2025 EU-wide Stress Test - Template Guidance PDF

2025 EU-WIDE STRESS TEST TEMPLATE GUIDANCE 5 JULY 2024 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE Contents List of tables 4 List of figures 4 List of boxes...

2025 EU-WIDE STRESS TEST TEMPLATE GUIDANCE 5 JULY 2024 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE Contents List of tables 4 List of figures 4 List of boxes 4 1. Introduction and general remarks 5 1.1 Objective of this guidance 5 1.2 Overview of the templates 5 1.3 Data input and formats 8 1.4 Supervisory reporting standards 9 2. Template specific guidance 10 2.1 General information 10 2.1.1 Input template 10 2.2 Credit Risk 11 2.2.1 CSV_CR_SUM 11 2.2.2 CSV_CR_SCEN 12 2.2.3 CSV_CR_SECTOR 17 2.2.4 CSV_CR_REA 21 2.2.5 CSV_CR_REA_OF 23 2.2.6 CSV_CR_REA_IRB 24 2.2.7 CSV_CR_REA_STA 25 2.2.8 CSV_CR_COVID19 25 2.2.9 CSV_CR_SEC_SUM 27 2.2.10 CSV_CR_SEC 29 2.2.11 CSV_CR_NPL 30 2.3 Market Risk, CCR losses and valuation 33 2.3.1 CSV_MR_SUM 33 2.3.2 CSV_MR_FULL_REVAL 34 2.3.3 CSV _MR_OPT_REVAL 44 2.3.3 CSV_MR_RESERVE 45 2.3.3.1 FuVA reserves 46 2.3.3.2 Liquidity and model uncertainty reserves 46 2.3.4 CSV_MR_PROJ 50 2.3.5 CSV_MR_CCR 51 2.3.6 CSV_MR_REA 53 2.4 NII 54 2 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE 2.4.1 CSV_NII_SUM 54 2.4.2 CSV_NII_CALC 57 2.4.3 CSV_NII_CALC_FUNDING_MATCH 65 2.5 Conduct risk and other operational risks 66 2.5.1 CSV_OR_GEN 66 2.5.2 CSV_OR_CON 67 2.6 Non-interest income, expenses and capital 68 2.6.1 CSV_REA_SUM 68 2.6.2 CSV_NFCI_DIV 69 2.6.3 CSV_ONEOFF 70 2.6.4 CSV_CAP 73 2.6.5 CSV_P&L 77 2.6.6 CSV_MDA 83 2.6.7 CSV_LR_MDA 85 2.6.8 CSV_CAPMEAS 88 2.6.9 CSV_ORAE 88 2.7 Use of pro-forma data in the stress test 89 Annex I: Market risk 92 3 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE List of tables Table 1: Overview of the CSV Templates 6 Table 2: Overview of the TRA Templates 7 Table 3: Reporting requirements for the main credit risk variables 14 Table 4: List of NACE sectors 19 Table 5: List of energy-intensive manufacturing activities 21 Table 6: Example DTA calculation (not arising from temporary differences) 81 Table 7: Example DTA calculation (arising from temporary differences) 82 Table 8: List of constraints for which the bank may be allowed to use pro-forma data in case of major events affecting its business model 90 Table 9: Balance sheet items at partial or full fair value and the reporting of their impact 92 List of figures Figure 1: Colour-scheme of different cells in the templates 9 List of boxes Box 1 Steps to perform the FULL REVALUATION matching items with hedges 35 Box 2: The bucketing approach 43 Box 3: Application of the liquidity and model uncertainty methodology 47 Box 4: Examples on client revenues 50 4 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE 1. Introduction and general remarks 1.1 Objective of this guidance 1. The purpose of this document is to provide technical guidance, together with the Methodological Note, to the participating banks for populating the set of templates for the 2023 EU-wide stress test. This document will not provide any definitions or requirements that go beyond the ones given in the Methodological Note. If there are cases where the guidance contradicts the requirements from the Methodological Note, the latter prevails. 2. Each of the Methodological Note’s chapters has a subchapter on the scope of application, on the definitions used in the chapter and on reporting requirements. Most of the information needed for the population of the templates is included in the Methodological Note - in particular in the sub sections on the scope of application, on the definitions used in the chapter and on reporting requirements. Their content will thus not be restated in this guidance. This document should therefore be read in conjunction with the Methodological Note. 3. Any abbreviations used in this document are defined in the Methodological Note. 4. The first section of this document covers general topics such as template types, data input and formats and supervisory reporting standards applied. The remainder of this document is structured following the order of the templates according to the file ‘2023 EU-wide Stress Test- Templates’. 5. Each template is covered in a separate section containing a summary of the purpose and data of the template, followed by a description of its structure, i.e. what information is contained in rows and columns. If any specific definitions or requirements are applicable to this template this is then covered in the following paragraph. Finally, links of the template with other templates are outlined. 1.2 Overview of the templates 6. The 2023 EU-wide Stress Test templates are grouped into the following template types: Instructions: Template which gives general information on how to populate the templates and also indicates the version number of the file; 5 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE Input: Template into which banks are requested to enter basic information such as the bank’s name, material countries and country / currency combinations, currency breakdowns, and prescribed NFCI growth rate parameters; Calculation Support and Validation data (CSV): Templates which, with the exception of certain summary templates, are to be populated by the participating banks and in some areas contain the stress test calculation. These templates will be used to populate the transparency templates; Transparency (TRA): Data on stress test outcomes to be disclosed on a bank-by-bank basis along with the publication of the stress test results. The TRA templates are populated automatically. 7. Table 1 and Table 2 below include an overview over all the templates. Banks will have to populate the Input table and all CSV templates, except the CSV optional templates as explained below in paragraph 9. Table 1: Overview of the CSV Templates Section or topic Template name Description Summary of templates and colour code N/A Instructions applied Input of bank name and relevant countries for credit risk, country/currency pairs for NII and currency breakdowns of N/A Input Other remaining administrative expenses and NFCI, and prescribed NFCI growth rate parameters Credit risk CSV_CR_SUM Credit risk – Summary Credit risk – Scenarios (projection for Credit risk CSV_CR_SCEN credit risk losses) Credit risk CSV_CR_SECTOR Exposures by sector of economic activity Credit risk CSV_CR_REA Credit risk – REA Credit risk CSV_CR_REA_IRB REA – IRB approach floor Credit risk CSV_CR_REA_STA REA – STA floor S-TREA for IRB exposures moving to the Credit risk CSV_CR_REA_OF STA approach for output floor calculation Credit risk CSV_CR_COVID19 Credit risk – COVID-19 public guarantees Credit risk CSV_CR_SEC_SUM Securitisations – Summary Credit risk CSV_CR_SEC Securitisations Credit risk CSV_CR_NPL NPL calendar Market risk, CCR losses and CVA CSV_MR_SUM Market risk – Summary 6 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE Section or topic Template name Description Market risk, CCR losses and CVA CSV_MR_FULL_REVAL Market risk – Full revaluation Market risk, CCR losses and CVA CSV_MR_RESERVE Market risk – Revaluation of reserves Market risk – Projection of client revenues Market risk, CCR losses and CVA CSV_MR_PROJ of items held with a trading intent and their related hedges Market risk, CCR losses and CVA CSV_MR_CCR Market risk – Counterparty defaults Market risk, CCR losses and CVA CSV_MR_REA REA – Market risk NII CSV_NII_SUM NII – Summary NII CSV_NII_CALC NII – Calculation CSV_NII_CACL_MATCH NII – Calculation for cases with funding NII _FUNDING matches Conduct risk and other CSV_OR_GEN Conduct and other operational risk losses operational risks Conduct risk and other CSV_OR_CON Material conduct risk losses operational risks Non-interest income, expenses CSV_REA_SUM REA – Summary and capital Non-interest income, expenses CSV_NFCI_DIV Evolution of net fee and commissions and capital income, dividend income Non-interest income, expenses CSV_ONEOFF Adjustments for non-recurring events and capital (one-offs) Non-interest income, expenses CSV_ORAE Breakdown of other remaining and capital administrative expenses Non-interest income, expenses CSV_P&L Evolution of P&L and capital Non-interest income, expenses CSV_CAP Capital and capital Non-interest income, expenses Calculation of potential distribution and capital CSV_MDA restriction following breach of the MDA trigger level Non-interest income, expenses Major capital measures and material CSV_CAPMEAS and capital losses Table 2: Overview of the TRA Templates Section or topic Template name Description Summary adverse or baseline scenario N/A TRA_SUM (stress test results) Credit risk TRA_CR_STA Credit risk (loss projection) STA Credit risk TRA_CR_IRB Credit risk (loss projection) IRB 7 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE Section or topic Template name Description Credit risk – Securitisations (REA Credit risk TRA_CR_SEC projection) Non-interest income, expenses TRA_REA REA (projection) and capital Non-interest income, expenses TRA_P&L P&L (projection) and capital Non-interest income, expenses TRA_CAP Capital (projection) and capital Non-interest income, expenses Major capital measures and material TRA_CAPMEAS and capital losses 8. Besides the instructions template, the majority of the transparency templates are not addressed in this document as they are automatically populated and are presented only for informational purposes. 9. There are two CSV templates whose population is optional: If a bank has not individuated any capital measure or material loss within the time horizons requested in the template CSV_CAPMEAS, it does not have to populate the CSV_CAPMEAS template; If a bank has no eligible one-off events to be taken into account in the stress test it does not have to populate the CSV_ONEOFF template. 1.3 Data input and formats 10.No changes should be made to the sheets or the structure of the file, i.e. the only edits should be the input of data. In particular, the password-protection of the sheets should be left intact, and no columns or rows should be inserted, (re)moved or replaced. Sheets whose password- protection has been decrypted and after some modifications encrypted again cannot be processed by the EBA and will therefore be rejected. 