|
OPDU Report 24
- May 2008
With defined benefit pension provision reducing, particularly in the
private sector, the debate on whether there are viable options to
the main alternative method of defined contribution, which passes
all the risks to the member, is gathering pace. Two decades ago pension regulation would rely on the self-discipline
and social responsibility of the stake-holders of defined benefit
occupational pension schemes. But, pension assets grew so fast in relation
to Gross Domestic Product that the impact on the profits and losses
of the sponsoring companies became disproportionately large. Besides the pension industry was confronted with
unfortunate events like the pilfering of the pension fund’s assets
by Robert Maxwell and a series of pension wind ups, leaving thousands of beneficiaries without pension. In an attempt to turn the tide, the call for improved risk sharing is advocated in order to make the generous DB schemes affordable and risk sustainable. It is pointed out that the strict UK pension regulation stands in the way of possible efficient transfers of risk and reward, leaving too much risk with a single stake-holder. Experiences in Switzerland and the Netherlands reveal that conditional indexation can be considered a commendable and highly efficient approach to resolve the issue. However, this form of risk sharing is not the panacea, as it will typically not be enough to reduce the risk incurred by the plan sponsor and/or future generations. In this article we would like to draw on our experiences with pension reform and reflect on the strengths and weaknesses of the various pension systems. We hope that this article will be helpful in a balanced debate on pension reforms that are much needed, not only in the UK but in other countries as well. The need for reform Allegedly, members of DB pension schemes bear no risk. The pensions are protected against inflation (up to some level) and payments are for life. Despite the perceived desirability of DB, or perhaps because of it, the system has come under great pressure. The most important reasons for the closure of DB plans are the very large pension risks for sponsors, partic-ularly for public companies, in connection with unfavourable trends in the equity markets, increased costs due to ageing and longevity, and a growing administrative burden. The easy solution has been to abandon the DB schemes and change to occupational defined contribution (DC) and personal pensions. This certainly relieves the sponsors from pension risk. But it cannot make risk disappear as it is transferred almost entirely to the individual member. Personal pensions may have worsened the pension problem even further, notably because assets of individual schemes often perform badly relative to institutional portfolios. We believe that DB and personal pensions are both unsustainable due to huge inefficiency and new risk sharing initiatives should be adopted, for instance conditional indexation. Why is Defined Benefit (DB) no longer the perfect solution? Longevity and ageing Longevity and ageing are often put forward as the main reasons for the non-sustainability of DB and the subsequent demand for pension reform. Obviously, it is extremely costly to finance the increase in the retirement period without a corresponding increase in the length of the working life. For example, retirement a year earlier can increase the pension cost of an individual by about ten per cent. Nevertheless, we believe that longevity in itself is not a sufficient reason for the non-sustainability of DB. Longevity risk could be effectively dealt with by applying a constant ratio between working and retirement periods, e.g. one year of retirement for every two years of work. Pension accounting and pension regulation The IAS 19 accounting rules as well as tightened pension regulatory frameworks are mentioned as the key drivers behind the need to abandon DB systems. It would be extremely unfortunate if this were to be true. Without doubt, the new standards and rules leave plenty of room for improvement. For example, according to the International Financial Reporting Standards (IFRS), a pension system that is not completely defined contribution should be classified as defined benefit. Such a strict classification is fundamentally wrong. For instance, in Switzerland and in the Netherlands payments that a company has to make are often limited in a formal agreement between the fund and the plan sponsor. This clearly reveals that these pension systems are not purely defined benefit. However, IFRS does not recognise formal risk-sharing agreements. Classification of the schemes as full DB pension systems creates an unwarranted pressure to change to personal pensions. Funding status Is a deficit in real or even nominal terms a valid argument for question-ing the sustainability of DB? We do not think so. We believe that the deficit is likely to be overstated, because risk premiums that can be expected to be earned in the future are neglected in funding status definitions. This in itself is not wrong at all, but deficits become much less severe if there is time to recover taking into account even very prudent future risk premiums. Therefore, we feel that the current funded status is not enough reason to reform DB either. The most impor-tant issue is risk. Conditional Indexation The compensation for inflation in DB schemes is largely responsible for the cost and the risk incurred by the plan sponsor. The Limited Price Indexation regime in the UK already shifts risk from the sponsor to the