A sound, solvent and solidarity-based pension system

Date 30 September 2004
Time 14:40 pm
Location Event Centre, Aalsmeer
Speaker Mr. Dirk Witteveen, Executive Director of De Nederlandsche Bank NV

Speech by Mr. Dirk Witteveen, Executive Director of De Nederlandsche Bank NV for the 'Zwitserleven Pensionday 2004', in the Event Centre, Aalsmeer.

A sound, solvent and solidarity-based pension system

Those who know me, know me in various roles. Today I will be taking you from a rigorous programme to the conclusion. I have been asked to speak about pensions. It is worth noting that today is 30 September, two years to the day since the PVK sent the pension funds a letter which caused quite a stir.

Sound pensions: realizable expectations?

The primary task of a sound pension system is to guarantee its commitments. Expectations are raised which must be met; this means they must be realistic and realizable. Pensions have come under scrutiny. This has technical, financial, economic and legal aspects.

  • Technically: how can we maintain a sound and sustainable pension system based on solid social and actuarial foundations?
  • Financially and economically: what will it cost, and how will it affect the economy in general, and labour costs, labour mobility, purchasing power and savings behaviour in particular?
  • Legally: how do we formulate a transparent and statutorily unambiguous pension system so that it is comprehensible for all stakeholders, from those heading the pension funds to the participants, i.e. nearly the entire adult population? 

You are all aware that a new pensions act is currently being prepared, and that the underlying financial assessment framework is in the process of being rounded off. At the eleventh hour, there are also new Government plans to reckon with. These will undoubtedly put extra pressure on the pension funds’ capacity for implementation and communication. In addition, we all know – the supervisory authorities in particular – that the pension funds are not operating on the basis of a tranquil and balanced financial position.

At the press conference held on the occasion of the presentation of the PVK’s Annual Report for 2003, I drew attention to various concerns. It was observed that the future indexation of pensions has not been fully funded by the pension funds. This is not required under law; indexation may be financed on a pay-as-you-go basis or from extra return. My worry is that this is not always communicated realistically to the participants. I would be in favour of a structurally higher coverage ratio, which would also give more scope for containing fluctuations in pension premiums. It must be clear to everyone that sound pensions come at a price. I am saying this not just as a professional doomsayer, but also as a central banker concerned with stability, in the realization that stability and prosperity are underlain by confidence.

The warnings issued six months ago came with several rough calculations:

  • retiring at 62 is 40-100% more expensive than retiring at 65;
  • an indexation of 2% per annum costs 20-60% more premium than a basic (nominal) pension;
  • the financial possibilities for absorbing further adversity remain limited and
  • by 2030, every per cent indexation costs 3.6% of the wage and salary bill (1990: 2.6%) (assuming full average pay). 

The representatives of the pension funds’ umbrella organisations had their reaction ready. They called the PVK’s pronouncements out of proportion, pointing out that the pension funds had weathered the storm and had landed in calmer waters. Their arguments boiled down to this: that the costs of a possible pre-pension scheme were already included in the premium, and that indexation did not necessitate extra premium increases.

All I can say is: I wasn’t fully convinced then and I still am not now; I simply hope that our message has got through. One could of course see the reassuring and optimistic mind-set of the pension funds as a positive sign. In any case, the main conclusion ought to be that what we, as a nation, consider an appropriate pension, befitting the prosperity level of our society, is no longer as self-evident as it used to be.

Sound pensions: a matter of certainty

The soundness of a pension system is basically a matter of certainty. Participants must be able to count on commitments being realised. Speaking in technical terms: what is the maximum acceptable chance that a pension fund is under-funded and needs to take restoration measures? After all, if a shortfall is not supplemented or remedied, pension commitments will be in jeopardy. Politically, that certainty was fixed at 97.5%. In other words, a pension fund has a chance of 1 in 40 of having to take restoration measures within one year’s time. It is then given a long period (of up to 15 years) to restore its funding situation up to the statutory requirements. This means that, effectively, the chance of a pension fund facing a reserve or coverage shortfall is much larger, because it is more vulnerable during a period of restoration. It is only when it has returned to its mandatory level, that the pension fund again runs the original risk. This is a fine principle, so long as we abide by it. Let me assure you that we are addressing this issue in the consultations on the formulation of the desired measure of certainty. We will subsequently apply this stipulation flexibly but with determination.

