This is the analysis promised in our previous Blog covering the supplementary matters required under the proposed Regulations, contained within Schedule 2 of the Regulations, and on which no discussion was invited by the Consultation on the proposed Regulations.

These are some of the key details required for practical implementation of the Act and Regulations, but with no discussion or consultation being offered, which is a source of significant concern given the far-reaching consequences of what is being proposed.

This blog is intended principally for distribution within Parliament, as we believe that the Regulations go far beyond anything which may be inferred from the Pension Schemes Act 2021, and by virtue of that require a full debate within Parliament.  We follow the same conventions and practices employed in earlier blogs in this series, links to previous blogs in the series may be found at the end of the blog.

---

The Schedule unfortunately raises more questions than it answers.

Schedule 2

Statement of strategy – supplementary matters

1. For the purposes of section 221B(2)(d) of the Act, the supplementary matters are set out in paragraphs 2 to 19.

Maturity

2. The actuary’s estimate of the maturity of the scheme as at the effective date of the actuarial valuation to which the funding and investment strategy relates, as set out in that valuation.

As the ‘maturity’ of a scheme is not a simple concept, it requires more definition. Duration is a very poor measure of ‘maturity’. Moreover, the maturity of schemes which are open to future accrual or to new members may render any information as to the current ‘maturity’ questionable as a decision aid. Where the metric employed to measure ‘maturity’ depends upon the discount rate employed, the assumptions surrounding the determination of that discount rate further degrades the information contained within the metric employed.

It is worth noting that duration reduces to a simple average of the time to the projected pension cash outflows when the discount rate is zero, otherwise known as the average life. This illustrates the point that any term chosen as the point of ‘significant maturity’ is arbitrary. For any one member, the amount at risk relative to the employer’s insolvency is at its greatest at the point that the pension comes into payment. For example, if the member’s pension starts at 65, and we assume the member’s life expectancy is 20 years, as the member is paid his or her pension the amount at risk declines over time to zero at the end of year 20 in our simplified example.

However, the risk exposure is the likelihood of the employer’s insolvency, the level of funding of the scheme at the time the employer becomes insolvent and the cost of annuities. At times of high interest rates, the cost is low (and vice versa).  Furthermore, the amount of the pension in payment  has the benefit of the 100%  protection from the Pension Protection Fund once the member has reached the normal pension age applicable to his pension. Only the future increases are subject to reduction. That protection is paid for by the PPF levies which are ultimately born by the employers.

So an individual has no interest in the significant maturity of the scheme. It is the individual’s own risk level that matters to the individual.

The risk and uncertainty that is being managed is a scenario

(1) when the scheme is selling a significant  level of its investments each year  to pay pensions,

AND  (2) the market is moving against the scheme at that time,

AND (3) the employer covenant is insufficient to  make up the shortfall and the employer becomes insolvent

AND (4) the buyout market price is high

AND (5) the scheme assets are insufficient to buy out the benefits in full.

Looked at this way, the approach of the Regulations, as we have drawn out in earlier blogs, is to adopt a gold plated approach to managing this risk without any regard to the cost to the taxpayer (employer contributions to make up the higher funding cost are tax deductible) or to the employer or to the investment by the employer in future growth of its business.  By way of reminder, we estimate the cost to the taxpayer of these regulations as in excess of £ 38 billion.

3. For a scheme which has not reached the relevant date, how the maturity of the scheme is expected to change over time.

This is not feasible without many heroic and often pointless assumptions. If the metric employed is duration, we might employ the rule of thumb that duration declines by one year with the passage of three years’ time elapsed for typical long-closed schemes, with no change in the discount rate, but we would also see that the increase in discount rates over the past year has resulted in duration declining by four years – a twelve-fold difference in the rate of change. To satisfy this requirement would require a projection of discount rates over the future life of the scheme, together with projections of future member longevity and future inflation. Note that increasing longevity and inflation decrease the ‘maturity’ of a scheme, that is to say these increase the duration of the scheme, while increasing the projected pensions payable.

