers have found it hard to enforce the risk sharing they legally had. When they did withhold payouts they lost popularity.”
Routledge believes this problem could also apply to collective DC schemes, even though all risk is borne by the members. “It all rests on the actual benefits paid to members when the time comes. If returns haven’t been good, and a target benefit has not been reached, then there could be a court case.”
As for implementation, Routledge says that communication and industrial relations are the two key issues.
“The proposed change to pensions is something that can make the unions relevant to their members,” says Routledge. “But I suspect the unions would prefer these changes to a wholesale switch to DC. However, they will want more input on how schemes are run, when indexation is triggered off and on, and how much every member puts in, making running pension plans even more difficult.”
Daniel Vassiliades, principal at Punter Southall, says: “Everyone is searching for a half-way house, where employers have some control over costs, and employees have some control as to what they get out. Under these proposals, the efficiency of pension provision would be maintained, as companies can ride out any downside, rather than having to pass stringent tests all the time.”
Vassiliades continues: “The major problem in the UK is the lack of flexibility and the burden of measurement. Pension schemes are measured every year both for the sponsor company’s accounts and for funding purposes. Because there is an annual measurement, if the scheme goes backwards on a year-to-year basis, someone gets the blame. This has led to a conservative investment approach, so the scheme can be expected to cost more than if it had been invested in a return-seeking way.”
Vassiliades stresses: “In practice, someone has to define the tests which you have to pass in order to pay the pension income or cut it back. That is adding a new layer of bureaucracy to a creaking UK system.”
He states that removing the capped RPI on pension increases would constitute a reduction in members’ pension benefits. “What pension funds could do is use a discount to pay a lower guaranteed increase, such as 1-2%, instead of inflation,” he says. “The jury’s still out as to whether it’s working in the Netherlands, as it’s not been going long.”
Vassiliades adds: “It has come too late to stop the DB closures. All we can hope for is that by allowing the risk-sharing approach to happen, these sorts of schemes are encouraged. And it might lead on to other things such as deregulation.”
Kocken is slightly pessimistic about slowing down DB closures. “Although it would be beneficial to employers, employees and retirees alike, the chances of success seem to be lower in the UK,” he says. “But the biggest impact will come from applying conditional indexation to existing liabilities, as has happened in the Netherlands. And relationships between employers and unions may differ substantially between the UK and the Netherlands, reducing the probability of success in the UK.”