Sound pensions: the price of certainty

Opting for less certainty does not mean that pensions will, on average, become cheaper; after all, pension premiums represent the cash value of future commitments. A low solvency standard does mean, however, that pension under-funding will occur more often and that it will be more serious, thus making for a higher frequency of premium increases and other measures. A low degree of certainty will thus unintentionally contribute to fluctuating premiums. Such volatility is bad for macro-economic stability. It will furthermore benefit some generations and prove detrimental for others. There are more reasons for wishing to prevent effects of this sort. As the population ages, support for measures to make up for premium shortfalls, and for pay-as-you-go systems will wane. The dependency rate, i.e. the ratio of those aged 65 and over to the cohort of those aged between 15 and 65, is expected to double over the next three decades. As a result, the above measures will put growing pressure on the wage and salary bill. It seems to me that it is for the pension funds to make wise decisions on this point. Here soundness and solidarity are at loggerheads. I will say more about this later. The ageing of the population is illustrated by the development of the dependency rate. The chart shows that this ratio will rise from 21% to 44% over the next 35 years. As the Government rightly points out, labour participation is of crucial importance: it is essential that the share of the potential labour force, i.e. the entire cohort of those aged 15 to 65, actually take part in the labour process. The chart also shows that we aren’t the only country facing this problem. Thanks to the high level of capital funding of our pension system, we do, however, have a relatively sound starting position.

Sound pensions: how have they been funded so far?

The cost of pensions has been insufficiently addressed in the past. How come? Over the past decades, the Dutch pension system was continually being restructured, from fixed amounts to average wages to final pay, and more recently: from greater flexibility and early retirement back to a longer working life. Before, the standard pension came out at 70% of final pay. The pension burden was, however, mitigated, via the so-termed franchise, by the development of state old age pensions into full-fledged basic pensions. Then economic conditions deteriorated, and in 1982, the Wassenaar Accord was signed, under which wage moderation was exchanged for shorter working hours.

The moderate wage movements made for a lighter pension burden, and at the end-1980s, the pension funds showed sound balance sheets. This attracted the attention of the Government and in the early 1990s they began to wonder whether the hidden tax revenues in pension fund assets could not be collected earlier.

The social partners and the pension funds’ management were perhaps overly eager to act on this hint. Another mistake was the undue focus on the premiums paid annually, rather than actual cost, which had been rising. Pension assets were consequently no longer hoarded, and ample opportunity was given for lower premiums, premium holidays and even reversed transfers. Moreover, some of these resources were used to build up very generous supplementary pension schemes. And the attending costs were not fully passed on in labour costs. In brief: is it fair to say that we have recently been faced with material increases of the pension burden or should one say that the earlier alleviation of the burden has come to an end?

Coverage from an historical perspective

A prudential supervisor looks at solvency rather than at premium-setting. Yet the main directives for supervision on pension funds also contain stipulations on premium-setting. More about this later. As explained earlier, the PVK’s main task is to ensure that pension commitments can be honoured. There is nothing wrong with highlighting the costs of such commitments. Notably since September 2002, we have therefore been drawing extra attention to our solvency requirements. This has yielded results which I hope are appreciated. Let me give you a brief overview.

But before I do, let me point out that there is no such thing as the “right” coverage ratio of assets to liabilities. So far, in practice, the 4% standard was always used: the ratio of assets, marked to market, to liabilities, consisting of nominal pension commitments, converted into present value at an actuarial rate of interest of 4%. The difference in valuation bases is now about to change. The official 4% figures from the past show that developments were fairly calm between 1986 and 1995: coverage was around 110%, in a period when share investments were relatively minor. This average ratio then rose to over 150% in 1999, only to drop to just below 110% again within three years’ time. These percentages are averages, and differ significantly from one fund to the next. It was notably the enlarged proportion of share investments which took its toll, but on closer inspection this was not the only factor of importance.