The risk faced by scheme members does not necessarily increase as scheme maturity decreases. Indeed, the risk represented by the overall scheme, which is closed, declines as pensions are paid. The risk faced by members, as illustrated in the example above, is also discontinuous around the point of retirement, falling sharply at that point. With the retirement of the last member of the scheme,  there is no further risk to the amounts of the pensions of scheme members in payment after the respective normal pension ages applicable to their pensions.

Those amounts are fully covered by the Pension Protection Fund (as long as it remains solvent). Only the future increases are subject to reduction. Note also that the risk bearing capability of a sponsor employer is usually some fixed monetary amount or proportion of its balance sheet, while the risk faced by members in aggregate declines far more rapidly than the rate of decline of its total projected (undiscounted) liabilities.

Investment risk

4. The current level of risk in relation to the investment of the assets of the scheme.

How is ‘risk’ to be defined and measured? Is this to be an absolute or relative measure? Is the metric to be based on historic developments; if so, over what period? If the metric is to be prospective, how is uncertainty to be provided for in the estimation of ‘risk’ and this is non-trivial in the context of long-term liabilities stretching over decades? Over the remaining lives of most closed schemes, uncertainty rather than measurable ‘risk’ is the dominant consideration. The current level of ‘risk’ in the sense of asset price volatility is often the result of highly uncertain future developments.

There is a further problem in that measured risk, such as it is, is perceived risk, which may be contrasted with the actual risk, the underlying hidden risk which goes undetected until it materialises in the sharp adverse movements of crises. Andrew Crockett, a former head of the BIS, has cautioned on this:

“The received wisdom is that risk increases in recessions and falls in booms.” Indeed, this is exactly what the volatility statistics would imply, and: “In contrast, it may be more helpful to think of risk as increasing during upswings, as financial imbalances build up, and materialising in recessions.”

In other words that risk actually falls as bubbles burst and prices fall, though a measure such as volatility reports it as increasing, usually sharply. This topic is dealt with extensively in Jon Danielsson’s book, https://yalebooks.yale.edu/book/9780300234817/the-illusion-of-control/ and see also https://thedecisionlab.com/biases/illusion-of-contro. l. We are engaging in another Emperor’s New Clothes exercise providing tailoring opportunities to a wide range of consultants and model makers at considerable expense for an illusion.

5. For a scheme which has not reached the relevant date—

 (a) the level of risk the trustees or managers of the scheme intend to take in relation to the investment of the assets of the scheme as it moves along its journey plan;

This is not feasible. Rational trustees would predicate their current risk-taking on the covenant strength, both willingness and ability, of the sponsor – but projecting this over the potentially long term to reach the relevant date would be exercises in fair-ground fortune-telling. We challenge the Regulator to provide a fully worked example of how they might do this for some specific scheme in practice rather than in abstract theory to see if the Emperor has any clothes.

              (b) how this complies with the principles in paragraph 4 of Schedule 1; and

Paragraph 4 of schedule 1 states:

4.—(1) The principles set out in sub-paragraph (2) relate to the level of risk that can be taken by the trustees or managers of a scheme in relation to the investment of the assets of the scheme as it moves along its journey plan.

(2) The principles are that the level of risk that can be taken—

(a) is dependent on the strength of the employer covenant (so that more risk can be taken where the employer covenant is stronger and less risk can be taken where the employer covenant is weaker);

(b) subject to sub-paragraph (a), depends on how near the scheme is to reaching the relevant date (so that, subject to the strength of the employer covenant, more risk can be taken where a scheme is a long way from reaching the relevant date and less risk can be taken where a scheme is near to reaching the relevant date).

We do not believe that it is feasible to comply with these principles when applied to prospective situations. It may be possible to comply with the principles in sub-paragraph (2) at the current time of valuation, but there is a problem in that the principles are procyclical and conflict with the Regulator’s objective to minimise any adverse impact on the sustainable growth of an employer as set out in the Pensions Act 2004, Section 5(1)(cza). Indeed, the draft Regulations do not acknowledge the existence of this objective.

