Final Results - Part 8 of 13
Risk and balance sheet management (continued)
Market risk
Market risk arises from changes in interest rates, foreign currency, credit spread, equity prices and risk related factors such as market volatilities. The Group manages market risk centrally within its trading and non-trading portfolios through a comprehensive market risk management framework. This framework includes limits based on, but not limited to, value-at-risk (VaR), stress testing, position and sensitivity analyses.
At the Group level, the risk appetite is expressed in the form of a combination of VaR, sensitivity and stress testing limits. VaR is a technique that produces estimates of the potential change in the market value of a portfolio over a specified time horizon at given confidence levels. For internal risk management purposes, the Group's VaR assumes a time horizon of one trading day and a confidence level of 99%. The Group's VaR model is based on a historical simulation model, utilising data from the previous two years.
The VaR disclosure is broken down into trading and non-trading portfolios. Trading VaR relates to the main trading activities of the Group and non-trading VaR reflects reclassified assets, money market business and the management of internal funds flow within the Group's businesses.
As part of the ongoing review and analysis of the suitability of the Group's VaR model, a methodology enhancement to the ABS VaR was approved and incorporated into the regulatory model in 2010. The credit crisis in 2007-2009 caused large price changes for some structured bonds and the spread based approach to calculating VaR for these instruments started to give inaccurate risk levels, particularly for bonds trading at a significant discount to par. The methodology enhancement harmonised the VaR approach in the US and Europe by replacing the absolute spread-based approach with a more reliable and granular relative price-based mapping scheme. The enhancement better reflects the risk in the context of position changes, downgrades and vintages as well as improving the differentiation between prime, Alt-A and sub-prime exposures.
The VaR model has been approved by the FSA to calculate regulatory capital for the trading book. The approval covers general market risk in interest rate, foreign exchange, equity and limited commodity products and specific risk in interest rate and equity products.
As the VaR model is an important market risk measurement and control tool and is used for determining a significant component of the market risk capital, it is regularly assessed. The main approach employed is the technique known as back-testing which counts the number of days when a loss (as defined by the FSA), exceeds the corresponding daily VaR estimate, measured at a 99% confidence interval. The FSA categorises a VaR model as green, amber or red. A green model is consistent with a good working model and is achieved for models that have four or less back-testing exceptions in a 12 month period. For the Group's trading book, a green model status was maintained throughout 2010.
Risk and balance sheet management (continued)
Market risk (continued)
The Group's VaR should be interpreted in the light of the limitations of the methodology used, as follows:
|
· |
Historical simulation VaR may not provide the best estimate of future market movements. It can only provide a prediction of the future based on events that occurred in the 500 trading day time series. Therefore, events more severe than those in the historical data series cannot be predicted. |
|
|
|
|
· |
The use of a 99% confidence level does not reflect the extent of potential losses beyond that percentile. |
|
|
|
|
· |
The use of a one day time horizon will not fully capture the profit and loss implications of positions that cannot be liquidated or hedged within one day. |
|
|
|
|
· |
The Group computes the VaR of trading portfolios at the close of business. Positions may change substantially during the course of the trading day and intra-day profits and losses will be incurred. |
These limitations mean that the Group cannot guarantee that profits or losses will not exceed the VaR.
Risk and balance sheet management (continued)
Market risk (continued)
The table below details the Group's trading portfolio, segregated by type of market risk exposure, and between Core and Non-Core, Counterparty Exposure Management (CEM) and Core excluding CEM.
