JP Morgan Chase 2008 Annual Report Download - page 115

JPMorgan Chase & Co./ 2008 Annual Report 113
Trading VaR includes substantially all trading activities in IB.
Beginning in the fourth quarter of 2008, the credit spread sensitivi-
ties of certain mortgage products were included in trading VaR. This
change had an insignificant net impact on the average fourth quarter
2008 VaR. However, trading VaR does not include: held-for-sale fund-
ed loan and unfunded commitments positions (however, it does
include hedges of those positions); the debit valuation adjustments
(“DVA”) taken on derivative and structured liabilities to reflect the
credit quality of the Firm; the MSR portfolio; and securities and
instruments held by corporate functions, such as Corporate/Private
Equity. See the DVA Sensitivity table on page 115 of this Annual
Report for further details. For a discussion of MSRs and the corporate
functions, see Note 4 on pages 141–155, Note 18 on pages
198–201 and Corporate/ Private Equity on pages 73–75 of this
Annual Report.
2008 VaR results
IB’s average total trading and credit portfolio VaR was $202 million
for 2008, compared with $106 million for 2007, and includes the
positions from the Bear Stearns merger since May 31, 2008. The
increase in average and maximum VaR during 2008 compared with
the prior year was primarily due to increased volatility across virtually
all asset classes. In addition, increased hedges of positions not
specifically captured in VaR – for example, macro hedge strategies
that have been deployed to mitigate the consequences of a systemic
risk event and hedges of loans held-for-sale – significantly increased
the VaR compared with the prior period.
For 2008, compared with the prior year, average trading VaR diversi-
fication increased to $108 million from $77 million, reflecting the
impact of the Bear Stearns merger. In general, over the course of the
year, VaR exposures can vary significantly as positions change, market
volatility fluctuates and diversification benefits change.
VaR backtesting
To evaluate the soundness of its VaR model, the Firm conducts daily
back-testing of VaR against daily IB market risk-related revenue,
which is defined as the change in value of principal transactions rev-
enue (less Private Equity gains/losses) plus any trading-related net
interest income, brokerage commissions, underwriting fees or other
revenue. The daily IB market risk-related revenue excludes gains and
losses on held-for-sale funded loans and unfunded commitments and
from DVA. The following histogram illustrates the daily market risk-
related gains and losses for IB trading businesses for the year ended
2008. The chart shows that IB posted market risk-related gains on
165 of the 262 days in this period, with 54 days exceeding $120
million. The inset graph looks at those days on which IB experienced
losses and depicts the amount by which 99% confidence level VaR
exceeded the actual loss on each of those days. During the year
ended December 31, 2008, losses were sustained on 97 days; losses
exceeded the VaR measure on three of those days compared with
eight days for the year ended 2007. The Firm would expect to incur
losses greater than those predicted by the 99% confidence level VaR
estimates once in every 100 trading days, or about two to three
times a year.
99% Confidence Level VaR
IB trading VaR by risk type and credit portfolio VaR
As of or for the year ended 2008 2007 At December 31,
December 31,(a) (in millions) Average Minimum Maximum Average Minimum Maximum 2008 2007
By risk type:
Fixed income $ 181 $ 99 $ 409 $ 80 $ 25 $135 $253 $106
Foreign exchange 34 13 90 23 9 44 70 22
Equities 57 19 187 48 22 133 69 27
Commodities and other 32 24 53 33 21 66 26 27
Diversification (108)(b) NM(c) NM(c) (77)(b) NM(c) NM(c) (152)(b) (82)(b)
Trading VaR $ 196 $ 96 $ 420 $ 107 $ 50 $188 $266 $100
Credit portfolio VaR 69 20 218 17 8 31 171 22
Diversification (63)(b) NM(c) NM(b) (18)(b) NM(c) NM(c) (120)(b) (19)(b)
Total trading and credit
portfolio VaR $ 202 $ 96 $ 449 $ 106 $ 50 $178 $317 $103
(a) The results for the year ended December 31, 2008, include five months of heritage JPMorgan Chase only results and seven months of results for the combined JPMorgan Chase and
Bear Stearns; 2007 reflects heritage JPMorgan Chase results only.
(b) Average and period-end VaRs were less than the sum of the VaRs of its market risk components, which is due to risk offsets resulting from portfolio diversification. The diversification
effect reflects the fact that the risks were not perfectly correlated. The risk of a portfolio of positions is therefore usually less than the sum of the risks of the positions themselves.
(c) Designated as not meaningful (“NM”) because the minimum and maximum may occur on different days for different risk components, and hence it is not meaningful to compute a
portfolio diversification effect.