11.The templates have a common colour code to flag different categories of cells, using the logic described in the figure below. Cells in light blue are used for the processing the stress test data provided by the banks, but have no direct relevance for banks. 8 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE Figure 1: Colour-scheme of different cells in the templates Input cell to be filled by participating banks Calculation cell within a sheet Links between sheets Not to be filled in Row header or column header Additional identifiers to be used for the data extraction - cells hidden with the exception of row and column numbers 12.If a field requires a text input, a drop-down menu is implemented in the respective template. The only exception to this rule is template CSV_ONEOFF where some fields are free-text fields as they require descriptive information. 13.Monetary amounts should be reported in million euros (rounded to two decimal places) if not specifically indicated otherwise. When originally accounted in a currency different from euro the same exchange rates should be applied as for the COREP/ FINREP reporting. 14.Percentage data should be reported in the format ‘X.XX%’, i.e. not in decimals. 15.If the value required in a field amounts to 0, the respective field should be populated with 0 and not be left blank nor be populated with negligible amounts. 1.4 Supervisory reporting standards 16.All templates used in the 2025 EU-wide stress test refer to the specific version of supervisory reporting requirements in place as of 31 December 2024. This means, for all templates, the use of FINREP and COREP standards as for EBA reporting framework 3.2 (applicable for reports until 31 December 2024). Banks are required to restate some starting points and compute projections in accordance with CRR3/CRD6, as explained in relevant sections of the Methodological note and this document. In the case of resubmission of FINREP and COREP reports, templates should be filled in with the most updated data. When needed, banks should prove that a re-submission of FINREP/COREP is in process and explain the differences in the explanatory note. 9 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE 2. Template specific guidance 2.1 General information 2.1.1 Input template 17.This template contains general information on the bank participating in the stress test. In this template, banks are required to select their bank’s name and the most material countries for the reporting of credit risk data, the country / currency combinations for NII, as well as the currency breakdown of other remaining administrative expenses and NFCI, and the prescribed NFCI growth rate parameters. 18.The fields LEI and country of the selected bank will be populated automatically. 19.While in general the definitions of the Methodological Note apply, specific definitions to be highlighted for the use of this template comprise: The country fields in the credit risk table have to be populated according to section 2.3.4 of the Methodological Note; The country/currency fields in the NII table have to be filled according to section 4.3.6 Box 21 of the Methodological Note; The breakdown of other remaining administrative expenses has to be filled with the top 15 currencies in EUR amounts12 net of average projected one-off adjustments (if any).3 However, if 95% of the sum of all currencies have already been covered before filling the top 15 currencies, the bank has the option to report the remaining part in the “Other” (currencies) category. Banks using pro-forma data for other remaining administrative expenses shall provide a breakdown based on such data and include the (aggregate) difference between supervisory and pro-forma figures in the input sheet under cell E45 (while for all other banks this cell should be filled-in with zero); and 1 By contribution to the specific item and based on the relevant exchange rate at the cut-off date. 2 Also if operating with only one currency. 3 For example, if a bank has other remaining administrative expenses for GBP 100mn and a projected one-off adjustment (on other remaining administrative expenses) of GBP 10mn per year, it should fill the table with GBP 90mn denominated in EUR amounts as per applicable exchange rate. 10 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE The breakdown of NFCI has to be filled with the top 15 currencies in EUR amounts.45 However, if 95% of the sum of all currencies has already been covered before filling the top 15 currencies, the bank has the option to report the remaining part in the “Other” (currencies) category. 20.The Top 50 country / currency combinations in this template are linked via formula from this template to the CSV_NII_CALC template, the Top 10 country exposures are linked via formula from this template to the CSV_CR_SCEN and CSV_CR_REA templates. The Top 3 country exposures are linked via formula from this template to the CSV_CR_COVID19. Furthermore, the currency breakdown of NFCI is linked to the CSV_NFCI_DIV template, while the currency breakdown of other remaining administrative expenses is linked to the CSV_P&L template. 21.The exchange rate variations in this template are linked to the evolution of interest income and expenses under CSV_NII_CALC and CSV_NII_CALC_FUNDING MATCH, to the floor in other remaining administrative expenses under CSV_P&L and to the evolution of NFCI in CSV_NFCI_DIV. The exchange rate variations will be provided in the templates for all countries that are covered by the macroeconomic scenario. This applies to the ‘Other’ currency as well, where provided exchange variations in the templates shall be used according to the share of exposures in scope of ‘Other’. Detailed information on these calculations shall be included in the explanatory note. Furthermore, banks shall populate their presentation currency in the INPUT template. The exchange rate variation will be computed accordingly. 22. The fields for the prescribed NFCI growth rates are to be filled with the bank specific growth rates transmitted to banks. The growth rates represent for each year of each of the baseline and adverse scenarios the prescribed cumulative path of NFCI relative to the starting point. In the adverse scenario, the three-year cumulative growth rate of NFCI is subject to a maximum and minimum reduction (floor/cap) constraints. In the baseline scenario, the three-year cumulative growth rate of NFCI is subject to maximum reduction (floor) constraint. These constraints, if binding, apply to each year of the corresponding scenario horizon and are already reflected in the prescribed growth rate parameters. 2.2 Credit Risk 2.2.1 CSV_CR_SUM 23.This template shows the credit risk P&L impact for all positions excluding securitisations. It is automatically populated (from CSV_CR_SCEN) and no data needs to be entered in this template. 4 Same as footnote 1 5 Also if operating with only one currency. 11 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE 24.The rows of the template show end of year information (e.g. distribution of exposures to stages, stock of provisions, coverage ratios), flows between stages, impairments and credit risk parameters. The columns include the breakdown by year and scenario. 25.The CSV_CR_SUM template calculates the total impairment losses, which is linked via formula from this template to the CSV_P&L template. 2.2.2 CSV_CR_SCEN 26.In this template, banks are required to provide historical and projected credit risk information on exposure, LTV ratio, funded collateral, provisions and credit risk parameters broken down by asset class, regulatory approach, country of exposure, year and scenario. Moreover, the main credit risk stress test calculations are incorporated in this template. Banks are requested to provide “restated” data for 2024 by allocating their exposures as of 31 December 2024 to the exposure classes in accordance with CRR3. The restatement aims to align the starting point exposure class allocation with the exposure class allocation in the projections horizon, which considers the application of CRR3. 27.The columns of the template include all variables to be reported and are grouped into the sections ‘Beginning of year – Stocks’, ‘Within year - Flows and Parameters’, and ‘End of year – Stocks’. 28.The rows of the template are grouped by year, scenario, regulatory approach (A-IRB, F-IRB, STA), asset classes and geographical breakdown of the country of exposure. For the population of those fields the following applies: The asset classes refer to the ones described in section 2.3.3 of the Methodological Note. To consider the application of CRR3 for the stress test projections, the actual exposure classes differ from restated and adverse and baseline scenario exposure classes. Projections are performed based on the restated exposure classes; The geographical breakdown field contains the following expressions: (i) ‘Total’, (ii) the Top 10 countries automatically populated based on the data entered in worksheet ‘Input’, and (iii) ‘Other’ for the residual not attributed to any country. In contrast to data required for the projection horizon, for historical values a lower number of columns have to be populated with data from 2023 and 2024. Historical information (actual) considers the allocation of exposure classes according to the CRR version applicable on 31 December 2024. Restated information considers the allocation of exposure classes according to the CRR version applicable on 1 January 2025. These columns can also be found in section 2.3.4 of the Methodological Note (Table 2 and Table 3 for the actual historical information and Table 4 and Table 5 for the restated information). 