It is interesting to compare the Dutch corporate pension industry with its US equivalent, as they have chosen different routes to cope with the increasing burden of ageing: whereas the Americans have individualised their pension system to a large extent, collectivity was preserved in the Netherlands.

In evaluating these different routes, I would emphasise four key messages. First, I would argue that a clear separation between a sponsor and its pension fund is a desirable feature of any pension system. Second, I would point at the added value of collective pension plans over individual ones. Third, I would argue that this collective nature saddles pension fund asset managers with an important social mandate. And fourth, I would also point at an opportunity in the pension delivery market, which I challenge asset managers to seize.

The US pension system is characterised by a relatively small government-imposed pay-as-you-go system. This benefit constitutes only 45% of total retirement income in the US.

Here, the Netherlands and the US are comparable. Indeed, the Dutch state pension scheme is also fairly small, as it provides the average employee with about 50% of his total pension benefit. At the same time, about 55% of total US retirement benefits is provided through a funded system.

In the Netherlands, too, funded pensions are fairly popular, constituting about 50% of total retirement benefits. Notwithstanding these similarities, there are also differences. In particular, pension benefit replacement rates differ widely between the two countries: whereas the US has an average pension replacement rate of 51%, this rate amounts close to 84% in the Netherlands.

Also, US pension funds have shifted much more risk towards their participants while the Dutch pension system provides its participants with a relatively high degree of certainty on the expected pension benefits.

Within the class of funded US pension benefits, we need to distinguish between two cases. First, there are the traditional company-linked defined benefit pension schemes, which are losing popularity. The decline of these defined benefit pension schemes has been reflected in a rise of the so-called 401(k) plans.

Under such plans, the employee chooses to transfer a certain portion of his wage to a 401(k) account, from which he can draw after retirement. These 401(k) plans are typically of a defined contribution nature and at times depend heavily on the investment capabilities of the individual concerned.

Traditional US company pension plans clearly resemble Dutch company pension plans. Both are collective in nature, both invest pension assets on their participants' behalf, and both provide a defined pension benefit.

However, the legal setting in which these company pension funds have to operate, varies between the two countries.

In the US, company pension funds are rarely financially separate from the sponsor company, so that pension assets are ill-protected against employer bankruptcy.

The Dutch system, on the other hand, is characterised by legal separation between the sponsor and the pension fund, so that bankruptcy of the sponsor does not necessarily have fatal consequences for the fund.

To me, this legal separation between sponsor and pension fund has two crucial advantages. First, it excludes insidious conflicts of interest. After all, if the sponsor and the pension fund are not separated, it is not entirely clear for whose benefit a pension fund is run. For shareholders or participants?

In this respect, experience in the US has shown that in some cases pension fund surpluses were channelled back to the sponsor, while deficits were borne by pension fund participants via premium increases or entitlement cuts.

This criticism admittedly also applies to the behaviour of some Dutch pension funds in the 1990s. But the new Financial Assessment Framework (FTK), which became active as of 1 January this year has made the situation more balanced by explicitly spelling out the modalities under which contribution holidays can take place.

Second, a crucial advantage of the legal separation between a sponsor and its pension fund is that not all the employees' eggs are put in a single basket.

In the absence of such a separation, the system creates concentration risk for employees as it makes both their human and financial capital dependent upon one and the same firm.

Consequently, if a firm were so unfortunate as to go bankrupt, its employees would lose not just their jobs, but their pension entitlements as well.

This concern is increasingly problematic in modern dynamic economies where company lifetimes are decreasing. With the introduction of new accounting rules, which require companies explicitly to report their pension liabilities on their balance sheets, many companies discovered that they had actually been transformed into asset managers and life insurers, at great distance from their core business.

For example, with pension liabilities amounting up to $11bn (€7.9bn), General Motor's pension obligations roughly equalled its market capitalisation, which amounted to $12bn at the end of 2005. In fact, some US companies (including some airline companies) had to renege on their pension promises by entering into Chapter 11 insolvency protection. Via this route, they were able to transfer their pension obligations to the pension benefit guarantee corporation and to continue their business, with lower implied pension benefits for retirees.

