Written evidence from the Pensions & Investment Research Consultants Ltd (PIRC) LDI0063

1              Summary

Problems with the accounting standards which have led to the proliferation of LDI’s include:-

• international accounting standards treating pension funds as if investment to cover pensions obligations is a value based “hedge” (using short run values for assets and liabilities), when in fact a pension fund needs to be a cashflow based “hedge” (using long run cashflows to meet net cash obligations). See Para 2 below.

• pension funds being required to use an arbitrary bond based discount rate; whereas life insurers acquiring DB pension schemes can use a more favourable discount rate by using an expected rate of return from the assets they chose to hold, as they follow a different accounting standard

• taking point-in-time market values of gilts, which is not consistent with outcomes where gilts are held as long term investments for held to maturity cash-flows

• the use of low discount rates - such as when gilt yields are low – creates extreme valuation sensitivity, thus magnifying liabilities in a “hall of mirrors” effect.

• issues from the Penrose Enquiry following the collapse of Equitable Life Assurance Company were not resolved in setting accounting standards

• the international accounting standards system is a mish-mash of illogical assertions and inconsistent within itself (see Irish Court case. Para 7 of this paper).

Macroeconomic effects from the accounting standard for pension funds has led to

• pensions funds no longer being significant providers of capital to UK companies

• LDI driven investments proliferating, which are based on a false perception of risk (short term volatility) rather than the long-term cash flow risk.

Pension scheme funding levels and accounting standards should revert to the pre-IAS 19 actuarial basis, whereby funding is calculated based on expected portfolio returns - with a margin for error – and expected obligations, taking account of the fact that relatively small changes in contribution from employers and employees can have a significant effect on future funding.

2              Liabilities “going down as interest rates rise” – a baseless mantra

Justification of LDI for pension funds largely rests on the baseless mantra that liabilities go down as bond yields go up.  That is false and is a function of using inappropriate discounting at inappropriate discount rates, rather than a fact of real world money.

The mantra was thankfully challenged by Nigel Mills MP who correctly identified (Q143 in the  session on 14 December 2022[1]) that actual pension obligations (real cash outflows at a future date) are likely go up in circumstances of interest rate rises - given the link between inflation and interest rates rising in response to inflation). Even if indexed linked bonds are used, the obligations still go up. Any investment approach based on ill-conceived numbers is unlikely to be wise.

The problem flows from the accounting standards International Accounting Standard 19 “(IAS19”) and the UK equivalent which had their genesis at the turn of this century. IAS 19 creates an arbitrary number for what is called a pension “liability” that may bear little relation to the actual pension obligation.

I cover later how and when the concept of discounting for accounting purposes arose (paragraph 8). It is not a natural conclusion that any liability should be discounted in a set of accounts. Rather, discounting began from a desire to reduce liabilities to take credit for expected returns on assets. Some of the language around the use of discounting in accounts isn’t actually a rational description of the economics of discounting, but rather a language of loaded terms to sell it as a concept.

Pension accounting is actually an anomaly against the IFRS[2] system itself. IFRS identifies two different types of hedges, one type is to hedge values, the other type is to hedge cash flows.

The objective of a long-term pension fund investment strategy should be to hold investments as a hedge to match cashflows of investments and obligations. Unfortunately IAS 19 treats a sponsoring employer’s pension fund as if the objective was to hedge point in time values of investments and liabilities. IAS 19 is calculating pension funding by subtracting one wrong number (bond rate discounted liabilities) from another wrong number (marked to market portfolios).

3              Abracadabra – now you see it now you don’t

There is a clear anomaly between the rates insurers use and the rates pension funds use to discount pension obligations.

When an insurer buys out a Defined Benefit Pension Scheme the accounting rules applicable to insurers, International Financial Reporting Standard 17 (“IFRS 17”) allows them (if that’s what they have done before) to use a higher non-bond rate for discounting their obligations. The difference between a yield of 2% (essentially a price to earnings multiple of  50) and 4% (a price to earnings multiple of  25) is a 100% difference. Put another way a change in discount rate one way can halve a liability, and the other way it can double it.

So exactly the same pension obligations reduce enormously when expressed as accounting “liabilities” when in the hands of insurers. That is a reason why a bought out DB Pension Scheme when in the hands of e.g. Legal and General, doesn’t need to have an LDI strategy, using L&G’s LDI offering, because the favourable accounting standard magics a large part of what IAS19 gives as a “liability” – and hence a “funding deficit” - away.

So insurers are selling products to pensions funds that they themselves don’t need when they have then bought these pension funds out because – as set out below – they have negotiated with standard setters a better deal (see para 10 later). In any other commercial setting that would be seen as a significant unfair competition issue.

4              Not a hedge but a poor investment strategy which undermines real economies

LDIs are not actually a hedge but a misguided investment strategy, because LDIs are aiming to hedge the accounting numbers, whilst the accounting isn’t dealing with the true nature (i.e. business model) of a pension fund. LDI driven investments are treating pension funds as something that needs to have the value at a point in time hedged when the issue at stake is the long-run return on an investment portfolio.

