## 1phlea2022002 - EXECUTIVE SUMMARY ## Source details **Canonical URL:** [1phlea2022002 - EXECUTIVE SUMMARY](https://www.imf.org/-/media/files/publications/cr/2022/english/1phlea2022002.pdf) ## Other formats - [Markdown version](/-/media/files/publications/cr/2022/english/1phlea2022002.pdf.md) - [Structured JSON version](/-/media/files/publications/cr/2022/english/1phlea2022002.pdf.json) --- ### Macrofinancial setting and system structure - Total assets of the financial system amount to 126 percent of GDP. - The banking system holds about 94 percent of the system’s assets. - Bank credit is just over 50 percent of GDP and mostly goes to nonfinancial corporates (NFCs). - Only a third of adults have formal accounts. - Forty-six universal and commercial banks (UKBs) hold over 94 percent of bank assets. - 60 percent of UKB assets are held by the top five banks (all domestic). - D-SIBs, including branches of some Global-SIBs, hold about 80 percent of the banking sector. - Foreign bank subsidiaries and branches hold seven percent of bank assets. - About 500 small thrift banks (TBs) and rural and cooperative banks (RCBs) exist. - Banks follow a traditional commercial banking business model, relying on deposits and lending mostly to large NFCs; 80 percent of the loans go to NFCs. - Bank balance sheet structure (In percent of total assets): - Cash, 15 - Security, 24 - Interbank, 2 - Loan, 55 - Other asset, 5 - Liability structure (In percent): - Peso deposit, 62 - FX deposit, 12 - Bill and bond, 8 - Other liability, 5 - Equity, 13 - Banks are liquid with nearly 40 percent of their assets in securities and central bank reserves. - Fintech ecosystem is nascent: a quarter of adults made or received at least one digital payment in the preceding year (2017 Global Findex). ### COVID-19 shock, macro impact, and recovery projection - Real 2020 GDP contracted by 9.5 percent. - Government containment measures resulted in 12 percent (half a year on half a year, seasonally adjusted.) real GDP contraction in the first half of 2020. - Recovery quarterly growth (seasonally adjusted, quarter-on-quarter): - Third quarter 2020: real GDP increasing by 8.0 percent. - Fourth quarter 2020: real GDP increasing by 5.6 percent. - The economy is expected to grow 6½ percent in 2021 (document: “expected to grow 6½ percent in 2021”). - Fund projection (January 2021 WEO): 2021 real GDP growth to be 6.6. percent. - The Philippines’ 2020 revision to real GDP growth rate forecast: -15.8 (difference between October 2019 and January 2021 WEO forecasts). ### Stress-testing framework and methods used - Bank solvency and liquidity stress tests follow the standard FSAP method (top-down, quasi-static balance-sheet; 3-year horizon). - Estimated models gauge second-round effects of bank solvency results on GDP, focusing on the credit channel via a bank-by-bank panel credit growth model and an SVAR macro-financial model. - Scenario-based NFC stress test focused on 151 sample firms (147 publicly listed) accounting for about 44.4 percent of outstanding NFC debt and 46.1 percent of banks’ total loan portfolio (net of reverse repos). - Joint bank–NFC cash-flow liquidity stress test used to gauge effects of loan moratoria and rollover behaviors. - Stress tests do not incorporate mitigating policy effects such as credit guarantees, regulatory responses, and loan moratoria (except for system-wide liquidity analysis). ### Macroeconomic scenarios used in stress tests - Four common macroeconomic scenarios considered: - Baseline: - Follows October 2020 WEO; sharper 2-year cumulative GDP contraction than AFC; equivalent to a three-standard deviation shock to the pre-COVID WEO forecast (as of January 2020) showing growth rates near potential (6½ percent per year for 2020–24). - Upside: - Incorporates national authorities’ forecast as of September 2020 with slightly more optimistic 2020 GDP. - Adverse: - Real GDP growth would contract by 11 percent in 2020; first 2-year cumulative growth equals a ⅔ standard deviation shock to the baseline (nearly a four standard deviation shock to the January 2020 WEO forecast). - Severe adverse: - First 2-year cumulative growth would reach -6½ percent, equivalent to a 1⅓ standard deviation shock to the baseline. ### NFC corporate sector stress and key metrics - Pre-pandemic (end-2019): - Median ICR about 3.6; debt-weighted ICR about 4.9. - Median cash ratio in 2019 about 24 percent. - Share of NFC debt held by firms with ICR below one rose from about 1.7 percent in 2016 to about 5.3 percent in 2019. - Share of sample firms with ICR below one in 2019 was about 22 percent. - Debt-weighted-average interest coverage ratio (ICR) (preserved phrasing from source): - 4.9 percent at end-2019 - 1.3 percent in the upside/baseline (context: “decline from 4.9 percent at end-2019 to 1.3, below one, and 0.2 percent in the baseline, adverse, and severe adverse scenarios, respectively”) - below one in the adverse scenario - 0.2 percent in the severe adverse scenario - NFC stress-test outcomes (end-2020 estimates): - Under baseline: - Median ICR expected to decline to about 0.9 in 2020 (from 3.6 in 2019). - Debt-weighted mean ICR expected to decline to about 1.3 in 2020 (from 4.9 in 2019). - Share of NFC debt-at-risk would jump to 55 percent in 2020 (from 5.3 percent in 2019). - Share of firms with ICR below one expected to rise to about 48 percent of total sample firms in 2020 (from 18.3 percent in 2019). - Under stress scenarios: - Debt-at-risk: - about 45 percent in the upside scenario - about 76 percent in the adverse scenario - about 81 percent in the severe adverse scenario - Cash positions and cash ratios: - System-wide cash-to-assets ratio would drop to 8.7 percent in 2020 (without mitigating policies), down from 9.7 percent in 2019. - Median cash ratio would drop to about 23 percent in 2020 from 24 percent in 2019. - Implication: NFC distress could significantly increase NPLs without private and public sector actions or economic recovery; loan moratoria can help firms survive temporary liquidity shocks but may only delay bankruptcies if crisis is prolonged or business models are fundamentally challenged. ### Bank solvency stress-test results and systemic risk - Coverage: 542 banks (46 UKBs, 49 TBs, 447 RCBs) representing 100 percent of the banking system’s assets. - System aggregate total capital adequacy ratio (CAR) — figures at the end of the stress test horizon (2022): - 15.6 percent at baseline starting point (Latest actual) - falls to 11.7 percent by 2022 in the baseline scenario - falls to 9.3 percent in the adverse scenario - falls to 4.9 percent in the severe adverse scenario - Regulatory minimum (hurdle) for CAR used: 10 percent. - Bank failure counts (number of banks not meeting minimum CAR by end-2022): - Baseline October: Total 185 - Upside: Total 178 - Adverse: Total 201 - Severe Adverse: Total 214 - Capital shortfalls (system total, percent of GDP): - Baseline October: 1.0 - Upside: 0.5 - Adverse: 1.9 - Severe Adverse: 3.9 - UKBs: - CAR UKB under scenarios: - Latest actual: 15.3 - Baseline October: 11.0 - Upside: 13.1 - Adverse: 8.5 - Severe Adverse: 3.7 - CET1R UKB under scenarios: - Latest actual: 12.7 - Baseline October: 8.9 - Upside: 10.8 - Adverse: 6.5 - Severe Adverse: 2.1 - CET1 regulatory minimum applied: 6 percent (applies only to UKBs for this exercise). - Heterogeneity and vulnerabilities: - Median CAR among UKBs: 21.5 percent; interquartile range: 15.5 percent to 70.6 percent. - D-SIBs median CAR: 15 percent; non-D-SIBs median CAR: 34 percent. - State-owned banks (three banks representing 15 percent of assets) fail even in the upside scenario; CAR drops from 13.7 percent to 1.8 percent under the baseline and to -0.9 percent under the adverse scenario. - System-wide capital depletion from the starting point amounts to 3.9 percent. ### Liquidity profile and stress-test findings - Reserve requirement: 14 percent at end-2019; reduced to 12 percent when COVID-19 hit (increased usable buffer modestly). - High-quality liquid assets (HQLA) for LCR calculation are mostly reserves and sovereign securities. - Funding structure: - Retail and wholesale deposits are main funding sources; customer deposits account for the majority of funding. - Dollarization: 15 percent of deposits are in FX; 11 percent of loans are in FX. - LCR and NSFR outcomes: - System all-currencies LCR (Basel 2013): 157 percent. - System FX LCR: 239 percent. - Median individual bank total currency LCR: 207 percent. - Median individual bank FX LCR: 88 percent. - Under combined severe shock, system all-currencies LCR falls to 69 percent and system FX LCR falls to 80 percent. - NSFR: exceeds 100 percent in most banks; some banks slightly below 100 percent; very few (mainly foreign branches) further below. - Cash-flow stress test (CFST) results: - Only a few banks show net funding gaps; most gaps are marginal and largest at 1-month horizon. - Only one bank shows a shortfall of 1.8 percent of GDP. - All banks have enough required reserves to cover their shortfalls; if banks are allowed to use all reserves, all universal and commercial banks can survive severe cash-flow stress for months. - System appears more resilient to FX liquidity shocks than local currency liquidity shocks. - Recommendation: the BSP could consider introducing FX LCR. ### Bank–NFC liquidity linkage and loan moratoria effects - Direct effect of loan moratoria: improves NFC cash balance while reducing bank cash inflows and liquid assets. - Key behavioral parameters: - Moratorium utilization rates examined: 0 percent, 50 percent, 70 percent, 90 percent. - Moratorium lengths examined: five months (already took place) and 12 months (potential extension). - Bank rollover rate central assumption: 90 percent; pessimistic assumption: 50 percent. - Selected NFC liquidity outcomes (preserving figures as presented): - Under 12-month moratoria, 70 percent moratorium utilization, and 90 percent rollover: - NFC Cash-to-AssetsRatio (In percent of end-2019 total assets): 9.7; 8.8; 9.0; 10.6; 10.8 (series shown in source figures). - Under 12-month moratoria, 70 percent moratorium utilization, and 50 percent rollover: - NFC Cash-to-AssetsRatio: 9.7; 5.3; 6.3; 9.6; 9.8. - Key implications: - If banks continue to roll over maturing NFC loans, NFC liquidity balances recover with or without moratoria. - For NFCs, moratoria substantially improve liquidity when banks’ rollover rate is low but less so otherwise. - For banks, moratoria effects on liquidity depend critically on whether NFCs have alternative financing sources; if NFCs withdraw deposits, banks may experience broadly the same cash-flow effects irrespective of moratoria. - At system level, moratoria often reallocate liquidity between banks and NFCs rather than changing aggregate system liquidity. - Recommendation: the Bangko Sentral ng Pilipinas (BSP) should monitor banks’ and NFCs’ contingent financing plans to better gauge systemwide effects. ### Second-round effects on GDP and counterfactual policy simulation - Empirical bank–macro linkage findings: - Long-run elasticity: a sustained one percentage point increase in the rate of credit growth leads to a 0.07 percentage point increase in first-year GDP and a 0.16 percentage point increase in two-year cumulative GDP (SVAR results). - In the adverse (severe adverse) scenario, the banking sector CAR declines by nearly 8 (12) percentage points. - These CAR declines could reduce real GDP level by