7.The independent risk management function must conduct a regular back-testing program and profit and loss attribution program including but not limited to comparison of the risk measure and profit or loss values generated by the model against actual daily changes in portfolio value, as well as hypothetical changes based on static positions.
8.A Bank’s back-testing program must cover a minimum period of 250 business days.
9.A Bank’s back-testing program must include a formal evaluation of instances where trading outcomes are not covered by the risk measures (termed ‘exceptions’) on at least a quarterly basis, using the most recent twelve months modelled results and profit data. The Bank must document all exceptions generated from its ongoing back-testing program, including an explanation for the exceptions. A Bank must have the capacity to perform back-testing analysis both at the level of the whole portfolio and at the level of sub-portfolios or books that contain material risk.
10.A Bank must perform back-tests using both actual trading outcomes and hypothetical trading outcomes. Hypothetical trading outcomes are calculated by applying the day’s price movements to the previous day’s end-of-day portfolio. When performing back-tests using actual trading outcomes, a Bank must use clean trading outcomes, i.e. actual trading outcomes adjusted to remove the impact of income arising from factors other than market movements alone, such as fees and commissions, brokerage, additions to and releases from reserves which are not directly related to market risk (such as administration reserves).