Guidelines on Climate Stress Testing 2025

Guidelines on Climate Stress Testing 2025 Annexure 1 of FSD Circular No.1 of 2025 Financial Stability Department Guidelines on Climate Stress Testing 2025

Guidelines on Climate Stress Testing 2025 Annexure 1 of FSD Circular No.1 of 2025 The Team Shaukat Ali Executive Director Amer Hassan Director Dr. Jameel Ahmed Additional Director Sajawal Khan Senior Joint Director Muhammad Moaiz Siddiqui Joint Director Anum Naeem Deputy Director Mohammad Abdul Rehman Ansari Deputy Director Ali Inayat Assistant Director For queries and feedback, please contact: stresstesting@sbp.org.pk

Guidelines on Climate Stress Testing 2025 Annexure 1 of FSD Circular No.1 of 2025 Contents Definitions..........................................................................................................................................................1 Introduction........................................................................................................................................................2 Importance of Climate Change ...................................................................................................................2 5.1: Physical Risk Analysis for Banks, DFIs and MFBs...............................................................................4 Sensitivity Analysis for Physical Risk..........................................................................................................5 Physical Risk: Credit Shocks ........................................................................................................................5 Physical Risk Credit Shock One (PRCS-1): Losses Similar to Floods of 2022 ................................6 Physical Risk Credit Shock Two (PRCS-2): Losses Similar to Floods of 2025................................6 Physical Risk Credit Shock Three (PRCS-3): Losses on Vulnerable Districts and Sectors – Flood Hazard Ranking..............................................................................................................................6 Physical Risk Credit Shock Four for Banks and DFIs (PRCS-4): The Impact of sector-specific vulnerabilities to climate change on credit-worthiness of borrowers................................................8 Physical Risk Liquidity Shock (PRLS): Withdrawal of Deposits in High Flood Hazard Areas.........8 Assessing the Financial Impact of Physical Risk Analysis.......................................................................9 5.2: Transition Risk Analysis for Sample D-SIBs..........................................................................................9 Transition Risk Scenario – Imposition of Carbon Tax............................................................................9 Assessing Financial Impact for Transition Risk for Banks ...................................................................10 Interest Coverage Ratio ..........................................................................................................................10 Macro Stress Testing (MST) / Scenario Analysis for Sample D-SIBs.................................................11 Annexure – A: List of Sectors with Vulnerability Score ............................................................................12 Annexure B1 – List of Flood Affected Districts (PRCS-1 & PRCS-2)...................................................15 Annexure B2 – List of Flood Affected Districts (PRLS)...........................................................................16 Annexure – C: Carbon Emission Intensity by Sector ................................................................................17

Guidelines on Climate Stress Testing 2025 1 Definitions1 Domestic Systemically Important Banks (D-SIBs): D-SIBs are the banks designated under D-SIBs Framework 2018 (BPRD Circular No.04 of 2018). However, keeping in view the risk profile of an individual bank and market dynamics, SBP may identify additional bank(s) as D-SIB(s) for the purpose of these guidelines. Carbon Tax: A carbon tax (or energy tax) generally refers to a tax levied on the carbon content of some goods and services. The purpose is to reduce CO2 emissions by increasing the price of these goods and services. It is one of the main types of tools currently used in climate change policies around the world. NPLs: For the purpose of these guidelines, NPLs refer to exposures classified in Stage 3 under IFRS 9 or overdue as per the timeline prescribed in relevant prudential regulation2 . The estimated increase in NPLs under the shock shall be treated as “Loss” category for calculating the provisions / allowances. Emissions Intensity: A ratio that measures the amount of pollution, typically Greenhouse gases (GHGs) such as carbon dioxide (CO2), relative to a specific unit of activity or output. 1 FIs may also refer to SBP’s Regulatory Framework for Effective Management of Climate-Related Financial Risks, 2025 for other climate related definitions. 2 See, Annexure-I of BPRD Circular Letter No. 24 of 2021

Guidelines on Climate Stress Testing 2025 2 Introduction

1. State Bank of Pakistan (SBP) issued its first detailed guidelines on Stress Testing (ST) in 2005, 3 which were revised in 20124 and further updated in 20205 to account for the developments in global supervisory practices, availability of new techniques and changes in local banking business and regulations. Importance of Climate Change
2. Climate change and its related risks have emerged as the leading global challenge for economic and financial stability, with dire implications for the welfare of affected communities and parties. As encompassing as these risks are, measuring the adverse impacts on the performance and stability of the financial sector has become crucial. The climate related risks, viz., physical risk and transition risk, can transmit to the traditional risks faced by the financial institutions such as credit, market, liquidity, and operational risks. (See Table 1) Table 1: Potential effects of climate risk drivers (physical and transition risks) Risk Effects Credit Credit risk increases if climate risk drivers reduce borrowers’ ability to repay and service debt (income effect) or banks’ ability to fully recover the value of a loan in the event of default (wealth effect i.e. damages to collateral). Market Reduction in value of financial asset, including the potential to trigger large, sudden and negative price adjustments where climate risk is not yet incorporated into prices. Climate risk could also lead to a breakdown in correlations between assets or a change in market liquidity for particular assets, undermining risk management assumptions. Liquidity Banks’ access to stable sources of funding could be reduced as market conditions change. For instance, climate risk drivers may cause banks’ counterparties to draw down deposits and credit lines. Operational Increasing legal and regulatory compliance risk associated with climate-sensitive investments and businesses. Also includes damages to bank’s own infrastructure, e.g., branch building, IT equipment etc. Reputational Due to changing market or consumer sentiment, increasing reputational risk to such FIs, which are not adequately sensitive to the impacts of their products on climate. Source: BIS (2021). Climate Related Risk Drivers and Transmission Channels
3. Given that Pakistan ranks as the 5th most vulnerable country to extreme weather events like floods, heatwaves and droughts6 , the significance of climate change and the country’s high vulnerability to associated risk, the SBP has adopted climate change as one of the strategic themes in its Strategic Plan i.e., Vision 2028. 7 The Vision, inter alia, aims to promote allocation of resources towards development of a green economy, support the buildup of resilience against climate risks and preserve financial stability. As part of annual scenario analysis (macro-stress testing), SBP has been covering the likely impact of severe weather event (e.g. flooding) on solvency of banking sector. For instance, Financial Stability Review (FSR) 2018 was the first document to include climate-related risks in the hypothetical 3 BSD Circular 5 of 2005 4 BSD Circular 1 of 2012 5 FSD Circular No. 01 of 2020 6 Global Climate Risk Index 7 SBP Vision 2028: https://www.sbp.org.pk/SBPVision/Index.html

