IORP II: towards a new solvency framework?
Jacqueline Lommen reviews the two recent CEIOPS papers and calls on the pensions industry to make an active contribution to the upcoming IORP II public consultation
At the end of March 2008, the Committee of European Insurance and Occupational Pensions Supervisors (CEIOPS) published two papers. The first paper describes the problems encountered when the IORP Directive is put in practice.(1) This inventory has been prepared in view of the upcoming evaluation of the IORP Directive. The second paper is a comparative assessment of the financial requirements for IORPs in the EEA countries.(2) This survey is part of the potential further development of the European solvency framework for IORPs, informally referred to as "IORP II".
By sharing this material with the broad public, CEIOPS is extending its reputation as a transparent body towards the pension sector. This is good news for anybody interested in cross-border pension opportunities in Europe. A decision to integrate two or more local pension schemes into a centrally managed cross-border IORP can now be taken in possession of more material facts. CEIOPS' information confirms that far more is happening in the cross-border pensions market than many are yet aware of.
Does Belgium really allow a 6% discount rate to be applied? Is the Netherlands indeed "expensive", pricing itself out of the market and being an unattractive domicile for cross-border IORPs? Is the funding position of Anglo-Saxon pension trusts weak compared to the IORPs in other countries? Is Solvency II, the prudential regime for insurance companies, really to be applied to IORPs? The CEIOPS reports provide valuable insights, reveal the true facts and help remove persistent misunderstandings. The fundamental differences between the various local prudential regimes for IORPs are carefully laid out. At the same time the papers detail the parameters which are used for the actuarial calculations in each country.
In comparing the local financial requirements for IORPs, CEIOPS makes a clear distinction between on the one hand technical provisions and on the other, security mechanisms.
With regard to the calculations of the technical provisions, the CEIOPS survey shows that only four countries apply the current risk-free market interest rate as the discount rate: The Netherlands, Denmark, Sweden and Portugal. In eight countries, much higher discount rates are allowed. The benchmarks here are the expected returns from the pension schemes' investment portfolio. Countries in this group include Belgium, Ireland, Luxembourg (with the CAA IORP vehicle), Liechtenstein and the UK. In contrast, there are also states that embed extra prudence in the technical provisions by setting a relatively low discount rate, even substantially below the risk free market rate, according to rules derived from insurance regulations. Countries which follow this practice are Germany, France and Norway. Finally, there are member states, like Luxembourg (with the CAA IORP (insurance) vehicle) that set a fixed maximum discount rate which is unrelated to market developments.
In some EEA member states there are additional discount rate requirements due to the local social and labour law. These financial requirements apply only to local pension schemes and are not applicable to the foreign pension schemes being managed by an IORP based in these member states. This is a major source of confusion. An example is Belgium, where a discount rate is prescribed in respect of pension value transfers.
The impact of mortality tables in explaining international differences in financial requirements tends to be underestimated. To the extent that the variations result from different longevity risk profiles of local scheme members, the variation obviously does not matter in an international comparison. What is relevant though, is the degree of prudence that member states require. Not all states, for instance, appear to be using the most up to date mortality tables, nor is the inclusion of steady improvement of life expectancies or incorporation of additional safety margins universal. The limited up-to-date nature of tables in use in Liechtenstein, Luxembourg and the UK is striking. In Germany, France and Norway, where the pensions sector is traditionally dominated by insurers, prudence in setting longevity risk assumptions scores relatively high, though.
In addition, when comparing the minimum required technical provisions in the EEA member states, the overwhelming influence of indexation is often overlooked. For instance, the CEIOPS paper reveals that the UK, Ireland and Luxembourg make it a legal obligation for IORPs to account for price inflation and/or wage inflation in the technical provisions. A prudent approach.
In contrast, several countries have a nominal framework. The Netherlands, Germany, France, Norway, Sweden, Finland, Liechtenstein and Malta lie in this group. These countries do not require by law that the technical provisions take into account indexation of rights. However, some of these countries put in practice other methods to protect member's purchasing power, for example by including the inflation protection in the solvency buffers. Belgium, Denmark and Italy, take a flexible approach. These countries only make it compulsory to account for inflation and salary increases in the technical provisions if the terms agreed on in the specific pension scheme require so.