additional 4 (9) percentage points by 2021 (estimate of second-round effects). - Counterfactual policy analyzed: one-time write-off of NPL worth 30 percent of LLP stock in 2021 financed by available excess capital (e.g., limiting dividends, owner family injections). - Estimated policy costs and benefits (2019 real GDP = 100): - Benefit 1 (Maximum difference in the level of real GDP during 2020-22): - Baseline: 1 - Adverse: 1.62 - Severe adverse: 2.86 - Benefit 2 (Sum of differences in the level of GDP from 2020 to 2022): - Baseline: 2.31 - Adverse: 3.20 - Severe adverse: 4.13 - Cost (30 percent of loan-loss provision stock as of 2021, one time): - Baseline: -1.56 - Adverse: -2.16 - Severe adverse: -2.88 - Interpretation: single-year benefits roughly match costs; benefits accrue over multiple years and appear to exceed costs when considering 2021–22 and outer years. ### Policy recommendations and selected near- and medium-term actions - Short-term (ST) recommendation: - Limit bank dividend distributions while downside risks remain high and be ready to take additional measures to strengthen banks’ capital if the risks materialize to continue providing credit to the economy. (Timing: ST; Agency: BSP, FSCC members) - Medium-term (MT) recommendations: - Consider introducing FX LCR. (Timing: MT; Agency: BSP) - Develop macro-scenario based microprudential stress test of banks. (Timing: MT; Agency: BSP) - Develop macroprudential stress test of banks. (Timing: MT; Agency: BSP) - Develop tools that jointly examine banks, NFCs, and the real economy to enhance systemic risk assessment. (Timing: MT; Agency: BSP) - Enhance collaboration within the BSP to conduct essential macroprudential risk analyses, including macro scenario stress tests of banks. (Timing: MT; Agency: BSP) - Improve data sharing arrangements within the BSP. (Timing: MT; Agency: BSP) - Further strengthen NFC information sharing between the BSP and SEC. (Timing: MT; Agency: BSP, SEC) - Continue the effort to construct an NFC database beyond listed firms. (Timing: MT; Agency: SEC) - Collect more granular data on banks’ credit risk (PD, LGD, LTV), leveraging IFRS 9. (Timing: MT; Agency: BSP) - Improve household survey data and credit registry to develop borrower-based credit risk indicators. (Timing: MT; Agency: BSP) - Consider monitoring contingent financing plans of large banks and NFCs. (Timing: MT; Agency: Government, BSP) - Additional operational recommendations: - Allow forbearance measures to lapse as scheduled and avoid introducing new ones; forbearance hampers banks’ ability to support the economy and can undermine financial stability. - Be ready to facilitate sale and recovery of bad assets and to raise additional capital starting with conglomerate owner families and private sector funding; public funding only as a last resort. - Enhance liquidity stress test toolkit and data collection (maturity structure, split retail/wholesale deposits, loans split into households and NFCs). ### Institutional capacity, stress-testing governance, and data gaps - Current organization: supervision sector implements all bank-related analysis; Office of Systemic Risk Management (OSRM) focuses on non-financial sectors and their link to banks. - No unit conducts macro-scenario stress testing despite staff capacity; BSP should start such exercises. - Data gaps and recommended actions: - Leverage IFRS 9 (introduced in 2018) for calculation of regulatory capital and start collecting more granular credit risk parameters such as PD, loss-given-default (LGD), and loan-to-value (LTV) ratio. - Enhance NFC information sharing between the BSP and Securities and Exchange Commission (SEC). - Continue effort to construct an NFC database beyond listed firms. - Over the medium term, enhance household survey and credit registry data to develop more granular borrower-based credit risk indicators. - BSP supervisory practice: - BSP conducts semi-annual microprudential single-factor solvency stress tests and other targeted tests (market risk, real estate, ad hoc exercises including loan moratoria impact). ### Caveats, sensitivities, and limitations - Tests do not incorporate mitigating policy effects such as credit guarantees, regulatory responses, and loan moratoria (except system-wide liquidity analysis). - Credit risk projections are based on historical relationships between macro variables and PD; these relationships may not hold in the unusual COVID-19 context. - Behavioral assumptions in the standard FSAP approach (e.g., most NPLs remain on banks’ balance sheets during the entire stress-test horizon; NPL restructuring, write-offs, and sales to special purpose vehicles are assumed out) may yield relatively pessimistic estimates under extremely large shocks. - Bank portfolios are assumed constant throughout the horizon, disregarding potential optimal portfolio adjustments by banks. - Results are sensitive to LGD, cure rate, and interest margin shock assumptions; sensitivity analyses show notable variation in CAR and capital shortfalls when LGD and cure rates vary. *Source: EXECUTIVE SUMMARY (content up to February 2021) contained in 1phlea2022002 - EXECUTIVE SUMMARY.* ### EXECUTIVE SUMMARY __________________________________________________________________________ 6 ### 1phlea2022002 - EXECUTIVE SUMMARY ### Macrofinancial setting and system structure - Total assets of the financial system amount to 126 percent of GDP. - Bank credit is just over 50 percent of GDP and mostly goes to nonfinancial corporates (NFCs). - Only a third of adults have formal accounts. - Non-bank financial institutions and capital markets—especially bond markets—are substantially less developed than banks. - The Fintech ecosystem is nascent. ### COVID-19 shock, macro impact, and recovery projection - GDP contracted by 9½ percent in 2020. - The economy is expected to grow 6½ percent in 2021. - Policy support and easing of containment measures contributed to recovery starting in the second half of 2020. - Authorities implemented time-bound regulatory relief and forbearance measures; the scale of loan moratoria and credit guarantees has been relatively limited. ### Stress-testing framework and methods used - Bank solvency and liquidity stress tests follow the standard FSAP method. - Estimated models gauge second-round effects of bank solvency results on GDP, focusing on the credit channel. - Scenario-based NFC stress test focused on listed firms. - Joint bank–NFC cash-flow liquidity stress test to gauge effects of loan moratoria. ### NFC corporate sector stress and key metrics - Debt-weighted-average interest coverage ratio: - 4.9 percent at end-2019 - 1.3 percent in the upside/baseline (context: “decline from 4.9 percent at end-2019 to 1.3, below one, and 0.2 percent in the baseline, adverse, and severe adverse scenarios, respectively” — preserved as in source) - below one in the adverse scenario - 0.2 percent in the severe adverse scenario - Debt-at-Risk: - 5 percent at end-2019 - about 45 percent in the upside scenario - 80 percent in the severe adverse scenario - NFC distress could significantly increase NPLs without private and public sector actions or economic recovery. - Loan moratoria can help firms survive liquidity shocks if the crisis is temporary but may only delay bankruptcies if the crisis is prolonged or business models are fundamentally challenged. - Support channels that could mitigate contagion to bank solvency: wealthy owner families of large conglomerates, fiscal aid, and credit guarantees for smaller firms. ### Bank solvency stress-test results and systemic risk - System aggregate total capital adequacy ratio (CAR): - 15.6 percent at baseline starting point - falls to 11.7 percent by 2022 in the baseline scenario (still above the ten percent minimum requirement) - falls to 9.3 percent in the adverse scenario - falls to 4.9 percent in the severe adverse scenario - Second-round effects from distress could reduce the real GDP level by an additional 4 to 9 percentage points in adverse scenarios. - CARs start to recover in 2022 as the economy recovers (per model projections). ### Caveats on solvency stress-test interpretation - Tests do not incorporate mitigating policy effects such as credit guarantees, regulatory responses, and loan moratoria. - Credit risk projections are based on historical relationships between macro variables and probability of default (PD); these relationships may not hold in the unusual COVID-19 context. - Behavioral assumptions in the standard FSAP approach—used for cross-country comparability—may yield relatively pessimistic estimates under extremely large shocks. - Most NPLs are assumed to remain on banks’ balance sheets during the entire stress-test horizon; NPL restructuring, write-offs, and sales to special purpose vehicles are assumed out. - Bank portfolios are assumed constant throughout the horizon, disregarding potential optimal portfolio adjustments by banks. ### Policy simulation and capital measures - A one-time counterfactual policy to write off NPLs early using available excess capital could improve projected GDP for several years and yield net benefits above the cost. - Write-off could be financed by limiting banks’ dividend distribution immediately as a precautionary measure, with readiness to take additional measures to strengthen banks’ capital should risks materialize. - Recommendation: limit bank dividend distributions while downside risks remain high and be ready to take additional measures to strengthen banks’ capital to continue providing credit to the economy. ### Liquidity profile and stress-test findings - Banks have sufficient buffers to withstand severe liquidity shocks, in part supported by high levels of the reserve requirement. - High-quality liquid assets (HQLA) for LCR calculation are mostly reserves and sovereign securities. - Banks rely mainly on retail and wholesale deposits. - System appears more resilient to FX liquidity shocks than local currency liquidity shocks. - The net stable funding ratio and cash-flow analysis show similar outcomes. - At end-2019, most reserves originated from a 14 percent reserve requirement ratio (RR-ratio). - The BSP reduced the RR-ratio to 12 percent when the COVID-19 crisis hit, which increased system-wide and individual banks’ usable buffer modestly. - If banks are allowed to use all the reserves, all universal and commercial banks can survive severe cash-flow stress for months. - Since total currency LCR replaced FX liquidity requirements for foreign currency unit, the BSP could consider introducing FX LCR. ### Bank–NFC liquidity linkage and loan moratoria effects - Liquidity stress to NFCs from lower earnings can spill over to banks; loan moratoria can complicate linkages. - Direct effect of loan moratoria: improves NFC cash balance while reducing bank cash inflows and liquid assets. - Without moratoria, NFC cash balance declines for debt service, which improves bank liquidity conditions unless NFCs withdraw deposits to meet payment obligations. - If banks continue to roll over maturing NFC loans, NFC liquidity balances recover with or without moratoria. - For NFCs, moratoria substantially improve liquidity when banks’ rollover rate is low but less so otherwise. - For banks, moratoria effects on liquidity depend critically on whether NFCs have alternative financing sources; if NFCs withdraw deposits, banks may experience broadly the same cash-flow effects irrespective of