Critical accounting estimates

The Group’s accounting policies are set out on pages 165 to 173. Certain of these policies, as well as estimates made by management, are considered to be important to an understanding of the Group’s financial condition since they require management to make difficult, complex or subjective judgements and estimates, some of which may relate to matters that are inherently uncertain. The following accounting policies include estimates which are particularly sensitive in terms of judgements and the extent to which estimates are used. Other accounting policies involve significant amounts of judgements and estimates, but the total amounts involved are not significant to the financial statements. Management has discussed the accounting policies and critical accounting estimates with the Board Audit Committee.

Fair value of financial instruments

Some of the Group’s financial instruments are carried at fair value through profit or loss such as those held for trading, designated by management under the fair value option and non-cash flow hedging derivatives.

Other non-derivative financial assets may be designated as available for sale. Available for sale financial investments are initially recognised at fair value and are subsequently held at fair value. Gains and losses arising from changes in fair value of such assets are included as a separate component of equity. The fair value of a financial instrument is the amount at which the instrument could be exchanged in a current transaction between willing parties, other than in a forced or liquidation sale. Financial instruments entered into as trading transactions, together with any associated hedging, are measured at fair value and the resultant profits and losses are included in net trading income, along with interest and dividends arising from long and short positions and funding costs relating to trading activities. Assets and liabilities resulting from gains and losses on financial instruments held for trading are reported gross in trading portfolio assets and liabilities or derivative financial instruments, reduced by the effects of netting agreements where there is an intention to settle net with counterparties.

Valuation methodology

The method of determining the fair value of financial instruments can be analysed into the following categories:

• (a) Unadjusted quoted prices in active markets where the quoted price is readily available and the price represents actual and regularly occurring market transactions on an arm’s length basis.
• (b) Valuation techniques using market observable inputs. Such techniques may include:
•      – using recent arm’s length market transactions;
•      – reference to the current fair value of similar instruments;
•      – discounted cash flow analysis, pricing models or other techniques commonly used by market participants.
• (c) Valuation techniques used above, but which include significant inputs that are not observable. On initial recognition of financial instruments measured using such techniques the transaction price is deemed to provide the best evidence of fair value for accounting purposes.

The valuation techniques in (b) and (c) use inputs such as interest rate yield curves, equity prices, commodity and currency prices/yields, volatilities of underlyings and correlations between inputs. The models used in these valuation techniques are calibrated against industry standards, economic models and to observed transaction prices where available.

The following tables set out the total financial instruments stated at fair value as at 31st December 2007 and those fair values are calculated with valuation techniques using unobservable inputs.

Download table as excel file

			Unobservable
			inputs	Total
			£m	£m
Assets stated at fair value
Trading portfolio assets			4,457	193,691
Financial assets designated at fair value:
– held on own account			16,819	56,629
– held in respect of linked liabilities to customers
under investment contracts			–	90,851
Derivative financial instruments			2,707	248,088
Available for sale financial investments			810	43,072
Total			24,793	632,331
			Unobservable
			inputs	Total
			£m	£m
Liabilities stated at fair value
Trading portfolio liabilities			42	65,402
Financial liabilities designated at fair value			6,172	74,489
Liabilities to customers under investment contracts			–	92,639
Derivative financial instruments			4,382	248,288
Total			10,596	480,818

Various factors influence the availability of observable inputs and these may vary from product to product and change over time. Factors include for example, the depth of activity in the relevant market, the type of product, whether the product is new and not widely traded in the market place, the maturity of market modelling and the nature of the transaction (bespoke or generic).

To the extent that valuation is based on models or inputs that are not observable in the market, the determination of fair value can be more subjective, dependant on the significance of the unobservable input to the overall valuation. Unobservable inputs are determined based on the best information available, for example by reference to similar assets, similar maturities, appropriate proxies, or other analytical techniques.

The effect of changing the assumptions for those financial instruments for which the fair values were measured using valuation techniques that are determined in full or in part on assumptions that are not supported by observable inputs to a range of reasonably possible alternative assumptions, would be to provide a range of £1.2bn (2006: £0.1bn) lower to £1.5bn (2006: £0.1bn) higher than the fair values recognised in the financial statements.

The size of this range will vary over time in response to market volatility, market uncertainty and changes to benchmark proxy relationships of similar assets and liabilities. The calculation of this range is performed on a consistent basis each period.