members. About ten years ago, we introduced another form of conditional indexation in the Netherlands. In so-called conditional indexed DB schemes accrual of pension rights to make up for inflation is conditional on the funding status of the plan, not on the level of inflation. Typically, retained indexation is reimbursed as soon as the funding status has sufficiently recovered. Consequently, the ability of the plan to recover is significantly improved, because risk premiums are automatically used to improve the funding status of the scheme if necessary. Besides, because equity returns and inflation are negatively correlated in the short-term (but positively in the long-term) conditional indexation enables the fund to avoid the negative impact of this correlation. Maturity and solidarity Conditional indexation is a very efficient instrument for reducing risk for the sponsor because of enhanced diversification. But will it be enough? We fear not. The most important flaw in DB schemes is the reliance on the willingness and ability of the plan’s stakeholders to absorb the risk that unavoidably accompanies investments. If risk cannot be sustained by the stakeholders, clearly the pension plan should not be exposed to the amount of investment risk and the scheme should be reformed. Can risk be managed such that DB schemes have a future? For the current DB systems in the UK the answer to the question is for the most part no. This is because of the increasing maturity of pension plans and the resulting accumulating pension risk for sponsors and/or future members. In DB schemes, the proportion of liabilities relative to salaries increases over time. By rule of thumb, 10 per cent of the pension liabilities of a DB scheme that started 10, 25, or 50 years ago is equal to approximately 10, 25, respectively 40 per cent of salaries. This means that if a mature fund would have to make up a 10 per cent deficit, it would require an amount equal to 40 per cent of salaries. Funding deficits using contributions would put too large a burden on young people to support the old people. We see less possibility and propensity in contemporary society to do so. As a result, reform of the DB system is ineluctable. Why personal pensions are not the answer? The faulty arguments mentioned above have globally led to a strong movement from DB toward purely individual DC systems, initially in the US and more recently in the UK and the Netherlands as well. From the sponsor’s perspective, moving to a personal pension system seems to solve the main shortcoming of DB because the contributions are fixed. However, personal pensions, in which employees bear virtually all the risk, have even more serious shortcomings than DB Investments Individuals typically make very poor investment decisions; they trade too often and too early or too late, and naturally suffer from a lack of economies of scale. It is estimated that individuals could underperform well managed, low-cost pension schemes by five percentage points annually (according to a study by John Bogle in 2005). If you consider that one percentage point less return over an entire life cycle is equivalent to approximately twenty five per cent lower pensions, the performance results of individual investors can be potentially disastrous. Rather than being a solution, personal pensions are potentially the basis of a severe pension crisis. Non-mandatory pension saving Contributing to a personal pension is often voluntary. People can decide how much they choose to save for their retirement. Unfortunately, according to a recent survey by Brewin Dolphin around ten per cent of investors or around 2.4 million people planned to stop, reduce or interrupt their pension contributions during the next 12 months. According to the Pensions Policy Institute around 20 million people in the UK are not currently accruing any rights in a private pension. The dramatic result of this insufficient pension saving in combination with unprofessional pension investing is that large numbers of people will have insufficient retirement income to pay for their basic requirements Time diversification The next serious flaw of a personal pension is timing risk. For example, individuals who started saving for retirement in the early fifties of the last century would have accrued forty years later approximately half the pension capital compared to those who began saving in the late fifties. Robert J. Shiller writes in his book “The New Financial Order: Risk in the 21st Century”: “The essence of finance is reducing the impact of risks by spreading them around to many people, and not just by diversifying an investment portfolio. We must share the risk of generations in a constructive way. Making the generations depend on the success of their investments for their own retirement is not risk management”. Therefore we feel that remedying the DB pension problem by moving to personal pensions will put employees in a position of insufficient savings that are subject to significant invest-ment performance issues and considerable timing risk. The move to personal pensions might solve the weakest point of DB for a sponsor, but ultimately it will only worsen the pension problem for society, rather than solving it. Occupational defined contribution (DC) Because pure DB is not sustainable due to maturing plans and personal pensions have considerable short-comings that cannot be disregarded, a system called occupational defined contribution schemes seem to be a good alternative. These schemes are typically mandatory. Investing is carried out