Guest Viewpoint: Philip Shier - Society of Actuaries in Ireland
One of the objectives of the European Commission in the revision of the IORP Directive was to enhance the information provided to prospective members, members and beneficiaries of IORPs. This led to the inclusion of Articles 36-44 in the Directive that came into effect in January.
The original Commission proposal envisaged 21 new articles, but the Council and the Parliament felt that these were too prescriptive, and the final text provides more flexibility for member states to transpose the requirements of the Directive in a way which is appropriate for their national specificities.
Article 38 requires that IORPs provide members, at least annually, with a pensions benefit statement and article 39 sets out the minimum information that must be provided. This includes:
“(d) information on pension benefit projections based on the retirement age as specified in point (a) and a disclaimer that those projections may differ from the final value of the benefits received. If the pension benefit projections are based on economic scenarios, that information shall also include a best-estimate scenario and an unfavourable scenario, taking into consideration the specific nature of the pension scheme.”
Recital (63) states that the unfavourable scenario “should be extreme but plausible”.
There was some debate within the ECON committee of the Parliament on the wording related to the projection of benefits and one proposal was that the “worst case scenario” should be illustrated, but this was rejected in favour of the wording above.
The Actuarial Association of Europe proposed that three possible outcomes should be illustrated: best estimate, unfavourable and favourable, as this would provide a better balance between risk and reward, and would also be more likely to encourage individuals to save for retirement.
Leaving aside the possibility that a member state will decide that the projections are not based on economic scenarios, the Directive permits member states to require three (or more) projections, provided they include the best estimate and unfavourable scenarios.
The projections will, of course, depend on the assumptions made about future investment returns, interest rates, inflation, wage inflation, mortality and other parameters. Article 38 (4) requires that: “Member states shall set out rules to determine the assumptions of the projections referred to in Article 39 (1)(d). These rules shall be applied by IORPs in order to determine, where relevant, the annual rate of nominal investment returns, the annual rate of inflation and the trend of future wages.”
Member states will have considerable flexibility in how they determine these ‘rules’ as there is nothing in the Directive or its recitals to constrain or guide them, although one would expect that EIOPA will encourage consistency of approach.
This contrasts with the development of the Regulatory Technical Standards (RTS) on the Key Information Document (KID) for Packaged Retail and Insurance-based Investment Products (PRIIPs) which has run into difficulty. The current proposal from the Commission would require four scenarios to be projected – moderate, favourable, unfavourable and a stress scenario. The derivation of the assumptions to be used in the projections are also specified in the RTS and the proposal from the Commission is to assume that the moderate return for all asset classes will be zero. The favourable and unfavourable scenarios reflect the tenth and ninetieth percentile outcomes of projections undertaken in accordance with detailed formulae set out in the RTS and reflect the volatility of returns on the relevant investments over the preceding five years.
In my view this approach has two significant drawbacks:
• Assuming the same central return for all asset classes does not illustrate fairly the balance between risk and reward; the customer is likely to choose the product with the lower downside risk without appreciating that taking greater risk would generally give higher returns
• Basing the volatility on recent past performance implies that future performance will be a continuation of recent trends, whether that is a bull or a bear market. It would be preferable not to use past performance (which as we know may not be a guide to the future) but to use some objectively determined assumptions appropriate for the future. If it is felt necessary to use past performance, this should be considered over a longer period which would ideally capture the full economic cycle.
While occupational pensions do not fall under the PRIIPs requirements, it can be argued that there should be a consistent approach to determining how future outcomes are illustrated under occupational pensions and long-term retail investment products.
The process of transposing the IORP Directive is in its early stages and member states will presumably consider various approaches to setting the “rules to determine the assumptions of the projections”. Possible approaches could include ‘hard coding’ the assumptions in legislation or regulation, setting out qualitative rules which IORPs must interpret, or delegating the detail to a body such as the national actuarial association. Whichever approach is used, the ultimate objective is to ensure that the information provided to members is up-to-date, clear and not misleading, as required under article 36.
Among the questions that need to be addressed as part of the process of implementing this requirement in national legislation are:
• How many scenarios should be shown (for example, the minimum of two specified in the Directive, or more)? How can the different levels of risk inherent in different investment strategies be best illustrated?
• How frequently should the assumptions be reviewed? If these are intended to be long-term assumptions, is it necessary that they be changed to reflect short-term market movements, or is consistency from one year to the next more important?
• Is it appropriate for the same assumptions to be adopted regardless of the term to retirement?
• How should allowance be built in for future expected changes to investment allocations eg, under a lifestyling strategy?
• How can more complex investment vehicles such as absolute return funds or hedge funds be illustrated?
• How should retirement income be illustrated? Is the member assumed to purchase an annuity or drawdown from the accumulated fund on some specified basis?
• How can the projected numbers be made meaningful? Should the projected outcomes be expressed in current money terms and, if so, does this mean deflating the figure by reference to wage inflation or price inflation?
While actuaries are able to develop models to provide projections of future benefits based on complex methodologies and detailed assumptions, it is important that policymakers have regard to two fundamental requirements:
• The information provided to the member must be comprehensible, so that he can use it to make informed decisions about contribution levels and investment choices.
• The cost of providing the information must be kept to a minimum, as any material additional costs are likely to be borne ultimately by the members.