Further information on the fair value of financial instruments is provided in Note 49 to the accounts.

The following summary sets out those instruments which use inputs where it may be necessary to use valuation techniques as described above.

Corporate bonds

Corporate bonds are generally valued using observable quoted prices or recently executed transactions. Where observable price quotations are not available, the fair value is determined based on cash flow models where significant inputs may include yield curves, bond or single name credit default swap spreads.

Mortgage whole loans

The fair value of mortgage whole loans are determined using observable quoted prices or recently executed transactions for comparable assets. Where observable price quotations or benchmark proxies are not available, fair value is determined using cash flow models where significant inputs include yield curves, collateral specific loss assumptions, asset specific prepayment assumptions, yield spreads and expected default rates.

Commercial mortgage backed securities and asset backed securities

Commercial mortgage backed securities and asset backed securities (ABS) (residential mortgages, credit cards, auto loans, student loans and leases) are generally valued using observable information. Wherever possible, the fair value is determined using quoted prices or recently executed transactions. Where observable price quotations are not available, fair value is determined based on cash flow models where the significant inputs may include yield curves, credit spreads, prepayment rates. Securities that are backed by the residual cash flows of an asset portfolio are generally valued using similar cash flow models.

The fair value of home equity loan bonds are determined using models which use scenario analysis with significant inputs including age, rating, internal grade, and index prices.

Collateralised debt obligations

The valuation of collateralised debt obligations (CDOs) notes is first based on an assessment of the probability of an event of default occurring due to a credit deterioration. This is determined by reference to the probability of event of default occurring and the probability of exercise of contractual rights related to event of default. The notes are then valued by determining appropriate valuation multiples to be applied to the contractual cash flows. These are based on inputs including the prospective cash flow performance of the underlying securities, the structural features of the transaction and the net asset value of the underlying portfolio.

Private equity

The fair value of private equity is determined using appropriate valuation methodologies which, dependent on the nature of the investment, may include discounted cash flow analysis, enterprise value comparisons with similar companies, price:earnings comparisons and turnover multiples. For each investment the relevant methodology is applied consistently over time.

Own credit on financial liabilities

The carrying amount of financial liabilities held at fair value is adjusted to reflect the effect of changes in own credit spreads. As a result, the carrying value of issued notes that have been designated at fair value through profit and loss is adjusted by reference to the movement in the appropriate spreads. The resulting gain or loss is recognised in the income statement.

Derivatives

Derivative contracts can be exchange traded or over the counter (OTC). OTC derivative contracts include forward, swap and option contracts related to interest rates, bonds, foreign currencies, credit standing of reference entities, equity prices, fund levels, commodity prices or indices on these assets.

The fair value of OTC derivative contracts are modelled using a series of techniques, including closed form analytical formulae (such as the Black-Scholes option pricing model) and simulation based models. The choice of model is dependant on factors such as; the complexity of the product, inherent risks and hedging strategy: statistical behaviour of the underlying, and ability of the model to price consistently with observed market transactions. For many pricing models there is no material subjectivity because the methodologies employed do not necessitate significant judgement and the pricing inputs are observed from actively quoted markets, as is the case for generic interest rate swaps and option markets. In the case of more established derivative products, the pricing models used are widely accepted and used by the other market participants.

Significant inputs used in these models may include yield curves, credit spreads, recovery rates, dividend rates, volatility of underlying interest rates, equity prices or foreign exchange rates and, in some cases, correlation between these inputs. These inputs are determined with reference to quoted prices, recently executed trades, independent market quotes and consensus data.

New, long dated or complex derivative products may require a greater degree of judgement in the implementation of appropriate valuation techniques, due to the complexity of the valuation assumptions and the reduced observability of inputs. The valuation of more complex products may use more generic derivatives as a component to calculating the overall value.

Derivatives where valuation involves a significant degree of judgement include:

– Fund derivatives

Fund derivatives are derivatives whose underlyings include mutual funds, hedge funds, indices and multi asset portfolios. They are valued using underlying fund prices, yield curves and available market information on the level of the hedging risk. Some fund derivatives are valued using unobservable information, generally where the level of the hedging risk is not observable in the market. These are valued taking account of risk of the underlying fund or collection of funds, diversification of the fund by asset, concentration by geographic sector, strategy of the fund, size of the transaction and concentration of specific fund managers.