12 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE For stages that have zero exposure at the beginning of the exercise but for which there are exposure transitions towards that stage during the exercise,6 credit risk parameters (TRs, LRs, LGDs) should be reported in line with the hierarchies of approaches outlined in section 2.4.1 of the Methodological Note. This implies in particular that if there is performing exposure (S1 or S2) at the beginning of the exercise all parameters should be reported. Consequently, parameters for stages where no exposure is reported for the starting point should be estimated and reported using observed or modelled equivalent values (e.g. same or comparable country and asset class), once duly adjusted for the macroeconomic variables corresponding to the year of estimation. Notwithstanding this, if there is only S3 exposure at the beginning of the exercise, parameters affecting S1 or S2 exposures do not need to be filled. As stated in section 2.4.1 of the Methodological Note, the overarching principle is the parameter’s suitability and comparability for projections. The parameters related to exposures or asset classes without associated credit risk (e.g, ‘other non-credit obligation assets’) shall reflect that absence of credit risk, i.e. 0% transitions rates, loss rates and LGDs shall be reported. According to paragraph 126 of the Methodological Note, ECB benchmark parameters have to be applied to an entire pivot asset class7 if the coverage of existing satellite models is very low (i.e. if the banks’ satellite models do not ensure the estimation of all the PD/TR and LR/LGD parameters, respectively, for a minimum of 10% of the pivot asset class exposure). The parameters reported at the level of a main asset class (e.g. Retail or Corporate) should be consistent with the parameters used in the respective pivot asset class. 29.While in general the definitions of the Methodological Note apply, specific definitions to be highlighted for the use of this template comprise: Credit risk relevant exposure is based on the CRR/CRD definition but should be always the amount before credit risk adjustments and should exclude all fair value positions (FVOCI and FVPL), which is in contrast to the REA-templates, and exposures subject to CCR. For the purpose of this template, securitisation positions are excluded (see section 2.2 of the Methodological Note). 30.While historical exposures and provisions have to be reported, the template calculates projected exposures and provisions out of starting point exposures / provisions and stressed 6 It is expected that this case can only occur in the starting point as transition rates should be non-zero during the exercise. 7 Pivot asset class refers to the lowest level of aggregation (e.g. ‘Corporates – SME - Secured by real estate property). 13 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE credit risk parameters. For this purpose, the CSV_CR_SCEN template contains the formulas to calculate provisions for the credit risk stress test (for all segments subject to the stress test, except securitisations) according to section 2.4.3 of the Methodological Note. As per paragraph 110 of the Methodological Note, management overlays that were created as a forward-looking provision before the cut-off date might be used to offset future expected losses during the stress test horizon.8 Nevertheless, no release of accumulated provisions for S3 exposures is permitted for any year or scenario and this restriction shall be applied at exposure level (management overlays should not be negative). In addition, starting points need to be suitable for the scenario projections (paragraphs 113 and 119 of the Methodological Note). Data is linked via formula from the CSV_CR_SCEN template to the CSV_CR_SUM template. Moreover, exposure and impairment data is linked via formula from this template to the TRA_CR_IRB and TRA_CR_STA templates. 31. 32.Table 3 summarises the reporting requirements for the main credit risk variables in each year of the template. Table 3: Reporting requirements for the main credit risk variables9 2023 BoY; Item 2023 EoY; 2024 EoY 2025 BoY 2025-2027 2024 BoY Exp S1 Actual with restated Exp S2 Actual Actual distribution across Exp S3 IFRS 9 stages10 Projected with Stocks Prov Stock S1 Actual ST definition11 Actual with restated Prov Stock S2 Actual Actual distribution across Prov Stock S3 Actual11 IFRS 9 stages10 8 To account for management overlays in the projections, provisions reported at the end of 2024 and the beginning of 2025 shall both be reported including management overlays. Thereby, possible conservatism in the provisioning at the cut-off date is considered in the calculation performed in the stress test templates as the total impairment losses (template CSV_CR_SUM) at the end of each period are computed as the difference with respect to the initial stocks of provisions. 9 The terms actual and restated on this table refer to the IFRS9 stage allocation and follow the logic of paragraph 56 of the Methodological Note. The actual and restated in this table, do not refer to the actual and restated exposure classes to account for the application of CRR3 on 1 January 2025, which should also be taken into account in the reporting. 10 Restatement to reflect the stress test definitions from paragraph 56 of the Methodological Note. 11 Formula-based fields. 14 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE Prov old S3-S3 Actual N/A Prov SX-S312 Actual Item 2023 2024 2025-2027 S2-S1 flows S1-S2 flows Flows Actual Actual Projected with ST definition11 S1-S3 flows13 S2-S3 flows13 Transition rates Modelled LGDs Parameters N/A (according to Modelled (stress test scenario) Cure rates par. 33) Loss rates 33.According to paragraph 112 of the Methodological Note, for the estimation of starting point parameters (rows covering the year 2024 of the CSV_CR_SCEN template), only the baseline scenario shall be taken into account. In line with paragraph 69 of the Methodological Note, the historical stock of provisions as of end-of-year 2024 (both “actual” and “restated” in line with the application of CRR3 as of 1 January 2025) shall be the actual one in accordance with the bank’s accounting framework (e.g. based on a multi-scenario approach applied under IFRS 9). The beginning-of-year 2025 stock of provisions is related to a historical period but should be allocated in line with the distribution of exposure by stage restated in order to reflect the stress test definitions from to paragraph 56 of the Methodological Note (see 34.Table 3). As a consequence, it is expected that the stock of provisions both as of end-of-year 2024 and beginning-of-year 2025 will not match the 2024 risk parameters calculated based on paragraph 112 of the Methodological Note. Contrarily, the projected stock of provisions for end-of-year 2025, 2026 and 2027 shall consider risk parameters calculated in accordance with the (baseline/adverse) scenario provided and under the perfect foresight assumption. In order to support the process of quality assurance performed by the competent authorities, banks should provide in the explanatory note details on the weighted scenarios employed under IFRS 9 and the impact of the assumptions from paragraph 112 of the Methodological Note to the starting point risk parameters and provision coverage ratios. 35.According to paragraph 108 of the Methodological Note, the two lines ("of which: Residential guaranteed loans (Prêts cautionnés) insured by an eligible residential property loan guarantor" and "of which: other than Residential guaranteed loans (Prêts cautionnés) insured by an eligible residential property loan guarantor") are intended to give details of the asset classes 12 Consider the provisions allocated to exposures that moved to S3 and remained in S3 at the end of the respective year. 13 Consider all transitions to S3 during the respective year (reported once in case it enters several times into S3). 15 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE “Retail Secured by residential real estate Non-SME” (A-IRB and F-IRB) and “Secured by mortgages on immovable property and ADC exposures Residential immovable property” (STA) regarding: (i) the exposures secured by residential real estate properties insured by an eligible residential property loan guarantor (first row); and (ii) the remaining exposures of the same asset class (second row). Under A- and F-IRB, the sum of the exposures and provisions relevant to these two "of which" rows should equal the exposures and provisions reported in the corresponding pivot level row “Retail Secured by residential real estate Non-SME”. Under STA, the sum of exposures and provisions of these two “of which” rows should also reconcile with the amounts reported in the corresponding pivot level row “Secured by mortgages on immovable property and ADC exposures Residential immovable property”. Hence, the sum of all “of which” rows under this asset class14 might not reconcile with the values reported in the corresponding sum level row “Secured by mortgages on immovable property and ADC exposures Residential immovable property”. If a bank has no Retail non-SME exposure secured by residential real estate or “Secured by mortgages on immovable property and ADC exposures Residential immovable property” and insured by an eligible residential property loan guarantor, these two “of which” rows should be left blank. These rows shall not include the exposures falling under the COVID-19 public guarantee schemes, which will be reported in template CSV_CR_COVID19. 36.For the asset class STA “Secured by mortgages on immovable property and ADC exposures - Residential immovable property and Commercial immovable property” (pivot rows), banks are requested to report more granular information under a set of "of which" rows. The sum of the exposures and provisions, respectively, of all "of which" rows may not reconcile with the values on the pivot line. However, it is expected that the sum of the "of which" rows for “Non IPRE (Secured)”, “Non IPRE (Unsecured)” and "IPRE” reconciles with the values reported under the respective pivot rows. Similarly, for the asset class (pivot line) “Secured by mortgages on immovable property and ADC exposures Other” the sum of the "of which" rows “Non IPRE” and "IPRE” is expected to reconcile with the values reported under the respective pivot row. 37.The column ‘average maturity (yrs)’ is aggregated in ‘sum’ asset classes as an average of ‘pivot’ asset classes weighted by the respective performing exposure (i.e. non-performing exposure is assumed to have no defined maturity). 38.The columns related to transition rates are aggregated in ‘sum’ asset classes as an average of ‘pivot’ asset classes weighted by the respective exposure (e.g. TR1-3 is weighted by the S1 exposure in the beginning of the period). 