|
|
Quarter ended |
|
Year ended |
||||||||||||||||
|
|
31 December 2010 |
|
30 September 2010 |
|
31 December 2010 |
|
31 December 2009 |
||||||||||||
|
|
Average |
Period end |
Maximum |
Minimum |
|
Average |
Period end |
Maximum |
Minimum |
|
Average |
Period end |
Maximum |
Minimum |
|
Average |
Period end |
Maximum |
Minimum |
|
Trading |
£m |
£m |
£m |
£m |
|
£m |
£m |
£m |
£m |
|
£m |
£m |
£m |
£m |
|
£m |
£m |
£m |
£m |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Interest rate |
64.0 |
57.0 |
83.0 |
47.6 |
|
50.5 |
74.3 |
74.3 |
38.6 |
|
51.6 |
57.0 |
83.0 |
32.5 |
|
57.0 |
50.5 |
112.8 |
28.1 |
|
Credit spread |
134.4 |
133.4 |
196.1 |
110.2 |
|
214.0 |
190.8 |
243.2 |
174.5 |
|
166.3 |
133.4 |
243.2 |
110.2 |
|
148.3 |
174.8 |
231.2 |
66.9 |
|
Currency |
15.2 |
14.8 |
25.6 |
8.4 |
|
15.4 |
16.7 |
26.2 |
9.3 |
|
17.9 |
14.8 |
28.0 |
8.4 |
|
17.9 |
20.7 |
35.8 |
9.2 |
|
Equity |
10.1 |
10.9 |
15.2 |
4.7 |
|
7.2 |
5.4 |
17.9 |
2.7 |
|
9.5 |
10.9 |
17.9 |
2.7 |
|
13.0 |
13.1 |
23.2 |
2.7 |
|
Commodity |
7.9 |
0.5 |
18.1 |
0.5 |
|
8.9 |
13.8 |
15.7 |
3.2 |
|
9.5 |
0.5 |
18.1 |
0.5 |
|
14.3 |
8.9 |
32.1 |
6.5 |
|
Diversification |
|
(75.6) |
|
|
|
|
(119.2) |
|
|
|
|
(75.6) |
|
|
|
|
(86.1) |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Total |
154.3 |
141.0 |
191.5 |
110.8 |
|
213.1 |
181.8 |
252.1 |
156.1 |
|
168.5 |
141.0 |
252.1 |
103.0 |
|
155.2 |
181.9 |
229.0 |
76.8 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Core |
99.2 |
101.2 |
121.0 |
58.3 |
|
123.8 |
115.0 |
153.4 |
99.6 |
|
103.6 |
101.2 |
153.4 |
58.3 |
|
101.5 |
127.3 |
137.8 |
54.8 |
|
CEM |
49.1 |
54.6 |
64.2 |
38.7 |
|
74.7 |
73.0 |
82.4 |
70.4 |
|
53.3 |
54.6 |
82.4 |
30.3 |
|
29.7 |
38.6 |
41.3 |
11.5 |
|
Core excluding CEM |
81.3 |
78.7 |
102.8 |
54.2 |
|
84.2 |
78.4 |
96.5 |
72.0 |
|
82.8 |
78.7 |
108.7 |
53.6 |
|
86.7 |
97.4 |
128.5 |
54.9 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Non-Core |
105.5 |
101.4 |
119.7 |
92.3 |
|
135.7 |
101.8 |
169.4 |
97.5 |
|
105.7 |
101.4 |
169.4 |
63.2 |
|
86.3 |
84.8 |
162.1 |
29.3 |
Risk and balance sheet management (continued)
Market risk (continued)
Key points
|
· |
The Group's period end VaR reduced as the exceptional volatility of the market data from the period of the financial crisis dropped out of the 500 days of time series data used in the VaR calculation. The credit spread VaR was particularly impacted as a result of this effect. |
|
|
|
|
· |
The Group's maximum and average credit and Non-Core VaR were higher in 2010, than in 2009 due to Non-Core exiting several highly structured positions which due to their complexity and layering, required unwinding with different counterparts over different periods. The timing of the unwind has led to increased VaR, until the exit was completed in October and the VaR then reduced back to the levels held earlier in the year. |
|
|
|
|
· |
CEM VaR was greater in 2010 than 2009 due to the novation of counterparty risk hedging trades from RBS N.V. to RBS plc. For RBS N.V. there is no local regulatory requirement for counterparty hedges to be included in VaR, as they are treated on a standardised basis but on novation to CEM in RBS plc, under UK regulatory requirements, the trades were captured by the VaR model resulting in an increase in VaR. |
|
|
|
|
· |
CEM trading VaR also increased as a consequence of the implementation of a discounting approach based on the real funding cost for the collateralised derivatives. |
|
|
|
|
· |
Commodity VaR decreased during the year since a significant part of the Group's interest in RBS Sempra Commodities JV. was sold during the year. |
Risk and balance sheet management (continued)
Market risk:
GBM traded revenue
Please see attached URL
http://www.rns-pdf.londonstockexchange.com/rns/7729B_-2011-2-24.pdf
Key points
|
· |
The average daily revenue earned from GBM's trading, balance sheet management and other trading activities in 2010 was £25.4 million, compared with £37.8 million in 2009. The standard deviation of these daily revenues was £22.0 million compared with £32.3 million in 2009. The standard deviation measures the variation of daily revenues above the mean value of those revenues. |
|
|
|
|
· |
An analysis of the frequency distribution of daily revenue shows that there were 22 days with negative revenue during 2010 compared with 16 days in 2009. The most frequent result is daily revenue of between £25 million and £30 million with 37 occurrences in 2010 compared with 26 occurrences in 2009. |
|
|
|
|
· |
The effect of any month end adjustments, not attributable to a specific daily market move, is spread evenly over the days in the month in question. |
|
|
|
|
· |
The graph of daily revenues for 2010 shows a narrower distribution of revenues compared to 2009. |
Risk and balance sheet management (continued)
Market risk (continued)
The table below details the Group's non-trading VaR portfolio, excluding Structured Credit Portfolios (SCP) and loans and receivables (LAR), segregated by type of market risk exposure and between Core and Non-Core.