Once interest rates have bottomed LDIs are not providing “hedging” but are guaranteeing losses to the extent gilts already purchased will fall in value on a mark to market basis when interest rates rise. The collateral calls would be better described as “IAS 19 loss funding”.

However, the total return of a portfolio of gilts already held when held to redemption does not change with market prices, the coupon (the fixed return) and the return of capital (the nominal value of a bond) are fixed cash amounts fixed from the point of purchase. Mark-to-market models of bond holding risk (volatility risk) may be appropriate for short-term bond traders, but not a pension fund. Pension funds are prohibited by tax law in the UK from engaging in trading activities and would lose tax exempt status were they found to be doing it.

5              Macro economic effects – and national security issues

There has been tremendous shifting of wealth in the UK away from pensioners and sponsoring companies, due to outsourcing accounting standard setting to unaccountable private sector bodies. The International Accounting Standards Board is self-perpetuating Delaware based foundation. The USA wisely does not use the standards.

Given that the UK had an unusually high DB pensions presence any harmful effect from IAS 19 would be expected to disproportionately harm the UK, and not, e.g. Russia or China. Yet Russia and China are treated as stakeholders on an equal basis by the IASB.

The interest of the Big 4 accounting firms in one set of global standards does not necessarily align with objectives of national interest.

It’s not surprising the UK has poor productivity when accounting standards and actuarial methods which are aping the accounting have created enormous shifts from a supply of funds to companies (via equities) to cheap funds for governments (bonds).

The statement from the former CEO of the Pension Protection Fund that “funding trumps covenant” is correctly criticised in Con Keating’s evidence . 

The public interest should be vested in healthy (wisely invested pension schemes) with the system being able to deal with a small number of sponsor collapses.

Instead, The Pensions Regulator (“TPR”) has fallen into the trap of “funding trumps covenant” and helped create systemically anaemic pension schemes, which due to poor bond returns act as a bigger drag on companies, in an anaemic economy.

The insurance element of the TPR and Pension Protection Fund’s (“PPF”) work would benefit from accepting that most companies will be going concerns. The “funding trumps covenant” line” takes an approach which essentially sees sponsors as “gone concerns” and treats DB pension funds as albatrosses around the sponsor’s neck. Ironically TPR has helped create albatrosses.

As an analogy the “funding trumps covenant” line, were it followed in motor insurance would lead to risk reduction by insisting all insured cars were to be kept parked in a garage and never driven.

That myopic approach of TPR/PPF is similar to the discredited pre-2007 Basle banking regime which placed too much stead on types of banking capital that could be used if a bank was no longer a going concern and thus put into solvent run-off - so as to be pay back interbank debt. The flaw with that thinking is that is that what might work for one bank isn’t a sensible approach for all banks. If a whole banking system is run on the basis it can be put into solvent run off, it risks creating an outcome where there isn’t a banking system left.

Endowment mortgage mis-selling occurred because actuaries assumed investment returns would remain high. Equitable Life collapsed as a result of guaranteed annuities doing the same.

LDIs are the converse of that. What is identical in each case is systemic harm from group-think. Outsourcing thinking to accounting standard setters can become a catalyst for group think.

6              Herd investment behaviour

The fall in gilt yields we saw from year 2000 onwards can partly be attributed to pension funds en masse switching to bonds in the early 2000’s. The effect of large numbers of pension funds buying bonds at the same time in response to the new accounting standard was identical to Quantitative Easing, only it was not the Central Bank doing it.

That is not proper investment led demand but herd created demand due to accounting standards and investment advisers as well as TPR using the same approach. (Related to that exotic credit products that were extremely risky – Collateralised Debt Obligations etc – which helped cause the banking crisis were themselves produced to satisfy demand, in the wake of the shift from equities).

Gilt yields were artificially low in recent years, firstly due to the herd effect of pension funds shifting to gilts, and then due to overt Quantitative Easing as a result of the Banking Crisis.

7              The emperor has no clothes

An Irish Court was clear that International Accounting Standards did not reconcile to first principles and were an injustice to logic[3] the judgment stated.

“Before going on to that question, it is necessary to pause to ask why one is being required to do any injustice to logic in the first place.  It would seem unlikely (or even impossible) to be the case that, if all matters were being dealt with by reference to the same set of first principles, any such illogicality could arise.  However, here not all matters are being dealt with by the application of the same first principles.

“The accounts of new ILP are required, as a matter of law, to be compiled in accordance with the relevant accountancy standards and not in accordance with first principles. There have been many areas in the past where deeming one thing to be another (doubtless for good reason) has had unintended consequences

“It is also obvious that the unusual facts of this case have demonstrated that there is a problem with the relevant accountancy standards, at least insofar as they apply to circumstances such as those which have arisen in this case.”

8              How discounting of liabilities started – as late as 1987

The history of having any discounting is relevant to the current debate. It is clear from the proceedings of a working party under the Institute of Actuaries in 1987 [4] that the concept of discounting financial liabilities was novel.