by the collective rather than by the individual plan members. The group of course benefits from economies of scale and prevents individual members from taking poor decisions, which solves two of the most important shortcomings of personal pensions. Contributions are fixed and there are no statutory retrospective contributions. A fixed contribution is not to be understood as a fixed amount or as a fixed percentage of salaries. It can be a “fair value” contribution that varies with changing interest rates or it can change in line with assumptions about future investment returns. However, the contribution agreement does not entail any further liability to the sponsor. Consequently, in many cases the mandatory DC is superior to both pure DB and individual pension schemes. True cost of DC Transforming a DB into an occupational DC system, which usually comes at the request of the employer, trustees reasonably ought to demand a fair price for the risk transfer. In practice, even in cases where pension risk for the sponsor is at present significantly restricted, the price of this transfer often equals thirty per cent of the actuarial cost of the DB scheme. Employers have to consider carefully if this ‘dowry’ is an efficient (risk-adjusted) allocation of capital. There is a significant inefficiency of occupational DC if risk is not fairly priced. Hybrid conditionally indexed DB and DC Occupational DC amends important shortcomings of the pure DB and individual systems. But in spite of its advantages over DB and personal pensions, DC is no panacea for all perceived pension problems. A fair transition is costly and it lacks some of the advantages of conditional indexed DB. Proponents of DB have argued that eliminating DB would squander an important compensation and nego-tiation tool for the employer. In ageing societies where workers become increasingly scarce, such a policy instrument is essential to attract and retain a strong workforce. The argument is supported by the remarkable observation that the pension scheme for board members of companies is often still final pay DB, whereas the company has moved to DC for the other employees. Clearly, DC schemes are less helpful in attracting talents. A combination of conditionally indexed DB and DC is just what we want. It brings together the strengths of DB as well as DC but taking out their weaknesses at the same time. In a typical hybrid DB / DC system, members accrue DB rights up to a certain salary level. For any salary above that level a DC plan is in place. The employer keeps participating in the pension scheme, but the amount of pension risk that he is able and willing to bear is restricted. Yet, because of the DB components, employees are provided an inflation-proof basic pension. Hybrid schemes outperform both DB and DC. For employers, hybrid schemes imply lower expected pension cost and constrained risk. At the same time, it enables them to use pensions as a compensation tool, which is less possible with DC. For employees, hybrid schemes are more in accordance with the fundamental objectives of providing retirement income, i.e. a solid basis. On top of that there is a high likelihood of an additional income and negative out-comes are significantly less painful than in pure DC. Summary and conclusions The pressing reason for pension reform is the maturity of the plans. Employers and employees are no longer able to bear the increased liability risk resulting from the high number of retirees. As is the case in most of the current pension issues this could and should have been foreseen decades ago; now, it has grown into an unavoidable problem that simply needs to be dealt with. One way of eliminating pension risk for the sponsor is to convert to a DC scheme or to abandon the plan entirely. For many reasons, this is the wrong remedy. First of all, the risk of the current employees will increase significantly, whereas the main cause of the problem is the relatively large amount of retirees and their pension liabilities. Secondly, there is little doubt that most individuals are unable and even unwilling to carry out pension management on their own account. Occupational DC pension schemes do not suffer from the shortcomings of individual retirement savings. Besides, the contribution contains no retroactive component, and as a result all pension fund risk is removed from the financial statements of the sponsoring companies. On the other hand, the collective aspect of DC guarantees that pension saving and investing is done jointly as a group, thereby avoiding the important pitfalls of personal pensions. Although DC signifies an important step forward, it is by no means a magic potion that cures all pension problems. First of all, in order to remove pension risk from their financial statements, sponsors ought to pay the fair price for “selling” pension risk to the members. This could increase the expected pension cost up to 30 per cent. In addition, sponsors will lose pensions as an important compensation tool, a development that might especially be detrimental in an ageing society where workers are increasingly becoming scarce. We therefore strongly plead for a hybrid solution that uses the strong points of DB and DC as building blocks. It offers a limited, conditionally indexed DB pension, and on top of that an unlimited professionally managed DC at low-cost. A hybrid scheme is a flexible and dynamic pension system that conforms in the most efficient and effective way possible to the desires and constraints of all stakeholders.
|
 |
|
|||||||||
|
|||||||||||