14 “of which: IPRE“, “of which: SME“ ,”of which: IPRE, of which: SME, "of which: Residential guaranteed loans (Prêts cautionnés) insured by an eligible residential property loan guarantor", "of which: other than Residential guaranteed loans (Prêts cautionnés) insured by an eligible residential property loan guarantor") 16 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE 39.The columns related to loss rates, LGDs and cure rates are aggregated in ‘sum’ asset classes as an average of ‘pivot’ asset classes weighted by the respective exposure multiplied by the respective transition rates (e.g. LGD1-3 is weighted by the S1 exposure in the beginning of the period multiplied by TR1-3). 40.The column ‘PD PiT (%)’ shall be subject to a technical minimum floor of 0.001% for the starting point year 2024 in all cases where there exists non-zero ‘Performing Exposure (Exp)’ (either S1 or S2) of ‘Beginning of year- Stocks’. 2.2.3 CSV_CR_SECTOR 41.In this template, banks are required to provide historical and projected credit risk information on exposure, LTV ratio, funded collateral, provisions and credit risk parameters broken down by year, scenario, country of exposure, and NACE sector. The scope of the asset classes of the exposures to be included in this template is described in the Methodological Note. 42.In general, the template follows the same calculation logic as implied in the template CSV_CR_SCEN. Therefore, unless specified otherwise in the following paragraphs, the same instructions provided for the template CSV_CR_SCEN apply. 43.In columns 1 "PD/TR - Percentage of exposures with projections based on sectoral models, e.g. via sensitivities by sector (%)" and 2 "LGD/LR - Percentage of exposures with projections based on sectoral models, e.g. via sensitivities by sector (%)" the banks should indicate the percentage of exposures for which the projections of sectoral risk parameters are obtained by using dedicated models that produce sector-specific risk parameters based on the sectoral dynamics prescribed by the scenario or by applying a reasonable loss distribution approach, e.g. sensitivities, to portfolio-level projections. The approach followed should be further described in the explanatory note. 44.The stress test scenario projects GVA only for the 27 EU countries, the euro area and the EU in aggregate terms. Since the list of countries reported in CSV_CR_SECTOR should match with CSV_CR_SCEN, this list might include non-EU countries depending on the geographical distribution of exposures of the bank. In this case, banks shall report exposures, provisions and parameters for the non-EU countries and are expected to document the approach used in the explanatory note. Banks should explain to what extent the assumptions and constraints used for sectoral projections for non-EU countries are consistent with the overall narrative of the ST macroeconomic scenario. 45.The rows of the template are grouped by year, scenario, country of exposure, and NACE sector of exposure. 17 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE a) The sectoral breakdown contains the following expressions: (i) ‘Total’, (ii) the list of economic activities based on NACE codes, (iii) a memorandum item for corporate exposures in scope of this template, but which are reported as ‘STA – Secured by mortgages on immovable property – and ADC exposures’ and are not reported within the SME exposures in CSV_CR_SCEN. b) The list of economic activities covers all NACE sections (1-digit level) at varying granularity for which banks need to provide both historical and projected information. Additionally, banks have to provide a more granular breakdown of historical information for selected NACE divisions (2-digit level), which may not necessarily be exhaustive15 with respect to the corresponding NACE sections (Table 4), Table 4. Under the breakdown ‘Energy-intensive activities’, banks shall report the NACE divisions listed in Table 4, whereas the breakdown ‘other’ shall encompass all activities in NACE section C excluding the aforementioned NACE divisions listed in Table 4. c) With regard to the NACE sections R-U (Arts, entertainment and recreation; other service activities; activities of households; activities of extra-territorial organisations and bodies) and for the starting points only, banks are required to provide a more granular breakdown of information for the NACE sections R-S (Arts, entertainment and recreation; other service activities) in the dedicated o/w rows. d) The template does not include a specific row for the exposures for which the sector is unknown. Banks shall report in the explanatory note the amount of exposures for which the attribution of a NACE sector was controversial, distinguishing the cases where multiple assignments were possible or those where the sector was unknown, and the criteria used to attribute a sector in the former cases. e) For the starting points, the total exposures and provisions at each country level reported in CSV_CR_SECTOR, including the distribution across IFRS stages, should be consistent with those reported in CSV_CR_SCEN. 15 As lists of the NACE divisions provided for these selected sectors are not meant to be exhaustive, the sum of the exposures and provisions reported for the NACE divisions for a given sector reported under “o/w” rows might not reconcile fully with the “pivot” row corresponding sector reported under the NACE section. 18 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE Table 4: List of NACE sectors NACE sections Selected NACE divisions A Agriculture, forestry and fishing B Mining and quarrying C Manufacturing C10-C11 Manufacture of food products and beverages C12 Manufacture of tobacco products C17 Manufacture of paper and paper products C18 Printing and reproduction of recorded media C19 Manufacture of coke and refined petroleum products C20 Manufacture of chemicals and chemical products C21 Manufacture of basic pharmaceutical products and pharmaceutical preparations C22-C23 Manufacture of rubber, plastic and other non-metallic mineral products C24-C25 Manufacture of basic metals and fabricated metal products, except machinery and equipment C26-C27 Manufacture of computer, electronic and optical products and electrical equipment C28 Manufacture of machinery and equipment (not elsewhere classified) C29-C30 Manufacture of motor vehicles, trailers, semi- trailers and other transport equipment D Electricity, gas, steam and air conditioning supply E Water supply, sewerage, waste management and remediation activities F Construction F41 Construction of buildings F42-F43 Civil engineering and specialised construction activities G Wholesale and retail trade, incl. repair of motor G46 Wholesale trade, except of motor vehicles and vehicles and motorcycles motorcycles 19 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE NACE sections Selected NACE divisions G47 Retail trade, except of motor vehicles and motorcycles H Transportation and storage H49 Land transport and transport via pipelines H50-H51 Water and air transport H52-H53 Warehousing, support activities for transportation, postal and courier activities I Accommodation and food service activities I55 Accommodation I56 Food and beverage service activities J Information and communication K Financial and insurance activities L Real estate activities M-N Professional, scientific and technical activities; administrative and support service activities O-Q Public administration and defence, compulsory social security; education; human health services and social work activities R-U Arts, entertainment and recreation; other service activities; activities of households; activities of extra-territorial organisations and bodies16 16 For the starting points, banks are requested to provide information on the NACE sections R-Arts, entertainment and recreation and S-Other service activities in the dedicated o/w rows. 20 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE Table 5: List of energy-intensive manufacturing activities NACE divisions C.10 Manufacture of food products C.11 Manufacture of beverages C.12 Manufacture of tobacco products C.16 Manufacture of wood and of products of wood and cork, except furniture; manufacture of articles of straw and plaiting materials C.17 Manufacture of paper and paper products C.18 Printing and reproduction of recorded media C.19 Manufacture of coke and refined petroleum products C.20 Manufacture of chemicals and chemical products C.22 Manufacture of rubber and plastic products C.23 Manufacture of other non-metallic mineral products C.24 Manufacture of basic metals C.25 Manufacture of fabricated metal products, except machinery and equipment C.28 Manufacture of machinery and equipment n.e.c. 2.2.4 CSV_CR_REA 46.In this template, banks are required to provide credit risk information on exposure values, regulatory risk parameters, REA, and expected losses broken down by year, scenario, regulatory approach and asset class (rows). CSV_CR_REA_STA and CSV_CR_REA_IRB templates source the information from the CSV_CR_REA template. Banks shall restate the “Actual” exposure classes on 31 December 2024 to “Restated” exposure classes to account for the application of CRR3 on 1 January 2025. Exposures reported under the adverse and baseline scenario should also be allocated in accordance with the exposure classes that account for the application of CRR3. 47.The rows show the combinations of geographical breakdown, scenario, year, regulatory approach and asset classes. The columns show the different end of year exposure values, regulatory risk parameters, REA, EL and stock of provisions. 