Strategic Shareholder Climate and Risk Financial Financial Barclays PLC 437 report information sustainability report Governance review review statements Annual Report 2022 Notes to the financial statements (continued) For the year ended 31 December 2022 For existing/historical exposures where origination point scores or data are no longer available or do not represent a comparable estimate of lifetime PD, a proxy origination score is defined, based upon: • back-population of the approved lifetime PD score either to origination date or, where this is not feasible, as far back as possible (subject to a data start point no later than 1 January 2015); or • use of available historical account performance data and other customer information, to derive a comparable ‘proxy’ estimation of origination PD. ii) Qualitative test This is relevant for accounts that meet the portfolio’s ‘high risk’ criteria and are subject to closer credit monitoring. High risk customers may not be in arrears but either through an event or an observed behaviour exhibit credit distress. The definition and assessment of high risk includes as wide a range of information as reasonably available, such as industry and Group-wide customer level data, including but not limited to bureau scores and high consumer indebtedness index, wherever possible or relevant. Whilst the high risk populations applied for IFRS 9 impairment purposes are aligned with risk management processes, they are also regularly reviewed and validated to ensure that they capture any incremental segments where there is evidence of credit deterioration. iii) Backstop criteria This is relevant for accounts that are more than 30 calendar days past due. The 30 days past due criteria is a backstop rather than a primary driver of moving exposures into Stage 2. The criteria for determining a significant increase in credit risk for assets with bullet repayments follows the same principle as all other assets, i.e. quantitative, qualitative and backstop tests are all applied. Exposures will move back to Stage 1 once they no longer meet the criteria for a significant increase in credit risk. This means that, at a minimum all payments must be up-to-date, the PD deterioration test is no longer met, the account is no longer classified as high risk, and the customer has evidenced an ability to maintain future payments. Exposures are only removed from Stage 3 and reassigned to Stage 2 once the original default trigger event no longer applies. Exposures being removed from Stage 3 must no longer qualify as credit impaired, and: a) the obligor will also have demonstrated consistently good payment behaviour over a 12-month period, by making all consecutive contractual payments due and, for forborne exposures, the relevant EBA defined probationary period has also been successfully completed or; b)(for non-forborne exposures) the performance conditions are defined and approved within an appropriately sanctioned restructure plan, including 12 months’ payment history have been met. Management overlays and other exceptions to model outputs are applied only if consistent with the objective of identifying significant increases in credit risk. Forward-looking information The measurement of ECL involves complexity and judgement, including estimation of PD, LGD, a range of unbiased future economic scenarios, estimation of expected lives (where contractual life is not appropriate), and estimation of EAD and assessing significant increases in credit risk. Credit losses are the expected cash shortfalls from what is contractually due over the expected life of the financial instrument, discounted at the original effective interest rate (EIR). ECLs are the unbiased probability-weighted credit losses determined by evaluating a range of possible outcomes and considering future economic conditions. The Group uses a five-scenario model to calculate ECL. An external consensus forecast is assembled from key sources, including HM Treasury (short and medium-term forecasts) and Bloomberg (based on median of economic forecasts), which forms the Baseline scenario. In addition, two adverse scenarios (Downside 1 and Downside 2) and two favourable scenarios (Upside 1 and Upside 2) are derived, with associated probability weightings. The adverse scenarios are calibrated to a broadly similar severity to the Group's internal stress tests and stress scenarios provided by regulators whilst also considering IFRS 9 specific sensitivities and non-linearity. The favourable scenarios are designed to reflect plausible upside risks to the Baseline scenario which are broadly consistent with the economic narrative approved by the Senior Scenario Review Committee. All scenarios are regenerated at a minimum semi-annually. The scenarios include key economic variables (including GDP, unemployment, House Price Index (HPI) and base rates in both the UK and US markets) and expanded variables using statistical models based on historical correlations. The upside and downside shocks are designed to evolve over a five-year stress horizon, with all five scenarios converging to a steady state after approximately seven years. The methodology for estimating probability weights for each of the scenarios involves a comparison of the distribution of key historical UK and US macroeconomic variables against the forecast paths of the five scenarios. The methodology works such that the baseline (reflecting current consensus outlook) has the highest weight and the weights of adverse and favourable scenarios depend on the deviation from the baseline; the further from the baseline, the smaller the weight. A single set of five scenarios is used across all portfolios and all five weights are normalised to equate to 100%. The same scenarios used in the estimation of expected credit losses are also used to inform Barclays' internal planning. The impacts across the portfolios are different because of the sensitivities of each of the portfolios to specific macroeconomic variables, for example, mortgages are highly sensitive to house prices, and credit cards and unsecured consumer loans are highly sensitive to unemployment. Definition of default, credit impaired assets, write-offs, and interest income recognition The definition of default for the purpose of determining ECLs, and for internal credit risk management purposes, has been aligned to the Regulatory Capital CRR Article 178 definition of default, to maintain a consistent approach with IFRS 9 and associated regulatory guidance. The Regulatory Capital CRR Article 178 definition of default considers indicators that the debtor is unlikely to pay, includes