moratoria. - Recommendation: the Bangko Sentral ng Pilipinas (BSP) should monitor banks’ and NFCs’ contingent financing plans to better gauge systemwide effects. ### Institutional capacity, stress-testing governance, and data gaps - Current organization: supervision sector implements all bank-related analysis; Office of Systemic Risk Management (OSRM) focuses on non-financial sectors and their link to banks. - No unit conducts macro-scenario stress testing despite staff capacity; BSP should start such exercises. - Options: several units could work jointly or different sections conduct distinct exercises depending on objectives. - Data gaps and recommended actions: - Leverage IFRS 9 (introduced in 2018) for calculation of regulatory capital and start collecting more granular credit risk parameters such as PD, loss-given-default (LGD), and loan-to-value (LTV) ratio. - Enhance NFC information sharing between the BSP and Securities and Exchange Commission (SEC). - Continue effort to construct an NFC database beyond listed firms. - Over the medium term, enhance household survey and credit registry data to develop more granular borrower-based credit risk indicators. ### Selected recommendations (as presented in Table 1) - Limit bank dividend distributions while downside risks remain high and be ready to take additional measures to strengthen banks’ capital if the risks materialize to continue providing credit to the economy. (Timing: ST; Agency: BSP, FSCC members) - Consider introducing FX LCR. (Timing: MT; Agency: BSP) - Develop macro-scenario based microprudential stress test of banks. (Timing: MT; Agency: BSP) - Develop macroprudential stress test of banks. (Timing: MT; Agency: BSP) - Develop tools that jointly examine banks, NFCs, and the real economy to enhance systemic risk assessment. (Timing: MT; Agency: BSP) - Enhance collaboration within the BSP to conduct essential macroprudential risk analyses, including macro scenario stress tests of banks. (Timing: MT; Agency: BSP) - Improve data sharing arrangements within the BSP. (Timing: MT; Agency: BSP) - Further strengthen NFC information sharing between the BSP and SEC. (Timing: MT; Agency: BSP, SEC) - Continue the effort to construct an NFC database beyond listed firms. (Timing: MT; Agency: SEC) - Collect more granular data on banks’ credit risk (PD, LGD, LTV), leveraging IFRS 9. (Timing: MT; Agency: BSP) _Source: EXECUTIVE SUMMARY (content up to February 2021) contained in 1phlea2022002 - EXECUTIVE SUMMARY._ ### introduction of IFRS 9 (67, 68, 69). ### introduction of IFRS 9 (67, 68, 69). ### Policy recommendations and near-term actions - Medium-term (MT) — BSP: Improve household survey data and credit registry to develop borrower-based credit risk indicators (69). - Medium-term (MT) — Government, BSP: Consider monitoring contingent financing plans of large banks and NFCs (205). - Timeframe notation used in source: Short-term (ST) = within one year; medium-term (MT) = one to three years. ### Financial system size and structure - Total assets of the financial system amount to 126 percent of GDP. - The banking system holds about 94 percent of the system’s assets. - Bank credit is just over 50 percent of GDP as banks hold substantial liquid assets. - Access to finance: only a third of adults have formal accounts (one-third). - Banking sector composition: - Forty-six universal and commercial banks (UKBs) hold over 94 percent of bank assets. - 60 percent of UKB assets are held by the top five banks (all domestic). - D-SIBs, including branches of some Global-SIBs, hold about 80 percent of the banking sector. - Foreign bank subsidiaries and branches hold seven percent of bank assets. - About 500 small thrift banks (TBs) and rural and cooperative banks (RCBs) exist. ### Bank business model and asset composition - Banks follow a traditional commercial banking business model, relying on deposits and lending mostly to large NFCs. - Eighty percent of the loans go to NFCs, an unusually high level compared to other countries. - Loan shares by bank type: - Loans represent about 55 percent of assets for UKBs and RCBs. - Loans represent close to 72 percent of assets for TBs. - Banks are liquid with nearly 40 percent of their assets in securities and central bank reserves. - Bank balance sheet structure (In percent of total assets): - Cash, 15 - Security, 24 - Interbank, 2 - Loan, 55 - Other asset, 5 - Liability structure (In percent): - Peso deposit, 62 - FX deposit, 12 - Bill and bond, 8 - Other liability, 5 - Equity, 13 ### Credit distribution and sector exposures - 80 percent of loans go to NFCs. - Exposure to real estate loans is relatively low due to a regulatory limit of 20 percent of total loans applicable to UKBs (raised to 25 percent upon COVID-19 permanently). - Real estate loans are largely commercial; TBs provide one-third of their loans to residential properties. - TBs and RCBs exposure to household consumption and agriculture loans: - TBs: 32 percent - RCBs: close to 50 percent ### Nonbank financial institutions and capital markets - NBFIs (insurers, mutual funds, pension funds) are much smaller than several Asian peers. - Insurance penetration: 1.3 percent (from 2013-2017). - Total industry premiums: 1.65 percent of GDP. - Total insurance industry assets: US$31.5 billion. - Domestic stock market capitalization: roughly 90 percent of GDP. - Bond outstanding: roughly 30 percent of GDP. - Debt market is dominated by government securities. - Informal nonbank micro-financial institutions (mostly pawnshops) are present; informal financing among family members is more significant to households than bank loans. ### Fintech and digital payments - Fintech ecosystem is nascent. - 2017 Global Findex: only a quarter of the adult population made or received at least one digital payment in the preceding year (a quarter). - Constraints include expensive bank charges and barriers to establishing IT and communication infrastructure for the archipelago of over 7,000 islands. ### International exposures and spillovers - Banks’ direct cross-border exposure is low at about 10 percent of bank assets and liabilities. - Dollarization: - 15 percent of deposits are in FX. - 11 percent of loans are in FX. - International remittance inflows: about eight percent of GDP annually. - Note: remittances can be credited to banks only in pesos in most cases. ### COVID-19 shock and economic performance - Real 2020 GDP contracted by 9.5 percent. - Government containment measures resulted in 12 percent (half a year on half a year, seasonally adjusted.) real GDP contraction in the first half of 2020. - Recovery quarterly growth (seasonally adjusted, quarter-on-quarter): - Third quarter 2020: real GDP increasing by 8.0 percent. - Fourth quarter 2020: real GDP increasing by 5.6 percent. - Fund projection (January 2021 WEO): 2021 real GDP growth to be 6.6. percent. - The Philippines’ 2020 revision to real GDP growth rate forecast: -15.8 (difference between October 2019 and January 2021 WEO forecasts), one of the worst-hit economies. ### Pre-pandemic macro-financial fundamentals and resilience - Economic growth over 6 percent during 2013–19. - Public debt steadily declined in the past 20 years. - Residential mortgages: only 4 percent of GDP. - By end-2019, NPL ratio was low by historical and EM comparator standards. - Total regulatory capital adequacy ratio (CAR) has been stable at about 15 percent in the past ten years. - Quality of capital is high. - Return on assets (ROA): about 1½ percent. - Bank-type NPL ratios: - TBs: around 6 percent. - RCBs: around 11 percent. - UKBs: about 1½ percent. - TBs and RCBs have higher capital ratios than UKBs. ### Dispersion and systemic considerations - Significant dispersion in financial soundness indicators across bank types and individual banks. - CAR statistics among UKBs: - Median CAR: 21.5 percent. - Interquartile range: from 15.5 percent to 70.6 percent. - Among UKBs, D-SIBs median CAR: 15 percent. - Non-D-SIBs median CAR: 34 percent. - TBs and RCBs exhibit larger dispersion in ROAs and NPLs, indicating pockets of vulnerability. - Failures of TBs or RCBs would not pose systemic risk in principle, but weaknesses could be problematic if resolution is not handled effectively. *Source: IMF staff summary of "introduction of IFRS 9 (67, 68, 69)."* ### 2020. The exchange rate appreciated slightly against the USD for 2020 as a whole, and gross ### 1phlea2022002 - 2020. The exchange rate appreciated slightly against the USD for 2020 as a whole, and gross ### Financial-sector developments and policy responses - Gross international reserves recovered by nearly US$20 billion to US$110 billion between end-April and end-year (11 months of import coverage). - The BSP cut policy rates and reserve requirements in contrast to the AFC. - Authorities launched a small (0.6 percent of GDP) credit guarantee program for loans to small- and medium-sized enterprises (SMEs) and the agricultural sector. - Moratoria (total of five months) expired at the end of 2020. - BSP issued time-bound regulatory relief and forbearance measures allowing banks to delay NPL recognition and provision over a maximum period of five years subject to BSP approval. - Uptake of forbearance appears limited given BSP’s tight approval criteria. - BSP efforts to keep track of credit quality information without policy measures to maintain transparency should help assessing the impact of some policy measures going forward. - Banks: lending standards have tightened, credit is contracting though the credit gap remains positive as GDP contracts. - NPL ratio rose from 2.1 at end-2019 to 3.4 percent in September 2020. - Share of past-due loans and restructured loans rose. - CAR rose over one percentage point since end-2019 (Table 4). - Figures may have optimistic bias under moratoria and forbearance measures. - Banks continued to receive new deposits, reducing the loan-to-deposit ratio noticeably. ### Key financial stability risks and transmission channels - Key risks stem from the COVID-19 crisis and bank-corporate linkages. - Lockdowns and social distancing expected to depress NFC earnings and could spill over to bank health through funding and credit exposures and ownership linkages of mixed conglomerates. - Standard macroeconomic policies and measures to support borrowers (credit guarantees, loan moratoria) could complicate transmission channels. - Regulatory responses, including forbearance, might have intended and unintended effects. ### FSAP stress testing approach and scope - FSAP stress tests aim to be macroprudential focusing on systemic risk; they start with macroeconomic scenarios and include feedback effects such as contagion in interbank and financial markets and solvency-liquidity linkages. - This FSAP: - Conducted bank solvency tests covering all banks using end-2019 data. - Conducted liquidity tests examining UKBs. - Applied new tools to assess bank–NFC and bank–economic linkages. - Performed NFC stress tests assessing earning shocks by industry using interest coverage ratio (ICR) and cash ratio. - Constructed macro-scenario stress testing models to gauge impact on ICR and cash ratios under the same macroeconomic scenarios used for bank stress tests. - Linked solvency test results to estimate second-round effects on GDP through credit growth channels. - Assumptions and exclusions: - Exercises do not take into account mitigating effects from already announced or prospective sector-specific policy