These principles would require de-risking of the asset portfolio at times when the employer is stressed, and with such de-risking would then require additional contributions from the sponsor employer at the very time the sponsor employer is least able to do so. This would appear to run  counter to the Government’s pro-growth policy.

It is to be expected that this will have a further deleterious effect on the presently moribund capital markets of the UK.

We also do not believe that these principles are sound. Following the best traditions of trying to understand the unknowable universe as set out by Einstein, a gedankenexperiment (thought experiment) is in order.

Let us suppose that the ‘relevant date’ coincides with the date of retirement of the last remaining deferred member and that this is a closed scheme, with inflation uplift which under scheme rules is limited price inflation subject to the same cap of 2.5% pa as the PPF offers (and that all of the member’s pensionable service is post 5^th April, 1997. This scheme then presents no risk to scheme members even if the sponsor subsequently fails.

It requires a very strange logic to justify having minimal exposure to the sponsor employer at the time that actual risk to members approaches zero.

The Act and these Regulations have all been presented in terms principally of risk to members. If the intention is to protect the Pension Protection Fund, which is the statutory duty of TPR (as well as the sustainable growth of the employer) this should have been clearly stated and the relevant discussion taken place.

              (c) how the trustees or managers intend to achieve compliance with the principle in paragraph 3(2)(b) of Schedule 1 by the relevant date.

Paragraph 3(2)(b) of Schedule 1 states:

(b) on and after the relevant date the scheme is subject to the requirement that its assets must be invested in accordance with a low dependency investment allocation.

Low dependency is introduced by Regulations 5 and 6:

Low dependency investment allocation

 5.—(1)For the purposes of these Regulations, a low dependency investment allocation, in relation to the investment of the assets of a scheme, means the requirements in paragraph (2) are satisfied in a way that complies with an objective that further employer contributions are not expected to be required to make provision for accrued rights to pensions and other benefits under the scheme.

 

(2) The requirements are—

(a) the assets of the scheme are invested in such a way that the cash flow from the investments is broadly matched with the payment of pensions and other benefits under the scheme; and

(b) the assets of the scheme are invested in such a way that the value of the assets relative to the value of the scheme’s liabilities is highly resilient to short-term adverse changes in market conditions. [Emphasis Added]

 

We have made the point elsewhere that a scheme funded at or above the present discounted value of the best estimate of projected liabilities, where that discount rate is the best estimate of the returns to the asset portfolio held, whatever that portfolio might be, does not expect to make further calls on the sponsor employer. It may or may not need to make such calls with equal likelihood. This is no more than the meaning of expectation in probability theory and actuarial science.

We have also discussed elsewhere the impossibility of identifying short-term adverse changes in market conditions ex ante. We also do not understand why the resilience should be limited only to the short term; permanent irrecoverable losses are far more damaging.

The requirement for matching cash flows broadly brings with it some major restrictions on the assets which may be held. This is a form of dedication, and the explicit matching of cash flows to defease liabilities is well-known to be the most expensive method of achieving this (so much for sustainable growth of the employer).

The requirement effectively limits portfolio assets to conventional and index linked gilts. Corporate obligations are unsound as they are subject to the same systemic, economy wide risks as the sponsor employer.

Their performance will be impaired by these systemic conditions at the same time as the sponsor employer. In addition to the systemic risk exposure, corporate obligations also carry idiosyncratic risk. Cash flow management requires the portfolio to consist of assets which have a finite term; the contractual return of principal at maturity/redemption constitutes the majority of the cash flows.

This precludes the use of equities and indeed UCITs, even those invested solely in gilts, as their values at future dates cannot be forecasted reliably.