Guidelines on Climate Stress Testing 2025 3 shocks to conduct resilience analysis of banks. Later on, more focused and detailed climate stress testing, covering both physical and transition risks, were also published in FSR 2023 and FSR 2024. 4. Incidentally, post issuance of SBP’s revised Stress Testing (ST) Guideline in 2020, the Basel Committee on Banking Supervision (BCBS) has also issued Principles for Effective Management and Supervision of Climate-Related Financial Risks in 2022 ‘to promote a principles-based approach to improving risk management and supervisory practices related to climate-related financial risks’. 5. In line with BCBS Principles for climate-related financial risks, 2022 and best practices of peer countries on climate stress testing, SBP is issuing the following Climate Stress Testing (CST) Guidelines for SBP regulated financial institutions, as a supplement/addendum, i.e., Section 5, of existing Guidelines on Stress Testing, 2020. The new section aims to help financial institutions assess the impact of climate-related physical and transition risks on credit, market, operational and liquidity risks, and to set minimum regulatory expectations on this subject. 6. Section 2 of Stress Testing Guidelines, circulated vide FSD Circular No. 01 of 2020, which outlines Objectives of Stress Testing, Governance Structure, Resource Adequacy, Data and IT Infrastructure, shall continue to apply on this newly introduced Climate Stress Testing Guidelines. The Climate Stress Testing exercise should be forward-looking and shall be the primary tool in risk identification and monitoring at the institution. Accordingly, the climate stress testing exercise should explore the impacts of climate change and transition to a low carbon economy on the strategy and business model of FIs, identify relevant climate-related risks, measure impact of climate risks on exposures of FIs and potential losses etc. The exercise results should be reported to and discussed in the Risk Management Committee (RMC) regularly and may be used as an input in formulating strategic objectives and decisions. Furthermore, the sample D-SIBs should incorporate climate risk while designing scenarios for macro￾stress testing (MST) exercise, as required vide FSD Circular No. 01 of 2020. 7. Climate related Data: For an effective stress-testing exercise, the availability of accurate, relevant, reliable and appropriately extensive data in a timely manner is quite crucial. Since the management of climate-related risks is a relatively new field and rapidly evolving, the availability of quality data is a prime consideration. Financial Institutions (FIs) are advised to collect necessary climate-related data that affect their business models via credit, market, operational and liquidity risks. The nature and extent of data should be commensurate with the size and complexity of FIs’ exposure to climate risks. FIs may refer inter alia to National Disaster Management Authority (NDMA), National Disaster Risk Management Fund’s Natural Catastrophe (NatCat) Model, and provincial disaster management authorities of respective provinces for geospatial data on flood-affected districts as updated from time to time. Pakistan Green Taxonomy (PGT) provides a classification system to identify green activities / assets8 . FIs may also collect relevant borrower-specific data for conducting climate stress testing such as outstanding exposures / financing, borrowers’ carbon emissions9 , list of climate vulnerable economic sectors etc. In addition to the afore-mentioned sources, FIs may use any reliable source for climate risk related data after duly assessing the reliability and suitability of the source and data. Furthermore, most of the data required for conducting Climate Stress Testing is also acquired and maintained in different units 8 SH&SFD Circular No. 06 of 2025 9 For transition risk assessment, FIs shall start compiling borrowers’ emission data that are required to disclose under SECP’s directive dated December 31, 2024, as amended from time to time. For emissions data of the remaining significant borrowers, FIs may refer to the ‘sectoral emission intensities’ as a proxy from reliable sources.