support measures, except for system-wide liquidity analysis. - Effects of credit guarantees and regulatory responses are not incorporated (size of credit guarantee appears small and does not influence thrust of assessment much). - Moratoria expired at the end of 2020. - Forbearance measures inconsistent with Basel III minimum requirements are not included, except system-wide liquidity analysis that incorporates announced and possibly enhanced loan moratoria (which are not forbearance). - Stress tests are not forecasting exercises; they target resilience under tail events with neutral assumed bank behaviors to ensure cross-country comparability. ### Macroprudential linkage and second-round effects methodology - Second-round effects estimated focusing on credit channels: - Bank-by-bank panel models project credit growth in response to changes in bank soundness indicators (outputs of stress tests, such as capital). - Aggregated credit growth shock fed into a structural VAR (SVAR) macro-financial model to gauge resulting changes to GDP growth beyond original macro scenarios. - Difference from Catalan-Hoffmaister approach: - This FSAP uses DSGE-generated macro scenarios and then estimates second-round effects by SVAR; the estimate is a “first-round estimate of the second-round effects,” implying some incompleteness and model uncertainty. - Counterfactual policy analysis: - Model used to evaluate effects of policies such as NPL write-off or accumulation of loan-loss provisions (LLPs). - Additional provisions needed to write-off NPLs are treated as total cost of counterfactual policy; benefits measured by improvement of real GDP over stress test horizon. - Funding for write-offs could come from shareholders, owner families, new shareholders, or the government as last resort. - Liquidity linkage analysis: - Focus on loan moratorium effects using a tool that links cashflow liquidity stress tests of NFCs and banks. - Loan moratoria reduce debt service cash inflows to banks and outflows from NFCs; overall impact on bank liquidity depends on runoff rates of NFC deposits. ### Macroeconomic scenarios used in stress tests - Four common macroeconomic scenarios considered across exercises, all showing real GDP paths more severe than the AFC but less severe than the 1984 political turmoil: - Baseline: - Follows October 2020 WEO forecast; factors in tight lockdown effects in first half of 2020 (-8½ GDP growth rate for the year), followed by visible recovery in 2021. - Shows a much sharper 2-year cumulative GDP contraction than AFC, equivalent to a three-standard deviation shock to the pre-COVID WEO forecast (as of January 2020) showing growth rates near potential (6½ percent per year for 2020–24). - Actual 2020 growth rate (-9.5 percent) turned out weaker than October WEO, but roughly the same as January WEO forecasting -9.6 percent for 2020 and 6.6 percent for 2021 (Table 3). - Upside: - Incorporates national authorities’ forecast as of September 2020 with a slightly more optimistic GDP growth rate in 2020 than the baseline. - Adverse: - Assumes prolonged lockdown measures throughout 2020 with some scarring effects in 2021. - Real GDP growth would contract by 11 percent in 2020, followed by a weaker recovery in 2021 than the baseline. - First 2-year cumulative growth amounts to a ⅔ standard deviation shock to the baseline (nearly a four standard deviation shock to the January 2020 WEO forecast). - Severe adverse: - Assumes prolonged and even more stringent lockdown in 2020 with more severe scarring effects in 2021. - First 2-year cumulative growth would reach -6½ percent, equivalent to a 1⅓ standard deviation shock to the baseline. - Note: Ex post, January 2022 WEO forecast (real GDP growth rate is -9.6 percent for 2020, 4.6 percent for 2021, and 6.3 percent for 2022) is revised down from October 2020 WEO (-8.3 percent for 2020, 7.4 percent for 2021, and 6.5 percent for 2022) but still above the adverse scenario. ### Corporate-sector (NFC) stress test: sample and key findings - Sample: - Data from Capital IQ, S&P Market Intelligence. - 151 non-financial firms as of end-2019, of which 147 firms are publicly listed. - Sample accounts for about 44.4 percent of the Philippines’ outstanding NFC debt and 46.1 percent of banks’ total loan portfolio (net of the reverse repo agreements with the BSP). - Pre-pandemic financial position (end-2019): - Median ICR about 3.6; debt-weighted ICR about 4.9, implying adequate debt service capacity. - Median return on assets about 6 percent for the median firm. - Median cash ratio in 2019 about 24 percent (enough to cover almost one-quarter of total liabilities coming due within a year). - Cash ratio declined from a peak of 37 percent in 2014 but remained above the ASEAN median of 23 percent in 2019. - Share of NFC debt held by firms with ICR below one rose from about 1.7 percent in 2016 to about 5.3 percent in 2019. - Share of sample firms with ICR below one in 2019 was about 22 percent. - Of the 5.3 percent of NFC debt held by firms with ICR below one in 2019, about 68 percent (or 3.6 percent of total sample NFC debt) was held by firms in the industrial sector (transportation, construction, heavy machinery and equipment). - ICRs significantly lower for smaller firms and those in energy, materials, and information technology. - Cash ratios below 20 percent observed for firms in materials, communication services, and real estate. - Stress test methodologies: - Two complementary approaches to estimate ICRs at end-2020: - Direct shock approach: apply shocks to subcomponents of ICR (operating income and interest payment) at end-2019. Operating income shock set from consensus earnings forecasts of market analysts; shocks to interest payments (exchange rate and interest payment shock) set in line with the baseline scenario. - Regression-based approach: predict ICRs at end-2020 using macroeconomic and global variables as explanatory variables, allowing estimation under upside, adverse, and severe adverse macroeconomic scenarios. ### Additional methodological notes and caveats - Bank solvency test does not examine additional effects from interbank exposures as such links are negligible (Figure 3). - Loan moratoria could reduce NFC liquidity stress and therefore drawdown of NFC’s liquid assets including bank deposits; while it reduces cash inflows to banks from loan repayments, banks may experience lower deposit withdrawal from NFCs. It will reduce bank capital if banks eventually need to write off moratorium-related restructured loans. - The exercise focuses on NPL write-off and LLP instead of capital since the first two show statistically significant explanatory power in the credit model while bank capital ratios do not. - There is model uncertainty, especially given the exercise was conducted in the middle of the COVID-19 crisis; figures should be interpreted carefully. *Source: IMF staff analysis as presented in the FSAP chapter.* ### section in addition to the baseline. ### section in addition to the baseline. ### Vulnerability of non-financial corporates (NFCs) — Baseline findings - Cash positions at end-2020 under the baseline scenario are estimated by adding expected cash flow during 2020 to cash balance at end-2019. - Cash flow from operations is assumed to decline in line with the operating income shock assumed for the ICR analysis. - Capital expenditure and debt refinancing are set at levels broadly consistent with the magnitude of the macroeconomic shocks, based on historical trends. - Under the baseline scenario: - Median ICR is expected to decline to about 0.9 in 2020, down from 3.6 in 2019. - Debt-weighted mean ICR is expected to decline to about 1.3 in 2020, down from 4.9 in 2019. - Share of NFC debt-at-risk would jump to 55 percent in 2020, up from 5.3 percent in 2019. - Share of firms with ICR below one is expected to rise to about 48 percent of total sample firms in 2020, up from 18.3 percent in 2019. - The deterioration in debt service capacity is expected to be driven by the fall in ROA, reflecting lower projected real GDP growth in the Philippines and the rest of the world. - The sharp increase in debt-at-risk is expected to be driven by large firms (indicated by larger decline in debt-weighted ICR than median ICR). - Industry variation: energy, consumer discretionary, and industrials expected to see the largest increases in debt-at-risk and firm-at-risk shares. Information technology is also vulnerable largely due to already low pre-pandemic profitability. - Note on classification: tourism-related industries (hotels, restaurants, leisure) fall under consumer discretionary per Capital IQ industry classification. ### Vulnerability of non-financial corporates — Stress scenarios - Under adverse macroeconomic scenarios: - Debt-at-risk share in 2020 is expected to increase to about 76 percent (adverse) and 81 percent (severe adverse), compared with 55 percent in the baseline. - Under the upside scenario, debt-at-risk share is expected to reach 45 percent. - Cash position effects: - System-wide cash-to-assets ratio would drop to 8.7 percent in 2020 (without mitigating effects from policies such as a moratorium on loan repayments, including principal and interest), down from 9.7 percent in 2019. - Median cash ratio would drop to about 23 percent in 2020 from 24 percent in 2019. - Across industries, consumer discretionary, energy, and real estate expected to experience the most severe cash shortages in 2019 due to initial thin cash buffers (real estate) or large expected declines in cash flow from operations (consumer discretionary and energy). ### Monitoring and data gaps for NFC vulnerabilities - The analysis relies on a relatively small set of Philippine NFCs that account for about 44.4 percent of total outstanding NFC debt in the economy. - More comprehensive data exist (compiled by SEC) but become available with significant lags, hampering timely monitoring. - SEC initiatives to digitalize more comprehensive NFC data are welcome. - Efforts to improve information sharing between the BSP and the SEC should continue. ### Bank solvency stress test — Coverage and framework - The solvency stress test (STeM in Appendix I) covers: - 46 UKBs (21 universal and 25 commercial banks) including D-SIBs, - 49 TBs, - 447 RCBs. - These 542 banks represent 100 percent of the Philippines banking system’s assets. - Scenario-based assessment uses balance sheet information and a three-year horizon under four scenarios; tests do not account for regulatory relief/forbearance and borrower-support measures not reflected in scenarios. - Tests assume a quasi-static balance sheet: - Allocation of assets and composition of funding sources remain the same as of the latest observation. - Gross exposures grow in line with nominal GDP growth. - Banks can build capital buffers only through retained earnings (no new equity issuance). ### Credit risk, PDs, and cure assumptions - Credit risk satellite models estimated for UKBs, TBs, and RCBs using quarterly data for 2005-2019; models selected based on in-sample fit and economic sign constraints. - UKBs’ PDs are very responsive to real GDP growth; TBs and RCBs more influenced by local factors and show higher historical volatility. - Under the baseline scenario, PD peaks: - UKBs: 13 percent - TBs: 10 percent - RCBs: 16 percent - Aggregate PD starting points (December 2019): - UKBs: 1.04 percent - TBs: 6.6 percent - RCBs: 9.7 percent - LGD assumptions (kept