– Commodity derivatives

Commodity derivatives are valued using models where the significant inputs may include interest rate yield curves, commodity price curves, volatility of the underlying commodities and, in some cases, correlation between these inputs, which are generally observable. This approach is applied to base metal, precious metal, energy, power, gas, emissions, soft commodities and freight positions. Due to the significant time span in the various market closes, curves are constructed using differentials to a benchmark curve to ensure that all curves are valued using the dominant market base price.

– Structured credit derivatives

Collateralised synthetic obligations (CSOs)are structured credit derivatives which reference the loss profile of a portfolio of loans, debts or synthetic underlyings. The reference asset can be a corporate credit or an asset backed credit. For CSOs that reference corporate credits an analytical model is used. For CSOs on asset backed underlyings, due to the path dependent nature of a CSO on an amortising portfolio a Monte Carlo simulation is used rather than analytic approximation. The expected loss probability for each reference credit in the portfolio is derived from the single name credit default swap spread curve and in addition, for ABS references, a prepayment rate assumption. A simulation is then used to compute survival time which allows us to calculate the marginal loss over each payment period by reference to estimated recovery rates. Significant inputs include prepayment rates, cumulative default rates, and recovery rates.

Allowances for loan impairment and other credit risk provisions

Allowances for loan impairment represent management’s estimate of the losses incurred in the loan portfolios as at the balance sheet date. Changes to the allowances for loan impairment and changes to the provisions for undrawn contractually committed facilities and guarantees provided are reported in the consolidated income statement as part of the impairment charge. Provision is made for undrawn loan commitments and similar facilities if it is probable that the facility will be drawn and result in recognition of an asset at an amount less than the amount advanced.

Critical accounting estimates

Within the retail and small businesses portfolios, which comprise large numbers of small homogeneous assets with similar risk characteristics where credit scoring techniques are generally used, statistical techniques are used to calculate impairment allowances on a portfolio basis, based on historical recovery rates and assumed emergence periods. These statistical analyses use as primary inputs the extent to which accounts in the portfolio are in arrears and historical information on the eventual losses encountered from such delinquent portfolios. There are many such models in use, each tailored to a product, line of business or customer category. Judgement and knowledge is needed in selecting the statistical methods to use when the models are developed or revised. The impairment allowance reflected in the financial statements for these portfolios is therefore considered to be reasonable and supportable. The impairment charge reflected in the income statement for these portfolios is £1,605m (2006: £1,809m) and amounts to 70% (2006: 87%) of the total impairment charge on loans and advances in 2007.

For larger accounts, impairment allowances are calculated on an individual basis and all relevant considerations that have a bearing on the expected future cash flows are taken into account, for example, the business prospects for the customer, the realisable value of collateral, the Group’s position relative to other claimants, the reliability of customer information and the likely cost and duration of the work-out process. The level of the impairment allowance is the difference between the value of the discounted expected future cash flows (discounted at the loan’s original effective interest rate), and its carrying amount. Subjective judgements are made in the calculation of future cash flows. Furthermore, judgements change with time as new information becomes available or as work-out strategies evolve, resulting in frequent revisions to the impairment allowance as individual decisions are taken. Changes in these estimates would result in a change in the allowances and have a direct impact on the impairment charge. The impairment charge reflected in the financial statements in relation to larger accounts is £701m (2006: £265m) or 30% (2006: 13%) of the total impairment charge on loans and advances in 2007. Further information on impairment allowances is set out on pages 100 to 101.

