Accounting for Pensions Survey 2001
Detailed analysis of Reports and Accounts   -   Disclosure


We analysed and ranked the disclosures of 79 companies, as listed in Appendix 2.

For this purpose, we excluded 14 companies who had no evidence of significant defined benefit provision and 3 companies which did not report pension costs under UK generally accepted accounting principles (Anglo American, BP Amoco and Shell).

We also excluded Billiton, Old Mutual and South African Breweries, which are essentially overseas companies who have chosen to be quoted on the London Stock Exchange.

We had to exclude Compass, whose report was not available.

We have analysed the disclosures in a very similar way to previous years. Virtually all companies disclose certain basic items. For example, although not all companies disclosed the assumed rate of investment return, all disclosed enough to know the return relative to salary growth.

We have made one modification to the scoring system this year, reflecting the increased use of market value methods. In order to score full marks, companies whom we suspected of using a market value method needed to make it absolutely clear that assets were indeed taken into account at market value. Where it was ambiguous as to whether this was the case, they were marked down slightly.

We believe that there are four key additional items that have to be disclosed for a full understanding. We set these out below, with an explanation of an attaching "score". All companies have earned a starting score of 10 for the basic items.
  Disclosure

 Score
Dividend
growth rate
or actuarial
asset value		 Asset valuations are generally based on the discounted income method, and therefore the dividend growth assumption has a very important impact on the results. We accepted either the absolute dividend growth rate or the rate relative to the investment return.

Seventeen companies did even better than this and quoted the actuarial value used. This is better because the dividend growth assumption is the main, but not the only, assumption in the asset valuation process.

For some companies it was clear that assets were taken at full market value. These and companies quoting the actuarial value of assets were all awarded a score of 5, as for companies disclosing the dividend growth rate.

As illustrated in section 6.2, a market value basis looks quite different to a traditional basis. As such, in some cases, we had strong suspicions that market values were used, but we were unable to draw a firm conclusion. These companies were awarded a score of 3.

90% of companies achieved a score of 5 for disclosure of the assumption used to value assets, compared to 75% last year.

 5 or 3
Pension increases The assumed rate of pension increases is necessary for a full understanding, but there is less variability of cost from this assumption. 94% of companies disclosed this item, which is up slightly on last year's survey (92%).

 1
Split of
regular cost
and variation		 The regular cost is the ongoing cost to the company of providing the benefits of the pension scheme ignoring surplus or deficit. Disclosure of this item is, therefore, very helpful. 47% of companies separately included this compared with 36% who did so last year. This again is a great improvement, even though it remains disappointingly low.

 2
Method of spreading The actual pension cost will be lower or higher than the regular cost, on account of spreading the surplus or deficit.

There are a number of ways this can be done, and each spreading method gives a different pattern for future pension costs. 57% of companies disclosed the spreading method compared to 45% last year.

This gives an aggregate ranking from 10 to 20. A full list of each company's ranking is set out in Appendix 2. The spread of ranking scores is set out below.

 2
Ranking Ranking

The top ranking (20) was achieved by an impressive twenty five companies this year; their details are listed in Appendix 2.

No companies scored the minimum ranking of 10 this year.

The average ranking this year was 1.4 higher than last year (and 2.3 higher than the year before that). We are pleased to see these continued improvements, even in the dying days of SSAP24, and hope it will have prepared companies for the rather more extensive disclosure requirements of FRS17.
Celltech Celltech are the only company to score less than 13 this year. That said, their pension cost is small relative to the size of the company and does not have a big impact on their reported profits. The transition to FRS17 will undoubtedly cause them to look more closely at pension cost disclosure.
Scottish Power Scottish Power gave a concise breakdown of disclosures for its five different defined benefit schemes to gain a maximum score of 20. It can be done!
WPP Last year we had to exclude WPP from our analysis as they quoted ranges of assumptions that were too wide to be meaningful. Perhaps as a result, this year, like Scottish Power, they have adopted a tabular format for the disclosure of their many schemes and suddenly all has become clear. As one would expect, the assumptions used for their Japanese schemes reflect very different financial conditions from those in the UK and the US.
Marconi Marconi's score has fallen from 16 last year to 14 this year. The pension cost is based on a revised actuarial valuation - no assumption for dividend growth is given and although the liability assumptions could well be market-related it is not stated whether assets have been taken into account at market value. In addition to quoting two discount rates for past service and a different rate for determining future contributions (to which we have no objection) they have quoted a fourth, "average" discount rate and it is not at all clear what this represents.
Reed Elsevier Reed Elsevier's disclosure looks very similar to their disclosure in the previous year, but their score has fallen from 20 to 18. Last year they disclosed full details of a traditional actuarial valuation with long-term assumptions and included a dividend growth assumption - all easily understood.

This year they disclose details of a new valuation and an investment return assumption that certainly looks market-related - and there is no dividend growth assumption. This suggests that assets have been taken into account at market value. Why then, do they talk about "actuarial" values of assets?
Spirent Spirent's disclosure initially looks good - the assumptions are set out fully and appear market-related. In addition, they state that a dividend growth assumption is "n/a", as we would expect for a market value method.

Unfortunately, their methodology is thrown into doubt when they say that "the level of funding represents the "actuarial" value of the assets expressed as a percentage of the value of the liabilities on the statutory MFR (minimum funding requirement)". This leaves the reader somewhat confused, as there is no concept of an actuarial value of assets in an MFR valuation and the MFR is of no relevance to pension scheme accounting.