Guidelines on Climate Stress Testing 2025 4 / divisions of the regulated FIs. The senior management of FIs shall ensure cross-functional integration and cooperation for availability of required data. 8. Climate Stress Testing and Risk Management: Climate stress testing is a rapidly evolving area of risk management and tools/techniques in this area are being refined to capture impact of climate risk accurately. It is faced with the challenges of data limitations and involves diversity and idiosyncrasies in the risk profile of FIs’ exposure to climate change due to unique circumstances of their borrowers. In this connection, these guidelines provide basic regulatory expectations for climate stress testing. In view of the complexity of environmental science, emerging climate change patterns, unique circumstances of each FI as well as its significant borrowers and exposures, the FIs may exercise due judgement while using the results of Climate Stress Testing to determine the materiality of climate￾related risks. Going forward, these guidelines will be further strengthened in the light of developments, experiences gained and the emerging best practices. In addition to these guidelines, FIs are encouraged to adopt more innovative and sophisticated approaches to model climate risk and conduct stress testing according to their risk profiles, size and availability of data. 9. The newly introduced Section 5 is divided into two parts: • 5.1 - Physical Risk Analysis for Banks, DFIs and MFBs covers scenarios for physical risk applicable on banks, DFIs and MFBs while • 5.2 – Transition Risk Analysis for Sample D-SIBs covers transition risk analysis applicable on the sample D-SIBs, as determined by the SBP. • Moreover, this section also requires the inclusion of climate risk in Macro Stress Testing / Scenario Analysis as conducted by sample D-SIBs. The relevant requirements for these exercises are as follows: Scope: Physical Risk: Banks (including Islamic banking operations of conventional banks), DFIs and MFBs; and Transition Risk: Sample D-SIBs. Balance sheet assumption: A static balance sheet shall be assumed in order to gauge the sensitivity of existing business models to potential shocks. Frequency: Physical and transition risk analyses will be conducted at least annually, based on end December data. Timeline: Physical and transition risk analyses will be conducted by second quarter of the following year. However, the first analysis may be conducted by end of Q3CY26 based on end-December 2025 data. 5.1: Physical Risk Analysis for Banks, DFIs and MFBs 10. Pakistan’s climate has experienced significant fluctuations in recent years, and flooding has emerged as a leading physical risk affecting vast agriculture plains, rural dwellings, urban centers and coastal areas of the country. Floods in Pakistan primarily occur during monsoon season damaging cash crops like cotton and rice and sometimes affect winter (Rabi) crops e.g. wheat due to prolonged water retention. Since its creation, Pakistan has faced major flood events almost every 3rd year, causing damages to the

Guidelines on Climate Stress Testing 2025 5 economy and carrying implications for the financial soundness of FIs. The significance of this risk is evident from the fact that prior to the unprecedented floods of 2010, the total cumulative damages caused to the national economy by the recurring floods amounted to around US$ 19 billion. Thereafter, the floods of 2010 alone inflicted losses of around US$ 10 billion, with further losses of US$ 30 billion caused by the 2022 pluvial floods. 11. In this backdrop, the physical risk analysis presented here attempts to estimate the incremental credit losses to the banking sector in case of flooding in different districts of the country in varying intensities. The scope of this exercise extends to the lending portfolio of banks (including Islamic banking operations of conventional banks), DFIs and MFBs. Sensitivity Analysis for Physical Risk 12. Historically, three episodes of intense floods have been recorded in Pakistan, which carry somewhat unique patterns and dynamics and affected different geographic areas of country. For instance, a. In 2010, the floods caused by heavy rains and melting of glaciers in Karakorum, Hindukush and catchment area of river Indus affected areas around Indus River basin. b. In 2022, heavier than usual monsoon rains in central and lower parts of the country, melting of glaciers and heatwaves brought about one of the worst floods, inundating one-third of the country land mass for prolonged period, displacing large swathes of population and taking a heavy toll on human lives, livelihoods and livestock. c. In 2025, unprecedented deluge in three rivers i.e. Ravi, Satluj and Chenab due to heavy rains in the Himalayas and adjacent lower ranges caused widespread destruction in both rural and some urban areas surrounding these rivers as well as in lower belt of Indus River. 13. In this backdrop, five shocks related to physical risk have been developed to test the resilience of banks, DFIs and MFBs to potential recurrence of another episode of such devastating floods: four credit risk shocks and one liquidity risk shock. 10 This exercise will enable FIs to assess the impact of climate risks from one year to another as the loan portfolio dynamics as well as vulnerability levels of hazard-prone areas are likely to change. Specifically, first two credit risk scenarios take into account the effects of temporal changes in loan portfolio, whereas, the remaining three scenarios also take into account the changes in vulnerability levels of hazard-prone districts in the light of changing climate patterns. The shocks will thus enable a regulated FI to assess resilience of its existing business model (loan portfolio) to either potential recurrence of historical stress events (e.g., floods) or changes in climate patterns. Physical Risk: Credit Shocks 14. The four credit risk scenarios can be broadly classified into two sets: a. First set of physical risk credit shocks (PRCS) include PRCS-1 and PRCS-2, which are applicable on banks, DFIs and MFBs. First two shocks are based on the actual (historical) losses experienced by FIs in the past and assess the impact on FIs’ asset quality (NPLs) and capital position if floods similar to 2022 and 2025 occur again in climate vulnerable sectors and locations. While, PRCS-3 is a hypothetical shock that encompasses the impact of all districts affected during flood episodes of 2022 and 2025. b. PRCS-4 (applicable only on banks and DFIs) stipulates banks/DFIs to design and apply an additional physical risk scenario in respect of large corporate borrowers for their material 10 Due to the nature of FIs’ business i.e. saving mobilization and lending, credit risk is the leading risk faced by them accounting for 67% of their required regulatory capital. Similarly, the liquidity risk carries significance due to its nature and potential to suddenly jeopardize the stability of FIs.

