DWP’s Consultation on Options for Defined Benefit schemes
Iain Clacher and Con Keating
In looking at this consultation, it suffers the same problems evident in the Impact Assessment published alongside the DWP’s draft Occupational Pension Schemes (Funding and Investment Strategy and Amendment) Regulations, being based on highly questionable statistics and models used by the Pensions Regulator (TPR).
The consultation paper states:
-
The Pensions Regulator (TPR) analysis [footnote 4] suggests that of the approximately 5,000 DB schemes, over 3,750 are in surplus on a low dependency basis [footnote 5], with a further 950 schemes approaching surplus [footnote 6] on a low dependency basis. For those schemes in surplus, there was in excess of £225 billion in aggregate surplus assets on a low dependency basis across the DB universe [footnote 7], roughly 17% of total DB assets. However, data suggests the amount of surplus accessed, relative to current surplus, is low. HMRC analysis suggests around £180 million in surplus has been extracted over the 5 year period between March 2018 to March 2023 [footnote 8].
The footnotes, which are highly informative, are:
-
Figures provided by TPR with a calculation date as at 30 September 2023.
-
For these purposes low dependency liabilities are calculated using a generic approach of gilts + 50 bps. Other bases might be suitable.
-
Schemes funded over 75% are considered as approaching surplus levels.
-
Based on the net funding position of schemes in surplus, on a buyout basis. Purple Book 2023.
-
This figure was estimated using Accounting for Tax (AfT) data on reported DB surplus payment charges divided by the 35% charge rate to estimate total payments, aggregated over a 5-year period.
The first point to note is that there is a mismatch of dates between TPR’s calculations as at 30th September 2023 and the Purple Book’s buyout figures, which are as at March 31st 2023. The PPF reported scheme assets as £1,404 billion in March 2023 and £1,330 billion in September 2023. TPR by contrast reported scheme assets as £1,415 billion in March 2023 and £1,372 billion in September 2023. Over this six-month period, scheme assets have fallen by 5.3% by PPF calculation but only 3.0% by TPR calculations.
The situation gets far worse when the estimates of buyout liabilities are considered. The PPF reports a decline of 40.4% in the cost of buy-out over the year to March 2023, from £2,105.0 billion to £1,254.9 billion. By contrast, TPR reports buy-out liabilities declining from £2,108 billion to £1,441.0 billion over this period, a decline of 31.6%. There is a disagreement of circa £186 billion. The buyout funding ratio of schemes using the PPF figures is 116.7% but by TPR’s figures, just 98.2%. Using the PPF figures rather TPR’s biases the results to the point that they are unusable.
As an aside, our own research suggests that the buy-out liabilities of the universe of schemes are £1,490 billion (a decline of 29.3%). However, we have very little confidence in this estimate, as it was based on conversations with three BPA insurers and one consultancy.
We are also treated to differences in the estimate of Technical Provision liabilities (TPs) being variously, £1,236 billion or £1,203 billion. We show in Table1, the cost of buyout relative to technical provisions estimates:
Table 1
By TPR’s figures, the cost of buyout lies between 16.6% and 19.8% above Technical Provisions, or only between 1.5% and 4.3% using the PPF’s figures.[1]
There is, as we have noted in many other commentaries, a significant discrepancy between the estimates of scheme assets from TPR (£1,415 billion) and PPF (£1,404 billion) and that of the ONS quarterly Financial Survey of Pension Schemes (£1,244 billion). This difference, respectively £160 billion (PPF) or £171 billion (TPR), is material. The TP funding levels, using then ONS March 2023 asset figure, are 100.6% or 103.4% depending on whether we use PPF or TPR liability estimate.
The low dependency figures are also suspect. TPR’s report to the Work and Pensions Select Committee show these liability values as £1,314 billion in March 2023 and £1,166 billion in September 2023. These estimates are respectively 9.2% and 8.3% respectively above the TP liability estimates, of £1,203 and £1,077 billion. The low dependency figures are based upon a single equivalent discount rate (SEDR) of gilts plus 0.5%. For schemes to have their low-dependency funding this close to Technical Provisions would imply that the average SEDR was close to gilts plus 100 basis points.
