british american tobacco p.l.c. annual report 2008 - Notes on the accounts: Note 12

Sub Navigation

12 Retirement benefit schemes

The Group’s subsidiary undertakings operate around 150 retirement benefit arrangements worldwide. These arrangements have been developed in accordance with local practices in the countries concerned. The majority of scheme members belong to defined benefit schemes, most of which are funded externally, although the Group also operates a number of defined contribution schemes.

The liabilities arising in the defined benefit schemes are determined in accordance with the advice of independent, professionally qualified actuaries, using the projected unit credit method. All schemes are formally valued at least every three years.

The principal pension schemes are in the UK, Germany, Canada, the Netherlands and Switzerland. Together these schemes account for over 86 per cent of the total obligations of the Group’s defined benefit schemes.

In addition, the Group operates significant schemes in Canada which provide employees with certain other retirement benefits such as health care. The liabilities in respect of these benefits are also assessed by qualified independent actuaries, applying the projected unit credit method.

As explained in note 1, the Group has changed the accounting policy for the recognition of actuarial gains and losses under IAS19 and adopted IFRIC14. The comparative figures have been restated accordingly for the following changes:

For the year ended 31 December 2007, the profit from operations and taxation were both reduced by £1 million and therefore the profit for the year was unchanged from that previously reported.

The impact of the accounting changes was to reduce the Group's total equity at 31 December 2008 by £817 million and increase the profit for the year by £4 million.

The total accumulated net actuarial losses taken through the Statement of recognised income and expense at 31 December 2008 were £591 million (2007: £44 million) for subsidiaries and £357 million loss (2007: £39 million gain) for associates and joint ventures.

The amounts recognised in the balance sheet are determined as follows:

In Jamaica, the pension scheme holds shares in Carreras Group Ltd (a Group subsidiary) with a fair value of £4 million (2007: £7 million). In Switzerland, a pension scheme owns a property with a fair value of £19 million (2007: £11 million), part of which is occupied by British American Tobacco Switzerland SA.

The amounts recognised in the income statement are as follows:

The above charges are recognised within employee benefit costs in 2008 and 2007 in note 3(a) and include £2 million in 2008 in respect of pension schemes reported as part of the restructuring costs charged in arriving at the profit from operations (note 3(e)).

The movements in scheme liabilities are as follows:

The actuarial gains in both years principally relate to increases in discount rates.

The movements in funded scheme assets are as follows:

The actuarial losses in both years principally relate to the decline in market value of scheme assets.

Contributions to defined benefit schemes are determined after consultation with the respective trustees and actuaries of the individual externally funded schemes, taking into account regulatory requirements. Contributions in 2009 are expected to be £180 million for pension schemes and £9 million for health care schemes, compared to £145 million and £6 million respectively, in 2008.

The movements in the unrecognised funded scheme surpluses are as follows:

Movements in unrecognised scheme surpluses are recognised in the Statement of recognised income and expense.

In 2007, a surplus apportionment scheme in South Africa was approved by the relevant regulatory authority, resulting in the recognition of a previously restricted surplus.

The principal actuarial assumptions (weighted to reflect individual scheme differences) used in the following principal countries are as follows:

For health care inflation in Canada, the assumption is 10 per cent reducing to 5 per cent by 2018.

For the remaining pension schemes, typical assumptions are that real salary increases will be from 0 per cent to 4 per cent per annum and discount rates will be from 1 per cent to 5 per cent above inflation. Pension increases, where allowed for, are generally assumed to be in line with inflation.

For health care inflation in Canada, the assumption is 10 per cent reducing to 5 per cent by 2013.

For the remaining pension schemes, typical assumptions are that real salary increases will be from 2 per cent to 4 per cent per annum and discount rates will be from 2 to 5 per cent above inflation. Pension increases, where allowed for, are generally assumed to be in line with inflation.

Discount rates are determined by reference to normal yields on high-quality corporate bonds at the balance sheet date. For countries where there is not a deep market in such corporate bonds then the yield on government bonds is used.

Mortality assumptions are subject to regular review. In Canada, Germany, Switzerland and the Netherlands the same tables were used for both years. In the UK, for post-retirement mortality assumptions at 31 December 2008 the PXA92 (year of birth) table rated up 3 years for active and deferred members and PXA92 (year of birth) table rated up 2 years for current pensioners, with a 1.2 per cent underpin on future improvements, all with the medium cohort effect, have been used. At 31 December 2007, the PXA92 (B=1965) and PXA92 (B=1935) tables were used, along with the medium cohort effect, however, active and deferred members were rated up by 1 year and current pensioners were rated up by 3 years. In Canada UP94 tables, in Germany Heubeck tables 2005G, in Switzerland EVK 2000 mortality tables and in the Netherlands GBM/V 2005-2050 tables have been used.

Based on the above, the weighted average life expectancy, in years, for mortality tables used to determine benefit obligations is as follows:

Use the scrollbar below to view the full content of this table  

The expected rates of return on scheme assets in the following principal countries are as follows:

For the remaining pension schemes, typical expected long-term real rates of return ranged from 1 per cent to 8 per cent.

For the remaining pension schemes, typical expected long-term real rates of return ranged from 2 per cent to 8 per cent.

For the remaining pension schemes, typical expected long-term real rates of return ranged from 2 per cent to 7 per cent.

Expected rates of return are determined taking into account the current level of expected returns on risk-free investments, the historical level of risk premium associated with other invested assets, and the expectations for future returns on such assets.

The major categories of assets as a percentage of the total fair value of scheme assets are as follows:

Valuation of post-retirement schemes involves judgements about uncertain future events. Sensitivities in respect of the key assumptions used to measure the principal pension schemes as at 31 December 2008 are set out below. These sensitivities show the hypothetical impact of a change in each of the listed assumptions in isolation, with the exception of the sensitivity to inflation which incorporates the impact of certain correlating assumptions such as salary increases. While each of these sensitivities holds all other assumptions constant, in practice such assumptions rarely change in isolation and the impacts may offset to some extent.

The effect on the profit before taxation reflects the impact on current service cost, interest cost and expected return on scheme assets where relevant, for 2009. The effect on scheme assets and liabilities is as at 31 December 2008.

A one percentage point change in health care inflation would have the following effects, which were similar in 2007:

The history of the present value of the scheme liabilities, fair value of the scheme assets, the scheme net deficits and experience gains and losses are as follows: