George Coats examines whether it is time to put politics aside and create one buffer fund that can provide for all generations and save money in the process

The downward trend of pension fund returns last year meant even those funds that came even close to zero return could see their asset allocation scrutinised by colleagues.

The 2008 results by Sweden's AP buffer funds have rekindled questions about their future. An independent committee affiliated to the finance ministry, the Expert Group on Studies of Public Economics (ESO), is expected to produce an in-depth report in September on their future, including how many of them there should be and how they should be managed.

The funds were launched as part of the 1998 pension reform, with a remit to achieve the highest possible return on their assets and to assist in smoothing out payments by the state income pension. But in 2008 the funds - AP1, AP2, AP3, AP4 and AP6 - reported negative returns.

AP1 posted a return of -21.9%, resulting in an investment loss of SEK48bn (€4.4bn) on the year; AP2's result was -24%, down SEK54.2bn; AP3 posted -19.8%, down SEK44.8bn; AP4 reported -21.0%, down SEK42.6bn; and AP6 recorded -16.6%, down SEK3.25bn (see table).

However, there are questions over the result of AP6, which invests in private equity. "How did it perform? That depends on which valuation method they use," says an observer. "AP1 to AP4 had some of the same private equity holdings and they have put totally different valuations on them. There is a particularly big difference between AP1 and AP3, where AP1 writes the loss off immediately, while AP3 employs some kind of smoothing."

Pension payments are based on a calculation of liabilities, including all future pensions to be paid out to retired and non-retired participants in the system, and assets, including all the money in the system multiplied by the average time a krona spends in it generating interest.

The AP funds are on the asset side of the balance, and their collective SEK190bn deficit last year is one factor contributing to the system's deficit. The other is that the average time that a krona spends in the system, currently 43 years, has dropped. Both factors mean that the asset side is shrinking. As a result, instead of balancing, as in a simple division of assets by liabilities (100/100 = 1), the result is around 0.96. The pension index is linked to average earnings growth, which is multiplied by pension system factor. So a total below one means that the full indexation is not passed on.

"As long as the pensions were higher every year, and the index has gone up eight or nine years in a row, things were fine," says Mikael Nyman, editor-in-chief of Stockholm-based newsletter Pensionsnyheterna. "In fact pensioners have been grossly over-compensated in comparison with inflation because the growth of the economy has exceeded the price index. But as things stand the average pension is going to be cut to SEK11,500 from SEK12,000 next year."

However, with a general election due next year nobody is anxious to see a cut.

At a recent seminar organised by Nyman, social security minister Cristina Husmark Pehrsson indicated that there would be a change from a one-year indexation link to a three-year average valuation to smooth out the balancing mechanism. "That way they might get through 2010 without anybody noticing," says Nyman. "But also, under such a method pensioners wouldn't get the topside because they would have a slower compensation when it starts growing again. And if I know the pensioners' organisation they'll want the money straight away."

On the margins of the same meeting Pehrsson's state secretary Bettina Kashef said there could be compensation for pensioners who lose money through a tax reduction or a rebate.

"Everybody knew that the AP funds would be emptied by the baby boomers," says Nyman. "But with the losses it is going to be quicker than we thought. This, in turn, has raised basic questions about how the AP funds are run. People are asking why there should be four of them, why they should have active AP fund allocation when in due time the outcome is like an indexed portfolio, and why they should pay enormous amounts for management if they do not achieve positive active returns, as they didn't last year. That's the debate now."

Consultants had recommended that there should be two funds; for fixed income and for equity investments. The decision to have multiple funds comes of political consensus that emerged from a committee of parliamentary parties and chaired by a forceful former social affairs minister Bo Könberg. "Everybody accepted that a buffer fund was needed to compensate for the different sizes of the generations when building a system like ours but the right-of-centre parties, especially the Liberals and the Moderates, were suspicious about the implications of a large state-owned investment vehicle, fearing re-nationalisation by the back door through investments on the Swedish stockmarket, and they insisted on a number of buffer funds," recalls Nyman.

But this fear turned out to be unwarranted. "What Könberg, a Liberal, did not know then was that those who worked in the AP funds had a financial education that said you had to have a global portfolio," says Nyman.

There were investment duplications. "AP1-4 were given identical investment rules, allowing them to invest in several different asset classes and requiring them to operate independently, on mutually competitive terms, and to formulate their own investment policies, corporate governance policies and risk management plans," notes an observer. "But the rules gave no incentive for the funds to deviate from each other.

"And they are doing active management," he adds. "If you look at the numbers, it is clear that the driver behind the active under-performance is the AP funds' internal management. Ironically, they were forced to undertake internal management because an earlier study said that their outsourcing fees were too expensive. But obviously if you have 180bps of under-performance like AP2, it is the most expensive thing you can do. So active managers have not delivered - but neither has internal management."

Former AP1 board member Harry Flam, professor of economics at Stockholm University, has long proposed just one AP fund. "I wrote this in an article almost three years ago proposing the funds be merged on cost grounds," he says. "The relative performance of the AP funds has not been very good, especially last year when the relative performance was somewhat of a catastrophe."

Flam has suggested the merger of the funds would save SEK1bn a year and a switch to pure passive management would save, in total, SEK1.6bn.