The general framework for occupational pension plans can be described as hybrid pension plans, with intra and inter- generational risk-sharing mechanisms between members and with sponsors. In this study, we define the degree of hybridity of a pension plan as the extent of risk-sharing with the sponsor. At one end of the spectrum lie traditional DB funds, where all the risk is for the sponsor as long as it can bear it and the pension benefit is independent of plan returns. In a DB scheme, the sponsor gives a guarantee to the fund, in exchange for the possibility that its initial cash contribution is reduced, so, the pension fund has an unfunded guarantee value. As the degree of hybridity of funds is linked to the degree of risk-sharing with the sponsor, collective defined contribution (CDC) funds, which are hybrid pension plans with collective risk-sharing and conditional indexation but without a sponsor, and DC funds, where the risk is entirely borne by individuals (and which we also call individual DC funds to distinguish them clearly from collective DC), both lie at the other end of the spectrum. However, because the market value of the pension rights in individual DC funds is always equal to the market value of the investment fund where all pension assets are invested, and because risk is individualised, they stand apart because they are not collective solutions.
Within the countries of interest, individual DC plans are the prerogative of the Commonwealth countries, and they have progressively replaced traditional DB funds in the UK and in the US. Continental European countries, and in particular the most innovative (the Netherlands) have opted for more risk-sharing in the form of hybrid funds – defined as collective pension plans that benefit from some (but varying) risk-sharing between participants and with the sponsor, and usually offer guarantees and conditional indexation.
Developed countries have on aggregate led the retirement system to rely less on unsecured sponsor guarantees. In fact, demographics imply that the unfunded part of the pension diminishes as a proportion of labour revenues – an analysis that also applies to pay as you go social security systems (A present-value framework would allow greater visibility on the future sacrifices that populations must make in order to balance public finances in the long run).
In the UK and the US, the rigidity of laws governing pension arrangements – conditional indexation is prohibited in the UK – has meant that the reduced aggregate reliance on sponsor guarantees has been obtained by a mix of DB and individual DC funds rather than by more hybrid funds.
In hybrid systems where participation is often mandatory (94% of the Dutch workers are covered by hybrid plans and only 3% by DC plans), the system can only adapt by greater hybridity, which means lower sponsor guarantees and lower overall (nominal) guarantees, as is currently being envisaged in the Netherlands.
The UK and the US mainly rely on individual DC funds and traditional DBs. Individual DC funds, as they are today, are usually deeply sub-optimal, as they are plagued by poor governance, sub-optimal default options, they are significantly more expensive than DB funds and, in the US, they do not offer annuitised income.
On the other hand, traditional DBs are often plagued by unhedged exposure to the risk that the sponsor goes bankrupt, and arguably the failure to transfer systematic longevity risk has been (and still is for remaining DBs) a very important risk over the long run and has contributed to exhausting the sponsor guarantee. On the whole, given current practices, the sustainability of traditional DB funds is not guaranteed.
In continental European countries, hybrid pension plans are professionally managed and cost-efficient vehicles. Participation in pension plans is usually mandatory and the large majority of the population is covered by hybrid pension plans. Yet, as they are regulated they must provide the same type of solution for all categories of investors. So an assessment of the type of guarantee they offer is necessary. Should hybrid pension plans offer nominal guarantees (as is usually done on invested premiums) and if so, to which participants should they offer these guarantees? Nominal guarantees may be extremely costly as in practice, and given the current yearly investment horizon, such guarantees may result in pension funds being locked in low interest rate yielding investments. This is particularly damaging in situations such as that currently being experienced, where quantitative easing may involve accelerating inflation in future years.
In the shift to individual DC funds in the UK and the US, individual DC funds run the risk of being acknowledged as inefficient vehicles, benefitting from no risk-sharing at all, being extremely underdiversified, lacking default (annuitisation in the US) and being dampened by high expansive.
So, it is urgent to rethink the DC framework and envisage a DC 2.0 that would try to compare or compete with collective solutions. There should be more involvement of states in the design of adequate solutions – or, more simply, in setting appropriate governance structures and organisations (i.e., appropriately defining roles and responsibilities to ensure an efficient delivery). The regulation of individual DC funds should be profoundly modified, and it should definitely make a departure from the inadequate retail fund regulation it is inspired by. Individual DC funds should be able to diversify their exposures and invest in illiquid securities and, in a nutshell, have the same flexibility that DB funds have today because ultimately, they must also provide retirement income.
