Managing the Risky Business of Pensions
On the face of it, pension schemes are quite simple. The scheme holds assets to cover its liabilities and, to the extent that their values move in tandem, the financial risk is broadly constrained. The complexity arises in selecting a basis on which to value the liabilities, where a number of approaches are not only possible but also common. Despite all these approaches reflecting a discounted present value of the scheme’s projected benefit payments, the valuation bases in use in the UK are far from consistent (see table below).
| Valuation | Impact | Theory |
|---|---|---|
| Accounting (IAS19/FRS17) | Company balance sheet and P&L | Scheme payments discounted at prevailing long-dated high quality (typically) AA-rated corporate bond yields. |
| Buy-out (aka solvency) | Cost to company of ‘exiting’ pension provision | Price of passing pension commitments to insurance company. Typically priced off yields available on gilts, swaps, corporate bonds and/or credit default swaps. |
| Economic | Best estimate of liabilities being met on ‘least risk’ basis | Scheme payments calculated using market consistent pricing of inflation, longevity, etc. and discounted at swaps rate. |
| PPF | Scheme levy paid to the Pension Protection Fund | Restricted scheme payments discounted at prevailing real and fixed gilt yields. |
| Scheme funding | Company cash payments to the scheme | Scheme payments discounted ‘prudently’ with regard to yields on scheme assets, bond yields, buy-out prices and a host of other things. |
The valuation basis on which the scheme’s sponsor and trustees focus will be an important determinant of its risk management policy and consequently the sort of investments that are needed for the scheme’s assets to best match movements in its liabilities. With so many competing and conflicting measures in play, retaining risk, or not tackling the problem, remain popular choices. The degree of liability hedging encountered in practice is demonstrated in the diagram below from a survey of 95 large pension schemes:
Unsurprisingly, pension risks can be reduced. Also, unsurprisingly, this normally comes at a cost. With the majority of de-risking measures relating to cash injections and/or changes in scheme asset allocation, the decisions to take boil down to trading risk for return. To properly consider any risk reduction solutions, some form of risk/return framework needs to be established.
Increasingly a value-at-risk (VaR) measure is used for this purpose, but stress testing or other methods can be applied. Assessing a 90% or 95% VaR (or conditional VaR) measure for the current position of the scheme is always instructive, e.g. the 1-in-10 or 1-in-20 worst case deterioration in a scheme’s surplus/deficit over the coming year on the chosen valuation measure.
Return will typically be measured as the relative return on scheme assets compared to the chosen liability valuation measure or, for a more tangible analysis, the probability of achieving a desired outcome over the next five or 10 years, e.g. the likelihood of being fully funded against the chosen measure in five years.
Against such a framework, the impact of different de-risking solutions for different valuation measures can be objectively analysed.
As ‘certain’ payments, those who adopt an economic approach would assert that the scheme’s liabilities are assessed as benefit payments discounted at a risk-free rate appropriate to the term until payment- usually either the rate available on government securities or the swap rate for a similar maturity. However, the two principal valuation measures (accounting and scheme funding) take very different approaches.
As mentioned above, the accounting valuation discounts the scheme’s benefit payments in line with AA-rated corporate bonds. Accounting standards and auditor guidance generally require some recognition of the scheme’s liability duration when choosing the appropriate bond yield.
The scheme funding valuation does not pretend to be an economic valuation- it is more akin to a cash flow budgeting exercise for scheme trustees and sponsors. For this purpose, scheme trustees will set a target level of funding, i.e. a level of assets which is deemed sufficient to cover future benefit payments, and determine how best to achieve this through cash contributions and investment returns. This must be done prudently, but can take account of numerous factors, for example:
The term ‘funding target’ above is important. The aim is to compare scheme assets against this target as a means of assessing cash contributions required to keep the scheme on track to pay its benefits. The funding target is regularly, and perhaps misleadingly, referred to as the trustees’ valuation of the scheme’s liabilities. As mentioned previously, the scheme’s liabilities ought to be a risk-free assessment, whereas a funding target is really a heady cocktail of investor preference, risk tolerance and prudent management.
It is therefore highly common to set a funding target that discounts scheme benefits at a rate including allowance for risk premia within the scheme’s investments. Widely varying views about the level and handling of variables, such as price inflation expectations and future pensioner mortality rates, add to the range of different approaches in use.
Returning to the process of reducing/managing pension risk through investment strategy or hedging products, the process involves measuring risks within the chosen framework, prioritising them, assessing the availability and cost of risk reduction solutions and deciding how best to employ the budget available for de-risking. The principal solutions currently available are aired below. All of them require some form of trustee consent before being undertaken.