What had occurred up to then is that regulators had allowed insurance companies to hold less ring fenced assets against liabilities - where the ring fenced assets were gilts - on the basis that the gilt coupon would give growth to meet the liabilities.

That regulatory concession was not matched by the accounting of that time, and didn’t need to  as assets are assets and liabilities are liabilities. Any attempt to increase assets, or decrease liabilities would be an artificial increase in the apparent strength of a balance sheet.

However, the post 1987 accounting treatment was to decrease liabilities by the process of discounting. That is a backdoor way of taking credit now for future investment returns. Put another way, insurers began to take accounting credit for interest that governments had not yet paid. That defies the accounting principle of prudence - not booking prospective profit

So the practice of discounting arose from having a pool of dedicated ring fenced assets, but since then international standards have allowed discounting even in industries – such as oil and gas decommissioning liabilities – where not only are there no ring fenced assets, but the assets may actually be worthless, stranded assets.

Penrose is clear that imprudence in accounting practice was at the behest of the insurers. Rather than accounting and  regulation based on first principles, which is the way judges tend to approach things when there is a corporate collapse, matters get muddied by lobbying principles. But it is not in the interest of creditors, policy holders, shareholders, or non-executive directors to overstate a balance sheet. Unfortunately it may be in the interest of executives.

9              Black Swans and clichés

Black Swan Events can be the inevitable outcome of quack theories and products. But the term “Black Swan Event” has become a cliché to make something that was very much predictable appear to be a rarity.

A poor regulator is worse than not having one at all. That’s because regulated parties left to their own devices are unlikely to all follow the same bad idea that a poor regulator foists on them. However, it is interesting to hear that some pension funds were pushing back against TPR’s coercive approach.

I also note neither the PRA or TPR has raised this, but if a pension fund is bought out by an insurer, protection to pensioners falls to the Financial Services Compensation Scheme instead of the Pension Protection Fund. Hence due to the anomalies in accounting standards the FSCS/PRA and the Pensions Regulator are looking at the same thing – pensions liabilities – in a different way. Both can’t be right.

10              Little has changed since the Penrose Enquiry Report of 2004

Lord Penrose said of insurance accounting in his report into the collapse of Equitable Life Assurance Company[5]:-

“The situation can be stated starkly. There are no UK accounting standards that specifically relate to insurance companies. So far as accounting standards are concerned, the insurance industry largely escaped regulation.”[6]

Lord Penrose set out how the insurance industry set its own guidance instead via a “Statements of Recommend Practice” (SORP) for which he damningly said, “The preparation and publication of the first SORP did not follow accepted consultative procedures.”[7]

He also said “the measurement of such liabilities using discount rates based on asset returns, which was inconsistent with certain prescribed accounting standards”[8].

And in terms of effect he said, “That practice had failed to keep up with industry developments and as a result exposed the investing public to the risks associated with inadequate information about the business to which they entrusted their savings”.[9]

International Accounting Standards were introduced in 2005 for the EU. However, the standard for insurance was not a standard, it was merely permission to use existing national practice. Which for the UK was essentially pre-Penrose use of what suited the insurance industry.

IFRS 17 was issued in 2017 requiring discount rates consistent with those for pension funds. That met - which a Freedom of Information request shows - to be intense insurance industry lobbying[10].  An HMT person was then seconded to the new UK Endorsement Board. IFRS 17 was then revised by the International Accounting Standards Board and has been endorsed by the UK. At least two UK Endorsement Board members have an interest in bulk annuity transfers (the buyout of pension funds).

11              Time bombs and accounting standards

Accounting standards have been implicated in the case of  banking collapses and as in the case of Carillon, construction companies. Also the sorry tale of Greensill Capital is a result of accounting standards encouraging “reverse factoring” a form of hidden financing.

It is not an exaggeration to say accounting standards have been putting time bombs under pension funds, banks, and construction companies. The common factor is that the IFRS system is weak in dealing with contracts that can take years (hence several sets of annual accounts) to reach a conclusion.

The lobbying letter from the large insurers correctly states that the USA has rejected international accounting standards, China has a selective approach to using them, and overall there is no global set of standards. It is anomalous that Russia and China have any involvement in UK accounting standard setting, in a way that would not be regarded as appropriate in any other area of national life.

LGPS Schemes didn’t suffer the same fate as corporate schemes, and hence maintained higher allocation to equities and didn’t need LDIs. They are not regulated by TPR.

12              Recommendations 

a)      Public good

There should be a standing committee of Parliament – perhaps a joint Lords Commons committee - to scrutinise accounting standards with particular attention to pension funds, banks, and insurance companies. At the moment accounting standards can have effects within the ambit of BEIS, the DWP and Treasury. BEIS takes the lead, but BEIS officials responsible for accounting standards do not appear to be equipped to deal with the issues. Problems with accounting standards are a recurring feature of Parliamentary enquiries. BEIS has done nothing tangible to resolve the problems that emerged from the collapse of Carillion – including “reverse factoring”.

This should include a Parliamentary led review of the use of discounting in any company accounts.

b)     National security

There should be a Parliamentary review of national security implications of international accounting standards dealing with matters of significant economic activity.

January 2023