constant across horizon and scenarios): - UKBs: 68 percent - TBs: 35 percent - RCBs: 66 percent - Cure rates with respect to NPLs (assumed fractions of historical rates): - UKBs: 10 percent - TBs: 22 percent - RCBs: 24 percent - These represent less than 25 percent of historical annualized cure rates. - Extra cure rate assumption: an extra 18 percent (total of 28 percent) with respect to New NPL (2020) assumed for UKBs in 2021. - Sensitivity analyses: - LGDs of 35 percent and 85 percent for all bank types (Appendix III-C). - Average historical annualized cure rates and zero-cure rate (Appendix III.C). ### Market risk, IRRBB, and interest margin shocks - Market risk module applies mark-to-market for AfS and HfT securities; modified duration formula for revaluation; HtM treated with credit risk approach. - Valuation changes in AfS securities affect OCI and capital one to one; HfT realized losses affect net income and are taxed. - Interest rate risk on the banking book (IRRBB) assessed using time-to-repricing buckets; maturity profiles assumed unchanged. - Interest margin shocks: - Shock calibrated as a fraction of the shock experienced during the AFC, which had a V shape with a peak of 80 percent in the second year. - Severe adverse scenario assumes a quarter of the AFC shocks; other scenarios assume milder shocks. - COVID scenarios consider more moderate margin shocks than the AFC due to BSP rate cuts and more benign financial conditions. - Net income projection components: - Net interest income accounts for balance sheet size, NPL increases, IRRBB, and interest margin shocks. - Loan loss provisions driven by credit risk evolution on loans and HtM securities. - Trading income accounts for HfT and FX-open position gains/losses. - Other income/expense assumed constant as a proportion of interest-earning assets. - Income tax rate set at 30 percent. - Dividend policy assumptions: - Dividends paid only if net income after taxes is positive. - UKBs’ dividend payout ratio anchored to 2019: out of 46 UKBs, 11 paid dividends with average payout ratio of 13 percent. - TBs and RCBs: payout ratios taken as individual bank 5-year averages; 2 out of 51 TBs and 32 out of 447 RCBs paid dividends in that period with average payout ratios of 0.8 percent and 0.9 percent, respectively. - Assumed no new share issuance or repurchases during the stress horizon. ### RWAs, hurdle rates, and stress test calibration - RWAs respond to credit risk changes under Basel III standardized approach: - Decrease in risk weights (to zero) from provisions related to new NPLs. - Increase in risk weights from the non-provisioned part of new NPLs (subject to 150 percent risk weight). - Changes as NPLs cure. - RWAs grow in line with balance sheet growth (nominal GDP growth). - Hurdle rates (Philippines minimums): - CET1R: 6 percent (applied only to UKBs) - T1R: 7.5 percent - Total Capital Ratio (CAR): 10 percent - UKBs also required to hold a 2.5 percent capital conservation buffer and, if applicable, a D-SIB buffer (1.5 or 2 percent) — these are not included in hurdle rates used here. - Presentation focuses on 10 percent CAR common across all banks. ### Bank stress test results — System outcomes - By 2022, system CAR evolution: - Starting CAR: 15.6 percent - Baseline: falls to 11.7 percent - Adverse: falls to 9.3 percent - Severe adverse: falls to 4.9 percent - Minimum CAR requirement: 10 percent - Even in the baseline, 185 banks (mostly RCBs), accounting for about a third of the system by assets, might not meet the 10 percent requirement. - In the adverse scenario, 201 banks could have capital shortfalls. - In the severe adverse scenario, 214 banks with three-quarters of the system’s assets miss the minimum CAR requirement. - Capital shortfalls are moderate—below four percent of GDP even in the severe adverse scenario. - UKBs: - More likely to meet CET1 requirement (6 percent) than CAR due to higher capital quality. - UKBs’ CET1 ratio remains above the hurdle under baseline, upside, and adverse scenarios but falls below national and Basel III CET1 requirement (4.5 percent) under severe adverse scenario. - CET1 ratios of 20 out of 46 UKBs fall below regulatory minimum under the severe adverse scenario, including D-SIBs. - The system-wide capital depletion from the starting point amounts to 3.9 percent. *Source: 1phlea2022002 - section in addition to the baseline. Canonical URL: https://www.imf.org/-/media/files/publications/cr/2022/english/1phlea2022002.pdf* ### 10.7 percent under the baseline and severe adverse scenarios, respectively. Mostly driven by ### 1phlea2022002 - 10.7 percent under the baseline and severe adverse scenarios, respectively. Mostly driven by ### Bank Solvency Stress Tests: Key Results (2022) — Aggregate outcomes - Capital ratios (CAR and CET1R) and capital shortfalls presented for scenarios: Latest actual, Baseline October, Upside, Adverse, Severe Adverse. - Latest actual: Total CAR 15.6, CAR UKB 15.3, CET1R 12.7, CAR TB 17.5, CAR RCB 19.4; capital shortfalls (in percent of GDP) all 0.0. - Baseline October: Total CAR 11.7, CAR UKB 11.0, CET1R 8.9, CAR TB 18.2, CAR RCB 14.2; total capital shortfall 1.0, UKB 0.9, TB 0.5, RCB 0.0, CET1R shortfall 0.0. - Upside: Total CAR 13.5, CAR UKB 13.1, CET1R 10.8, CAR TB 18.5, CAR RCB 14.3; total capital shortfall 0.5, UKB 0.5, TB 0.1, RCB 0.0, CET1R shortfall 0.0. - Adverse: Total CAR 9.3, CAR UKB 8.5, CET1R 6.5, CAR TB 17.3, CAR RCB 13.6; total capital shortfall 1.9, UKB 1.8, TB 1.1, RCB 0.0, CET1R shortfall 0.0. - Severe Adverse: Total CAR 4.9, CAR UKB 3.7, CET1R 2.1, CAR TB 15.7, CAR RCB 12.8; total capital shortfall 3.9, UKB 3.7, TB 2.8, RCB 0.0, CET1R shortfall 0.1. - Footnotes: - 1/ Figures at the end of the stress test horizon (2022). - 2/ Amount of money needed to bring CAR and CET1 to respective regulatory minimums. - Regulatory minima (hurdle rates for the stress tests): - UKBs and TBs and RCBs that are subsidiaries of UKBs: CAR 10 percent (Basel III 8 percent), CET1 ratio 6 percent (Basel III 4.5 percent), Tier 1 ratio 7.5 percent (Basel III 6 percent). Required buffers: 2.5 percent capital conservation buffer and, if applicable, a D-SIB buffer (of 1.5 or 2 percent). - Independent TBs and RCBs: CAR 10 percent and Tier 1 ratio 6 percent. They are not subject to buffer and leverage ratio requirements. ### Bank failures and asset shares by scenario - Number of banks not meeting minimum requirements (end of stress horizon, 2022): - Baseline October: Total 185, UKB 12, TB 8, RCB 6; share of failed banks’ assets in total 167 (Total), UKB 31.8 percent, TB 30.8 percent, RCB 24.1 percent, CET1R 0.5, CAR 0.5. - Upside: Total 178, UKB 9, TB 6, RCB 6; share of failed banks’ assets 163 (Total), UKB 25.2 percent, TB 24.2 percent, RCB 17.1 percent, CET1R 0.5, CAR 0.5. - Adverse: Total 201, UKB 18, TB 12, RCB 9; share of failed banks’ assets 174 (Total), UKB 59.9 percent, TB 58.7 percent, RCB 32.1 percent, CET1R 0.6, CAR 0.6. - Severe Adverse: Total 214, UKB 21, TB 20, RCB 12; share of failed banks’ assets 181 (Total), UKB 76.4 percent, TB 75.0 percent, RCB 64.3 percent, CET1R 0.8, CAR 0.6. ### Bank Solvency Stress Tests: UKBs Various Classifications (2022) — UKB subgroups - Capital ratios by subgroup (UBs vs KBs; DSIBs vs Non DSIBs; Top 10 vs Other; Conglomerate vs Other), across scenarios: - Latest actual: UBs 14.6, KBs 23.9, DSIBs 14.4, Non DSIBs 20.8, Top 10 14.3, Other 15.3, Conglomerate 14.5, Other 18.6. - Baseline October: UBs 10.5, KBs 17.6, DSIBs 10.5, Non DSIBs 14.5, Top 10 10.2, Other 11.0, Conglomerate 11.0, Other 11.0. - Upside: UBs 12.6, KBs 19.7, DSIBs 12.5, Non DSIBs 16.7, Top 10 12.3, Other 13.1, Conglomerate 13.0, Other 13.6. - Adverse: UBs 8.0, KBs 14.8, DSIBs 8.0, Non DSIBs 11.4, Top 10 7.8, Other 8.5, Conglomerate 8.7, Other 7.4. - Severe Adverse: UBs 3.2, KBs 9.9, DSIBs 3.3, Non DSIBs 6.2, Top 10 3.2, Other 3.7, Conglomerate 6.2, Other 1.5. - Recapitalization needs (in percent of GDP) by subgroup: - Latest actual: all 0.0. - Baseline October: UBs 0.8, KBs 0.1, DSIBs 0.7, Non DSIBs 0.2, Top 10 0.7, Other 0.9, Conglomerate 0.5, Other 0.4. - Upside: UBs 0.4, KBs 0.0, DSIBs 0.4, Non DSIBs 0.1, Top 10 0.4, Other 0.5, Conglomerate 0.3, Other 0.2. - Adverse: UBs 1.7, KBs 0.1, DSIBs 1.4, Non DSIBs 0.4, Top 10 1.4, Other 1.8, Conglomerate 1.1, Other 0.7. - Severe Adverse: UBs 3.5, KBs 0.3, DSIBs 3.1, Non DSIBs 0.7, Top 10 3.0, Other 3.7, Conglomerate 0.7, Other 1.2. - Footnotes: - 1/ Figures at the end of the stress test horizon (2022). - 2/ In this table conglomerate refers to domestic conglomerates. - UKBs regulatory minima reiterated: - CAR 10 percent (Basel III 8 percent), CET1 ratio 6 percent (Basel III 4.5 percent), Tier 1 ratio 7.5 percent (Basel III 6 percent). Required buffers: 2.5 percent capital conservation buffer and, if applicable, a D-SIB buffer (of 1.5 or 2 percent). ### Heterogeneity and vulnerability across banks - Banks’ stressed capital ratios vary considerably, driven by variations in starting capital ratios and initial PDs. - At the end of the stress test horizon, the median CAR is 15.08 percent under the adverse scenario, with an interquartile range from 4.4 percent to 26.3 percent. - Variations are more pronounced across UKBs due to larger variations in their starting CARs. - State-owned banks: - All three state-owned banks (representing 15 percent of the total banking system’s assets) fail the test even in the upside scenario. - Their CAR drops from 13.7 percent to 1.8 percent under the baseline and to -0.9 percent under the adverse scenario. - Vulnerability mainly explained by relatively low starting capital ratios and high initial PDs. - Foreign bank branches and subsidiaries: - Foreign bank branches (representing 6 percent of system assets) are resilient under all scenarios; starting CAR 28.8 percent and lower starting PDs help maintain CARs above regulatory minimums despite similar capital depletions. - Foreign subsidiaries are more vulnerable due to higher starting PDs; example: aggregate CAR falls from 18 percent to 7.7 percent under the baseline (continued discussion truncated in source). *International Monetary Fund — PHILIPPINES (Bank Stress Test results, 2022).