Companies regularly seek help in minimising the risk of increases in both pension contributions and the charge to profit and loss from pension arrangements. Invariably, it is difficult to do both, with accounting liabilities linked to corporate bond yields and trustee funding targets often, though by no means always, linked in some way to gilt yields. For example, over the six months to 31 March 2008, it was common to see these two liability measures diverge by up to 20-25% for a typical scheme.
Unsurprisingly when faced with this type of dilemma, many sponsors and trustees are giving serious consideration to transferring the whole pension scheme and its risks to another party, subject to a reasonable premium being charged.
Although alternative methods have recently been devised to enable a pension scheme and its risks to be passed on, the more traditional approach of securing annuity policies through a Financial Services Authority (FSA) regulated insurance company remains the preferred solution. Over recent years, increased competition in this market coupled with innovation has driven down bulk annuity prices, with the cost/benefit features of a scheme buy-out becoming increasingly attractive, even for those sponsors with limited risk management budgets or insufficient cash to meet an immediate premium in full.
However, budgetary constraints (both cash and accounting) remain a widespread barrier. Further, a full buy-out requires closure of the scheme and hence changes in employee terms and conditions in many cases. Also, by moving the scheme’s asset base off balance sheet the beneficial, if largely illusory, expected return on pension assets can no longer support the P&L result for companies with substantial schemes (although changes being considered to FRS17 and IAS19 over the next few years may reduce the perceived benefit).
Consequently, other partial hedging techniques are ordinarily up for consideration either on a standalone basis, or in combination. The principal approaches are covered below.
This involves trustees purchasing annuities as scheme assets for a section of members (with subsequent direct allocation of annuities to members possible) and is popular at present due to favourable annuity prices for scheme members already drawing a pension. The ability to significantly de-risk without closing the scheme for current employees is another welcome feature for many sponsors.
Hedging scheme cash flows, either accurately (insofar as this is possible for payments with limited predictability) or by duration, offers a viable means of reducing risk. Physical gilts/bonds or interest rate swaps can be used to achieve either aim and treasurers are typically familiar with the principles and execution aspects involved.
One significant elephant trap arises from the actuary’s discount rate in liability valuations. It remains common for actuaries to assess liabilities based around a long-dated gilt or bond index, despite such an index offering a poor reflection of the scheme’s benefit payments in terms of duration and convexity. Before entering into this type of a matching transaction, it is vital that an actuary is asked (usually by the trustees and sponsor) to value the liabilities by discounting payments in accordance with the relevant yield curve.
With a significant portion of UK pension benefits being linked to increases in the RPI, unanticipated increases in inflation pose a significant financial risk. Index-linked gilts or inflation swaps offer a means of limiting this risk, although both markets are rather thin at the 40+ year terms over which many pensions will be in payment.
A healthy debate exists over the existence or otherwise of an inflation risk premium in the pricing of these instruments and hence the efficiency of the hedge relative to retaining risk, particularly as pension RPI linkage is often capped (typically at 5% p.a.). Regardless of the rights and wrongs of this argument, again the actuary may hold the key. Looking into the future, RPI increases of 4% may well be painful to the balance sheet if the accounting figures previously produced by the actuary incorporate a long-term expectation of 3.5%.
There is burgeoning interest in this market, with numerous and varied parties looking to transact ‘longevity swaps’ or similar trades. In principle these may be based on actual scheme member mortality rates or the mortality rates in some benchmark population. With increases in projected life expectancy seemingly emerging with more severity and regularity, many sponsors may wish to pass this risk on. Relative to the use of insurer annuities, one perceived benefit is the ability of the sponsor to obtain longevity protection whilst retaining the asset base on balance sheet.
While the above comments relate largely to the UK, multinationals will have similar considerations in any other geographies where they have defined benefit obligations. Local market practices may differ substantially (e.g. a prevalence of unfunded pension obligations, lack of inflation linkage, restricted range of tradable hedging instruments), but these differences can offer opportunity.
Companies with pension obligations in a number of countries may be able to take certain risks in geographies whose markets and local regulation facilitate this, whilst avoiding some or all of the risks elsewhere. For example, influences might include the availability or otherwise of local pension guarantee insurance and/or tax relief on additional contributions. If done effectively using available techniques, this can enable sponsors (with the agreement of the respective fiduciaries) to run bespoke local risk profiles to arrive at their desired global risk allocation.
Despite inherent complexities, pension risks, like other business risks, need to be understood, quantified and managed if the value of the firm is to be optimised. Moreover, many firms outside of the financial sector may not feel they have the comparative advantage necessary to expect to generate excess economic returns from taking pension risk.
Although solutions are rarely simple and must be considered in relation to each scheme’s circumstances, if pension risks are of material concern, then techniques exist to address this and are increasingly being applied.