Coverage ratio marked to market retroactively

Suppose we had begun to consistently apply the market value approach earlier. Under this approach, the market value of investments is set off against nominal liabilities, which are also converted into cash at a capital market rate. The difference between the two coverage ratios speaks for itself: apart from a provision for general risks and buffers, it also contains the implicit scope for indexation. The PVK has calculated that this scope did not begin to decline materially in 2000, but in fact already in 1990. The funds’ assets no longer suffice to pay inflation-proof pension benefits. In other words, the decline in certainty regarding the conditional indexation of pensions has been going on for a long time already, but has been insufficiently recognised.

2003: restoration set in motion

Fortunately, the year 2003 saw a recovery of the share markets, following the booming 1990s and the creeping stock exchange crash at the start of the new millennium. The upturn is, however, not as rapid as the earlier fall, and this year has seen a discernible flattening of the positive trend. With hindsight, one can say that in the past investment results were not sufficiently subjected to anti-cyclical treatment. In other words, in fat years, one should save for the lean years which will surely follow. Almost inevitably, the price to be paid is that the fragile economic recovery is not stimulated by pension developments. Pension funding needs to be restored, but the economy is in fact unable to bear the attending extra burden. To save costs, pension plans are therefore now being cut back. One way of doing so is to raise the retirement age. Whatever measures are taken, pension funding needs to be restructured fast to boost the impaired confidence. This is also essential to support the recovery of the economy. In short: the current situation remains painful for the time being, even though the stock markets are recovering and capital market rates ceased to decline in 2003.

Improvement of average coverage ratio

In 2003, the pension funds’ average coverage ratio (using an actuarial rate of interest of 4%) improved by several percentage points. Another positive development is that not one of the 25 largest pension funds faces under-funding at this point in time, and that half of them no longer have a reserve shortfall. Of the funds confronted with under-funding at end-2002, three quarters no longer did so a year later. The restoration plans submitted show a further improvement for 2004. I would like to point out that the policy pursued by many pension funds is to remain outside the danger zone. And they have: even during the recent stock market crisis, only a small number of pension funds landed in difficulties, which incidentally could be remedied. On the whole, pension funds are not intent on seeking out the limits of what is considered acceptable. And luckily so, because this would only lead to serious intensification of pension supervision, which no-one really wants.

Cost-effective premiums

Risks can be covered through cost-effective premium-setting. Sharp fluctuations in premiums are, however, undesirable both economically and socially, as well as very difficult to explain to be public. As we have seen, unduly low premiums prove disastrous, but we should not, for the sake of stability, simply pay too much just to be on the safe side. The supervisory authorities can accept smoothing, which means reckoning with prudent long-term averages. However, such a policy should be transparent and yield solvent results. This is, after all, the overriding benchmark for prudential supervisors. Stable premiums are feasible, so long as they are in accordance with the principles of the Government’s main principles on financial supervision. This calls for forward-looking calculations and an analysis of long-term continuity.

Intergenerational solidarity

Looking ahead, we also need to review the essence of our pension system. An element which has been largely neglected is solidarity. Pension systems such as ours are made up of a complex structure of financial transfers and risk-sharing. Basically, everyone saves enough for his or her own pension (capital funding). Transfers are made at the end of the ride, and sometimes intermediately. Quantifying these transfer mechanisms is far from easy. Research commissioned by the Advisory Council on Government Policy has shown that collective pension funding on the basis of intergenerational solidarity is more efficient than individual pension funding. The downward risk for individuals is two to three times smaller thanks to the build-up of financial buffers and the possibility to spread risks among the generations. The possibility to adjust premiums, within certain margins, to the pension fund’s financial status is crucial to this system. This type of solidarity should therefore be treasured.

Solidarity is not self-evident

Although this is an important factor in terms of efficiency, the boundaries of intergenerational solidarity are clearly being reached as a result of demographic ageing. The declining proportion of younger workers cannot be asked to remedy shortfalls from the past, and mandatory participation in pension schemes may not be abused to this end. Future preparedness and hence the possibility to solve funding problems via premium adjustments will therefore decline. It must furthermore be remembered that the young are already asked to show relatively much solidarity with the elderly notably in terms of state old age pensions and health care.