An international comparison
CEIOPS has investigated the combined impact of the elements listed above. For 15 EEA member states where DB schemes play a major role, the required minimum size of the technical provisions is illustrated in a bar graph (see figure 1). A standardised example has been used. The individual effect of differences in discount rates, mortality tables, inflation protection and expenses is illustrated. This is a unique set of data and a first-time published comparison which contains valuable new insights.
It turns out, that for instance in Germany, Norway and France in particular, the minimum size of the technical provisions appears relatively higher than average. This is primarily because of the low discount rate applied. The UK, Ireland, Luxembourg and Spain also maintain sizeable provisions on the balance sheet of the pension trusts. This follows directly from the above mentioned approach towards inflation protection. Belgium, Denmark, Finland, Sweden and Greece, on the other hand, require relatively limited technical provisions. It is striking that when comparing technical provisions, The Netherlands is also among the latter group of "cheap" countries.
However, the financing requirements for IORPs are not limited to the technical provisions. The local prudential regimes encompass a wide range of supplementary security mechanisms. The IORP Directive itself for instance, requires own capital if the IORP bears biometric risks and/or provides (investment return) guarantees. In addition, several countries apply additional security mechanisms in order to protect pension entitlements if economic times follow a downturn. Security mechanisms are not always fully recognised when comparing standards internationally.
CEIOPS distinguishes between the following security mechanisms: regulatory own funds and supplementary solvency buffers, recovery periods, subordinated loans, plan sponsor support, national guarantee funds, conditional indexation, and, as a last resort, reduction of accrued pension rights. Which security mechanisms are being applied by which EEA member states is shown in the table below (see figure 2).
Ex-ante versus ex-post approach
The table shows that broadly two groups of countries can be distinguished. On the one hand, those countries which have incorporated ex-ante safeguards into their prudent regimes. The report identifies the Netherlands, Belgium, Germany, France and most Scandinavian countries as being in this group. In these member states, the starting point is the IORPs' balance sheet; a legal entity strictly separated from the plan sponsoring company. The IORP has to maintain mandatory solvency buffers. In some countries within the group it is allowed that the solvency buffers are partially financed via subordinated loans and facilitated through recovery periods whereas other countries act more strictly.
Some countries on the other hand, opt for ex post security mechanisms. These only come into force if the IORP reaches a situation of underfunding. Countries in this group include the UK, Ireland (to a partial extent), Germany, Belgium and Luxembourg. In these countries, the starting point is not the IORP, but the plan sponsor. The safeguards are incorporated into a legal obligation on the plan sponsor to provide additional contributions to cover shortfalls. Even so, recourse to a national guarantee fund, funded by the plan sponsors and/or members, is feasible. The first group of countries initially appears to be relatively "expensive". The ex ante security mechanisms have to be financed in advance. The solvency buffers have to be maintained within the IORP and remain beyond the reach of the sponsoring company. But this initial assessment completely turns if one also takes into account poor economic times. This is the point when the ex-post security mechanisms are activated in the second group of countries. The employer is obliged to support the IORP by means of additional contributions, and/or fees must be paid into the national guarantee funds. Ex post security mechanisms may even turn out not to work properly. For example, because the IORP's underfunding might coincide with financial difficulties on the part of the sponsoring company, therefore rendering the plan sponsor to serve as lender of last resort.
CEIOPS's study makes many other interesting details available, which go beyond the scope of this article. It describes, for example, the conditions under which an employer is obliged to support an IORP, the maximum permitted duration of recovery plans, the specific workings of national guarantee funds and the ways in which pension entitlements already accrued are allowed to be reduced.
Keep away from partial analyses
The CEIOPS papers show that 65 cross-border European IORPs had been set up by the end of February 2008. These are established in nine different EEA home member states. These 65 cross-border IORPS manage pension schemes originating in 18 different host countries.