 

6. For a scheme which has reached the relevant date, how the current level of risk complies with the principle in paragraph 3(2)(b) of Schedule 1. [Emphasis Added]

This is an onerous rule which requires far more than a simple statement of compliance; it requires a listing of assets held and their variability individually and collectively with a comparison with projected liabilities. The actuarial costs of compliance would be very substantial.

Liquidity

7. How the investments of the assets of the scheme comply with the principle in paragraph 6 of Schedule 1.

Paragraph 6 of Schedule 1:

6.—(1) The principle in sub-paragraph (2) relates to the liquidity of the assets of the scheme—

(a) as it moves along its journey plan; and

(b) on and after the relevant date.

(2) The assets of the scheme must be in investments with sufficient liquidity to enable the scheme to meet expected cash flow requirements and make reasonable allowance for unexpected cash flow requirements.

This regulation is excessive. If the assets held broadly match the expected cash flows of the pensions payable, as is required, their liquidity is irrelevant – the reliance is entirely upon the contractual terms of the assets.

It is simply not possible to project the liquidity of markets or assets over the timescales of pension schemes. Market liquidity is in large part determined by the monetary policy actions of the Bank of England or other relevant central bank; sometimes explicitly operating on asset prices as with the recent quantitative easing strategy and current quantitative tightening. It is also determined in part by the risk appetite of securities dealers, which is in part determined by their regulatory treatment, notably through capital requirements.

In the period from December 2021 to the end of August 2022, twenty-year conventional gilt prices fell by some thirty percent. The price of the 1/8^th% index linked gilt fell by almost 70%. Daily movements in price exceeded 12% on occasion. Conversations with Gilt edged market-makers have led us to believe that these declines can in part be attributed to distressed selling by pension schemes seeking to fund collateral calls on their derivatives and repo exposures. Note the endogenous nature of risk here and the resultant feedback amplification of actions.

It might be argued that these collateral calls were unexpected, but their scale indicates the impossibility of maintaining cash provisions sufficient for such large moves. The cost of such holding would be prohibitive. There is actually an argument here for the prohibition of any contracts which carry recourse to scheme assets, such as derivatives or repos. If provision for liquidity is desired, it can be easily and cheaply achieved through the use of standby lines of credit.

Funding level

8. The funding level of the scheme as at the effective date of the actuarial valuation to which the funding and investment strategy relates, as set out in that valuation.

This seems to be a standard actuarial valuation and not at all problematic.

9. For a scheme which has not reached the relevant date— (a) the assumptions used in specifying the funding level the trustees or managers intend the scheme to have achieved as at the relevant date; and

              (b) how these are different to the assumptions used in calculating the scheme’s technical provisions in the actuarial valuation to which the funding and investment strategy relates.

Note also the requirements of Regulation 6 here.

Low dependency funding basis

6.—(1) For the purposes of these Regulations, the liabilities of a scheme are calculated on a low dependency funding basis where they are calculated using actuarial assumptions which comply with the requirement in paragraph (2).

(2) The requirement is that further employer contributions would not be expected to be required to make provision for accrued rights to pensions and other benefits under a scheme (“S”), if the presumptions in paragraph (3) were satisfied in relation to S.

(3) The presumptions are—

(a) the funding level of S is 1:1, on a calculation of the liabilities of S which uses those actuarial assumptions; and

(b) the assets of S are invested in accordance with a low dependency investment allocation.

Our reading of these Regulations is that the funding level at and after the relevant date must satisfy (3) (a) and (b) above in order to satisfy the low dependency objective. For schemes which are open to new members and future accrual there may be no relevant date.

For schemes which are open to future accrual of existing (active ) members, the relevant date can be expected to shift further into the future with each year’s new awards. For schemes closed to new members and future accrual, this requires the actuary to generate assumptions with respect to liabilities and asset market conditions at the future date. Again, this is not a trivial exercise. It would also generate highly uncertain results.