48.Exposure values in this template are defined according to COREP definitions in line with paragraph 87 Methodological Note and thus might differ from the exposure data (Exp) reported in the template CSV_CR_SCEN. For the STA portfolio, exposure values should hence be provided net of provisions. In line with paragraph 114 of the Methodological Note, REA reported in the CSV_CR_REA template as well as provisions reported in the CSV_CR_SCEN template should exclude, if applicable, IFRS9 transitional arrangements for the reporting of the historical numbers. The impact of the latter is considered separately in the calculation of the transitional capital ratios in the CSV_CAP template. In line with paragraph 167 of the Methodological Note, the unfloored REA amounts for the “Restated” 2024 should be reported 21 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE net of transitional arrangements in accordance with Article 495d of the CRR3 but including other transitional arrangements. The impact of the transitional arrangements related to Article 495d of the CRR3 on the total REA will be considered separately within the CSV_CAP template. 49.‘Equity’ and ‘other non-credit obligation assets’ which are not treated under the Standardised approach should be reported in the rows of ‘A-IRB’ for the “Actual” starting point information. In line with paragraph 161 of the Methodological Note, banks that have been granted permission to use the IRB approach for the calculation of REA for equity exposures up to 31 December 2024 should continue to report these exposures under the IRB approach for the purpose of the “Restated” starting points and the stress test projections following the application of CRR3 on 1 January 2025. According to paragraph 168 of the Methodological Note, transitional arrangements for equity exposures, related to articles 495 and 495a of the CRR3 as well as the transitional adjustments related to articles 495b, 495c of the CRR3 should be considered for the projections of REA. However, banks should report the impact of the included transitional arrangements as an “of-which” item to the total U-REA in the CSV_CAP template. 50.For the asset class STA “Secured by mortgages on immovable property and ADC exposures Residential immovable property and Commercial immovable property” (pivot rows), banks are requested to report more granular information under a set of "of which" rows for the “Restated” starting points and projections. The sum of the exposures and risk exposures amounts, respectively, of all "of which" rows may not reconcile with the values on the corresponding pivot line. However, it is expected that the sum of the "of which" rows for “Non IPRE (Secured)”, “Non IPRE (Unsecured)” and "IPRE” reconciles with the values reported under the respective pivot rows. Similarly, for the asset class (pivot line) “Secured by mortgages on immovable property and ADC exposures Other” the sum of the "of which" rows “Non IPRE” and "IPRE” reconciles with the values reported under the respective pivot row. 51.The ‘stock of credit risk adjustments’ fields shall be populated by banks for “Actual” end-2024 in line with COREP, and the “Restated” end-2024 that reflects the application of the CRR3 on 1 January 2025. For the projections, the total stock of credit risk adjustments is linked to the variation of the stock of provisions from the CSV_CR_SCEN template but banks shall populate the stock of credit risk adjustments for defaulted assets (i.e. for assets in default according to the CRR3 applicable on 1 January 2025).17 The projected credit risk adjustments on non- defaulted assets are calculated in the template as the difference between total credit risk adjustments and credit risk adjustments on defaulted assets. 17 From the projected provisions on S3 exposures, banks are expected to provide in this field only those that are related to exposures defaulted according to the CRR. 22 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE 52.The respective exposure values and REAs are linked via formula from this template to the CSV_CR_REA_IRB and CSV_CR_REA_STA templates. 53.For exposures Banks should allocate and report the exposures to “Corporates - Specialised Lending” according to COREP as of 31 December 2024 for the “Actual” and according to applicable CRR3 definitions for the “Restated” 2024 and the projections.. If the bank applies the slotting criteria according to Article 153(5) of the CRR for the calculation of the capital requirements for specialised lending, the risk parameters (PDreg, LGDreg (default stock), LGDreg (non-defaulted assets) and ELBE) for the aggregated specialised lending should reflect the risk parameters of specialized lending excluding specialised lending under the slotting approach. All other fields have to be populated for specialised lending exposures reflecting the amounts for both specialised lending exposures excl. under the slotting approach and specialised lending under the slotting approach. 54.‘PD Reg – non-defaulted assets’ and ‘LGD Reg – non-defaulted assets’ are aggregated in ‘sum’ asset classes as an average of ‘pivot’ asset classes weighted by non-defaulted exposures. 55.‘LGD Reg – defaulted assets’ and ‘ELBE – default stock’ are aggregated in ‘sum’ asset classes as an average of ‘pivot’ asset classes weighted by defaulted exposures. 2.2.5 CSV_CR_REA_OF 56.The scope of the template covers only exposures which are under the IRB approach treatment as reported in CSV_CR_REA, at country level “Total”. For these exposures, banks are requested to report in the CSR_CR_REA_OF template, the equivalent exposure and REA values recalculated under the STA approach. The standardised REA amounts for the IRB exposures will serve as an input for the calculation of the standardised total REA (S-TREA) of the entity. 57.The template covers credit risk information on exposure values, risk exposure amounts application of CRR3), regulatory approach and asset class (rows). The recalculated amounts under the STA approach should be reported both i) using a STA asset classes breakdown and ii) using the original IRB asset classes breakdown. Thus, the totals for exposures and REA across all STA asset classes should fully reconcile with the totals across all the IRB asset classes. For the avoidance of doubt, only exposures reported under the IRB approach in CSV_CR_REA are to be reported in the CSR_CR_REA_OF template. The respective REA values are linked via formula from this template to the CSV_CR_REA_IRB template. 58.The same calculation logic and methodological constraints should be applied for the recalculation of IRB exposures under the STA approach as was the case for the exposures originally classified under STA. Thus, banks should simulate the impact of the application of the macroeconomic scenario on the exposures originally classified under the IRB approach, as if they were treated under the STA approach. 23 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE 59.No transitional adjustments according to Articles 465 and 495d of the CRR3 should be considered for the reporting of standardised credit risk REA at the starting point nor for the projections within the CSV_CR_REA_OF template. 2.2.6 CSV_CR_REA_IRB 60.This template contains risk exposure amounts per asset class and expected loss amounts for IRB exposures (expected loss separately for equity exposures and non-equity exposures). Fields are in general automatically populated out of data from CSV_CR_REA, except the REA for Equity and Other non-credit obligation assets, the additional value adjustments from Article 159 CRR in the starting point and the expected loss amount for Equity deducted to CET1 according to Article 36(d) CRR. Row ‘Standardised REA for IRB approach portfolios (fully loaded)’ is automatically populated from the CSV_CR_REA_OF template. 61.The rows show the breakdown of IRB REA by asset classes and total IRB expected loss, credit risk and additional value adjustments. The columns show the scenario and year split up into performing and non-performing exposure figures. Additional rows are included to consider the restatement of starting point exposures and projections in accordance with the updated exposure classes for the application of CRR3. 62.The expected loss amount for equity exposures should be reported in the ‘Memorandum item: Expected loss amount Equity deducted to CET1 according to Article 36(d) CRR’ only if the expected loss for equity exposures is included in the IRB shortfall calculation for COREP purposes. 63.The CSV_CR_REA_IRB template ensures that the REA floor (year-end 2024 REA) is applied for IRB exposures considering the “Restated” end-2024 REA. According to the Methodological Note, the floor implemented in CSV_CR_REA_IRB is based on REA without the application of the IFRS 9 transitional adjustments and arrangements related to Article 495d of the CRR3. The total IRB REA values after application of the floor are linked via formula from this template to the CSV_REA_SUM template in the calculation of total credit risk REA. In case regulatory risk- weight floors pursuant Article 458 of the CRR are in force, any floor that is in force as of 31 December 2024 must be assumed to remain in place for the whole projection horizon regardless of its expiration date. If regulatory risk-weight floors pursuant Article 458 in force on 31 December 2024 are modified for the application of CRR3 and the modification enters into force on 1 January 2025, the modified risk-weight floors pursuant Article 458 should be considered. The positive difference between such IRB REA floors and the projected IRB REA after the imposition of the stress test REA floor as per paragraph 162 of the Methodological Note (reported in CSV_CR_REA_IRB, “Total Risk Exposure Amount IRB exposures after floor”) shall be reported as “other REA” in the CSV_REA_SUM template. 24 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE 2.2.7 CSV_CR_REA_STA 64.This template contains risk exposure amounts under the STA (excluding securitisations). Fields are in general automatically populated out of data from CSV_CR_REA banks, except the REA for Equity and Others exposures. 65.The rows show the breakdown of STA REA by asset classes and total STA credit risk adjustments. The columns show the scenario and year split up into non-defaulted and defaulted exposure figures. Additional rows are included to consider the restatement of starting point exposures and projections in accordance with the updated exposure classes for the entry into force of CRR3. 66.The risk weighting when deducting software assets according to the RTS on the prudential treatment of software assets under Article 36 of Regulation (EU) No 575/2013 should be reported in the row “Other exposures”. 