* ### 3.2 under the adverse scenario. However, these subsidiaries represent less than one percent of the ### 1phlea2022002 - 3.2 under the adverse scenario. However, these subsidiaries represent less than one percent of the ### Key results and major caveats - Stress-test results should be interpreted with a high degree of caution given large economic and model uncertainty. - The credit risk projection is based on the historical relationship between macroeconomic variables and PD; this relationship may not hold given the unusual nature of the COVID-19 crisis. - Results are sensitive to micro-assumptions: LGD, the cure rate, and the interest margin shock. Appendix III presents variations to these parameters (not reproduced here). - Behavioral assumptions in the standard FSAP stress test (e.g., most NPLs remain on banks’ balance sheets throughout the horizon, write-offs/restructuring/sales to special purpose vehicles assumed zero) tend to give conservative estimates and are set to ensure cross-country comparability. ### Supervisory stress-testing practices at BSP - BSP’s supervision sector conducts semi-annual micro-prudential single factor solvency stress tests covering credit and market risks. - Credit risk exercise assumptions: - Write-off rates of 20 percent and 50 percent over total net loans for certain exercises. - Coverage: all UKBs and TBs. - Segments: various economic industries, types of consumer loans, and large exposures (conglomerates). - Market risk exercise assumptions: - Increases in domestic interest rates range from 300bps to 500bps. - Increases in U.S. interest rates range from 100bps to 300bps. - Exchange rate shocks range from 10 percent to 30 percent. - Coverage: all UKBs and their TBs subsidiaries, stand-alone TBs with total assets of at least 5 PHP billion or total capital of at least 1 PHP billion. - Quarterly stress tests on real estate exposures: - Shock equivalent to a 25 percent write-off on all real estate exposures; covers all UKBs and TBs. - BSP conducted ad hoc stress tests (e.g., impact of loan moratoria on bank solvency in summer 2020). ### Macroprudential vs microprudential stress testing - Supervision sector could start macro scenario–based micro-prudential stress tests (e.g., to support ICAAP and to compare BSP top-down results with bank-reported exercises). - BSP’s macroprudential unit currently focuses on systemic risk analysis of non-financial sectors (network analysis of NFCs and vulnerability analysis of NFCs and real estate developers) but does not conduct macroprudential stress testing of banks. - Macroprudential stress tests could: - Focus on systemic parts of the financial system (e.g., UKBs). - Include macro-financial feedback effects (second-round effects). - Link corporate sector stress test results to banks’ credit risks from NFC loans. - Inform potential contingent liabilities of government credit guarantee programs. - Be published at high level in BSP’s semi-annual Financial Stability Review once methodology is established. ### Data needs and IFRS 9 implications - Current supervisory data indicate only performing vs non-performing loans; BSP does not collect PD nor LGD data under standardized Basel II/III approach. - Additional data and indicators recommended: - PD and LGD indicators from banks and credit registries. - Collateral information, regularly updated collateral values, and loan-to-value (LTV) ratios. - Corporate and household survey data to identify appropriate DSTI levels. - Improved credit registry data and long-term credit registry when available. - IFRS 9 (adopted 2018) introduces ECL framework and opportunities: - ECL encourages forward-looking credit risk management using PD and LGD concepts. - BSP adopted IFRS 9-based accounting provisions to calculate regulatory capital from the beginning of 2018. - BSP should collect upgraded data to enhance top-down credit risk stress tests compatible with IFRS 9 while continuing collection under the previous format to preserve long-term series. ### Macro-financial linkage and second-round effects - Methodology: - A bank-by-bank panel credit growth model includes bank solvency test results as explanatory variables; model estimated using UKBs (2008Q1-2019Q3) which have about 90 percent of the banking sector’s assets. - Aggregate credit growth projection fed into a structural-VAR (SVAR) including credit, GDP, inflation, policy rate, and exchange rates; credit shock treated as exogenous. - Key empirical findings: - Increases in changes of NPL and loan loss reserve ratio reduce credit growth. - CAR did not show statistically significant impact on credit growth in the Philippines (hence excluded). - Long-run elasticity: a sustained one percentage point increase in the rate of credit growth leads to a 0.07 percentage point increase in first-year GDP and a 0.16 percentage point increase in two-year cumulative GDP. - Second-round effect magnitudes: - In the adverse (severe adverse) scenario, the banking sector CAR declines by nearly 8 (12) percentage points. - These CAR declines could reduce real GDP level by additional 4 (9) percentage points by 2021. - The effect may persist over the remaining test horizon. - Caveats on second-round estimates: - May overestimate effects because alternative financing (foreign financing, capital market financing) could fill bank credit gaps. - Potential double counting as initial macro scenario was generated by a DSGE model that includes banking and credit growth, while feedbacks are estimated with a separate SVAR. - The SVAR linear and symmetric structure may underestimate highly non-linear and asymmetric second-round effects. ### Counterfactual policy analysis (NPL write-off / timely loss recognition) - Counterfactual considered: one-time write-off of NPL worth 30 percent of LLP stock in 2021 financed by available excess capital. - Excess capital sources could include limiting dividend distribution, raising new capital from existing shareholders, owner families of conglomerates, new shareholders, and government as last resort. - Estimated policy costs and benefits (2019 real GDP = 100): - Benefit 1 (Maximum difference in the level of real GDP during 2020-22): - Baseline: 1 - Adverse: 1.62 - Severe adverse: 2.86 - Benefit 2 (Sum of differences in the level of GDP from 2020 to 2022): - Baseline: 2.31 - Adverse: 3.20 - Severe adverse: 4.13 - Cost (30 percent of loan-loss provision stock as of 2021, one time): - Baseline: -1.56 - Adverse: -2.16 - Severe adverse: -2.88 - Interpretation: - Single-year benefits roughly match costs, but benefits accrue over multiple years and appear to exceed costs when considering 2021–22 and outer years (e.g., 2023–24). - Results are consistent with the Philippines’ post-AFC experience where forbearance measures contributed to a prolonged “credit-less recovery.” ### Policy recommendations related to solvency and forbearance - Limit bank dividend distributions as a precautionary measure given significant downside risks. - Be ready to take additional measures to strengthen bank capital if downside risks materialize, including: - Policies facilitating sale and recovery of bad assets. - Raising additional capital starting with conglomerate owner families and private sector funding. - Public funding only as a last resort. - Allow forbearance measures to lapse as scheduled and avoid introducing new ones; forbearance hampers banks’ ability to support the economy and can undermine financial stability. - Continue using flexibility in accounting and Basel capital frameworks and develop/use macroprudential tools and buffers. ### Bank liquidity stress testing: scope, calibration, and results - FSAP top-down liquidity stress test used end-2019 regulatory data covering all 46 UKBs (including 11 branches of foreign banks); UKBs represent 92 percent of total banking sector assets. - Funding and liquidity characteristics: - Customer deposits are main funding; corporate and retail deposits each account for about half of total deposits. - Other forms of funding account for 18 percent of non-equity liabilities. - High quality liquid assets (HQLA) composition: cash and reserves and government securities account for half and 46 percent of HQLA, respectively. - Reserve requirement: 12 percent as of September 2020 and 14 percent at end-2019. - Banks’ direct cross-border exposure ≈ 10 percent of bank assets and liabilities. - Dollarization: 15 percent of deposits and 11 percent of loans are in FX. - Remittances ≈ eight percent of GDP per year. - LCR-based tests and calibrations: - Three scenarios: Basel 2013 calibration (actual latest observed LCR), wholesale funding shock calibrated for COVID-19-like shock, and combined shock (adds retail funding shock, haircuts on liquid assets, larger haircuts on inflows). - Wholesale shocks: 60 percent withdrawals of NFC deposits and 35 percent for operational deposits. - Retail withdrawals: 20 percent and 10 percent for less stable and stable deposits based on largest monthly single-bank past withdrawals of close to 11 percent. - Haircuts: 50 percent on corporate bonds. - Cash flow–based stress test (CFST): - Evaluated funding gaps for 1, 3, and 6 months using maturity ladder, simulated asset valuation shock akin to severe adverse solvency scenario. - Assumptions: one-month deposit withdrawals of 50 percent (institutional) and 20 percent (household); half of maturing cash inflow rolled over during initial month; liquidity shock persists up to 6 months with declining severity. - Two types of counterbalancing capacity (CBC): including and excluding required reserves to reflect liquidity release effects from reducing required reserves. - NSFR assessment used standard Basel III reported by banks. ### Liquidity stress test outcomes - LCR results: - System all-currencies LCR (Basel 2013): 157 percent. - System FX LCR: 239 percent. - Median individual bank total currency LCR: 207 percent. - Median individual bank FX LCR: 88 percent. - Most banks exceed 100 percent LCR; a few banks (mainly foreign bank branches) fall below requirement. - Under wholesale funding shock: - System all-currencies LCR falls to 101 percent. - System FX LCR falls to 128 percent. - Median LCR: 105 percent; median FX LCR: 71 percent. Several banks, especially foreign branches, have insufficient FX liquidity. - Under combined severe shock: - System all-currencies LCR falls to 69 percent. - System FX LCR falls to 80 percent. - Liquidity shortfalls most significant in several foreign bank branches. - CFST results: - Only a few banks show net funding gaps under the imposed stress; most gaps are marginal and largest at 1-month horizon. - Only one bank shows a shortfall of 1.8 percent of GDP. - All banks have enough required reserves to cover their shortfalls; lowering reserve requirements would allow meeting liquidity needs. - NSFR: - Exceeds 100 percent in most banks; some banks slightly below 100 percent; very few (mainly foreign branches) are further below. ### Liquidity policy and toolkit recommendations - Consider introducing an FX LCR if signs of FX liquidity risk build-up (FX liquidity currently concentrated in a few foreign bank branches). - Enhance liquidity stress test toolkit by advancing cashflow stress test and considering a system-wide test. - Increase granularity of collected maturity-structure information, including split between retail and wholesale deposits and split of loans into households and NFCs. - Enable a framework for system-wide liquidity stress testing that factors in endogenous liquidity pressures in the non-financial sector. ### Loan moratorium and bank–NFC liquidity linkages - COVID-19 and potential capital outflows could substantially weaken NFC liquidity; NFCs are largest takers of funding from abroad. - Liquidity contagion channels from NFCs to banks: - NFCs may liquidate domestic bank deposits, increasing corporate deposit runoff rates and affecting bank liquidity. - NFCs may seek additional funding (new loans or rollover), increasing banks’ outflow/inflow pressures if alternative funding sources are limited. - Loan moratoria (Bayanihan to Heal as One Act) introduce further complexity to linkages; moratoria initially applied regionally for 90 days and additional 60 days introduced in September (context truncated in source). *Source: IMF staff text from 1phlea2022002 — excerpted section provided.