Con Keating compares the current methods of hedging DB liabilities and de-risking with an idle attempt to fight fires.
One of the favourite narratives of consultants advocating complex strategies in response to the vicissitudes of current defined benefit (DB) accounting and valuation is that of the firefighter coming to our rescue. The techniques involved are well-known, and include liability driven investment and de-risking. They point to successes; schemes which have hedged their interest rate risk and much more, and are not now imperilled. The firefighter’s actions were audacious and heroic. It seems to have passed them by that hedging is costly.
Indeed, the only simple, perfect hedge of an asset is its sale. Such action denies the prime function of a pension fund: investment. The question examined here is whether this narrative is the stuff of pleasant dreams or traumatic nightmares. Investigation of counterfactuals to the historic development of pension schemes are the route chosen: what have the costs of these actions been?
The obvious place to start is with consideration of the overall portfolio asset allocation; the counterfactual scenario is that we had collectively maintained the equity-heavy allocations prevailing in 1995. To open this analysis, we assume for simplicity that the net cashflows into and out of the portfolio were unchanged and investment returns are the same as experienced by these asset classes. The result is that the 1995 allocation delivers £136bn of greater value than the actual portfolio employed. To give this perspective, the total portfolio was £510bn in 1995 and £1,513bn at the end of 2015.
According to the Office for National Statistics (ONS), special contributions have amounted to £98bn. These are, of course, a cost of the current regime. It would be tempting to conclude that all were unnecessary, but doubtless, if we considered the distribution of status of specific schemes, some would remain warranted. The more unrealistic assumption is that investment returns would have been unaffected by the unchanged allocations. The case of index-linked gilts is a prime example where change would almost surely have occurred. These are now 80% owned by pension funds, and over the period have seen their spreads close from RPI plus 230 basis points to RPI minus 170; a stellar performance.
Extension of academic studies suggests that this pension fund purchasing has accounted for a little over 200 basis points of this performance. Downward adjustment of fixed income returns by 50 basis points, and upward revision of UK equity returns by a similar amount seem plausible, and may be justified econometrically, increases the outperformance of the 1995 portfolio to £244bn.
The firefighters are prone to cite the purchases of index-linked gilts as a highly visible illustration of the success of their strategies, as their performance has been spectacular. However, can anyone really say that this was a rationally expected outcome. Moreover, it should be obvious that the current levels of prospective returns breach a fundamental consideration of rational investment, that the purchasing power of our savings be, at a minimum, retained and better, enhanced.
As has been shown in other papers and articles that interest rate hedging is a redundant action, arising from the liability measurement metric, let us then consider the liabilities arising from these hedging derivatives. According to the ONS these amounted to £290bn in 2014, and at recent growth rates would have approached £365bn in 2015. The perspective to that is that it represents 22% of the PPF’s current estimate of scheme liabilities. How long, I wonder, before some whizz begins suggesting the hedging of hedges? Let the infinite regress commence.
Over the period 1995 to 2015, inflation has been largely quiescent, the implicit forecasts of around 3.2% pa have not been breached in any sustained manner; this was the epoch of the “great moderation”. The actuarial forecasts have, if anything, been lowered, lowering liability projections. Experienced longevity has increased markedly, by some four years over the period. This will have inflated projected liabilities by some 30% over the 1995 to 2015 period; it amounts only to £150bn to £200bn in ultimate payments. This is an annual rate of growth of 1.27%; though it should also be noted that it appears to have moderated markedly in recent years. It cannot account for more than a small fraction of the six-fold increase in normal contributions that took place.
Liability management exercises may also be seen to be costly. Probably more so than assets as the perfect hedge is the purchase of offsetting instruments in more restricted and less liquid specialised markets. Closure to new members raises the cost of provision of the existing stock of pensions and the cost of future accrual for remaining active members.
The whole litany of these exercises bring with them further costs, and only too often, the greatest of costs for the employee: cessation of further provision of occupational DB pensions. Little, if any, of the costs of this activity has been presented in a fully transparent manner. The firefighter is a somnambulist; that vision a sleep-terror. The pension “crisis” is as unreal as that; it is time to wake and rise to the challenges of the day.
Con Keating is head of research at BrightonRock Group