STI 2018 Annual Report

61 In addition to VAR, as required by the Market Risk Rule, we calculate Stressed VAR, which is used as a component of the total market risk capital charge. We calculate the Stressed VAR risk measure using a ten-day holding period at a one-tail, 99% confidence level and employ a historical simulation approach based on a continuous twelve-month historical window selected to reflect a period of significant financial stress for our trading portfolio. The historical period used in the selection of the stress window encompasses all financial crises since January 1, 2008. Our Stressed VAR calculation uses the same methodology and models as VAR, which is a requirement under the Market Risk Rule. Table 23 presents VAR and Stressed VAR for the year ended December 31, 2018 and 2017, as well as VAR by Risk Factor at December 31, 2018 and 2017. Value at Risk Profile Table 23 Year Ended December 31 (Dollars in millions) 2018 2017 VAR (1-day holding period): Period end $2 $2 High 3 3 Low 1 1 Average 2 2 Stressed VAR (10-day holding period): Period end $42 $52 High 103 110 Low 25 22 Average 60 54 VAR by Risk Factor at period end (1-day holding period): Equity risk $2 $1 Interest rate risk 1 2 Credit spread risk 2 3 VAR total at period end (1-day diversified) 2 2 The trading portfolio, measured in terms of VAR, is predominantly comprised of four sub-portfolios of covered positions: (i) credit trading, (ii) fixed income securities, (iii) interest rate derivatives, and (iv) equity derivatives. In support of our comprehensive range of capital market activity, the trading portfolio also contains other sub-portfolios, including foreign exchange rate and commodity derivatives; however, these related trading risk exposures are not material. Our covered positions result primarily from underwriting and market making services for our clients, as well as associated risk mitigating hedging activity. The trading portfolio’s VAR profile, presented in Table 23, is influenced by a variety of factors, including the size and composition of the portfolio, market volatility, and the correlation between different positions. Notwithstanding normal variations in the VAR associated with individual risk factors, average daily VAR as well as period end VAR for the year ended December 31, 2018 remained largely unchanged compared to the same period in 2017. Stressed VAR remained within comparable historic ranges throughout the year ended December 31, 2018, reflecting typical fluctuations in portfolio composition and balance sheet usage, along with changes in risk factor levels. The trading portfolio of covered positions did not contain any correlation trading positions or on- or off-balance sheet securitization positions during the year ended December 31, 2018 or 2017. In accordance with the Market Risk Rule, we evaluate the accuracy of our VAR model through daily backtesting by comparing aggregate daily trading gains and losses (excluding fees, commissions, reserves, net interest income, and intraday trading) from covered positions with the corresponding daily VAR-based measures generated by the model. As illustrated in the following graph, there were two firmwide VAR backtesting exceptions during the twelve months ended December 31, 2018. These two backtesting exceptions were driven primarily by credit spread widening during the broader sell-off in equity and credit markets during the latter half of December 2018, which impacted our corporate credit trading portfolio of bonds and loans. The total number of firmwideVARbacktesting exceptions over the preceding twelve months is used to determine the multiplication factor for the VAR-based capital requirement under the Market Risk Rule. The capital multiplication factor increases from a minimum of three to a maximum of four, depending on the number of exceptions. There was no change in the capital multiplication factor over the preceding twelve months.