Guidelines on Climate Stress Testing 2025 6 exposure to major flood vulnerable areas. Furthermore, this scenario also encourages banks / DFIs to design and apply another additional physical risk scenario in respect of large corporate borrowers for their material exposure to other major physical risk hazards such as droughts, heatwaves etc. Physical Risk Credit Shock One (PRCS-1): Losses Similar to Floods of 2022 15. This scenario assumes recurrence of floods of 2022, considering the then affected districts. The list of affected districts of 2022 are placed at Annexure – B1. Two sub-scenarios are considered: a. Existing loan portfolio (fund based) in selected sectors and districts (see Annexure- A: Table￾1 for Banks, DFIs and Table-3 for MFBs) are stressed equivalent to the actual growth in NPLs for FIs observed during the one-year period following the floods of 2022, i.e., actual, post-floods sectoral growth of NPLs over four quarters i.e., Q3CY22 to Q2CY23. [Note: (i) Any flood-affected loans, which are rescheduled/restructured under any contingency regulatory relief schemes or under the existing flexibility in prudential regulations (e.g. R-5 of PRs for agri-financing for banks/DFIs) are to be added back in NPLs. Similarly, any flood￾affected loan that is directly written off by FIs should be added back into the loan and NPL portfolios to prudently assess the impact of flooding. (ii) In case the FI lacks data for the period or experienced a contraction in NPLs during the one-year period, the respective industry’s growth11 in overall NPLs may be used.] b. NPLs determined in (a) above is doubled. Physical Risk Credit Shock Two (PRCS-2): Losses Similar to Floods of 2025 16. This scenario assumes recurrence of floods of 2025, considering the then affected districts. The list of affected districts of 2025 are placed at Annexure – B1. Like previous flood scenario, two sub-scenarios are considered: a. Existing loan portfolio (fund based) in selected sectors and districts (see Annexure- A: Table￾1 for Banks, DFIs and Table-3 for MFBs) are stressed equivalent to the actual growth in NPLs for FIs observed over one-year period, post floods of 2025, i.e., actual sectoral growth of NPLs over four quarters i.e., Q3CY25 to Q2CY26. [Note given in PRCS-1 will also be applicable on PRCS-2] b. NPLs determined in (a) above is doubled. 17. FIs are encouraged to update the existing scenarios in case any flood event more severe than the floods of 2022 and 2025 occurs. Physical Risk Credit Shock Three (PRCS-3): Losses on Vulnerable Districts and Sectors – Flood Hazard Ranking 18. This scenario aims to check resilience of banking sector to floods in two dimensions: geographic and sectoral. For geographic vulnerability, it utilizes district-wise ‘vulnerability to floods’, as determined by latest available National Disaster Management Plan – 2025 of NDMA, as updated from time to time. For sectoral vulnerability, scoring is based on climate physical risk for the given sector (see Annexure￾A: Table-2 for Banks, DFIs and Table-3 for MFBs). This scenario, therefore, combines vulnerable districts and sectors (as shown in Table-2). 11 The banks, DFIs and MFBs may use their respective sectoral NPLs data as published in SBP’s Financial Soundness Indicators and Quarterly Compendium of the Banking Sector.

Guidelines on Climate Stress Testing 2025 7 19. This scenario considers districts with NDMP flood hazard ranking of Low, Medium or High12 that were flooded either in 2022 or 2025 and sectoral climate risk vulnerability scoring of Low, Medium or High. 13 20. In the first step, the Climate Vulnerable Exposure (CVE) is calculated for each district-sector combination based on the table below. For example, if the district vulnerability ranking is High and the sector vulnerability ranking is Medium then fund based exposure for given sector and district will be multiplied by the weight given in row 1 [High (100%)] and column 2 [Medium (30%)] of the table to get CVE for given district-sector combination. Table-2: CVE of Vulnerable Districts & Sectors Sectoral Vulnerability District Vulnerability High (100%) Medium (30%) Low (10%) High (100%) 1 0.3 0.1 Medium (50%) 0.5 0.15 0.05 Low (20%) 0.2 0.06 0.02 CVEi,j = αi,j × Fund Based Exposure where, i is the district (row in the matrix) and j is the sector (column in the matrix). The table is calculated based on the following workings. If the physical risk of a sector is ‘High’, then 100% of the Fund Based Exposure (FBE) of that sector will be considered. Further, if the sectoral physical risk score is ‘Medium’ or ‘Low’, 30% and 10% of FBE will be considered, respectively. These weights are further calibrated based on the vulnerability of the district. If the physical risk in the district is ‘High’, then 100% of the FBE in that district is considered. Further, if the district vulnerability is ‘Medium’ or ‘Low’, 50% and 20% of FBE will be considered, respectively. Hence, final weights of district-sector combination are calculated by multiplication of the respective vulnerabilities of given district and given sector. Once CVE is calculated for each district-sector combination, the Total CVE is calculated by adding CVEs for all sectors in all districts. In the second step, three scenarios are assumed for total CVE in vulnerable sectors and districts

1. 5 percent of CVE become non-performing
2. 10 percent of CVE become non-performing
3. 20 percent of CVE is become non-performing 12 See Annexure – B1 for district scoring, which is based on floods of 2022 or 2025. 13 See Table 2 in Annexure – A for sectoral scoring wherein physical risk vulnerability score of 1 denotes ‘Low’, physical risk vulnerability score of 2 denotes ‘Medium’ and physical risk vulnerability score of 3 denotes ‘High’.

