The Pension Fund Directive must be implemented in EU national laws by 23 September 2005. Its gestation was tortured and long – about 12 years – reflecting the considerable difficulties involved in obtaining agreement on the scope: what was included, and, equally significantly, what was not.
The directive deals with the management of funded occupational schemes, creating a framework for the prudential supervision of “institutions for retirement provision” (IORPs), enabling these, for the first time, to operate on a pan-European basis.
However, there are limits to the operation of the directive. It does not compel member states to set up funded occupational pension schemes; it ensures scrupulously that it does not encroach on the right of member states to decide for themselves how to structure their occupational pension pillar.
Therefore in France, for example, which has a comprehensive unfunded second pillar but very little funded provision, the directive will not, initially at least, have a significant impact. However, even those member states that have little funded second pillar provision must introduce the framework created by the directive.
In this article we propose to review some of the areas of potential difficulty in implementation, from the perspective of the IORP and the user – the sponsoring undertaking. This is against a background of one of the aims of the directive, in Recital 6: to make it possible for IORPs to operate across borders, encourage occupational pension saving and thereby contribute to economic and social progress.
The principal challenge, for the effective operation of the single market in the context of pension funds, is that the directive be implemented in a coherent way, enabling economies of scale to be achieved. This coherence is not certain, since the text gives member states many opportunities to interpret the directive’s requirements in a different manner.
Implementation of the directive will have an impact at a national level, and more widely. In the first instance, a member state will introduce legislation covering IORPs authorised by it (that is, ‘home state control’), and will take advantage of any discretions as it sees fit. Thus, for example:
pA state may decide that it will not implement Article 4 – that is, will not apply the directive to the occupational pension business of life insurers;
pIt may decide to introduce an exemption for small schemes (fewer than 100 members), either broadly or in respect of selected provisions of the directive (Article 5);
pIt may require all schemes to be authorised, rather than simply to register with the home state authority (Article 9);
pIt could decide that all schemes must hold regulatory own funds (Article 16(3));
pIt could lay down more detailed, quantitative investment rules (Article 18(5)), including rules “similar to” those in the life insurance directive.
A second aspect is the interpretation of articles where, on the face of it, there is no discretion available to the member states. An example can be found in Article 8, which requires legal separation between the IORP and the sponsoring undertaking. The term ‘legal separation’ is not defined, so can be interpreted with different degrees of rigour by the member states: a restrictive interpretation, requiring a separate company to be formed, would have cost consequences for the sponsor (and the IORP).
Equally, the full funding requirement in Article 16(3) in respect of cross border business may give rise to different interpretations, for the reason that, in practice, it is unlikely that an IORP can ensure that a scheme is fully funded “at all times”, or that the supervisor can, as it must, determine when to intervene without a level of supervisory oversight that might be considered disproportionate.
This is the case even if cross border schemes are ring-fenced. In theory, therefore, the interpretation of this provision could have a direct impact on the decision of a sponsor to use the services of an IORP on a cross border basis.
A further example is the use of “ring-fencing”. The directive does not define this term, but applies or refers to it in various contexts: between social security schemes and IORPs (Article 3); between life insurance operations and IORP activities of a life company (Article 4); between sponsor and IORP (Article 8); between assets and liabilities for funding purposes (Article 16(3)); and again between assets where quantitative host state rules apply (Article 18(7)).
Those provisions mainly affect an IORP in its home state. In two of the ring-fencing examples (Articles 16 (3) and 18(7)), however, local requirements can have an impact on an IORP offering its services in another state. Article 16(3) relates to full funding requirements in the situation where the IORP has cross border activities and may be underfunded as far as its home state liabilities are concerned; Article 16(3) enables the host state, through Article 21 (5), to ask the home state to require the assets and liabilities relevant to the host State to be ring-fenced.
As regards Article 18(7), a host state can require IORPs, including those authorised in another state and operating on its territory, to follow stricter quantitative investment rules. In both these cases, IORPs would be required to carve out a different practice for host state regulatory purposes. This would not only dilute the concept of home state supervision, but would inevitably add to costs for IORPs and, ultimately, for the sponsor.
One of the most obscure provisions, from the perspective of the IORP seeking to offer services across borders, is the meaning of “relevant social and labour law”. Host member states have the right to impose their relevant social and labour law on IORPs authorised in another state and offering their services on their territory. What does social and labour law mean, and how will it be interpreted in the member states? Some member states might interpret this provision so as to limit the extent of social and labour law as far as possible (Recital 37 gives the example of the definition and payment of benefits and conditions for transferability of rights). A cynic would suggest that another state might give as wide a definition as possible, which could reduce the attractiveness of that state to an IORP from another state.
Further, the provision of details of a host state’s social and labour law, to enable compliance by an IORP from another state, might be more or less specific and helpful. Those familiar with the application of host state ‘general good’ provisions in cross border insurance operations will recognise that there is a potential minefield in this area.
There are fora for discussion of implementation issues: Indeed, Article 21 of the directive requires member states to “ensure” uniform application of the directive through exchanges of information and experience. To this end, the Commission held a meeting in October, bringing together member states’ own views on interpretation of certain articles of the directive.
The Committee of European Insurance and Occupational Pensions Supervisors (CEIOPS) has created an occupational pensions committee with a mandate to develop a common understanding of the directive; facilitate interaction between supervisors on cross border matters; and “carry out preparatory work for dealing with issues”. Among the issues for the committee will be the preparation of a “protocol” whereby the supervisory authorities will cooperate and exchange information on the implementation of the directive, and monitoring technical provision calculations and adaptation of investment rules.
The directive is a framework. It could be nothing else, given the political sensitivity of the subject. It is only the beginning of the journey. Yet the issues discussed in this article indicate the kinds of difficulties that will confront sponsors and IORPs. Member states are required to take a coordinated approach on implementation; it is less obvious, however, that this coordination will consider the market issues alongside (perfectly understandable) prudential considerations. Therefore, while national implementing legislation is not yet fixed, interested market players may benefit from continued input to their supervisors on practical solutions which could contribute to fulfilling the single market objectives of the directive, albeit pending resolution of taxation issues.
Simon Arnot is a lawyer with law firm Steptoe & Johnson LLP in Brussels