In Table 2 below, we show TPR’s estimates of assets and liabilities for the period December 2021 to December 2023 and examine the funding ratios for both Technical Provisions and Low-Dependency portfolios (Lo-Dep).
Table 2
The growing disparity between ONS asset estimates and those modelled by TPR are shown in the row labelled TPR – ONS (highlighted in purple). It is worth emphasising that prior to June 2022, the differences between the ONS and TPR on assets is largely immaterial when we are dealing with an asset based that is well over £1 trillion. However, the December 2021 to March 2023 period difference is £201 billion, from -£30 billion to +£171 billion and was £231 billion at its widest.[2] This is consistent with the quarterly ONS surveys capturing margin calls and portfolio rebalancing activities which are not captured by TPR’s models.
We also see a pronounced decline in the cost of low dependency strategies relative to Technical Provisions, from a 17.4% premium to just 6.1%. (Lo-Dep Premium, purple) This is consistent with the definition of low dependency discount rate as a fixed spread above gilt yields of 50 basis points. As rates rise the significance of that spread declines. It also means that a scheme which is low-dependency at higher gilt yields, such as those we have experienced more recently, will not remain low-dependency if rates fall. If a scheme which was only just meeting low-dependency in December 2023, the same scheme would have an 11.3% shortfall should rates return to those prevailing in December 2021. Deficit repair contributions totalling £229 billion would need to be found from sponsor employers. Low dependency should not be defined as a fixed spread above gilts.
We show below in Table 3 the proportional changes in assets and liabilities and funding ratios. Over the full period TPR reports assets declining by just 15.2% and liabilities declining by 31.3% resulting in the Technical Provisions funding ratio improving by 23.3% from 103.3% to 127.3%. For the full period, TPR’s estimates of the low dependency liabilities decline by 37.9%, and the funding ratio improves by 36.4%, from 88% to 120.0% (see also Yable 2)
However, the story is very different when we use the ONS estimates of assets rather than those of TPR. For the period to March 2023 for which ONS asset estimates are available, we see declines in assets of 31.7% in contrast to a decline of just 21.0% for TPR’s asset estimate. The declines in liability values are 30.6% for technical provisions and 35.5% for the low dependency strategy.
Table 3
The funding ratio for technical provisions using TPR data shows an improvement of 13.9% over the period, but the funding ratio using the ONS asset estimates shows a decline of 1.5%.
The equivalents for low dependency are 22.4% and 5.9% improvements respectively. However, it should be noted that the improvement shown for the ONS low dependency strategy is considerably less than the improvement due to the arithmetic operation of a fixed spread above gilts.
Table 4 below shows the proportions of schemes in low dependency surplus using TPR’s figures and those of the ONS. We see that at March 2023, using the ONS asset value, just 31% of schemes are in low dependency surplus, while TPR report this as 64%.
This overestimation of the relative health of the DB system by TPR has already had consequences for policy assessments, notably the extremely low estimate of the cost of low dependency contained in the impact assessment for the draft Occupational Pension Schemes (Funding and Investment Strategy and Amendment) Regulations.
The table also shows the proportion of schemes in surplus on a technical provisions basis using ONS and TPR asset data.
Table 4
Final Thoughts
It is difficult to respond with any real confidence to DWP’s latest consultation given these large uncertainties as to the status quo. We will however end this note with a question: if schemes really are funded at 127% on the prudent technical provisions basis, why is the option of reopening schemes to new members not being actively considered as an option?
Notes
[1] This would be 23.9% using our buyout liability estimate.
[2] The latest version of the FSPS is due around late March 2024, and so we will then have sight of data covering the June and September numbers, and will update the analysis accordingly.