The Dutch pension system should therefore refrain from relying overly on intergenerational solidarity. This means that future risks must be made more manageable. The debate should therefore be diversified and not be limited to extremes of pure forms of defined benefit versus defined contribution. Defined benefits mean commitments, until recently based on final pay. In the case of defined contributions, only the contribution is set, and the participant bears the risk of the actual benefit. No-one will dispute that final pay systems – especially with guaranteed indexation – are an expensive facility, even though we seem to have only recently become aware of the price tag. Moreover it calls for a form of solidarity which is expensive if this system is maintained, but not strictly necessary. I am referring to the implicit solidarity with people who climb the career ladder late in their working life. This entails extra back-servicing, an additional burden which is borne by those who do not get promoted and, surprisingly, by those who make rapid promotion at the start of their career.

Defined benefits based on average pay as a manageable solidarity-based alternative
This year we consequently saw a major shift from final pay to average pay systems, especially in terms of the numbers of participants. In 2004, only about 12-13% of active participants still have final pay systems, against more than 50% in 2003. Obviously this development is largely accounted for by the vast numbers of participants of the large public servants’ pension funds. Average-pay systems allow of a more reasonable distribution of policy options, such as lower indexation. This is all the more the case now that indexation is becoming increasingly conditional. The burden is then shared by active participants, dormant participants and retirees alike. In an average-pay system, a certain measure of solidarity is combined with greater policy transparency and more opportunities for containing cost. I commend this system to you, rather than defined contribution schemes. In terms of risk management, defined contribution schemes are obviously highly attractive for pension funds and employers, because intergenerational solidarity is minimized and all risks are borne by the participants. But these schemes may have several unintended effects: employees may feel cheated when they first receive their pension benefits, while anxious participants may make unduly large provisions by saving. Changes will need to be considered carefully, also because – technically – defined benefit schemes cannot be easily converted into defined contribution plans; there are numerous intermediate forms.

More solidarity: early retirement schemes and pre-pensions

Used as we have become to prosperity in the  Netherlands , we have grown attached to our inflation-proof pensions. We have also grown fond of another pension facility, viz. the possibility of retiring before the age of 65. Both index-linking and pre-pensions are under pressure. I have already touched on indexation. It cannot have escaped your notice that the second issue, early retirement as we know it, is being increasingly questioned.

In my capacity as supervisor, I will refrain from voicing an opinion, but I do share some of the concerns put forward by the pension funds on this point. I would like to recall that, where early retirement is concerned, some of the measures called for have already been set in motion. Now that many early retirement schemes have been converted into pre-pension plans, the necessary contributions have been reckoned with in premium-setting. It will not be easy for the pension funds to calculate the individual claims built up and, if desired, to pay them intermediately. However, this does not detract from the need to review how much solidarity is befitting and acceptable in a society characterised by individualism, transparency, freedom of choice, assertiveness and a sense of individual responsibility.

In all this, a major role is played by increased transparency. Where insurance and pension products are concerned, especially in a competitive environment, it is essential to be open about one’s actions and to keep one’s word. Transparency is a sine qua non when it comes to results, costs and certainty. It is crucial to proper decision-making, insofar as there are choices to be made with regard to pensions. Only thus can optimum economic behaviour be attained.

The role of the PVK, soon to be DNB

I will round my speech off in my capacity as supervisor. The PVK, soon to become DNB, aims to exercise effective and efficient supervision. I have highlighted several essential aspects of supervision on pensions. The PVK seeks to safeguard the interests of stakeholders at all times. This is also one of the pillars of the key mission statement formulated by the PVK and DNB in the run-up to their merger: to strive for the stability of the financial system and of the institutions which form part of it.

There is no financial stability without the public’s confidence: confidence in general and confidence in the financial sector in particular. This confidence has been damaged in recent years, owing to, among other things, various accounting scandals. The pension funds, too, have faced difficulties, and still do to some extent. Their decreased solvency has led to considerable premium rises or limitation of indexation. It goes without saying that this hasn’t done confidence much good either. I have indicated that we intend to approach this issue with both greater transparency and in a more solidly founded manner. Proper supervision is not an objective per se; the objective is to contribute to stability and confidence and, at this particular point in time, to restore that confidence.