67.This template also ensures that the REA floor (year-end 2024 restated REA) is applied for STA exposures. The total STA REA after application of the floor is then linked via formula from this template to the CSV_REA_SUM template in the calculation of total credit risk REA. According to the Methodological Note, the floor implemented in CSV_CR_REA_STA is based on REA “Restated” end-2024 without the application of the IFRS 9 transitional adjustments and arrangements related to Article 495d of the CRR3. 2.2.8 CSV_CR_COVID19 68.In this template, banks are required to provide information regarding exposures subject to COVID-19 public guarantee schemes (PGS). This information includes exposure values, stocks of provisions, REA and credit risk parameters by year, scenario, regulatory approach and main asset classes affected by COVID-19 PGS. The scope of reporting is defined in paragraph 118 of the Methodological note. 69.Banks shall select the countries in scope of the CSV_CR_COVID19 template as per paragraph 119 of the Methodological Note via the drop-down menu of the dedicated column in the Input sheet. These countries will be automatically populated in sheet CSV_CR_COVID19. 70.The definitions of variables included in this template should follow the EBA Guidelines on COVID-19 measures reporting and disclosure (EBA/GL/2020/07). 71.Columns 1 to 24 refer to the exposures subject to public guarantee schemes according to the EBA/GL/2020/07 that are within the scope of CSV_CR_SCEN, except for columns 11 and 12 25 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE which should consider the ones within the scope of CSV_CR_REA.18 Exposures subject to public guarantee schemes as per the EBA/GL/2020/07 that are treated under the securitisation framework and reported in CSV_CR_SEC should also be reported in the CSV_CR_COVID-19 template.19 At the starting points, banks shall report exposures, provisions, and parameters restated in line with applicable regulation as at 1 January 2025 (CRR3). 72.In contrast to the reporting requirements for the other credit risk templates, in this template the exposures should be reported in asset classes before CRM substitution effects (i.e. the exposures should not be transferred to the asset class of the guarantor). This applies to both the P&L and REA perspectives of the template (i.e. exposure by stage and exposure value). The stocks of provisions and risk parameters should be consistent with the amounts and parameters reported in CSV_CR_SCEN and the amounts of REA should be consistent with the amounts reported in CSV_CR_REA, but all should be allocated proportionally in line with the breakdown of exposures20 by asset class before CRM substitution effects. 73.The rows of “Total” should cover all exposures subject to COVID-19 PGS, including the remaining exposures that are not allocated to the main asset classes with open fields in this template (i.e. not Corporates, Retail or non-SME Secured by mortgages on immovable property). For this reason, also the rows of “Total” are open fields to be reported by banks. 74.Should the exposures reported in the explicitly three reported countries not exhaust the entirety of exposures subject to COVID-19 PGS, then the “other” geography rows should be filled out to ensure that the values reported for the “Total” geography represent all exposures covered by the COVID-19 PGS. 75.The information regarding public guarantees covers loans and advances that, at the reference date, are subject to PGS that Member States introduced in response to the COVID-19 crisis.21 The EBA published a list of PGS issued, which can be accessed via the EBA website.22 Nonetheless, for the avoidance of doubt, the CSV_COVID19 template covers all exposures subject to COVID-19 PGS. 76.The field of “Total exposure reported under CRM substitution in Central Governments” should include the total exposures under COVID-19 PGS that were subject to CRM substitution effects 18 Columns 11 and 12 should consider both defaulted and non-defaulted exposures. 19 In line with paragraph 117 of the Methodological Note, the explanatory note should include information on the exposures reported in the template CSV_CR_COVID19 that are treated under the securitisation framework. The explanatory note should also include information regarding the risk parameters of securitised and non-securitised loans. 20 The stocks of provisions and risk parameters should be allocated in line with the exposure by stage, whereas REA should be allocated in line with the exposure value. 21 Including called public guarantees for which payment was not yet received from the guarantor. 22 https://www.eba.europa.eu/eba-publishes-overview-public-guarantee-schemes-issued-response-covid-19-pandemic 26 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE and are reported in the template CR_CSV_SCEN in the rows of ‘Central Governments”. Given the different perimeter (listed PGS for columns 2, 4, 6; CRM substitution for column 7), the sum of exposures in columns 2, 4 and 6 must be at least as large as the exposures reported in column 7. 77.The credit risk parameters should be the ones related to the sub-portfolios of exposures subject to COVID-19 PGS. These parameters should be consistent with the parameters reported in the template CSV_CR_SCEN but allocated proportionally in line with the breakdown of exposures by asset class before CRM substitution effects. 78.The parameters for loans and advances with public guarantees that cover loans which were also subject to EBA compliant COVID-19 moratoria should disregard the historical influence of moratoria. 79.For the starting point, the LGD parameters for exposures under COVID-19 PGS should be related only to the non-guaranteed part of the exposure (columns 22-24) while the remaining parameters should refer to the entire exposure (i.e. guaranteed and non-guaranteed). For the projections horizon, only the parameters that refer to the entire exposure should be reported. 2.2.9 CSV_CR_SEC_SUM 80.This template shows exposure values and REA for securitisation positions broken down by the regulatory approaches from (SEC-IRBA, SEC-SA, SEC-ERBA and SEC-IAA). This information is automatically populated from CSV_CR_SEC, except for the starting point. 81.The rows show the breakdown of REA and exposures by regulatory approach and total aggregated impairments. The columns show the scenario and the year. 82.In this template, only impairments for securitisation positions that are not subject to mark-to- market valuation are reported (i.e. excludes FVOCI and FVPL). This information is sourced from CSV_CR_SEC. 83.Banks should populate the columns ‘actual’ and ‘restated’ with the total exposure value, REA, reduction of REA due to the application of regulatory caps and REA on trading book positions subject to specific risk, in line with the regulation applicable as of 31 December 2024 and 1 January 2025, respectively. The projected values for the exposures subject to specific risk, should be stressed in line with section 2.7 of the Methodological Note. 84.The columns ‘actual’ and ‘restated’ should be populated with the REA, as of 31 December 2024, allocated to the regulatory approaches from CRR and reported without any mapping effect from the application of paragraph 184 of the Methodological Note. The ‘total’ REA from CSV_CR_SEC_SUM might therefore be different from the ‘total’ REA reported in CSV_CR_SEC 27 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE as the latter shows the sum of REA by credit quality step after mapping to the SEC-ERBA look- up table from Articles 263(3) and 264(4) of CRR (i.e. includes the mapping effect). 85.The rows “Risk exposure amounts Article 92(5)(a)(iii) (relevant for output floor and after REA floor)” serve for the purpose of standardised REA calculation for the securitization exposures. The scope of this block covers exposures which are reported under the SEC-IRBA and SEC-IAA approach treatment in CSV_CR_SEC. For these exposures, banks are requested to report the equivalent restated REA values assuming the same CQS allocation at the starting point and without considering transitional arrangements in accordance with the paragraph 13 of Article 465 of CRR3. The total standardised REA amount is collected in the row “Total” and will serve as an input for the calculation of the S-TREA of the entity. 86.While in general the definitions of the Methodological Note apply, definitions to be highlighted for the use of this template comprise: The definition of exposure which is referred to in section 2.7 of the Methodological Note (in line with Article 248 of CRR); Impairments for the use of this template exclude impairments for securitisation positions subject to mark-to-market valuation; Whenever a term is not defined specifically in the Methodological Note, the definitions of CRR apply. 87.Securitisation positions covered by Article 254(7) of CRR (“Other” securitisations with a RW of 1250% and not included in COREP in the four main regulatory approaches - SEC-IRB, SEC-SA, SEC-ERBA and SEC-IAA), shall be reported in the SEC-SA rows of this template. 88.The total securitisation impairments are linked via formula from this template to the CSV_P&L template. 89.The annual relative increase of the REA reported in the CSV_CR_SEC template is considered for the projection of REAs by regulatory approach. 90.The CSV_CR_SEC_SUM template applies the REA floors to total securitisation REAs separately for each regulatory approach. CSV_REA_SUM then sources the total floored securitisation REAs from the CSV_CR_SEC_SUM template. 91.Exposure values, REAs and impairments are linked via formula from this template to the TRA_CR_SEC template. 28 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE 92.In the case of synthetic securitisation positions as defined in Article 2(10) of CRR, the treatment of maturity mismatch provided by Articles 252(a) and 252(b) of CRR should be considered in the template CSV_CR_SEC_SUM in the row “Additional risk exposure amounts“. 2.2.10 CSV_CR_SEC 93.This template contains exposure values and REA for securitisation positions in SEC-IRB, SEC-SA, SEC-ERBA and SEC-IAA. In this template, banks are required to provide exposure values and cumulative credit risk adjustments for securitisations. The template will calculate REAs on an automated basis. 