* ### 2020. The direct effect of loan moratoria is to improve NFC cash balance while reducing bank cash ### 2020. The direct effect of loan moratoria is to improve NFC cash balance while reducing bank cash ### Mechanism and role of bank rollover rates - Direct effect: loan moratoria improve NFC cash balance while reducing bank cash inflows from loan repayments. - Rollover rate equivalence: Even without moratoria, if banks rollover 100 percent of repaid loans (including interest payment component), the resulting NFC and bank cash balances would be the same as the full moratorium case. - Interaction: With moratoria, if banks reduce the rollover rate of repaid loans, it weakens NFC cash position and increases banks’ liquid asset balance. Moratorium usage and banks’ rollover behavior act as substitutes. - Treatment of negative cash balances: Some firms’ cash balances decline to negative; the exercise considers whether banks provide additional financing so firms can restore their cash balance to 0. ### Scenario design and key assumptions - Number of scenarios: 8 scenarios. - Moratorium lengths considered: five months (already took place) and 12 months (potential extension). - Moratorium utilization rates considered: 0 percent (benchmark without moratoria), 50 percent, 70 percent (observed in April–May), and 90 percent. - Bank rollover rate assumptions: central assumption 90 percent (based on NFC data during past distress episodes); pessimistic assumption 50 percent (used for Basel III LCR one-month calculation). - Behavioral assumption note: During normal times, the observed rollover rate is above 100 percent, reflecting credit growth. ### NFC liquidity outcomes (cash-to-asset ratios and sensitivity) - Baseline (no moratoria, 70 percent moratoria usage counterfactual): assuming 70 percent of loans use moratoria, no new financing from banks other than rollover, and 90 percent rollover ratio, NFC’s cash-to-asset ratio falls slightly from 9.7 percent at end-2019 to 8.8 percent end-2020 without moratoria. - Low rollover outcome: when the rollover ratio is 50 percent, the cash-to-asset ratio declines to 5.3 percent end-2020 (under the same utilization and financing assumptions). - Moratorium extended (12 months) outcomes: - With 12-month moratoria and 90 percent rollover ratio, NFC cash-to-asset ratio improves to 10.6 percent (beyond the 2019 level). - With 12-month moratoria and 50 percent rollover ratio, NFC cash-to-asset ratio is 9.6 percent (similar to end-2019 levels). - Implication: as long as banks rollover most existing loans, only a relatively small share of NFCs need to use moratoria to maintain liquidity; with 90 percent rollover, even 50 percent moratorium utilization can bring NFC cash balance above 2019 level. ### NFC deposit withdrawal and bank liquidity linkages - Assumed liquidation order: NFCs liquidate all of their liquid assets proportionally (due to data limitation on composition of “cash and cash equivalents”). - Extreme deposit outflow scenario: if banks rollover only 50 percent of existing loans and there are no moratoria, NFC deposit withdrawal rate could be as high as 45 percent over 12 months. - Mitigated by moratoria: the 45 percent withdrawal rate drops to 13 percent with moratorium use of 50 percent. - High rollover case: when banks continue to rollover 90 percent of loans, corporate deposit withdrawal rate goes down to 9 percent without moratoria — even lower than some moratorium-use cases under tight credit conditions. - Overall bank cash (CBC) after indirect effects: once NFC deposit withdrawal effects are incorporated, banks’ remaining CBC after stress with NFC deposit outflow is about 110 percent of the end-2019 level across the four cases shown. - Reason: offsetting roles of NFC and bank behaviors; moratoria reallocate liquidity between banks and NFCs rather than changing system-wide liquidity, potentially nullifying policy effects at the system level. - Caveat: differences arise when NFCs rely more on nonbank and international financing or liquid assets, changing net liquidity available to the system. ### System-wide considerations and sensitivities - Alternative NFC financing: NFCs may have alternative sources including local bank deposits, deposits abroad, domestic and foreign securities, new financing from foreign banks, and domestic and international bond markets. Availability depends on global liquidity and country risk perceptions and domestic corporate bond market development. - Aggregate drivers: the overall effects critically depend on the share of borrowers that utilize moratoria and banks’ rollover decisions, as well as banks’ health, liquidity conditions, and risk tolerance. - Model constraints: exercise focuses on bank–NFC linkages and is a partial stress test covering only a subset of possible shocks; household deposit withdrawal is not considered in the exercise. The CBC balance remains above the 2019 level partly because the refinancing rate of banks’ own term borrowing is set at high (90 percent). ### Policy recommendations and operational actions for BSP and banks - Monitoring contingent financing plans: - BSP could monitor contingent financing plans of banks and NFCs to better gauge likely system-wide policy effects. - For banks, ICAAP under Pillar 2 of Basel III should include banks’ liquidity stress test results and contingent financing plans; resolution planning should require state-contingent financing plans, including Emergency Liquidity Assistance and other available support for much greater stress than ICAAP scenarios. - BSP could collect this information from each bank and assess likely banking sector-wide behavior. - Improve NFC information and disclosure: - Philippine firms face substantial disclosure requirements on financing details (e.g., currency composition), but these are usually footnotes. Standardizing disclosure format and converting to a database for quantitative analysis would be critical. - The Securities Exchange Commission’s initiative to create an NFC database is welcomed. - BSP could use its new powers under the BSP Act to request any information that matters for financial stability analysis from any economic sector. - BSP could consider discussing contingent financing plans of systemically important NFCs to refine behavioral assumptions in system-wide liquidity stress tests. - Institutional collaboration: - Strengthen collaboration between the macroprudential unit and the supervision sector, combining NFC and bank cash-flow stress tests. The BSP could consider arrangements similar to macroprudential stress tests discussed for bank solvency. *Source: IMF staff estimates and analysis contained in the provided chapter.* ### 1. NFCs may cash their liquid assets, including bank deposits, when their cash inflows from earnings are ### 1. NFCs may cash their liquid assets, including bank deposits, when their cash inflows from earnings are ### Mechanics of the bank–NFC liquidity linkage - NFC deposit withdrawal rate is measured by (change of NFC cash balance between end-2020 and end-2019)/(end-2019 cash balance), assuming NFCs liquidate all types of liquid assets proportionally. - If NFCs cash in assets other than bank deposits and find alternative financing (e.g., bonds), the withdrawal rate stays low. - Loan moratoria: - Help NFCs retain liquidity (indirect effect on NFCs). - Reduce cash inflows to banks (direct effect on banks). - Additional bank lending (rollover of repaid loans and new financing) shifts liquidity from banks to NFCs. ### Stress test scenarios and qualitative outcomes (from figures and narrative) - Common scenario parameters: - 12-month moratoria and stress test period. - Moratorium utilization scenarios: 70 percent; 50 percent. - Bank rollover rates: high (90 percent) and low (50 percent). - 12-month moratoria, 70 percent moratorium utilization by NFCs, and high rollover rate (90 percent): - "Moratoria improve NFCs’ cash balance to above the pre-stress level, but high refinancing rate limits the deterioration of cash position even without moratoria." - "While the direct effect of moratoria reduces banks’ liquidity buffer noticeably, the policy effects declines once the indirect effects from deposit withdrawal is accounted for." - 12-month moratoria, 70 percent moratorium utilization by NFCs, and low rollover rate (50 percent): - "Moratoria improve NFCs’ cash balance substantially compared to the levels without the measure." - "With lower refinancing rate, the direct effect of moratoria on bank liquidity increase substantially. However, the policy effects becomes muted once the indirect effects from deposit withdrawal is accounted for." - 12-month moratoria, 50 percent moratorium utilization by NFCs, and high rollover rate (90 percent): - "When bank loan rollover rate is high, NFCs can maintain their cash at the 2019 levels even with lower moratorium utilization below 50 percent." - "Lower moratorium utilization reduces banks’ liquidity loss slightly, but the overall impact including indirect effects of NFC deposit withdrawal remains the same as the case with higher moratorium utilization rate." - 12-month moratoria, 50 percent moratorium utilization by NFCs, and low rollover rate (50 percent): - "When banks’ rollover rate is low, NFCs would need to use moratorium more than 50 percent (indeed, close to 70 percent) to maintain the 2019 cash levels." - "Banks can withhold their liquidity by rolling over existing loans less. But the overall effects including deposit run-off remain the same." ### Key quantitative figures presented in the results (as shown) - NFC Cash-to-AssetsRatio (In percent of end-2019 total assets, 12 month, 70 percent moratorium utilization, and 90 percent rollover rate): - 9.7 - 8.8 - 9.0 - 10.6 - 10.8 - Bank Counter-Balancing Capacity (CBC) (12 month moratorium and test horizon; rollover rate = 90%, moratoirum take up = 70%; in percent of initial CBC): - 100 - 114 - 106 - 106 - 102 - 110 - 109 - 90 - 100 - 110 - 120 - 130 - 140 - 150 - NFC Cash-to-AssetsRatio (In percent of end-2019 total assets, 12-month, 70 percent moratorium utilization, and 50 percent rollover rate): - 9.7 - 5.3 - 6.3 - 9.6 - 9.8 - Bank CBC (rollover rate = 50%, moratoirum take up = 70%; in percent of initial CBC): - 100 - 147 - 110 - 106 - 111 - 111 - 110 - 90 - 100 - 110 - 120 - 130 - 140 - 150 - NFC Cash-to-AssetsRatio (In percent of end-2019 total assets, 12 month, 50 percent moratorium utilization, and 90 percent rollover rate): - 9.7 - 8.8 - 9.0 - 10.1 - 10.2 - Bank CBC (rollover rate = 90%, moratoirum take up = 50%; in percent of initial CBC): - 100 - 114 - 106 - 106 - 105 - 109 - 108 - 90 - 100 - 110 - 120 - 130 - 140 - 150 - NFC Cash-to-AssetsRatio (In percent of end-2019 total assets, 12-month, 50 percent moratorium utilization, and 50 percent rollover rate): - 9.7 - 5.3 - 6.3 - 8.4 - 8.7 - Bank CBC (rollover rate = 50%, moratoirum take up = 50%; in percent of initial CBC): - 100 - 147 - 110 - 106 - 122 - 111 - 109 - 90 - 100 - 110 - 120 - 130 - 140 - 150 ### Policy measures and context referenced - Notes on rollover rates and context: - "90 percent bank rollover rates are comparable to the distressed level observed during the past crises. During normal time, the rollover rates usually exceed 100 percent. 50 percent is the assumption from Basel III LCR." - The CBC balance remains above the 2019 level because the refinancing rate of banks’ own term borrowing is set at high (90 percent), and household deposit withdrawal is not considered in the exercise (even though loan moratorium and rollover of existing bank loans to the household are accounted for). - For banks, this exercise is a partial stress test exercise focusing on bank-NFC linkages, covering only a subset of possible shocks. ### Data, scope, and assumptions - Starting position: December 2019 (supervisory data on a 'solo basis'). - Institutions included in macro scenario tests: 542 banks (46 UKBs: 21 universal banks, 25 commercial banks; 49 TBs; 447 RCBs). - Market share covered: Nearly 100% of total banking sector assets. - Sources cited for figures: S&P Capital IQ; and IMF staff estimates. *Italic: Source — IMF staff estimates and figures as presented in the content unit.