Of course, there is no need to wait for regulatory change to improve practices, which should be ideally driven by a genuine aim to service investors’ needs rather than to blindly adhere to regulations, and to usual market practices. Today, DC funds dispose of significant means enabling them to avoid the pitfalls of short-term retail funds, as they can diversify their assets using international investments, corporate bonds, listed real estate, commodities, and even funds of hedge funds. In addition, today’s technology makes it possible to offer some guarantees in DC funds. Inflation guarantees or target inflation indexation are important for those who rely primarily on the income from DC pension funds, as is often the case now in the UK and the US. We recommend a pragmatic approach where the stochastic life-cycle (where market opportunities influence asset allocation) is a risk management facility made available on top of suitably designed building blocks: the PSP would be heavily diversified amongst asset classes; the LHP would have a significant anchor to inflation; the preference of investors towards guarantees (nominal guarantees, real guarantees or rather target indexation) should be the main choices that members should make, even if customisation is possible via supplementary funds. Transparency must also be increased. Nominal guarantees can foster confidence in DC arrangements, especially over the short run. However, the cost of financial guarantees must be made clear, for instance by showing by how much the participation to the upside is reduced (and by how much the downside is protected).
Traditional defined benefit funds, although on the decline, still have a place in the provision of retirement. To ensure that the remaining DB funds remain a sustainable option, these must be insured against the risk of default of their sponsor, and mortality should be hedged. The risk of greater longevity could be partly shared with employees (who in theory would need to work longer).
When it comes to hybrid plans, nominal guarantees on premiums are usually required in continental Europe. Yet for the younger participants, these guarantees can be reduced – as long-term investors they have little need for nominal guarantees if these comes at the expense of real (inflation) indexation. This reduction in nominal guarantees gives way to more flexible and performing funds and, on the whole, produces more sustainable hybrid plans. Nominal guarantees can be important for the older participants because, over the short-term, they can be considered a proxy for real ones. This also means that, ideally, the system should be flexible enough to accommodate different indexation policies, in line with the life- cycle approach where the risk and reward changes with age (younger participants are expected to have a leveraged exposure to the financial markets as they borrow against their human capital or lifetime income, and retirees that only have financial capital need more security).
However, the effort should not bear on regulators alone – practices need to evolve too, as there is empirical evidence that pension funds behave sub-optimally. Even though pension funds have been praised for their ability to diversify, empirical evidence suggests that they do not fully use their ability to diversify across asset types such as illiquid assets. They also often have a strong home bias, which is not only contrary to standard academic prescriptions that recommend diversifying assets geographically, but that also runs the risk of insufficiently protecting the purchasing power of retirees in maturing economies. The demographics should also be taken into account in investment policies, as it is an important driver for growth. It can be expected that at some point in the future – as has been the case in China and India over the last decade – the production of future retirement goods will be shifted towards countries with a growing workforce (thus emerging countries in terms of demographics) and adequate internal organisation and resources – infrastructure, education and democracy are often associated with economic growth. Lastly, the short-term behaviour shared by many pension funds ends up locking them in to minimum funding ratios and nominal strategies, a risk that can be avoided by an adequate definition of the investment strategy.
On the whole, even if regulations can be largely improved to ensure a sustainable framework for pension funds and give adequate long-term incentives, there is still much room for improvement in current practices, relative to academic prescriptions. Beyond standard portfolio construction theory, it is far from obvious that pension funds have adapted their investment policies to account for demographic changes.
Beyond the analysis of existing pension funds and products, the industry landscape is also evolving. An important objective of the IORP directive (EC 2003) is to allow an efficient provision of pension solutions at the European level, which goes along with the possibilities of creating trans-border pension plans, pooling pension assets and having multi-employer or pooled pension plans. Technically, such pooling is facilitated by DC funds because of the immediacy of the link between pension rights and asset value. Such pooling is also customary for life insurance products, but these are not natural for (very long-term) retirement savings.
The first country which these measures will impact is the UK, because the Pension Act 2011 is improving coverage with the auto-enrolment procedure. Furthermore, the default option is the NEST, which is raising the standard of DC funds, as it provides economies of scale, diversification and risk management options. Market players consequently need to reposition themselves so as to offer solutions that compete with that offered by the NEST.
Of course, we expect profound modifications in the European landscape of pension providers over the coming years. As UK firms that do not provide pension plans to their employees will need to do so, a market that provides pension plans to these firms is likely to emerge, and pooled, cost-efficient and professional solutions are likely to reap strong benefits. The need for efficient retirement solutions should also create a market for efficient retirement funds offered to existing DC pension plans. These offers require additional capacity on top of that of traditional asset managers who essentially provide investment funds to DB plans, and of that of fiduciary managers who essentially service existing DB plans, but do not deal directly with corporations. The consolidation of the provision of pension plans and associated risk and asset management should go along with a consolidation of traditional asset management companies.