Where we are is a great example of unintended consequences. Before the mid-1970s, private sector final salary pension schemes were mainly confined to oil companies and financial institutions. Most full-time public sector workers had access to such schemes and it is worth recording that NHS Pension Scheme had career average arrangements for practitioners as far back as 1948 (they are not a new concept).

We need to cover the benefits that had to, or could, be provided, consider how asset values suddenly changed and how the comparison approaches were amended. We should also think about post-retirement mortality and the largest problem of all

During the mid-1970s, there was a proliferation of private sector final salary pension schemes, many of which were set up in order to reduce National Insurance contributions as an alternative to “SERPS”. In those days, the only real controls were those imposed by Inland Revenue, designed to restrict tax relief. In 1975, under the Social Security Act 1973, early leavers started to become entitled to preserved benefits. However, that was the only externally imposed guarantee and their entitlements were initially limited to those aged at least 26 who also had 5 years’ pensionable service (since simplified to just 2 years’ pensionable service)..

From January 1985, preserved benefits became available to more early leavers and had to be increased until retirement. In 1986, early leavers also became entitled to transfer their pension rights elsewhere, with controlled minima being instituted. From April 1987, the Inland Revenue brought in “excessive surplus” controls (explained in pictures, old colour style). If it was decided that part of a surplus should be used for benefit increases, that could lead to post-retirement pension increases being retrospectively granted. From April 1997, partially indexed post-retirement pension increases became mandatory.

With effect from April 2006, pension rights were limited in an entirely different manner (explained in pictures, old colour style). Despite being called “simplification”, nearly 20 years later, removing those controls has turned out to be complex. The 2006 amendments achieved many pensioners suffering unexpected taxes and many company directors finding another reason to discontinue their pension scheme because they would no longer be as generously compensated – yet another nail.

Let’s now look at unexpected asset value changes. In July 1997, completing a smaller step along the same path taken by the previous UK Government in 1993, the new UK Government removed “advance corporation tax” relief for UK pension funds from dividends, which was a major change. At the time, this was thought to be worth around £5 b a year. At its simplest, that led to less money being receivable by pension fund trustees, which needed to be reflected in funding projections. Since 1997, we have also had the TMT collapse at the turn of the century and the banking crisis from 2007 plus the LLDI disaster of late 2022.

Since the mid-1960s, uniquely in the world, UK actuaries had been accustomed to value assets off market for pension funding purposes (I still believe that can be appropriate sometimes). With lower expected dividends from 1997 onwards (see above), that immediately increased funding costs, which was realistic.

For a few reasons, until the 1990s, comparing assets with accrued liabilities had not been seen as too crucial in the UK. First, past service was relatively short compared with future service, with any deficits being spread over future service. Secondly, the long-term expected returns used for funding were sometimes higher than realistic. Third, other than preservation, externally imposed guarantees were minimal. Each scheme would have had its own guarantees but they were known by, and presumably acceptable to, the scheme sponsor.

Since around 2000, UK actuarial opinion had been coalescing around taking assets at market value for all purposes. Associated with that was an ever-increasing pressure to switch assets from real assets to bonds. So far as I can recall, virtually none of the actuarial literature considered the long-term as opposed to the short term, something I am trying to redress though this website. A paper, entitled “Pension Fund Valuations And Market Values” (of which I was a co-author), was delivered to the Institute of Actuaries on 25 October 1999. Our principal question was how the liabilities should be treated IF the assets were taken at market value. Opinions varied considerably within the group but we were not recommending a switch to using market values.
.
Since then, marking to market has become the standard for corporate accounting (“FRS17”) and funding purposes. For the latter, this was essentially required by the 2003 EU directive which was mandatory. That we have legislation and accounting principles requiring a specific approach doesn't mean that it has ever been robustly supported by strong evidence.

Mortality improvements have often been cited as a major reason for funding problems. However, of the changes listed above, I believe that, on its own, the undue concentration on daily changes (“mark to market”) has been far, far more significant than longevity. That has led to many pension schemes being unnecessarily discontinued which is a terrible shame.

In reality, the crucial funding problem has been using the discount process, which is singularly ill-equipped to assist in understanding long-term financial institutions.