94.The rows show the combinations of scenario, year, regulatory approach and credit quality step. The columns show the exposures, credit risk adjustments, REA and risk weights broken down by risk bucket, STS classification, tranche seniority and maturity. 95.Exposure values and adjustments should be reported in the template differentiated by regulatory approach, credit quality step, risk bucket, STS classification, tranche seniority and maturity. The amount of exposures that are not subject to mark-to-market valuation should be reported separately in the dedicated “of-which” column of the template. 96.The column “Memo item: Exposure value (pre-floor)” serves for the purpose of collecting pre- regulatory-floor risk weight data of securitization positions at the starting point. The total securitization exposures shall be reallocated to credit quality steps corresponding to the risk weights before the application of the regulatory floors. 97.While in general the definitions of the Methodological Note apply, a specific definition to be highlighted for the use of this template is the definition of exposure which is referred to in section 2.7 of the Methodological Note (in line with Article 248 of CRR. Moreover, whenever a term is not defined specifically in the Methodological Note, the definitions of CRR apply. 98.The columns of exposure values are to be reported net of cumulative specific credit risk adjustments and in accordance with Article 248(1) of CRR. 99.The allocation of different securitisation positions to the risk buckets identified in paragraph 189 of the Methodological Note should consider, as a reference, the structured finance asset classes reported in the ESMA central repository (CEREP). Some examples of these asset classes are provided in the Committee of European Securities Regulators’ (CESR) Guidelines for the implementation of the CEREP. In particular, the following should be considered: (i) Asset- backed securities (ABS) are securities backed by non-mortgage financial assets including auto/boat/airplane loans, student loans, consumer loans, health care loans, manufactured housing loans, film loans, utility loans, equipment leases, credit card receivables, tax liens, non-performing loans, credit linked notes, recreational vehicle loans and trade receivables; (ii) 29 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE Residential mortgage backed securities (RMBS) includes Prime and non-Prime RMBS, and Home equity loans (HEL); (iii) Commercial mortgage backed securities (CMBS) includes asset types such as retail or office property loans, hospital loans, care residences, storage facilities, hotel loans, nursing facilities, industrial loans and multifamily properties; and (iv) Collateralised Debt Obligations (CDO) are securities backed by a portfolio of bonds and/or loans, including Collateralised Loan Obligations, Collateralised Bond Obligations, Collateralised Synthetic Obligations, single-tranche CDO and Collateralised Fund Obligations. The definition of asset- backed commercial paper (ABCP) should be in line with Article 2 (7) and (8) of Regulation (EU) 2017/2402. Re-securitisations of any of the above instruments should be allocated only to risk bucket 3. Synthetic securitisations, incl. arbitrage synthetic securitisations, qualify as "other positions" and thus should be allocated to Risk Bucket 3. On-balance sheet synthetic securitisations, however, may be allocated to other risk buckets according to equivalent traditional securitisations as long as the underlying risk is equivalent. In this case, banks should provide in the explanatory note supporting evidence for the equivalence in risk and the related allocation to another Risk Bucket. The country of issue of the instrument should be the domicile of the underlying assets and, according to paragraph 183 of the Methodological Note, mixed pools should be allocated to the bucket that covers the highest share of total REA within the tranche. 100. Securitisation positions covered by Article 254(7) of CRR (“Other” securitisations with a RW of 1250% and not included in COREP in the four main regulatory approaches - SEC-IRB, SEC-SA, SEC-ERBA and SEC-IAA), shall be reported in the SEC-SA rows of this template. 101. Senior securitisation positions in the qualifying traditional NPE securitisation positions treated according to paragraph 3 of Article 269a of CRR shall be reported in the dedicated rows of this template. 102. REA projections are automatically calculated in this template by applying prescribed stressed risk weights to the projected exposure values. 103. Exposure values, REA projections and credit risk adjustments are linked via formula from this template to the CSV_CR_SEC_SUM template. 2.2.11 CSV_CR_NPL 104. This template contains end-of-year exposure values, both for total non-performing exposures and those of which were originated or modified after 26 April 2019 as per Article 469a of CRR, as well as components of the actual loss coverage, minimum loss coverage requirements and amounts of insufficient coverage related to non-performing exposures in the scope of Art. 47c of CRR as regards minimum loss coverage. 30 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE 105. Since the calendar provisioning is a prudential measure, the scope of this template should be aligned with the template CSV_CR_REA. 106. The rows show the exposure values, the total minimum coverage requirements, the available coverage and the applicable amount of insufficient coverage, respectively for the unsecured part of NPE and those parts of NPE that are either secured by immovable property or by other funded or unfunded credit protection or guaranteed or counter-guaranteed by an eligible protection provider, reflecting the application of the CRR3 as of 1 January 2025. 107. The exposure value of a debt instrument shall be its accounting value measured without taking into account any specific credit risk adjustments, additional value adjustments, amounts deducted, other own funds reductions related to the exposure or partial write-offs made by the institution since the last time the exposure was classified as non-performing. The exposure value of a debt instrument that was purchased at a price lower than the amount owed by the debtor shall include the difference between the purchase price and the amount owed by the debtor. 108. The exposure value of a loan commitment given, a financial guarantee given, or any other commitment shall be its nominal value, which shall represent the institution’s maximum exposure to credit risk without taking account of any funded or unfunded credit protection. In particular, the nominal value of financial guarantees given shall be the maximum amount the entity could have to pay if the guarantee is called on and the nominal value of loan commitments shall be the undrawn amount that the institution has committed to lend. Regulatory haircuts shall apply pursuant to Title II of Part Three of the CRR. 109. As per paragraph 565 of the Methodological Note, no forbearance measures shall be assumed during the stress test horizon. Banks shall consider forbearance measures in the starting points as per COREP 35.03. Exposures subject to forbearance measures should be reported according to the timing when the first measures were applied after the classification of the exposure as non-performing. 110. The amounts of available insufficient coverage shall correspond to the total provisions and adjustments or deductions at individual exposure level. In the template, banks shall provide the (uncapped) amount of available coverage by i) specific credit risk adjustments, ii) additional valuation adjustments, iii) other own funds reductions, iv) IRB shortfall, v) difference between the purchase price and the amount owed by the debtor, and vi) amounts written-off by the institution since the exposure was classified as non-performing. It is assumed that no write-offs should take place within the three-year horizon of the exercise, only amounts written-off by the institutions prior to the start of the exercise should be reported. Additionally, banks shall provide the total amount of available coverage capped by the minimum coverage requirements at exposure level, based on which the applicable amount of 31 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE insufficient coverage is determined as the positive difference between the minimum coverage requirements and the capped amounts of available coverage. The capped amount shall be calculated separately for each exposure as the lower amount between the minimum coverage requirement for this exposure and the total provisions and adjustments or deductions for the same exposure. The total applicable amounts of insufficient coverage are linked via formula from this template to the CSV_CAP template. 111. The columns provide further breakdowns based on the time passed since the exposures in scope of the NPL calendar were classified as non-performing, both for the starting point and the years over the projection horizon, as per Article 47c of CRR. 32 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE 2.3 Market Risk, CCR losses and valuation 2.3.1 CSV_MR_SUM 113. This template shows the impact of the adverse scenario on market risk positions according to chapter 3 of the Methodological Note (i.e. full balance sheet revaluation, stress impact on valuation reserves, CCR losses and NTI projections), with almost all of the data sourced from the CSV_MR_FULL_REVAL, CSV_MR_OPT_REVAL, CSV_MR_RESERVE, CSV_MR_CCR and CSV_MR_PROJ templates. 114. The rows cover the different items subject to the adverse scenario stress (e.g. balance sheet full revaluation, reserves, counterparty credit risk loss, NTI projections). The columns cover the years and a breakdown of 2025 figures by accounting treatment. 115. Apart from the memorandum items on defined pension plans (application of market risk stress on these items covered in section 6.4.6 of the Methodological Note) and discontinued operations (paragraph 208 of the Methodological Note), the following items should be directly populated by banks: The maximum FRTB-ASA REA for SBA & RRAO (as reported in COREP 91 from row 0020 to 0080 and from 0120 to 0130, column 190) of the last four quarters. , ‘VaR23 as of end of year 2024’. ‘75th percentile of daily VaR in 2024’ should be populated by the participating banks for 2024. If the 75th percentile of the 2024 VaR figures is not a single data point, the 75th percentile value should be obtained by taking the average between the two adjacent data points. 