Guidelines on Climate Stress Testing 2025 8 Physical Risk Credit Shock Four for Banks and DFIs (PRCS-4): The Impact of sector-specific vulnerabilities to climate change on credit-worthiness of borrowers 21. PRCS-4 is applicable only for banks and DFIs. All banks and DFIs shall design and apply an additional physical risk scenario in respect of large corporate borrowers for their material exposure to major flood vulnerable areas (see Annexure – B1 for list of districts). For the purpose of this scenario, large corporate borrower means corporate borrower having fund-based outstanding exposures greater than 5% of the bank or DFI’s equity as defined in Prudential Regulations for Corporate / Commercial Banking14 . For calculating the impact under this scenario, the banks/DFIs may take into account: i- The borrower’s idiosyncratic characteristics, which determine resilience against major flooding. ii- Factors that can materially affect (directly and indirectly) the credit worthiness and repayment capacity of the borrower in a major flooding event. These factors may include, for instance, the geographical location of the borrower’s operation including the consideration of any difference between place of loan origination (e.g., a major city) and place of utilization (e.g., operations in vulnerable area); damages to infrastructure and capital/operating assets in the vulnerable area; breakdown of supply chain including damage to raw material, inputs, stocks, crops, livestock, farmland, (as applicable), etc. 22. However, banks and DFIs are also encouraged to design and apply additional physical risk scenario(s) in respect of large corporate borrowers for their material exposure to other major physical risk hazards such as droughts, heatwaves etc. (for selection of hazard prone areas, please refer to National Disaster Management Plan - NDMP). 23. Banks/DFIs shall calculate the incremental Allowance for Expected Credit Losses (ECL) arising from the deterioration in the creditworthiness of the stressed large corporate borrowers, in accordance with the instructions prescribed under BPRD Circular No. 3 of 2022, BPRD Circular Letter No. 16 of 2024, and any other directives issued by SBP from time to time. The impact of such additional or incremental ECL (net of tax) shall be reflected in the Capital Adequacy Ratio (CAR) of banks/DFIs as a part of the final results of the stress test scenario. Physical Risk Liquidity Shock (PRLS): Withdrawal of Deposits in High Flood Hazard Areas 24. This Liquidity Risk Shock (LRS) is applicable for all banks, DFIs and MFBs. The LRS assumes that in districts with flood hazard score of 5 (High) as per the NDMA’s latest available NDMP Flood Hazard Plan15 (see Annexure – B2), demand for cash increases leading to substantial deposit withdrawals over a period of three days as follows: a. Withdrawal of 5 percent deposits in the affected area on Day 1; b. Withdrawal of 5 percent deposits on Day 1 and additional 5 percent on Day 2; c. Withdrawal of 5 percent on Day 1, 5 percent on Day 2 and an additional 5 percent on Day 3. 25. For guidance, FIs may also refer to physical risk scenarios outlined in Box 4.1: Climate Risk Scenario Analysis (FSR 2023) 14 Prudential Regulations for Corporate / Commercial Banking 15 Currently available Plan can be found at: NDMP Plan 2025.

Guidelines on Climate Stress Testing 2025 9 Assessing the Financial Impact of Physical Risk Analysis 26. To carry out physical risk analysis, FIs may use district-wise, fund-based outstanding exposures / financing in vulnerable sectors (list of sectors at Table-1, 2 and 3 of Annexure-A for different shocks) and list of flood affected districts given for floods of 2022 and 2025 (Annexure – B1) and National Disaster Management Plan 2025 (Annexure – B2) 27. Credit Risk: Under all credit scenarios, FIs are required to: i. Calculate post-shock increase in NPLs for different physical risk scenarios ii. Calculate additional provisions / allowances (assuming incremental NPLs as loss), iii. Charge the additional provisions / allowances (net of taxes) to the capital and iv. Estimate the post-shock capital adequacy ratio (CAR). The post-shock CAR CARPS = Capital − (1 − τ)∆Prov RWA − ∆Prov where, τ is tax rate 28. Liquidity Risk: The amount of withdrawal of the deposits should be deducted from the liquid assets and the level of remaining liquid assets needs to be re‐calculated after each day. The post shock liquid assets to total assets ratio may then be calculated. 29. Apart from the scenarios prescribed above, banks, DFIs and MFBs are encouraged to explore other scenarios for physical risk such as heatwaves, droughts, landslides, smog, earthquake, snowfall, hailstorm etc., which impacts their strategic objectives, business model, operations and loan and investment portfolios over relevant time horizons. Moreover, changing frequency and intensity of climate hazards over time may result in losses higher than historical experiences; accordingly, banks, DFIs and MFBs are also encouraged to explore the impact of additional forward-looking shock scenarios of physical risk on credit risk, market risk (for example, variation in liquidity across assets exposed to climate-risk and variation in the speed at which exposures could be impacted), operational risk and liquidity risk. 5.2: Transition Risk Analysis for Sample D-SIBs 30. Transition risks, inter alia, entails financial risks that result from moving to a low carbon economy. They are driven by changes in policies, technology, market sentiment or customer behaviors. Transition Risk Scenario – Imposition of Carbon Tax 31. In line with the international commitments and climate risks, Pakistan and other countries are working towards low-carbon and resource efficient economy. 16 For this purpose, policymakers use different incentive structures such as imposition of carbon tax, stricter emission limits, introduction of carbon 16 As per latest NDC (Sep-2025), Pakistan has committed to reduce projected emission up to 50 percent by 2035.