These considerations lead to the first important message I would like to emphasise, namely that the separation between a sponsor and its pension fund should be a key feature of any pension system, to be fervently preserved and promoted.

By contrast, in the US these problems are mitigated in another way. There, companies introduced individual 401(k) plans, which do not appear on their balance sheets as a liability. This switch has the advantage that employees' pension assets are no longer exposed to employer bankruptcy; on the downside, however, benefits are now of a defined contribution nature. The risks are then placed squarely on participants' shoulders. In this respect, a further development of risk-pricing would help gain deeper insight into risk differences between pension systems.

A second key issue in the design of corporate pension plans is how to decide on investment policy and who should bear the residual risks. Is a shift to defined contribution schemes the optimal way to reduce pension risks for employers? I doubt it. In 401(k) plans, the individual decides what part of his wage he wishes to transfer to his 401(k) account. Moreover, notwithstanding recent initiatives that offer participants responsible investment-packages as a default, the majority of the plans still give contributors a substantial say in the investment policy.

This all means that many 401(k) plans rely heavily on individual responsibility as well as financial literacy. However, not everyone is able to understand the complex world of pension finance, as the average employee is not as well educated in finance as pension fund asset managers probably are.

Let me clarify this by asking two basic questions, which were also put to a large group of US workers: Suppose you have €100 in a savings account and the interest rate is 2% per year. After five years, how much do you think that you would have in the account if you left the money to grow- more than, less than, or exactly €102?

Now imagine that the interest rate on your savings account is 1% per year and inflation 2% per year. Would you then be able to buy more than, exactly the same as, or less than today with the money in the account?

Research has shown that over 50% of US pension plan participants could not answer these two simple questions correctly. And, as pension expert David Blake has noted more than once, that 50 % of the individuals do not even know what 50% is!

Moreover, the man in the street is not an expert on risk diversification: in for example the well-known Enron case, employees had invested 60% of their 401(k) assets in Enron stock. On average about one-fifth of all 401(k) assets are invested in their own company stock. Although this is understandable since an employee is always relatively confident of his own firm's prospect, this again leads to the undesirable situation where human and financial capital are tied to the wellbeing of one and the same firm. Additionally, under 401(k) plans, it is up to participants to decide what amount to reserve for pension purposes. Here, the problem is that the typical 401(k) participant seems to have commitment problems: estimates indicate that the typical 401(k) participant approaching retirement has saved less than $50,000 (€36,287), instead of the $300,000 he would have accumulated under a defined benefit system. Procrastination behaviour - based on thoughts such as "I will start saving tomorrow" - as well as limited rationality seem to abound.

Since individuals only seem to be rational up to a point, there is considerable value in preserving the collective nature of our system. In such a system, participants automatically contribute enough towards their old age financing while professional asset managers make sure that the contributions are well invested from a risk-return perspective. This ensures that every individual is provided with a reasonable pension benefit after retirement. Moreover, collective pension systems are cost efficient, as they exploit economies of scale, and enable intergenerational risk-sharing which is welfare-enhancing to risk-averse individuals.

The collective nature of our system brings important fiduciary responsibilities for pension fund asset managers. Compulsory participation implies that participants are not free to choose which pension fund to join, unless they move to another company or industry. In this way, asset managers' power to raise contributions is very similar to the government's power of taxation, from which one normally also cannot escape. Consequently, pension fund asset managers bear an important social responsibility - even more so than ‘ordinary' companies.

Therefore, pension fund management requires good governance as well as transparency to allow pension plan participants to see what is being done with their money on their behalf. In addition, the compulsory nature of participation implies that you should not only look at private returns, but also try to take social returns into account, as society as a whole has given pension fund asset managers the mandate to raise their funds. In this light, I applaud the recent initiatives that discourage socially irresponsible investments.

In most western countries, pension systems have already been individualised to some degree as part of the shift from defined benefit towards defined contribution systems.

To a lesser degree, this has also occurred in the Netherlands, with the introduction of conditional indexation and life course arrangements.

In this context, I foresee market opportunities for new pension products that are tailored to individuals, but maintain the benefits of collectivity. This involves both assisting individuals with complex choices and making risk pooling instruments more readily available.

This is an edited version of an address by Nout Wellink gave at Hoofdkantoor Shell at the Hague in June.