Goodwill

Management have to consider at least annually whether the current carrying value of goodwill is impaired. The first step of the impairment review process requires the identification of independent cash generating units, by dividing the Group business into as many largely independent income streams as is reasonably practicable. The goodwill is then allocated to these independent units. The first element of this allocation is based on the areas of the business expected to benefit from the synergies derived from the acquisition. The second element reflects the allocation of the net assets acquired and the difference between the consideration paid for those net assets and their fair value. This allocation is reviewed following business reorganisation. The carrying value of the unit, including the allocated goodwill, is compared to its fair value to determine whether any impairment exists. If the fair value of a unit is less than its carrying value, goodwill will be impaired. Detailed calculations may need to be carried out taking into consideration changes in the market in which a business operates (e.g. competitive activity, regulatory change). In the absence of readily available market price data this calculation is based upon discounting expected pre-tax cash flows at a risk adjusted interest rate appropriate to the operating unit, the determination of both of which requires the exercise of judgement. The estimation of pre-tax cash flows is sensitive to the periods for which detailed forecasts are available and to assumptions regarding the long term sustainable cash flows. While forecasts are compared with actual performance and external economic data, expected cash flows naturally reflect management’s view of future performance. The most significant amounts of goodwill relate to the Absa and Woolwich acquisitions. The goodwill impairment testing performed in 2007 indicated that none of the goodwill was impaired. Management believes that reasonable changes in key assumptions used to determine the recoverable amounts of Absa and Woolwich goodwill would not result in impairment.

Intangible assets

Intangible assets that derive their value from contractual customer relationships or that can be separated and sold and have a finite useful life are amortised over their estimated useful life. Determining the estimated useful life of these finite life intangible assets requires an analysis of circumstances, and judgement by the Bank’s management. At each balance sheet date, or more frequently when events or changes in circumstances dictate, intangible assets are assessed for indications of impairment. If indications are present, these assets are subject to an impairment review. The impairment review comprises a comparison of the carrying amount of the asset with its recoverable amount: the higher of the assets’ or the cash-generating unit’s net selling price and its value in use. Net selling price is calculated by reference to the amount at which the asset could be disposed of in a binding sale agreement in an arms-length transaction evidenced by an active market or recent transactions for similar assets. Value in use is calculated by discounting the expected future cash flows obtainable as a result of the asset’s continued use, including those resulting from its ultimate disposal, at a market-based discount rate on a pre-tax basis. The most significant amounts of intangible assets relate to the Absa acquisition.

Retirement benefit obligations

The Group provides pension plans for employees in most parts of the world. Arrangements for staff retirement benefits vary from country to country and are made in accordance with local regulations and customs. For defined contribution schemes, the pension cost recognised in the profit and loss account represents the contributions payable to the scheme. For defined benefit schemes, actuarial valuation of each of the scheme’s obligations using the projected unit credit method and the fair valuation of each of the scheme’s assets are performed annually in accordance with the requirements of IAS 19.

The actuarial valuation is dependent upon a series of assumptions, the key ones being interest rates, mortality, investment returns and inflation. Mortality estimates are based on standard industry and national mortality tables, adjusted where appropriate to reflect the Group’s own experience. The returns on fixed interest investments are set to market yields at the valuation date (less an allowance for risk) to ensure consistency with the asset valuation. The returns on UK and overseas equities are based on the long-term outlook for global equities at the calculation date having regard to current market yields and dividend growth expectations. The inflation assumption reflects long-term expectations of both earnings and retail price inflation.

The difference between the fair value of the plan assets and the present value of the defined benefit obligation at the balance sheet date, adjusted for any historic unrecognised actuarial gains or losses and past service cost, is recognised as a liability in the balance sheet. An asset arising, for example, as a result of past over-funding or the performance of the plan investments, is recognised to the extent that it does not exceed the present value of future contribution holidays or refunds of contributions. To the extent that any unrecognised gains or losses at the start of the measurement year in relation to any individual defined benefit scheme exceed 10% of the greater of the fair value of the scheme assets and the defined benefit obligation for that scheme, a proportion of the excess is recognised in the income statement.

The Group’s IAS 19 pension surplus across all pension and post-retirement schemes as at 31st December 2007 was a surplus of £393m (2006: £817m deficit). This comprises net recognised liabilities of £1,501m (2006: £1,719m) and unrecognised actuarial gains of £1,894m (2006: £902m). The net recognised liabilities comprises retirement benefit liabilities of £1,537m (2006: £1,807m) relating to schemes that are in deficit, and assets of £36m (2006: £88m) relating to schemes that are in surplus. The Group’s IAS 19 pension surplus in respect of the main UK scheme as at 31st December 2007 was £668m (2006: £495m deficit). The estimated actuarial funding position of the main UK pension scheme as at 31st December 2007, estimated from the triennial valuation in 2004, was a surplus of £1,200m (2006: £1,300m). Cash contributions to the main UK scheme were £355m (2006: £351m).

Further information on retirement benefit obligations, including assumptions is set out in Note 30 to the accounts.