Barclays 2003 Annual Report Download - page 63

Barclays PLC Annual Report 2003 61
Risk control
In Barclays Capital, the Head of Market Risk is responsible for the market
risk governance and control framework. Day-to-day responsibility for
managing exposure to market risk lies with the senior management of
Barclays Capital, supported by the Global Market Risk Management Unit
that operates independently of the trading areas. Daily DVaR utilisation
reports are produced across the main business areas and the five main
risk factor categories, namely interest rate, credit spread, foreign
exchange, equity and commodity risk.
Any DVaR excess at the business level, risk factor level or total level,
along with the relevant background information and proposed way
forward, is reported to the senior management of Barclays Capital and
the Group Market Risk Director. The Group Market Risk Director
presents these DVaR excesses to the Group Risk Oversight Committee.
As DVaR does not provide a direct indication of the potential size of
losses that could arise in extreme conditions, Barclays Capital uses a
number of complementary techniques for controlling market risk. These
include revenue loss triggers and stress tests. The latter are based on
both historical and hypothetical extreme movements of market prices
and are reviewed as part of the detailed market risk presentation at the
fortnightly Traded Products Risk Review meetings. The attendees at this
meeting include senior managers from Barclays Capital and Group
Central Functions.
If the potential loss indicated by a stress test exceeds an agreed trigger
level, then the positions captured by the stress test are reviewed and
discussed by Barclays Capital Market Risk and the respective Business
Head(s). The minutes of the discussion, including the merits of the
position and the appropriate course of action, are then sent to the
Group Market Risk Director.
Risk measurement
Barclays Capital calculates DVaR using the historical simulation method
with a historical sample of two years. As stated above, the calculation
assumes a one-day holding period and is performed to the 98% level
of confidence.
The DVaR methodology allows the interest rate risk (due to changes
in the government interest rates) to be measured separately from
credit spread risk (due to changes in credit spreads). The credit spread
is the premium for holding a non-government investment, and is the
difference between the total interest rate and the appropriate
government interest rate, for the same maturity.
In total, the DVaR methodology maps interest rate risk into eight
categories. These are government, interest rate swaps and six credit
grades for non-government exposures. This categorisation allows the
measurement process to discriminate between the market risk of
holding bonds with different credit qualities, for example AAA grade
securities as against non-investment grade securities. In particular,
it shows the effectiveness of hedging strategies such as shorting
government bonds or swaps against non-government bond portfolios.
The DVaR numbers shown in the table below are all based on the above
methodology.
Analysis of market risk exposures
Barclays Capital’s market risk exposure increased in 2003. The credit
businesses incurred additional credit spread risk, primarily due to
growing client business in corporate bonds and credit derivatives.
The daily average, maximum and minimum values of DVaR were
estimated as below.
Barclays Capital DVaR: Summary table for 2003 and 2002
Twelve months to Twelve months to
31st December 2003 31st December 2002
Average High(a) Low(a) Average High(a) Low(a)
£m £m £m £m £m £m
Interest rate risk 21.0 34.1 13.6 21.7 34.5 10.0
Credit spread risk 16.2 29.2 8.9 9.4 12.5 6.0
Foreign exchange risk 2.3 5.0 1.0 2.9 4.4 1.9
Equities risk 2.6 4.9 1.5 3.6 5.4 2.1
Commodities risk 4.4 7.0 2.2 1.8 3.3 0.8
Diversification effect (20.6) n/a n/a (16.2) n/a n/a
Total DVaR(b) 25.9 38.6 17.6 23.2 35.7 13.4
Notes
(a) The high (and low) DVaR figures reported for each category did not necessarily occur on the same day as the high (and low) DVaR reported as a
whole. Consequently, a diversification effect number for the high (and low) DVaR figures would not be meaningful and is therefore omitted from
the above table.
(b) The year-end total DVaR for 2003 was £37.2m (2002: £25.8m).