Guidelines on Climate Stress Testing 2025 10 markets etc. that affect the performance and financial soundness of firms. Moreover, changes in technology as well as the sentiments of customers, investors and market could lead to economic dislocation and a reassessment of the value of a variety of financial assets. It is important to note that firms can be exposed to both direct impact e.g. imposition of carbon tax in the country as well as indirect impact e.g. carbon tax on importers in EU under Carbon Border Adjustment Mechanism can impact the exports and revenues of domestic firms. For assessing the impact of transition risk on firms and financial institutions, a scenario of imposition of carbon tax has been suggested. The carbon tax reduces the bottom line of firms, thus affecting their financial performance as well as the repayment capacity to honour obligations to financial institutions. 32. Transition risk scenario is mandatory for sample D-SIBs; other banks and DFIs are encouraged to incorporate transition risk scenario in their climate risk analysis. For transition risk scenario, imposition of a carbon tax on corporates/firms based on their contribution to greenhouse gas (GHG) emissions is assumed. The tax would impact their profitability and hence the repayment capacity. In particular, the pre- and post-tax imposition Interest Coverage ratio (ICR) of corporate borrowers are assessed. Specifically, depending on which sector the firm belongs, the emissions may be estimated as 𝐸𝑚𝑖𝑠𝑠𝑖𝑜𝑛 = [𝑆𝑎𝑙𝑒𝑠 ÷ 𝑇ℎ𝑟𝑒𝑠ℎ𝑜𝑙𝑑] × 𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦, where, ‘Threshold’ is the PKR equivalent of one million US dollars. 33. For calculation of emissions, FIs shall start compiling borrowers’ emission data that are required to be disclosed under SECP’s directive dated December 31, 2024, as amended from time to time. For emissions data of the remaining corporate borrowers, FIs may use ‘sectoral emission intensities’ from any reliable source as a proxy. For reference, Standard & Poor’s data17 on GHG emissions of global industries is given at Annexure – C 34. Based on the emission levels of the corporate/firm, the Carbon Tax can be calculated as: 𝐶𝑎𝑟𝑏𝑜𝑛 𝑇𝑎𝑥 = 𝐸𝑚𝑖𝑠𝑠𝑖𝑜𝑛 × 𝑇𝑎𝑥 𝑅𝑎𝑡𝑒 The indicative tax rates may range from USD 5/tCO2e (i.e. US$ 5 per metric ton of carbon dioxide equivalent) to USD 50/tCO2e (in equivalent PKR). Incidentally, IMF (2021) suggests a floor of USD 25/tCO2e for lower income emerging countries.18 Assessing Financial Impact for Transition Risk for Banks Interest Coverage Ratio 35. To calculate the shock, FIs would use financial statements data of corporate borrowers, adjust the earnings before interest and taxes (EBIT)19 for different levels of carbon tax expenses, i.e., 5, 25 and￾50 US dollars per tCO2e, and calculate post shock ICRs of corporate borrowers as: 17 See S&P (2021), “Transition Risk: Historical Greenhouse Gas Emissions Trends for Global Industries” for sectoral emission intensities. In case of unavailability of borrower-level carbon emissions, for now, FIs may use sector-level emissions provided by S&P in “Transition Risk: Historical Greenhouse Gas Emissions Trends for Global Industries, 2021” (or an updated version) as a proxy to calculate borrower emission intensities. For reference, Carbon Emissions by Sector published by S&P Global is also provided at Annexure-C. 18 World Bank (2023). State and Trends of Carbon Pricing Dashboard. 19 Keeping in view the fact that any carbon tax may also affect the market competitiveness including sales and operating expenses of the firm, banks / DFIs are encouraged to incorporate these potential effects into EBIT calculations.

Guidelines on Climate Stress Testing 2025 11 ICRps = EBIT − Carbon Tax Interest Expenses 36. Map fund based outstanding exposures / financing with the post-tax ICRs. For the purpose this shock, the exposures with ICRps < 1, may be considered as NPLs. Finally, adjust the capital for incremental NPLs and calculate post-shock CAR. 37. Segregate the exposures into sectors, identify vulnerable sectors and the sectors affecting bank’s solvency due to potential transition to low carbon economy. 38. Encouraged Set of Stress Tests: All Banks / DFIs are also encouraged to perform transition risk stress tests; they may also use other techniques such as estimating stressed Probabilities of Default (PD) due to imposition of Carbon tax and calculate the Expected Credit Losses (ECLs). In this regard, banks may refer to Box 4.1: Climate Risk Scenario Analysis (FSR 2023) and Box 4.1: Climate Transition Risk and Financial Stability in Pakistan (FSR 2024) and other SBP analysis on Climate Stress Testing published from time to time. Furthermore, all banks / DFIs are encouraged to conduct climate stress testing over relevant time horizons by considering any scenarios identified by Network for Greening the Financial System. 39. Apart from the scenarios prescribed above, banks / DFIs are encouraged to explore other scenarios for transition risk such as policy changes, technological advancements, market shifts, evolving consumer preferences, increased supervision and asset repricing risks etc. which impact their strategic objectives, business model, operations and loan / financing and investment portfolios over relevant time horizons. Furthermore, banks / DFIs are encouraged to explore the impact of transition risk on credit risk, market risk (for example, variation in liquidity across assets exposed to climate-risk and variation in the speed at which exposures could be impacted), operational risk and liquidity risk. Macro Stress Testing (MST) / Scenario Analysis for Sample D-SIBs 40. The Sample D-SIBs under FSD Circular No. 1 of 2020, shall design scenarios that will invariably include climate-related risks in their MST exercises by considering the potential impact of climate change on the domestic economy when determining the magnitude / severity of stress applied to risk drivers. 20 20 See Chapter 4 – Resilience of the Banking Sector in various editions of SBP FSR for guidance.