|
|
Quarter ended |
|
Year ended |
||||||||||||||||
|
|
31 December 2010 |
|
30 September 2010 |
|
31 December 2010 |
|
31 December 2009 |
||||||||||||
|
|
Average |
Period end |
Maximum |
Minimum |
|
Average |
Period end |
Maximum |
Minimum |
|
Average |
Period end |
Maximum |
Minimum |
|
Average |
Period end |
Maximum |
Minimum |
|
Non-trading VaR |
£m |
£m |
£m |
£m |
|
£m |
£m |
£m |
£m |
|
£m |
£m |
£m |
£m |
|
£m |
£m |
£m |
£m |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Interest rate |
8.0 |
10.4 |
10.8 |
5.3 |
|
9.1 |
4.4 |
20.5 |
4.4 |
|
8.7 |
10.4 |
20.5 |
4.4 |
|
13.0 |
13.9 |
26.3 |
7.7 |
|
Credit spread |
17.0 |
16.1 |
21.8 |
15.4 |
|
22.6 |
19.4 |
26.4 |
19.4 |
|
32.0 |
16.1 |
101.2 |
15.4 |
|
81.7 |
100.3 |
131.5 |
39.7 |
|
Currency |
2.3 |
3.0 |
3.7 |
1.3 |
|
2.8 |
2.0 |
6.1 |
1.5 |
|
2.1 |
3.0 |
7.6 |
0.3 |
|
1.4 |
0.6 |
7.0 |
0.2 |
|
Equity |
2.9 |
3.1 |
4.6 |
0.3 |
|
0.4 |
0.4 |
1.7 |
0.3 |
|
1.2 |
3.1 |
4.6 |
0.2 |
|
3.3 |
2.2 |
5.8 |
1.6 |
|
Diversification |
|
(15.9) |
|
|
|
|
(6.8) |
|
|
|
|
(15.9) |
|
|
|
|
(20.4) |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Total |
16.2 |
16.7 |
21.3 |
13.7 |
|
23.8 |
19.4 |
29.1 |
19.4 |
|
30.9 |
16.7 |
98.0 |
13.7 |
|
80.4 |
96.6 |
126.9 |
46.8 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Core |
15.6 |
15.6 |
21.3 |
12.8 |
|
23.6 |
19.3 |
29.3 |
19.3 |
|
30.5 |
15.6 |
98.1 |
12.8 |
|
78.4 |
95.9 |
126.9 |
46.8 |
|
Non-Core |
2.8 |
2.8 |
4.1 |
0.2 |
|
0.7 |
0.3 |
2.0 |
0.2 |
|
1.3 |
2.8 |
4.1 |
0.2 |
|
3.5 |
1.9 |
16.9 |
- |
Key points
|
· |
The non-traded credit spread, Core and total VaR have decreased significantly due to the implementation of the relative price-based mapping scheme in the VaR methodology discussed above and the sale of available-for-sale securities in the US mortgage business. |
|
· |
The business model for the US mortgage business has focussed its activity on client facilitation flow trading during 2010. This has encompassed the disposal of a large portfolio of illiquid available-for-sale securities that were sold throughout the year, resulting in the non-traded VaR reducing. In parallel, the risk management of the business has been significantly enhanced to ensure that the business remains focussed on client facilitation flow trading of liquid assets. Tools have been implemented to monitor the liquidity of trading volumes, asset aged inventory controls have been tightened and granular asset concentration risk limits imposed, to complement the existing value-at-risk and stress testing market risk frameworks. |
VaR is not always the most appropriate measure of risk for assets in the non-trading book, particularly for those in Non-Core which will diminish over time as the asset inventory is sold down. In order to better represent the risk of the non-traded portfolios, the table above analyses the VaR for the non-trading portfolios but excludes SCP in Non-Core. These assets are shown separately on a drawn notional and fair value basis by maturity profile and asset class and are managed on both an asset and RWA basis. Also excluded from the non-traded VaR are the LAR products that are managed within the credit risk management framework. Consequently, these portfolios have been excluded from non-trading VaR and prior period data has been revised accordingly.
