Barclays 2009 Annual Report Download - page 304

302 Barclays PLC Annual Report 2009 www.barclays.com/annualreport09
Notes to the accounts
For the year ended 31st December 2009
continued
50 Fair value of financial instruments continued
Other credit products
Sensitivity of synthetic CDOs is determined by using valuations based on different assumptions for recovery and tranche mapping. The sensitivity
represents the impact of the application of different modelling assumptions. These model assumptions change the distribution of losses on the trades.
The sensitivity of valuations of the illiquid CDS portfolio is determined by applying a +/- 0.5% stress to the DV01 for each underlying reference asset.
The stress is based upon the average bid offer spreads observed in the market for similar CDS.
Derivative exposure to Monoline insurers
The main unobservable input in the valuation of the derivative exposure to Monoline insurers is the credit quality of the Monoline insurers. The approach
to determine downside sensitivity is dependent on the credit quality of the Monoline insurer. For higher quality Monoline insurers a shift in internal credit
ratings has been applied. For lower quality Monoline insurers the impact has been assessed by a shift to default and recovery rates. The recovery rate was
set to 10% and risk of default to 100%. To assess the upside risk to Monoline insurers, the underlying assets were flexed by 3%, based on the average
monthly move for the LCDX index over the preceding 12-month period.
Non-asset backed debt instruments
The sensitivity on convertible bonds, is determined by applying a +/- 1% shift for each underlying value. The shift is based upon the bid offer spreads
observed in the market for similar bonds.
The sensitivity on corporate bonds portfolio is determined by applying a +/- 1% shift for each underlying value. The shift is based upon the average bid
offer spreads observed in the market for similar bonds.
The sensitivity for fixed and floating rate notes is calculated using a +/- 1% shift in credit spreads.
Equity products
The sensitivity is estimated based on the statistical spread of consensus data services either directly or through proxies. The estimate has been calculated
using 1 standard deviation of the consensus returns.
Private equity
The relevant valuation models are sensitive to each of the key assumptions, such as projected future earnings, comparator multiples, marketability
discounts and discount rates. Valuation sensitivity is estimated by flexing such assumptions to reasonable alternative levels and determining the impact
to the outcome of the valuation, for example comparator multiples for a range of similar companies.
Fund and fund-linked products
The sensitivity measure is based on observing the largest monthly move in the main hedge fund index (HFRX) over the prior two-year period.
FX products
The sensitivity is based on the statistical spread of consensus data services. The estimate has been calculated using consensus data service statistical
standard deviation and this represents 2 standard deviations of the mid correlations.
Interest rate products
For inflation products, the sensitivity is calculated by stressing the correlation parameters. The sensitivity was determined by applying a +/- 8% shift to
the correlation risks and this was based on the historical observation of correlation levels over the preceeding month.
For base rate derivatives, the sensitivity is based on bid offer spreads of base rates swaps. The sensitivity was determined by applying a +/- 0.125% shift
to the PV01 for each underlying position.
Commodity products
The sensitivity is determined by applying values based on historic variability over two years. The estimate has been calculated using data for shortdated
forward volatility curves to generate a best and worst case likely scenario. The sensitivities are based on a 25% probability of occurrence over two years,
and taking 50% of the difference between the best and worst multiplied by exposure to the relevant risk factor.
Other
The sensitivity for fixed rate loans is calculated using a +/- 1% shift in credit spreads.
Valuation control framework
The independent price verification process is a key control in ensuring the material accuracy of valuation. Price verification procedures cover all fair value
positions. Data sources that are most representative of the market and readily available are used. The data sources are assessed against the following
characteristics: independence; reliability; consistency between sources and evidence that it represents executable levels. This control process assists in
the determination of whether market levels represent forced transactions which should not be considered in the assessment of fair value.