Guidelines on Climate Stress Testing 2025 12 Annexure – A: List of Sectors with Vulnerability Score Table-1: List of Vulnerable Sectors for Banks / DFIs for PRCS-1 and PRCS-2 S. No. Sector Code (Based on ISIC 4 Classifications of Private Sector Business) Name of Sector 1 41000000000 Agriculture, Forestry and Fishing 2 41100000000 Mining and Quarrying 3 41210000000 Manufacture of food products 4 41213000000 Manufacturing – Textiles 5 41214000000 Manufacture of wearing apparel 6 41215000000 Manufacturing - Leather and Related Products 7 41216000000 Manufacture of wood and of products of wood 8 41217000000 Manufacturing - Paper and Paper Products 9 41500000006 Construction 10 41749000000 Land transport and transport via pipelines 11 41855000000 Accommodation 12 42000000000 Real estate activities 13 60200000000 Consumer Financing Note: This dataset of advances on ISIC-IV sectors with district-wise detail is compiled and reported by regulated FIs to SBP via DWH DAP-portal on a quarterly basis.

Guidelines on Climate Stress Testing 2025 13 Table-2: List of Sectors for Banks / DFIs for PRCS-3 and PRCS-4 21 S. No. Sector Code (Based on ISIC 4 Classifications of Private Sector Business) Name of Sector Vulnerability Score Physical Risk Transition Risk 1 41000000000 Agriculture, Forestry and Fishing 3 3 2 41100000000 Mining and Quarrying 2 3 3 41210107200 Manufacturing - Sugar 2 2 4 41213000000 Manufacturing - Textiles22 5 41212000000 Manufacturing - Tobacco Products 2 2 6 41215000000 Manufacturing - Leather and Related Products 2 2 7 41217000000 Manufacturing - Paper and Paper Products 2 2 8 41220000000 Manufacturing - Chemicals and Chemical Products 2 2 9 41222000000 Manufacturing - Rubber and Plastics Products 2 2 10 41224000000 Manufacturing - Basic Metals 2 3 11 41232000000 Manufacturing - Others 2 3 12 41300000000 Electricity, gas, steam and air conditioning supply (Other than Thermal & Coal Based) 2 2 13 41335351020 Electricity (Thermal) 2 3 14 41335351030 Electricity (Coal Based) 3 3 15 41500000006 Construction 3 3 16 41600000000 Wholesale and retail trade; repair of motor vehicles and motorcycles 2 1 17 41700000000 Transportation and storage 2 3 18 41900000000 Information and communication 1 1 19 42000000000 Real estate activities 2 2 20 60200000000 Consumer Financing 3 2 21 70000000000 Others 2 1 Note: This dataset of advances on ISIC-IV sectors with district-wise detail is compiled and reported by regulated FIs to SBP via DWH DAP-portal on a quarterly basis. 21 The vulnerability scores are based on feedback received from the industry. Vulnerability scores of 1, 2 and 3 denote ‘Low’, ‘Medium’ and ‘High’ risk, respectively. 22 Banks / DFIs are encouraged to use higher vulnerability score for transition risk in respect of such sub-segments of textile sector that involve significant use of steam/thermal fuels, high emissions, strict buyer requirements, high liquid effluent, complex use of chemicals etc.

Guidelines on Climate Stress Testing 2025 14 Table-3: List of Sectors for MFBs for PRCS-1, PRCS-2 and PRCS-3 S. No. Reporting Chart of Account (RCOA) Code Sector Physical Risk Vulnerability Score 1 061200001 Enterprises 2 2 061200002 Agriculture 3 3 061200003 Livestock 3 4 061200004 Long Term Housing Finance 2 5 061200005 Consumer Lending 3 6 061200006 Others 2 Note: MFBs report sector-wise data of advances with district-wise breakdown to SBP vide DWH-DAP on a quarterly basis (ISIC￾IV). This district-wise data may be mapped along above-mentioned sectoral categories.