Risk and balance sheet management (continued)
Market risk: Structured credit portfolio (continued)
|
|
Drawn notional |
|
Fair value |
||||||||
|
|
CDOs |
CLOs |
MBS (1) |
Other ABS |
Total |
|
CDOs |
CLOs |
MBS (1) |
Other ABS |
Total |
|
|
£m |
£m |
£m |
£m |
£m |
|
£m |
£m |
£m |
£m |
£m |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
31 December 2010 |
|
|
|
|
|
|
|
|
|
|
|
|
1-2 years |
- |
- |
- |
47 |
47 |
|
- |
- |
- |
42 |
42 |
|
2-3 years |
85 |
19 |
44 |
98 |
246 |
|
81 |
18 |
37 |
91 |
227 |
|
3-4 years |
- |
41 |
20 |
205 |
266 |
|
- |
37 |
19 |
191 |
247 |
|
4-5 years |
16 |
- |
- |
- |
16 |
|
15 |
- |
- |
- |
15 |
|
5-10 years |
98 |
466 |
311 |
437 |
1,312 |
|
87 |
422 |
220 |
384 |
1,113 |
|
>10 years |
412 |
663 |
584 |
550 |
2,209 |
|
161 |
515 |
397 |
367 |
1,440 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
611 |
1,189 |
959 |
1,337 |
4,096 |
|
344 |
992 |
673 |
1,075 |
3,084 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
30 September 2010 |
|
|
|
|
|
|
|
|
|
|
|
|
1-2 years |
- |
- |
- |
58 |
58 |
|
- |
- |
- |
50 |
50 |
|
2-3 years |
84 |
19 |
46 |
66 |
215 |
|
79 |
18 |
34 |
63 |
194 |
|
3-4 years |
- |
35 |
29 |
211 |
275 |
|
- |
31 |
27 |
183 |
241 |
|
4-5 years |
19 |
7 |
6 |
57 |
89 |
|
17 |
7 |
4 |
52 |
80 |
|
5-10 years |
99 |
366 |
404 |
485 |
1,354 |
|
86 |
324 |
265 |
414 |
1,089 |
|
>10 years |
519 |
793 |
591 |
548 |
2,451 |
|
177 |
627 |
379 |
368 |
1,551 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
721 |
1,220 |
1,076 |
1,425 |
4,442 |
|
359 |
1,007 |
709 |
1,130 |
3,205 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
30 June 2010 |
|
|
|
|
|
|
|
|
|
|
|
|
1-2 years |
- |
- |
- |
67 |
67 |
|
- |
- |
- |
61 |
61 |
|
2-3 years |
75 |
20 |
43 |
85 |
223 |
|
70 |
18 |
31 |
80 |
199 |
|
3-4 years |
30 |
37 |
19 |
298 |
383 |
|
23 |
32 |
18 |
239 |
312 |
|
4-5 years |
20 |
11 |
38 |
59 |
128 |
|
17 |
10 |
33 |
53 |
113 |
|
5-10 years |
90 |
439 |
394 |
548 |
1,470 |
|
80 |
390 |
255 |
455 |
1,180 |
|
>10 years |
624 |
1,004 |
689 |
607 |
2,925 |
|
233 |
810 |
420 |
387 |
1,850 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
839 |
1,511 |
1,183 |
1,664 |
5,196 |
|
423 |
1,260 |
757 |
1,275 |
3,715 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
31 December 2009 |
|
|
|
|
|
|
|
|
|
|
|
|
1-2 years |
- |
- |
- |
81 |
81 |
|
- |
- |
- |
68 |
68 |
|
2-3 years |
40 |
- |
- |
19 |
59 |
|
24 |
- |
- |
18 |
42 |
|
3-4 years |
19 |
18 |
42 |
99 |
178 |
|
16 |
17 |
31 |
76 |
140 |
|
4-5 years |
17 |
47 |
36 |
332 |
432 |
|
3 |
41 |
29 |
275 |
348 |
|
5-10 years |
107 |
685 |
424 |
521 |
1,737 |
|
90 |
594 |
251 |
394 |
1,329 |
|
>10 years |
594 |
1,114 |
820 |
573 |
3,101 |
|
193 |
896 |
468 |
325 |
1,882 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
777 |
1,864 |
1,322 |
1,625 |
5,588 |
|
326 |
1,548 |
779 |
1,156 |
3,809 |
Note:
|
(1) |
Mortgage-backed securities (MBS) include sub-prime residential mortgage-backed securities (RMBS) with a notional amount of £471 million (30 September 2010 - £477 million; 30 June 2010 - £562 million; 31 December 2009 - £682 million) and a fair value of £329 million (30 September 2010 - £316 million; 30 June 2010 - £350 million; 31 December 2009 - £415 million), all with residual maturities of greater than 10 years. |
The SCP is within Non-Core. The risk on this portfolio is not measured or disclosed using VaR, as the Group believes this is not an appropriate tool for the banking book portfolio comprising of illiquid debt securities. The main driver of the reduction in drawn notional is the asset sales from a portfolio within an unwound securitisation arbitrage conduit. The impact of disposals on portfolio fair value has been partially offset by an increase in residual average price to 75% (2009 - 68%).
This information is provided by RNS