In this article, Keating and Clacher consider corporate insolvencies and their impact on the PPF – they call for a sense of proportion.
Iain Clacher and Con Keating
Last year DB pension schemes paid £65 billion in pension benefits to pensioners; the PPF paid £775 million, just 1.2% of the total.
In the 15 years since the foundation of the PPF, some 17% of the then outstanding 8,000 schemes have suffered sponsor failure and joined the PPF. This is an annual rate of a little over 1% – seemingly alarming since this is much higher than the rate of insolvency in the population of companies at large.
If this rate were representative, it would imply that few employers would survive long enough to see the discharge of all their pension liabilities – though that runs counter to the marketplace evidence. Far more schemes have wound up through buy-out and merger than entered the PPF, almost 1.5 times as many.
Share of liabilities in a shrinking universe
In this setting, the most important metric is arguably the amount of liabilities acquired by the PPF. Including schemes in assessment, these liabilities amount to 1.7% of the total of liabilities outstanding. While by number of members, the PPF has covered a slightly higher 2% of members. The past two years have not been kind to the PPF; schemes in assessment make up 29% of total liabilities. From these statistics the PPF intake has been overwhelmingly of small DB schemes, which of course are usually associated with small employers.
After considering pensions paid in those years, the insolvency rate by liabilities is just 0.1%. This is better that the experience of the German PensionSicherungsVerein (PSV) and the Swedish PRI-Pensiongaranti (PRI-PG) where it averages just under 0.3%. It is also markedly better than the UK corporate population, where it varies cyclically around a 0.6% average. However, that figure is greatly influenced by the high failure rate of young companies – over 50% do not reach their fifth birthday, and so are entirely absent from the DB pensions universe. The highest insolvency rate, 1.6%, was reported in the 1991/92 recession; we should expect it to be exceeded in the wake of coronavirus.
With no new DB schemes being created, the PPF universe is finite and shrinking. As this universe is non-refreshing, its aggregate credit standing will be the result of individual company’s credit-standing and the migration of that over time. These empirically derived transition matrices show that the strong tend to weaken over time, but also that the weak who survive tend to strengthen. As we are concerned with the aggregate standing of the remaining universe, we should note that scheme failures would tend to improve, while the schemes that are wound up would tend to have been those which were well funded with strong employers, so their removal would tend to weaken the credit quality of the remaining companies.
It is true that we cannot reliably predict the credit standing of most companies beyond about ten years as the uncertainty is too great. However, it would be a mistake to assume that they have rising likelihoods of failure – an assumption that the Regulator seems to believe. The evidence is that insolvency rates are cyclical around a steady average. The experience internationally of PSV and PRI-PG add support to this for universes of DB schemes.
Schöpferische Zerstörung
In much of this there is some confusion around the concept of ‘creative destruction’, the process by which the old fail and are replaced by more efficient newcomers: a key feature of capitalism and economic development. In recession, the process is more one of capital destruction without more efficient replacement. By construction companies are sempiternal institutions, but their sustainability over the long term depends critically on their ability to refresh and renew themselves through investment in human and technological capital. It is therefore concerning that a majority of companies in the latest CBI survey of their members reported that their DB funding costs are constraining their investment in both people and technology.
It is against this background, that the Pensions Regulator’s proposals that funding should increase to reduce or eliminate scheme dependence upon employer need to also be considered. It is true that some of this extra funding would find its way into capital markets, but those markets are unavailable to the majority of sponsors in the PPF universe. The extra funding required of schemes would come at the cost of reduced sustainability of the sponsor employer.
Concluding thoughts
As we have pointed out in earlier blogs, the risks to scheme member pensions are small, far smaller than the funding being required to resolve them. The risks to members are in any case limited to the amount of the PPF ‘haircuts’, which from published pension payment and valuation data, appear to be of the order of 15% of members’ promised pensions. It is also perfectly possible for the PPF to pay full benefits, as is the case with the PSV[i] and PRI-PG, and the overwhelming majority of their universes of schemes are entirely unfunded.
It is worth noting that the aggregate amount of projected benefits of the PPF universe is already declining as so many schemes are closed to new members and future accrual, which means that the risk exposure of the PPF and sponsors will decline as time passes. However, this has been masked by falling discount rates.
Put another way, loss rates by value of DB pensions are around 0.1% per annum, so 99.9% is working just fine. Moreover, the major issue is that of sponsor insolvency, which is the trigger for the PPF to step-in, and nothing has been done about that, as it is simply something that cannot be dealt with – companies fail. However, assuming all companies fail is patently wrong and is a faulty and expensive premise to start from.
KPMG have estimated the costs of the proposed funding regime to be £100 billion – our own work suggest that this may be a low estimate, but as the above shows, the risks from sponsor insolvency are small and basing a funding regime on it are excessively high
[i] Actually the PSV has a maximum pension cap of €8,400 per month, a sum which would be far in excess of the UK’s lifetime allowance.
his comment is written in response to an email received directly from a reader of this blog.
To clarify. The effect of removing sponsor failures is to enhance the credit-standing of the residual universe, all else being equal.
Transition Matrices. In any year, we expect no failures from schemes in the highest category, we expect 5% to be from schemes in the second category, 20% to be from the third category and 75% from the worst. This is predicated on the rating system being consistent and properly calibrated at an annual frequency. There are also movements across the ratings categories. Improvements and deterioration in credit ratings.
Finally, the 0.1% rate is of liabilities entering the PPF. The loss rate incurred by the PPF is a fraction of this, determined by the level of funding of schemes. Last year was unusually poor with funding at just 68% of liabilities. This implies that the schemes failing were extremely weakly funded. The more usual level of funding is around 80%, though there have been years in which the loss absorbed by the PPF has been lower than that borne by members as haircuts.
From the subsequent behaviour of PPF liabilities it appears that the valuation of the liabilities of schemes in assessment is extremely conservative, that they are being ‘kitchen-sinked’.