Guidelines on Climate Stress Testing 2025 15 Annexure B1 – List of Flood Affected Districts (PRCS-1 & PRCS-2) District NDMP Score Vulnerability District NDMP Score Vulnerab ility District NDMP Score Vulnerab ility District NDMP Score Vulnerab ility District NDMP Score Vulnerab ility District NDMP Score Vulnerab ility Abbottabad 3 Low Ghanche 3 Low Khairpur 4 Medium Lower Dir 4 Medium Okara 3 Low Sohbatpur 5 High Astore 3 Low Ghizer 3 Low Khanewal 3 Low Mansehra 4 Medium Panjgur 3 Low Sujawal 4 Medium Badin 3 Low Ghotki 5 High Kharan 3 Low Mastung 3 Low Pishin 3 Low Sukkur 5 High Bahawalpur 3 Low Gujranwala 5 High Khuzdar 3 Low Matiari 4 Medium Poonch/Rawalakot 4 Medium Swabi 4 Medium Bahwalnagar 3 Low Harnai 3 Low Killa Abdullah 3 Low Mianwali 5 High Quetta 4 Medium Swat 5 High Barkhan 2 Extremely Low Hyderabad 4 Medium Killa Saifullah 3 Low Mirpur Khas 3 Low Rahim Yar Khan 5 High Tando Allahyar 3 Low Batagram 3 Low Jacobabad 4 Medium Kohistan 4 Medium Multan 4 Medium Rajanpur 5 High Tando Muhammad Khan 4 Medium Bhakkar 4 Medium Jaffarabad 5 High Kohlu 3 Low Musakhel 3 Low Sahiwal 3 Low Tank 5 High Chaman 3 Low Jamshoro 4 Medium Kolai Palas 3 Low Muzaffargarh 5 High Sanghar 3 Low Thatta 4 Medium Charsadda 4 Medium Jhang 5 High Kurram 4 Medium Nagar 4 Medium Sargodha 3 Low Umer Kot 3 Low Chiniot 5 High Kachhi/Bolan 3 Low Lakki Marwat 3 Low Narowal 4 Medium Shaheed Benazirabad 4 Medium Upper Chitral 5 High D.I.Khan 5 High Kalat 3 Low Larkana 5 High Nasirabad 5 High Shangla 4 Medium Upper Dir 4 Medium Dadu 5 High Kambar/Shahadad Kot 5 High Lasbela 4 Medium Naushahro Feroze 5 High Sheikhupura 4 Medium Usta Muhammad 5 High Der Ghazi Khan 5 High Karak 3 Low Layyah 4 Medium Neelum 4 Medium Shikarpur 5 High Vehari 3 Low Dera Bugti 3 Low Kashmore 5 High Loralai 3 Low Nowshera 5 High Sialkot 4 Medium Zhob 3 Low Diamer 4 Medium Kasur 3 Low Lower Chitral 5 High Nushki 3 Low Sibi 3 Low Ziarat 4 Medium Source: NDMA, PDMA List of Flood Affected Districts in 2022 Floods along with NDMP Flood Hazard Score District NDMP Score Vulnerab ility District NDMP Score Vulnerab ility District NDMP Score Vulnerability District NDMP Score Vulnerab ility District NDMP Score Vulnerability BAHAWALNAGAR 3 Low HYDERABAD 4 Medium KHANEWAL 3 Low NAROWAL 4 Medium SHEIKHUPURA 4 Medium BAHAWALPUR 3 Low JACOBABAD 4 Medium KOHLU 3 Low NASIRABAD 5 High SHIKARPUR 5 High CHINIOT 5 High JAFFARABAD 5 High LARKANA 5 High NAUSHAHRO FEROZE 5 High SIALKOT 4 Medium DADU 5 High JAMSHORO 4 Medium LEHRI 4 Medium OKARA 3 Low SIBI 3 Low DERA BUGTI 3 Low JHAL MAGSI 5 High MANDI BAHAUDDIN 4 Medium PAK PATTAN 3 Low SUJAWAL 4 Medium FAISALABAD 4 Medium JHANG 5 High MATIARI 4 Medium RAHIM YAR KHAN 5 High SUKKUR 5 High GUJRANWALA 5 High KACHHI(BOLAN) 3 Low MIRPUR 2 Extremely Low RAJANPUR 5 High TANDO ALLAH YAR 3 Low GUJRAT 4 Medium KALAT 3 Low MULTAN 4 Medium SANGHAR 3 Low TOBA TEK SINGH 4 Medium HAFIZABAD 4 Medium KASHMORE 5 High MUZAFFARGARH 5 High SHAHDAD KOT 5 High VEHARI 3 Low HARNAI 3 Low KASUR 3 Low NANKANA SAHIB 3 Low SHAHEED BENAZIRABAD 4 Medium List of Flood Affected Districts in 2025 Floods along with NDMP Flood Hazard Score Source: Districts where more than 3% of Land Areas was Flood Affected according to Satellite Data

Guidelines on Climate Stress Testing 2025 16 Annexure B2 – List of Flood Affected Districts (PRLS) Districts Provinces Flood Hazard Score Flood Hazard Ranking Bajaur KP 5 High Chiniot Punjab 5 High Chitral KP 5 High D.G. Khan Punjab 5 High D.I. Khan KP 5 High Dadu Sindh 5 High Ghotki Sindh 5 High Gujranwala Punjab 5 High Jaffarabad Balochistan 5 High Jhal Magsi Balochistan 5 High Jhang Punjab 5 High Kashmore Sindh 5 High Larkana Sindh 5 High Mianwali Punjab 5 High Muzaffargarh Punjab 5 High Naseerabad Balochistan 5 High Naushero Feroze Sindh 5 High Nowshera KP 5 High Peshawar KP 5 High Qambar Shahdadkot Sindh 5 High Rahim Yar Khan Punjab 5 High Rajanpur Punjab 5 High Shikarpur Sindh 5 High Sohbatpur Balochistan 5 High Sukkur Sindh 5 High Swat KP 5 High Tank KP 5 High Usta Muhammad Balochistan 5 High List of Flood-Affected Districts for Physical Risk Liquidity Shock (PRLS)*

• List of Flood Affected Districts is based on districts with flood hazard score of 5 as per National Disaster Management Plan 2025, as updated from time to time.

Guidelines on Climate Stress Testing 2025 17 Annexure – C: Carbon Emission Intensity by Sector Carbon Emissions by Sector published by S&P Global23 Sector Emission intensities per Million USD revenue* Sector Emission Intensity (Million USD) ** Auto 16 Cement 888.0 Chemical 888.0 Food 111.7 Fuel & Energy 2036.0 Paper Products 888.0 Sugar 111.7 Textile 21.6 Others *** 25.0 Source: Standard & Poor’s Global *Emission intensities in Metric tons of Carbon dioxide per million US dollars. ** Sectoral Carbon Emissions Intensity is based on S&P (2021), "Transition Risk: Historical Greenhouse Gas Emissions Trends for Global Industries" *** For ‘Others’ category, estimates are based on median of different sectors in S&P (2021). 23 See S&P (2021), “Transition Risk: Historical Greenhouse Gas Emissions Trends for Global Industries” for sectoral emission intensities.