The actuarial valuation today would reflect the trajectory under which the scheme wishes to arrive at the required funding level at the relevant date, and the asset allocation for that is limited by the earlier principles that the scheme must take lower risk as the scheme approaches the relevant date. Of course, any agreed future deficit repair contributions would feed into the projected trajectory but be absent from the current valuation.

Actuarial valuations and modelling such as would be required here is likely to prove costly. The results would also be highly uncertain. Unless the model maker and the model user understand the model’s limitations, there is a very material risk that those involved in the decision taking (including those at the Pensions Regulator) will come to believe the model without appreciating its limitations – the illusion of control[i].

We would also state that we believe that Regulation 6(2) and its associated presumptions will induce a short-term fixation in sponsors and trustees, to the detriment of the long-term contrary to the aspiration that UK pension funds are sources of long-term patient capital to fuel innovation and power growth as illustrated by the aspirations in this report: https://www.britishpatientcapital.co.uk/wp-content/uploads/2021/11/BPC_Annual-report_2020.pdf

10. For a scheme which has reached the relevant date, the assumptions used in the actuary’s estimate of the funding level of the scheme as at the effective date of the actuarial valuation to which the funding and investment strategy relates.

This serves absolutely no purpose. It is simply a historic valuation and has no relevance to any risk now faced by members.

Technical provisions

11. The discount rate or rates and other assumptions used in calculating the scheme’s technical provisions in the actuarial valuation to which the funding and investment strategy relates.

In principle this is a sensible disclosure. In practice, it is orders of magnitude more onerous than its current equivalent. It will involve generating estimates of the future returns of an asset portfolio which varies in riskiness and associated returns. This is likely to be costly.

12. How the trustees or managers of the scheme expect the discount rate or rates to change over time.

This is simply ridiculous but does open the door for a new, if short-lived, growth business – selling crystal balls or another tailoring opportunity for the makers of the emperor’s new clothes.

Risk in relation to calculation of liabilities

13. How the level of risk taken in determining the actuarial assumptions used for the purposes of calculating the liabilities of the scheme complies with the principles in paragraph 5 of Schedule 1.

Paragraph 5 of Schedule 1 is reproduced below:

Risk in relation to calculation of liabilities on journey plan

5.—(1) The principles set out in sub-paragraph (2) relate to the level of risk that can be taken by the trustees or managers of a scheme in determining the actuarial assumptions used for the purposes of calculating the liabilities of the scheme as it moves along its journey plan.

(2) The principles are that the level of risk that can be taken—

(a) is dependent on the strength of the employer covenant (so that more risk can be taken where the employer covenant is stronger and less risk can be taken where the employer covenant is weaker);

(b) subject to sub-paragraph (a), depends on how near the scheme is to reaching the relevant date (so that, subject to the strength of the employer covenant, more risk can be taken where a scheme is a long way from reaching the relevant date and less risk can be taken where a scheme is near to reaching the relevant date).

We are somewhat surprised by this Regulation as it seems to imply that it is envisaged that different levels of risk estimations may be utilised in liability valuations from those associated with the asset portfolio, which was covered by the earlier discussion of paragraph 5 (b) earlier. The reservations we expressed there are also relevant here and a failure to distinguish between risk and uncertainty.

Employer covenant

14. An assessment of the strength of the employer covenant.

This is not problematic as it is an assessment of the current strength and already being conducted by most trustees in our experience, albeit with a job creation scheme for those holding themselves out as being able to assess the employer covenant strength – predominantly a case of creating the illusion of control

15. How long it is reasonable to rely on this assessment.

This is feasible but actually the wrong question. It would be better to have an assessment of the likelihood of sponsor failure over the remaining life of the scheme. A sponsor’s covenant assessment would typically have only a short useful life for decision-making, but the likelihood of sponsor failure may not be materially affected.

16. Any changes in the strength of the employer covenant since the last review of the statement of strategy.

This is not problematic, though it does place a requirement on trustees to have undertaken a covenant review along with reviews of the statement of strategy and it is dependent on the “…other factors which are likely to affect the performance or development of the employer’s business, as set out in a Code…” which is not yet available. And every time you do such an assessment it costs money and consumes time both of which are finite resources with the cost borne by the employer as to about 80% and the taxpayer by about 20% assuming the employer is making taxable profits.

