Source: https://www.prarulebook.co.uk/pra-rules/market-risk-internal-model-approach-crr/01-01-2027?p=1 · captured 2026-09-28 · checked 2026-09-28 · b65a8e4d

140101/01/2027
1402
140312.
1404
1405For each risk factor, an institution shall keep a record of at least one year of the results of their modellability assessment, including the documentation referred to in points (a) to (g) of paragraph 11. For risk factors for which one year of results is not yet available, an institution shall keep the maximum available track record of results.
1406
1407[Note: Paragraphs 1 and 2 of this rule correspond to Article 325be(1) and (2) of CRR as it applied immediately before revocation by the Treasury]
1408
140901/01/2027
1410
1411Article 325bf Regulatory Back-testing Requirements and Multiplication Factors
1412
14131.
1414
1415For the purposes of this Article, an ‘overshooting’ means a one-day change in the value of a portfolio composed of all the positions assigned to the trading desk that exceeds the related value-at-risk number calculated on the basis of the institution's internal model in accordance with the following requirements:
1416
1417(a) the calculation of the value at risk shall be subject to a one-day holding period;
1418
1419(b) scenarios of future shocks shall apply to the risk factors of the trading desk's positions referred to in Article 325bg(3), including risk factors that are considered non-modellable in accordance with Article 325be;
1420
1421(c) data inputs used to determine the scenarios of future shocks applied to the risk factors shall be calibrated to historical data referred to in point (c) of Article 325bc(4); and
1422
1423(d) unless stated otherwise in this Article, the institution's internal model shall be based on the same modelling assumptions as those used for the calculation of the expected shortfall risk measure referred to in point (a) of Article 325ba(1).
1424
142501/01/2027
1426
14272.
1428
1429An institution shall count daily overshootings on the basis of back-testing of the hypothetical and actual changes in the value of the portfolio composed of all the positions assigned to the trading desk.
1430
143101/01/2027
1432
14333.
1434
1435An institution's trading desk shall be deemed to meet the back-testing requirements where the number of overshootings for that trading desk that occurred over the most recent 250 business days does not exceed any of the following:
1436
1437(a) 12 overshootings for the value-at-risk number, calculated at a 99th percentile one tailed-confidence interval on the basis of back-testing of the hypothetical changes in the value of the portfolio;
1438
1439(b) 12 overshootings for the value-at-risk number, calculated at a 99th percentile one tailed-confidence interval on the basis of back-testing of the actual changes in the value of the portfolio;
1440
1441(c) 30 overshootings for the value-at-risk number, calculated at a 97.5th percentile one tailed-confidence interval on the basis of back-testing of the hypothetical changes in the value of the portfolio; or
1442
1443(d) 30 overshootings for the value-at-risk number, calculated at a 97.5th percentile one tailed-confidence interval on the basis of back-testing of the actual changes in the value of the portfolio.
1444
144501/01/2027
1446
14474.
1448
1449An institution shall count daily overshootings in accordance with the following:
1450
1451(a) it shall base the back-testing of hypothetical changes in the value of the portfolio on a comparison between the end-of-day value of the portfolio and, assuming unchanged positions, the value of the portfolio at the end of the subsequent day;
1452
1453(b) it shall base the back-testing of actual changes in the value of the portfolio on a comparison between the end-of-day value of the portfolio and its actual value at the end of the subsequent day, excluding fees and commissions; and
1454
1455(c) it shall count an overshooting for each business day for which the institution is not able to assess the value of the portfolio or is not able to calculate the value-at-risk number referred to in paragraph 3.
1456
145701/01/2027
1458
14595.
1460
1461An institution shall calculate, in accordance with paragraphs 6 and 7, the multiplication factor (mc) referred to in Article 325ba for the portfolio of all the positions assigned to the trading desks for which it has been granted an IMA permission.
1462
146301/01/2027
1464
14656.
1466
1467An institution shall calculate the multiplication factor (mc) as the sum of the value of 1.5 and an add-on between 0 and 0.5 in accordance with Table 3. For the portfolio referred to in paragraph 5, the institution shall calculate that add-on on the basis of the number of overshootings that occurred over the most recent 250 business days as evidenced by the institution's back-testing of the value-at-risk number calculated in accordance with point (a) of this paragraph. The institution’s calculation of the add-on shall be subject to the following requirements:
1468
1469(a) an overshooting shall be a one-day change in the portfolio's value that exceeds the related value-at-risk number calculated by the institution's internal model in accordance with the following:
1470(i) a one-day holding period;
1471
1472(ii) a 99th percentile, one tailed confidence interval;
1473
1474(iii) scenarios of future shocks shall apply to the risk factors of the trading desks' positions referred to in Article 325bg(3) and which are considered modellable in accordance with Article 325be;
1475
1476(iv) the data inputs used to determine the scenarios of future shocks applied to the modellable risk factors shall be calibrated to historical data referred to in point (c) of Article 325bc(4);
1477
1478(v) unless stated otherwise in this Article, the institution's internal model shall be based on the same modelling assumptions as those used for the calculation of the expected shortfall risk measure referred to in point (a) of Article 325ba(1);
1479
1480(b) the number of overshootings shall be equal to the greater of the number of overshootings under hypothetical and the actual changes in the value of the portfolio.
1481
1482Table 3
1483
1484Number of overshootings
1485
1486Add-on
1487
1488Fewer than 5
1489
14900.00
1491
14925
1493
14940.20
1495
14966
1497
14980.26
1499
15007