[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries
.

Pensions: A sizeable problem

The management of a pension scheme that is large compared with the size of its sponsoring employer poses particular problems. It can be that the employer’s maximum reasonable ability to contribute barely covers the scheme’s running expenses and PPF levies, with nothing left over for paying off a deficit. An employer is not necessarily in this position simply because it is unprofitable. It may be that the employer has declined substantially in size since its heyday and the rump employer is left with a large historic scheme to sponsor. The scheme may simply be way too big for the employer to cope with.

To keep things simple, let’s consider a scheme in which there is no benefit accrual. Take the assets of the scheme and, following an ‘employer covenant review’ exercise, add the present value of the maximum contributions the employer can afford to pay, net of expenses and PPF levies. This is the maximum resource available to the trustees.

If the scheme were well-enough funded for a buyout to be a possibility, that is very likely to be what you would plan for. You would secure the members’ benefits when you could and eliminate a massive risk for the employer. A scheme with a relatively small employer is unlikely to be in that happy position. If buy-out is beyond the realms of possibility, we need a plan to run the scheme on an ongoing basis for the long term, using best endeavours to deliver the promised benefits.

The trustees’ ongoing objective is to deliver the benefits in full, and they have a combination of investment and net contribution income to meet the benefit outgo. There will be a lower bound on the expected investment return, below which the trustees cannot expect to pay all the benefits. I have given a simple numerical example in Figure 1 (below). The liabilities valued at 4% discount rate are 133, and at 8% discount rate the liabilities are 63. At 5.5% discount rate, the value of the liabilities and the value of the assets and contributions are equal: both are 100.

In this example, the trustees need to have an investment strategy that is expected to return at least 5.5% pa. Any investment strategy with a lower expected return is not expected to provide enough income to pay the benefits in full.

The trustees’ task is to set an investment strategy that generates sufficient asset income to meet the benefit outgo. Note that, for this purpose, setting an expected return on non-bond assets using the ‘gilt yield plus something’ method simply will not do because it makes no attempt to assess the income expected from the assets. Rather, we need the rate of return that discounts the asset income cashflows and gives their market value as the answer — that is, the internal rate of return — for each of the main asset classes.

Suppose for the sake of a simple illustration we limit ourselves to two broad asset classes. At the valuation date, ‘equities’ have a dividend yield of 3.5%, expected real dividend growth of 1.5% pa and expected RPI of 3% pa, for an 8% pa expected return. ‘Bonds‘ have a yield of 4% pa. An asset allocation of 37.5% equities and 62.5% bonds produces the required 5.5% pa expected return needed to expect to deliver the benefits in full (excuse the simple addition of the percentages in this paragraph, rather than proper compounding).

One mantra is that a poorly funded scheme with a weak sponsoring employer should minimise volatility and put the assets in bonds. In my example, if the assets were all in bonds, which yield 4% pa, the investment return is expected to fall short of that required to deliver the benefits in full by 1.5% pa. The funding level is 103/133 = 78%. Members would not receive their benefits in full with a high degree of certainty. Which is lower risk — investing 37.5% in equities and 62.5% in bonds, targeting 100% of benefits plus or minus (say) 20%, or investing 100% in bonds targeting 78% of benefits plus or minus (say) 10%? The lower volatility of the second option is outweighed by the lower expected outcome, therefore the first option is the lower risk. The volatilities quoted here are purely illustrative, they are not the result of any modelling.

I do not see the sense in planning to fail by investing so much in bonds at the outset that there is no realistic hope of obtaining sufficient investment income to pay the benefits in full. We should at least try to deliver benefits in full by adopting an appropriate investment strategy. You never know, the actual outcome might be what we expected, or better, and the benefits paid in full.

The assumptions for valuing the liabilities should be as realistic as possible. The consequence of building in prudent margins is to drive the asset allocation away from bonds into equities. I am not suggesting ignoring the legal requirement for the demographic assumptions to be prudent, but bear in mind when setting the assumptions that excessive prudence in one part of the funding strategy will drive more risk-taking in the investment strategy. Having identified the lower bound on the required investment return, should the trustees invest to produce that return and no more? Or should they adopt a strategy that pursues a higher return?

Taking the first option, having set an asset allocation targeting the lower bound, the trustees may then review the scheme’s funding annually and adjust the asset allocation as required to maintain equality between the value of the maximum resource and the value of the liabilities. Doing this maximises the proportion of the assets in bonds at all times, and therefore minimises the exposure to assets of less certain return. The funding of the scheme is on a best-estimate basis, but the investment strategy is the most prudent possible, commensurate with providing for all of the benefits.

I do wonder about the opportunity cost of not investing more highly in equities at the outset. But my thinking here is that the performance of equities is uncertain. If the short-term performance of equities is below expectations, then having more than is strictly required at the outset is damaging. On the other hand, if the short-term performance of equities is above expectations, then having more than is strictly required at the outset is unnecessary.

If the scheme did eventually fail to deliver its benefits in full, it seems better to be able to say to the members that, though the scheme failed, the trustees took the least risk possible all the way through commensurate with trying to pay the benefits in full, than be in the position of having to explain that the scheme failed after the trustees had taken more risk than they needed to.

Another mantra is that the trustees of a scheme with a weak sponsoring employer should set a high value for technical provisions. Let’s say that a discount rate of 5.5% pa is deemed unsatisfactory for technical provisions, and a discount rate of 4.5% pa is required. We know that discounting the liabilities at 5.5% pa equates them to the value of the assets and future contributions. If the benefits falling for payment after the recovery period are to be discounted at 4.5% pa, then the discount rate during the recovery period must be more than 5.5% pa to maintain the equality of liabilities to the available resources. Consequently, the asset allocation must be shifted away from bonds and into equities with the objective of achieving a higher return during the recovery period. If successful, the assets may be rearranged at the end of the recovery period to be mostly bonds, since only a 4.5% pa return is required thereafter.

To summarise, where an employer’s ability to contribute to a scheme is limited, there may be a lower bound on the return to be achieved by the assets, below which the trustees cannot expect to provide the benefits in full.

Considering the range of investment strategies above the lower bound, I pose a question. Is it better to take the minimum investment risk at all times, by determining the most cautious asset allocation that is expected to deliver the required return, and reviewing that allocation regularly? Or, following the implications of the ‘weak employer means high technical provisions’ mantra, is it better to take more investment risk now in order to take less risk later?

Risk aversion and prudent management suggest only investing more in equities when circumstances compel it, and not before. But this conflicts with the requirement to set a high figure for technical provisions, which drives more investment risk-taking now in the hope of taking less investment risk later. What do readers think?

_______________________________________________________________

Derek Benstead is a scheme actuary with 12 years’ experience. The opinions expressed in this article are his own, and not those of his employer, First Actuarial