We note that Regulation 7 covers “Strength of the employer covenant”

(2) The strength of the employer covenant means—

(a) the financial ability of the employer in relation to the scheme to support the scheme;

 

(4) For the purposes of paragraph (2)(a), the matters to be considered in assessing the financial ability of the employer in relation to the scheme to support the scheme are— (a) the cash flow of the employer, as set out in a Code;

(b) the likelihood of an insolvency event, within the meaning of section 121 of the Pensions Act 2004 (insolvency event, insolvency date and insolvency practitioner)(a), occurring in relation to the employer; and

(c) other factors which are likely to affect the performance or development of the employer’s business, as set out in a Code

We have considered and discussed this is our response to Question 4 of the Consultation in Blog 2 of this series, which is available here:

We would urge readers of this to read that also.

The final three elements, below, appear to be no more than boilerplate statement requirements.

General

17. The extent to which the funding and investment strategy is or remains appropriate.
18. Confirmation that the trustees or managers have consulted the employer in relation to the scheme in the preparation or revision of Part 2 of the statement of strategy.

Though not relevant to this requirement, we were alarmed to hear recently one leading independent trustee make the following statement when referring to sponsor employers and deficit repair contributions:

“If the law gives that stick, then we will have to use it”

Hardly our idea of a consultation. We would also observe that consultations with the employer are more likely to be fraught with professional trustees, than with member nominated trustees.

19. Any comments that the employer in relation to the scheme has asked to be included in Part 2 of the statement of strategy.

Any? It is also important to draw out that, if the employer currently only has the right to be consulted on the valuation assumptions (as distinct from having the right to agree the assumptions), the Government is handing another blank cheque to the trustee board (which will encourage reckless conservatism)- see the Occupational Pension Schemes (Scheme Funding) Regulations 2005, Schedule 2, paragraph 9 and the amendment to Section 229(1) of the `Pensions Act 2004 made by the Pension Schemes Act 2021, Schedule 10, paragraph 6 to add in a new sub-paragraph (za).

To expand on this point, if  a scheme, under the current funding regime, does not, like the Universities Superannuation Scheme, need employer consent to the actuarial assumptions, schedule of contributions and recovery plan, then new sub-paragraph (za) automatically flows through to the add the new funding and investment strategy to the list of financially very significant events that do not need employer agreement. Will this be another/ the final nail in the coffin of defined benefit pension accrual for current and future active members of the Universities Superannuation Scheme?

Will it yet again put that scheme’s trustee board in the position of making another poor decision to follow its decision on the scheme’s valuation as at 31^st March, 2020? And we should not forget that the greater the required contributions to that scheme, the fewer the resources available to teach the students of today, at the universities who are employers in relation to the USS, who are amongst the growth and wealth creators of tomorrow.

---

Links to previous blogs

Blog 1: https://henrytapper.com/2022/08/10/con-keating-and-iain-clacher-appalled-by-dwps-proposed-funding-regulations/

Blog 2: https://henrytapper.com/2022/08/16/keating-and-clacher-explain-the-threat-from-the-proposed-dwp-funding-regulations/

Blog 3: https://henrytapper.com/2022/09/01/the-theoretical-and-analytical-basis-for-the-dwps-funding-regulations-are-not-fit-for-purpose/

Blog 4: https://henrytapper.com/2022/09/06/comply-or-explain-becomes-comply-or-else-keating-and-clacher-on-dwp-funding-regs-4/

Blog 5: https://henrytapper.com/2022/09/14/dwp-funding-regs-suffer-from-recency-bias-keating-and-clacher/

 

---

[i] See Jon Danielsson, “The Illusion of Control  Why Financial Crises Happen and What We Can (and Can’t) Do about it.” Yale University Press 2022