116. The reported daily VaR figures listed above shall be calculated based on all positions in the scope of the regulatory trading book using the regulatory VaR model. In case a regulatory VaR model is not available or covers only parts of the items in the regulatory trading book, internal models shall be used, to the extent possible, for the remainder of the regulatory trading book. in the explanatory note, banks should report parts of the regulatory trading book covered neither by regulatory nor by internal models. Regulatory VaRs shall be reported without using regulatory multiplication factors. 117. The drop-down menu in RowNum 1 of this template allows the specification of the approach followed. The differentiation of the three approaches, ‘comprehensive approach advanced’ (CA-adv), ‘comprehensive approach’ (CA) and ‘trading exemption’ (TE), for the 23 Considering a 10-day holding period. 33 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE purpose of this template is covered under section 3.3.1 of the Methodological Note. The definition of the scaling factor for the CA can be found in paragraph XX of the Methodological Note. 118. Balance sheet full revaluation projections are sourced from the CSV_MR_FULL_REVAL template, the stress impact on valuation reserves from CSV_MR_RESERVE, projected counterparty credit risk losses from CSV_MR_CCR, and NTI projections from CSV_MR_PROJ. 119. Depending on the accounting treatment, the adverse scenario stress projections are linked via formula from this template to the CSV_P&L template, except for the projections of NTI which are linked via formula to the CSV_MR_PROJ template, and the FVOCI impact projections which are linked to the CSV_CAP template. 2.3.2 CSV_MR_FULL_REVAL 120. The template contains the inputs and results for the full revaluation of positions under partial or full fair value measurement that should be subject to the market risk scenario as defined under section 3.4 of the Methodological Note. In the template, banks are requested to provide the following information as of the reference date: a. notional values of their positions, broken down by hedged items and hedging instruments b. fair values of their positions, broken down by hedged items and hedging instruments c. Granular first order sensitivities (‘delta’) of the positions to interest rate, FX, equity, commodity, and inflation risk (only CA and CA-adv banks) d. Granular second order sensitivities (‘gamma’ and ‘vega’) of the positions to interest rate, FX, equity, commodity, and inflation risk (only CA and CA-adv banks) 121. Banks should project in the adverse scenario in 2025 the following information: a. the overall gains & losses of the adverse scenario on P&L and OCI, broken down in columns by hedged items and hedging instruments. b. Breakdown of the sensitivities by risk factor (all banks, including TE banks) 122. To perform the full revaluation, banks should comply with the following procedure: 34 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE Box 1 Steps to perform the FULL REVALUATION matching items with hedges A. Re-organise items and related hedges according to the row structure of the CSV_MR_FULL_REVAL template a. Hedges accounted as FVPL (HfT) but related to items NOT in FVPL (HfT) should be considered in the rows of the items they refer to and reported in the appropriate columns (ColNums 3 to 5, 14 to 16, 26 to 28, 37 to 39), and excluded from the rows of FVPL (HfT) (RowNums 81 to 106). The re-allocation of FVPL (HfT) hedges should be such that no double counting occurs between these columns and rows. Example: Portfolios to be considered by row broken down in the appropriate columns Items Hedges Item: Amortised cost i1 i1 h1 Item: FVOCI i2 i2 h2 Total: FVPL (HfT) H H – (h1+h2+hn) hn (Of which: FVPL (HfT) hedges h1, h2 related to Item 1 and 2) (Of which: FVPL (HfT) hedges hn related to FVPL (HfT) positions) b. Micro-hedges should be considered in the rows related to the items they refer to. c. Hedges related to more than one item and portfolios of hedges should be considered and matched with those items they refer to according to i) the existence of an hedge-accounting relation under either IFRS 9 or IAS 39, and ii) the operational rationale of the hedges. Under the operational rationale of the hedges, banks are free to reallocate hedges in different rows following the criteria that best lead to a correct and realistic representation of the hedge adjustments. B. Perform the full revaluation according to the market risk scenario. C. Report notional amounts, FV, gains and losses, and sensitivities in the CSV_MR_FULL_REVAL template. Notice that notional amounts, FV, and gains and losses should be reported broken down in columns by hedged items and hedging instrument. Sensitivities should be reported in separate rows for items and hedges. 35 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE a. For the columns other than notional amount relative to hedges, amounts are to be reported by row net of asset and liabilities and net of long and short positions, i.e. the reported amount must be a correct, realistic, and consistently re-allocated image of the hedge amount/adjustment, which summed up to the values contained in the items columns, results in a correct and realistic FV and gains and losses by row. 123. The rows of the template show combinations of balance sheet position, IFRS 9 measurement type, type of instrument and related hedges, and type of counterparty. A detailed description of the row structure is provided in Table 9 of Annex I. 124. In the rows referring to FVPL (HfT) items (RowNum 81 to 106), banks should report only those items and related hedges that are entirely managed within the trading book. If hedges in the trading book relate to items not accounted in the trading book, the notional amount, FV, and gains and losses of such hedges should be reported in the rows relative to the items they refer to (and they should not be reported RowNums 81 to 106). The columns related to hedging instruments (ColNum 3 to 5, 14 to 16, 26 to 18) are expected to sum to the total amount of hedges reported by the bank and no double counting should occur between these columns and the rows related to FVPL HfT (RowNum 81 to 106). 125. Sensitivities should be reported by combinations of balance sheet position, IFRS 9 measurement type, and type of instrument. Sensitivities should be broken down between the part of the sensitivities corresponding to hedged items and that of hedging instruments. Sensitivities must also be reported separately for FVOCI Debt Instruments to General Governments. 126. For items held with a trading intent and their related hedges (from RowNum 81 to 106), the net between long and short positions must be reported. Depending on the instruments, long and short positions should be identified as follows: For linear instruments (equities, bonds, etc.) a long position refers to a position where an investor benefits from a rise in the price of the security. The opposite applies for short positions. For interest rates futures and forwards, a ‘short position’ means a position in which an institution has fixed the interest rate it will pay at some time in the future while a ‘long position’ means a position in which an institution has fixed the interest rate it will receive at some time in the future (in line with art. 328 point 2 of the CRR) 36 2025 EU-WIDE STRESS TEST – METHODOLOGICAL NOTE For plain vanilla swaps the criteria should be based on the fair value impact (positive or negative) coming from a change in the interest rates. In this regard, a fixed receiver swap instrument, where the fair value of the swap increases after a reduction in the interest rates, should be reported as a long position. On the other hand, a fixed payer swap instrument, where the fair value of the swap decreases after a reduction in the interest rates, should be reported as a short position. For reporting purposes swaps and derivatives should not be split in two legs. The notional, should be reported in full in the same row (either under long or short, depending on how it is classified) and should not be split in two halves. For basis swaps in the same currency and referring to the same index but with different maturities (i.e. swap 3m vs 6m Libor) the longer maturity leg have to be considered. The swap will be reported as “long” if the longer leg is received, or “short” if it is paid. For basis swaps with the two legs of the same tenor but referencing to two different indexes, banks are free to choose whether to consider the sensitivities as “long” or “short”. The criteria for the interest rates futures and forwards also apply to Credit Default swap (CDS). This means that if an increase in the credit spread would lead to a gain in the position, the CDS position should be reported as a "short position", and the corresponding sensitivity should be reported with a positive sign. Conversely, if an increase in the credit spread would lead to a loss for the CDS position, it should be reported as a "long position" and the corresponding sensitivity should be reported with a negative sign. For securities or cash denominated in a foreign currency (e.g.: forex forwards and futures), in case a depreciation of the foreign currency versus Euro would lead to a gain for the position, the corresponding FX sensitivity should be reported as "long position" with positive sign. If otherwise a depreciation of the foreign currency versus Euro would lead to a loss for the position, the corresponding FX sensitivity should be reported as "short position" with negative sign. In case of an appreciation of the foreign currency versus Euro the rules apply with opposite directions and signs. Cross currency swaps (XCC) with one leg denominated in Euro, should be reported as “long” if the fixed leg is received in EURO, while they should be reported as