Accounting for Pensions Survey 2000
Main findings 


Significance
of pension
cost		 Pension cost as a proportion of pre-tax profit can be very high. For British Airways, pension costs equated to 49% of profit and the assets of their pension schemes are worth nearly twice as much as their own market capitalisation. This was exceptional - the average pension cost for all companies was 4% of pre-tax profit. At the other extreme, cost can be low or even be an addition to profits; BG (British Gas) and Reckitt Benckiser showed pension credits (additions to profits) of over 5% of total profits.
Some
improvement		 Our seventh survey shows significant improvement in disclosure, which is encouraging. Whereas a year ago nine companies disclosed every item we believe to be necessary (up from five the year before that), this year thirteen companies did so.
Glaxo /
Carlton		 Glaxo and Carlton were highlighted in our report last year as having very poor disclosure. Both companies have made excellent efforts to change this. Glaxo now achieves our highest ranking for disclosure. Carlton scores 18 out of a possible 20 and has set out a complex pension position covering three major pension schemes in a neat table, picking up on best practice by other companies.
Schroders Schroders last year scored only 11 out of 20 but have produced almost "model" disclosure this year and gained the maximum score.

Whether or not these improvements were a direct response to our survey, we welcome the improvements made
Overall
standards
remain low		 Nevertheless, many FTSE 100 companies do not make anything like adequate disclosure of pension cost assumptions. We believe that a minimum disclosure level is that represented by a score of "16" on our ranking. One third of the relevant companies scored lower rankings than this. It is a pity, therefore, that any improvement may be "too little, too late" - FRED20 heralds major changes which just might have been forestalled if companies had been generous in their pension cost disclosures over the twelve years since SSAP 24 came into effect.
Sun Life &
Provincial		 Sun Life & Provincial score highly on the disclosure ranking, yet the way the company discloses its pension cost prompts us to describe it as a "model of confusion". Firstly they quote the pension cost as being increased by £10 million because of a surplus, when precisely the opposite would be expected. The total cost is described in money amounts but the breakdown between major schemes is expressed as percentages of payroll figures, which are not disclosed. Finally, the cost breakdown misuses the term "regular cost" to mean the cost allowing for surplus, but in fact it is defined as the cost excluding the effect of surplus.
BT In last year's report we criticised BT for confusing disclosure and possible incorrect treatment of very large redundancy costs. The same would apply to the accounts for BT's year ending 31st March 1999, covered in this survey. However our criticisms last year were published after the March 1999 report was produced. BT took on board the criticisms and their accounts to 31st March 2000, now published, have changed their disclosure to fully overcome our objections.

BT have not needed to account for redundancy costs - about £500 million for the two years to 31st March 1999 - because the surplus on an accounting basis was adequate to cover these costs. Nevertheless they made special contributions to the pension schemes of £200 million in the year to 31st March 1999, and £230 million the following year, in part because of the redundancy costs.

Under the FRED20 proposals BT would have had to account for the cost of the redundancy pensions even though the cost was covered by surplus. In BT's case, where extra cash was paid, this may not have had a significant impact. For other companies when the cost of benefit improvements is in fact met out of surplus the FRED20 requirement to charge the entire cost to profits for that year is seen by many as likely to deter companies from making benefit improvements.
ACT All companies would have been well aware last year of the potentially large effect of the ACT change, ie a little under a 20% fall in dividend income from UK equities. Most UK schemes have about a half of their investments in UK equities. Both the income on existing assets and the expected future investment return would be reduced. One would expect this to have a large effect on costs.

However our survey reinforces the results of last year's survey to show that the ACT cost has been deferred by altering actuarial assumptions. In particular the dividend growth assumption has been increased (relative to investment return) by about ½%, which has the effect of increasing the value of equity investments for actuarial purposes by as much as 15%.

This broadly compensates for the ACT change on past service benefits. (It still leaves an increase in cost for future service as the ACT change affects the assumed future investment return).

In practice the ACT change has to some extent resulted in a reduction in the margins used where these were significant. However, some companies (such as Cable & Wireless and HSBC) disclosed that the ACT change had a direct effect.
Market value
methods		 Eleven companies now use a "market value" method to determine their pension cost under SSAP 24. This is an increase over previous years, and might be taken to reflect the move towards market value inherent in the new standard proposed by FRED20.

In some ways the market value methods used are very consistent, in particular in their method of arriving at the assumed rate of inflation (see below). However, the key element in the methods used is the discount rate used to value the liabilities, which is generally taken to be the discount rate applicable to gilts at the time of valuation plus some margin to reflect the extra return expected on equities. This margin varied greatly and there was no clear pattern at all.

The market value methods used bear little relation to that proposed by FRED20. The new standard would discount liabilities using corporate bond rates, with the return on equities used to obtain a short-term "expected return on assets" as an offset to the growth in liabilities.
Inflation
assumption		 The method used to determine the assumed rate of inflation in market value methods is generally to compare the returns on fixed interest and index-linked gilts. If returns on fixed interest long-dated gilts are, say, 3% pa higher than the returns on similarly dated index-linked gilts, then it follows logically that the market is expecting inflation to be 3% pa.

In fact the prices of fixed interest and index-linked gilts are determined by the laws of supply and demand. Expected inflation is a factor in the relative pricing but only one factor - more important is the demand for each type of stock which is heavily determined by other factors such as pension funds wishing to match MFR liabilities and insurance companies matching liabilities.
Abbey
National		 Abbey National's accounts do not disclose a dividend increase assumption. Neither do they say that a market value was used. Consequently they score no marks in our ranking for asset values. However the assumptions used look likely to be for a market value method. If Abbey National use a market value method it is a pity they did not make this clear.
Balance sheet
entries		 The pension cost under SSAP 24 and the actual company contributions to the pension arrangement can differ considerably. Any excess of cost over contributions accumulates to form a provision in the balance sheet; conversely, any excess of contributions over cost gives rise to a prepayment. Most companies disclosed some provision or prepayment; in a few cases they were very substantial.
Prepayments Large prepayments, ie assets in the balance sheet, appear for BP Amoco ($2,542 million), Shell (£1,159 million), Diageo (£672 million) and Lloyds TSB (£647 million).
Provisions Large pension provisions were held by Unilever (£1,170 million) and BT (£953 million).
Back to top