RBS 2012 Annual Report Download - page 248

246
Business review Risk and balance sheet management continued
Market risk: Risk measurement continued
Additionally, Group Risk Analytics (GRA) assess the appropriateness of
all new or amended models prior to their introduction. Existing approved
models are re-assessed on a periodic basis to ensure they remain fit-for-
purpose, for example, following significant market developments or
portfolio changes. The models required to be reviewed by GRA (in
relation to market risk) include VaR, SVaR, IRC, APR and economic
capital. The independent validation review process will consider some or
all of the following areas as appropriate:
x Test and challenge the logical and conceptual soundness of the
methodology;
x The assumptions underlying the model will be tested, where feasible
against actual behaviour. The validation report will judge the
reasonableness and stability of the assumptions and specify which
assumptions, if any, should be routinely monitored in production;
x Compare model results with independent model replication;
x Compare outcome with results from alternative methods;
x Test parameter selection and calibration;
x Ensure that model outputs are sufficiently conservative in areas
where there is significant model uncertainty;
x Confirm applicability of tests for accuracy, and stability; recalculate;
and ensure that results are robust; and
x Ensure appropriate factor sensitivity analysis has been performed
and documented.
Stress testing*
The Group undertakes daily stress testing to identify the potential losses
in excess of VaR. Stress testing is used to calculate a range of trading
book exposures which result from severe and extreme market events.
Stress testing measures the impact of exceptional changes in market
rates and prices on the fair value of the Group’s trading and available-for-
sale portfolios. The Group calculates sensitivity analysis, historical stress
tests and bottom-up stress testing.
Sensitivity analysis measures the sensitivity of the current portfolio of
positions to defined market risk factor movements. These stresses are of
a smaller magnitude compared to historical or bottom-up stress testing
and are subject to the Group Market Risk limit framework.
Historical stress tests calculate the changes in the portfolio valuations
that would be generated if the extreme market movements that occurred
during significant historical market events were repeated. Historical stress
tests also form part of the Group Market Risk limit framework.
Bottom-up stress testing requires analysis of the market risk exposures
by risk factors and different liquidity horizons, to identify the key risks.
Stresses for these risks are then designed following consultation with risk
managers, economists and front office. The tests may be based on an
economic scenario that is translated into risk factor shocks by an
economist or by risk managers and front office as a means of assessing
the vulnerabilities of their book.
The Global Market Risk Stress Testing Committee reviews and discusses
all matters relating to market risk stress testing. Stress test exposures are
discussed with senior management and relevant information is reported
to the Group Risk Committee, the ERF and the Board. Breaches in the
Group’s market risk stress testing limits are monitored and reported.
Reverse stress testing is designed to assess the plausibility of scenarios
derived by stressing market risk factors until the loss reaches a given
threshold. Market Risk contributes to the firm wide, cross risk reverse
stress tests.
In addition to VaR and stress testing, the Group calculates a wide range
of sensitivity and position risk measures, for example interest rate ladders
or option revaluation matrices. These measures provide valuable
additional controls, often at individual desk or strategy level.
Pricing models*
Pricing models are developed and owned by the front office. Where
pricing models are used as the basis of books and records valuations,
they are subject to oversight and approval by asset level modelled
product review committees. These committees prioritise models for
independent validation by GRA taking into consideration both the
materiality of risk booked against the model and an assessment of the
degree of model risk (i.e. valuation uncertainty arising from choice of
modelling assumptions). GRA review aims to quantify model risk by
comparing model outputs against those of alternative independently
developed models, the results of which are used by Market Risk to inform
risk limits and by Finance to inform model reserves.
Marking-to-market
To ensure that the risks associated with trading activity are reflected in
the financial and management statements, assets and liabilities in the
trading book are measured at their fair value. Any profits or losses on the
revaluation of positions are recognised in the income statement on a daily
basis.
The fair value is the amount at which the instrument could be exchanged
in a current transaction between willing parties. The fair values are
determined following IAS 39 ‘Financial Instruments: Recognition and
Measurement’ guidance, which requires banks to use quoted market
prices or valuation techniques (models) that make the maximum use of
observable inputs.
When marking-to-market using a model, the valuation methodologies
must be approved by all stakeholders (trading, finance, market risk,
model development and model review) prior to use for profit and loss and
risk management purposes.
Traders are responsible for marking-to-market their trading book
positions on a daily basis. Traders can either:
x directly mark a position with a price (e.g. spot foreign exchange); or
x indirectly mark a position through the marking of inputs to an
approved model, which will in turn generate a price.
*unaudited