What is the best domicile for a cross-border IORP? The CEIOPS paper emphasises that careful work needs to be done when comparing the prudential regimes of IORPs in different European countries. An analysis should not be carried out on the basis of partial elements. It is unfair to compare countries on the basis of the level of the discount rate only or the presence or absence of mandatory solvency buffers. There is a danger of comparing apples with pears, which leads to half truths and hasty conclusions. An integrated approach is required. CEIOPS highlights that technical provisions and supplementary security mechanisms frequently counterbalance each other. In countries where technical provisions are set prudently, the need to finance supplementary security mechanisms is often minimised. Elsewhere, the opposite is true.
Market experience supports the CEIOPS findings. Nobody is well served by IORPs that use regulatory arbitrage to select their home country. For employees, the pension entitlements are not well protected by a race to the bottom in IORPs' financial requirements. For employers, financial requirements are frequently just one of the considerations.
Market insight shows that considerations that are specific to a particular employer and a particular IORP are decisive. Which of a company's existing local IORPs is best suited to serve as the base for an consolidated multi-country European IORP platform? Where is good governance best available? Where is ring-fencing allowed and can the extent of solidarity be set in a flexible way? Which solution will meet the greatest support from both the plan sponsor and the members? A careful analysis, based on true facts and involving all stakeholders, is a prerequisite for a successful approach towards the establishment of a cross-border IORP.
Solvency II versus IORP II
The CEIOPS survey serves as a base for a potential European convergence of the financial regimes for IORPs. This discussion has been triggered by Solvency II, European prudential framework for the insurance companies, currently in the making. In recent years, the question has often been raised as to whether IORPs should also be regulated by the new Solvency II regime. An intensive public debate on this issue is currently underway.
CEIOPS has now clearly stated in its report that in its view, "…an integral application of Solvency II requirements to pension institutions…the material differences between pension funds and insurance companies in many countries…suggest this is not an appropriate course to pursue. Such action could lead to excessive costs and thus bears the risk of threatening the continued provision of defined benefit schemes." The general aim of IORP II is stated as being, "the safeguarding of pension beneficiaries' claims at reasonable cost".
The next step in the IORP II process is a broad, public consultation. This will take place in the second half of 2008. All pension industry players will be invited by the European Commission and CEIOPS to provide input.
If a more convergent EU prudential regime for IORPs will be developed, an integrated approach will have to be taken, as described above by CEIOPS. If the risk-free market interest rate and solvency buffer principles, as common in Basel II and Solvency II, are copied wholesale, the pension sector will suffer unduly. The specific characteristics of IORPs and their pension promises made would be neglected.
Moreover, the indexation method needs to be harmonised. Inflation protection, if provided by the pension scheme, must either be incorporated into the technical provisions ór in the solvency buffers, uniquely in every state. Without a more harmonised method of calculating technical provisions, there will be a risk of mandatory solvency buffers overshooting the mark, making DB schemes too costly.
The less widely applied security mechanisms, like the plan sponsor covenant and the guarantee fund, must become an integral part of the potential new regime. This could potentially be achieved by means of a valuation of these alternative security mechanisms and deducting the value from newly required mandatory solvency buffers. Moreover, long transition periods would be recommendable. This applies also to subordinated loans and generous recovery plans, which are employed in several countries. It is essential that these possibilities also remain intact or are otherwise discounted for in the potential IORP II.
It is high time for the European private pension industry to think pro-actively and in a structured fashion about
IORP II, in close cooperation with the European Commission and CEIOPS. Who will take up the gauntlet? The better the alternative, the greater the chance that Solvency II is definitely kept at distance from the IORPs.
1 Initial review of the key aspects of the implementation of the IORP Directive, 31 March 2008, CEIOPS
2 Survey on fully funded, technical provisions and security mechanisms in the European occupational pension sector, 31 March 2008, CEIOPS.
Jacqueline Lommen is senior international employee benefits consultant at Hewitt Associates. She is a former officer of the Dutch pension supervisor (DNB) where she was closely involved with pensions affairs at CEIOPS