* ### 2. Channels of ### 2. Channels of Risk Propagation ### Methodology and Models - IMF Solvency Stress Test Workbox (Balance-sheet model). - Satellite models for macro-financial linkages: - Credit Risk: Satellite models per bank type to estimate loan losses. Regression model for logit transformed PDs. Regressors include the lagged dependent variable and contemporaneous and lagged macroeconomic variables: GDP growth, short term rates, term spread, unemployment, stock price, and exchange rate. - Market risk: valuation losses for HfT and AfS securities are calculated using a Mark to Market (MtM) approach. Valuation losses for held-to-maturity (HtM) securities are calculated using a credit risk approach. As a sensitivity analysis, an MtM approach is used for the HtM securities. - Net interest income: A gap analysis is conducted based on granular data on asset/liability structure of individual banks broken into types of funding sources and time to re-pricing buckets. Interest margin shocks vary per scenario. - Pre-impairment income for banks: Income in the absence of shocks is assumed to stay at the level observed for 2019 with the additional feature that non-performing loans will not generate any income. - No effects from sector-specific mitigation policies are incorporated. Government credit guarantee scale is 0.6 percent of GDP and given only to SMEs and the agricultural sector; moratoria (introduced twice) expired at the end-2020. End-2019 data are used instead of 2020 data. - Stress test horizon: 3 years (2020-2022). - Domain Assumptions: Top-down by FSAP Team. ### Scenario analysis (Tail shocks) - Three macro scenarios: baseline, adverse, and severe adverse. Shocks primarily affect real economic activities; financial conditions remain relatively benign with limited exchange rate pressures and central bank rate cuts supported by ample global liquidity. - Baseline scenario: - Follows October 2020 WEO. - Shows sharper GDP contraction in 2020 than the AFC but stronger medium-term growth in line with potential growth ~6½ percent. - Two-year cumulative growth in 2021 is 14.8 percentage points lower compared to January 2020 WEO, corresponding to a three standard deviation shock using data from 1990-2019. - Unemployment: 11.3 percent at end-2020; returns to pre-COVID levels by 2021. - Upside scenario: - Faster recovery in second half of 2020. - Real GDP contracts by -6.7 percent in 2020 (vs -8.4 percent under baseline). - Unemployment returns to pre-COVID levels by end-2020. - Adverse scenario: - Prolonged containment throughout 2020 and scarring in 2021. - BSP cuts policy rates by 290 basis points in 2020. - Stock prices decline by over 14 percent in 2020-21. - Net interest margin declines by 15 percent at the worst point in the three years. - Unemployment: 14.8 percent by 2020 and 8.6 percent by 2021; returns to pre-COVID levels by 2022. - Severe adverse scenario: - Prolonged and more stringent containment in 2020; scarring in 2021 equivalent to a 3.8 standard deviation shock to two-year cumulative growth. - BSP cuts policy rates subject to zero lower bound, reducing short term interest rates 290 basis points in 2020. - Stock prices decline by over 14 percent in 2020-21. - Net interest margin declines by 20 percent at the worst point in the three years. - Unemployment: 18.4 percent by 2020 and 11.9 percent by 2021; returns to pre-COVID levels by 2022. - Adverse macro scenarios are constructed using the DSGE model developed for the climate scenario. - Domain Assumptions: Top-down by FSAP Team. ### Risks, Behavioral Assumptions, and Buffers - Risks/factors assessed: - Credit risk (provision costs) - Market risk, including FX risk - Stress on pre-provision profits, including interest margin - Behavioral adjustments: - Balance sheet growth assumption: Quasi-Static—balance sheet size/GDP remains constant. - Balance sheet composition remains constant over the stress test horizon. - Banks can only accumulate capital through retained earnings. - Dividend policy: Banks pay dividends only if net income after taxes is positive, with payout ratio consistent with 2019 ratios for UKBs and historical experience 2014-2019 for TBs and RCBs. - Out of 542 banks, 45 paid dividends: 11 UKBs, 2 TBs, and 32 RCBs. - Average payout ratios: 13 percent for UKBs, 0.8 percent for TBs, and 0.9 percent for RCBs. - Tax rate: 30% (corporate tax rate). ### Calibration of Risk Parameters and Regulatory Standards - PDs: proxies based on actual and estimated new NPL flows over performing loans. PDs for banks with limited credit information taken as weighted average PD of other institutions. - LGDs: 68 percent for UKBs, 35 percent for TBs, and 66 percent for RCBs, based on average historical provision coverage ratio (provisions/NPLs). - Cure rates (with respect to NPL(t-1)): 10 percent for UKBs, 22 percent for TBs, and 24 percent for RCBs per year (equivalent to a quarter of historical averages). In addition, an extra annual cure rate of 18 percent on New NPLs (2020) is assumed for UKBs in 2021 (totaling 28 percent). - Regulatory/accounting framework: Basel II standardized approach. - Hurdle rates based on minimum capital requirements: - 6 percent for Common Equity Tier 1 (applies only to UKBs) - 7.5 percent for Tier 1 - 10 percent for total capital (T1+T2) - RWAs evolve with credit growth, net of increases in provisions, and are adjusted by new NPLs net of provisions to reach weight of 150% required by regulation. - Domain Assumptions: Top-down by FSAP Team. ### Reporting Format for Results - Outputs include: - Capital shortfalls per bank type - Number of banks and percentage of assets that fail to meet hurdle rates per bank type - Evolution of capital ratios under the scenario horizon per bank type and for various bank classifications - Decomposition of drivers of changes in capital ratios per bank type - Distribution of capital ratios per bank type over the scenario horizon --- ### Banking Sector: Second Round (Bank–Macro Transmission) ### Institutional Perimeter and Data - Institutions included: 46 UKBs (21 universal banks, 25 commercial banks). - Market share: 90.8% of total banking sector (gross) loan. - Data and baseline date: Solvency stress test results for Baseline (October WEO), Adverse and Severe Adverse scenarios. Starting position: End of 2019 data. - Domain Assumptions: Top-down by FSAP Team. ### Methodology and Satellite Models (Bank–Macro Transmission) - Simplified application of Catalan and Hoffmaister (2020). - Elasticities of individual bank loan growth to macroeconomic variables and bank-specific characteristics; elasticities of macroeconomic variables to aggregate bank loan. - Credit Growth Model: - Panel model for UKBs with individual bank credit growth as dependent variable. - Final model regressors: lags of real credit growth, contemporaneous and lags of macro variables (real GDP and change in policy rate), and bank-specific factors (change in NPL ratio and loan loss reserve ratio). - Estimation period: 2008Q1-2019Q3. - CAR and distance to regulatory minimum were excluded (not significant). - SVAR macro-financial model: - Five equations capturing interactions of real credit, real GDP, inflation, real policy rate, and nominal exchange rate. - Cholesky decomposition for contemporaneous relations; real credit enters first. - Estimation period: 200Q4-2019Q3. - Domain Assumptions: Top-down by FSAP Team. ### Counterfactual Policy Experiment and Adjustments - Counterfactual: reducing LLP stock by 30 percent only in 2021 (interpreted as using bank capital and income to write off NPLs by closing the gap between NPL and LLP (1-LGD)). - Adjustments and assumptions: - UKBs' aggregate loan growth approximates overall banking sector loan growth. - Baseline real credit growth assumed equal to baseline real GDP growth. - Real credit growth in adverse scenarios calculated from baseline plus deviations due to macro and bank-specific variable differences. - Horizon: 3 years (2020-2022). ### Reporting Format for Results - Second Round outputs: GDP growth path and GDP level path (relative to 2019 GDP). - Policy Simulation outputs: - Benefit: (1) maximum difference in the level of GDP (based on cumulative GDP growth), and (2) sum of differences in the level of GDP (based on cumulative GDP growth). - Cost: amount of funds injected into the banking sector to implement the policy. --- ### Banking Sector: Liquidity Risk ### Institutional Perimeter and Data - Institutions included: - 40 UKBs for modified LCR tests (90 percent of the banking system) - 46 UKBs for cashflow tests and NSFR (92 percent of the banking system) - Data and baseline date: - Supervisory data on a solo basis - September 2019 (to be updated using March 2020 data) - Domain Assumptions: Top-down by FSAP Team. ### Methodology and Tests - LCR (one month) - Cashflow test (one, three, and six months) - NSFR (one year) ### Risks and Buffers - Risks: Funding liquidity, Market liquidity, Moratoria effects. - Buffers: - LCR: (stressed) cash inflows and HQLA - Cashflow: (stressed) cash inflows and counterbalancing capacity - N.A. ### Tail shocks and Parameters - LCR run-off rates: - Household deposits: 20 percent for less stable deposits and 10 percent for stable deposits. - Institutional deposits: 60 percent for corporate deposits, 35 percent for operational deposits, 20 percent for less stable, and 10 percent for stable retail deposits. - LCR haircuts: 50 percent haircut to Level 2a assets (corporate bonds). - Cashflow run-off rates: - 1 month—50 percent on institutional deposits and 20 percent on household funding; roll-off rate of 50 percent on cash inflows. - Rates decline over 1-3 months and 3-6 months to reach 30, 10, and 10 percent. - Cashflow haircuts: 20 percent on unencumbered eligible collateral and 50 percent on equities. - NSFR: Same as Basel III. ### Regulatory Standards - LCR: 100 percent (liquidity shortfall by bank). - Cashflow: net funding gap (shortfall) by bank. - NSFR: 100 percent. --- ### Non-Financial Corporate Sector (NFC) ### Institutional Perimeter and Data - Institutions included: 151 non-financial firms (147 listed and 4 non-listed). - Market share: - 44 percent of total NFC debt - 70 percent of total market capitalization - 46 percent of total bank loans - Data and baseline date: Capital IQ, S&P Global Market Intelligence; consolidated balance sheets as of end-2019. - Test horizon: 2020. - Domain Assumptions: Top-down by FSAP Team. ### Methodology and Tail Shocks - Debt service capacity: ICR (one-year). - Cash flow: cash ratio = cash and cash equivalent / current liabilities (one-year). - ICR shocks: - Interest payment shock: 0 percent. - Exchange rate shock: 4 percent appreciation against the USD. - Operating income shock (percent of operating income in 2019, varying across industries): - Baseline: between -35 percent and -137 percent - Upside: between -26 percent and -103 percent - Adverse: between -39 percent and -151 percent - Severe Adverse: between -47 percent and -185 percent - In all scenarios, least affected industry: utilities; most affected: consumer discretionary. - Cash flow assumptions: - Capex2020 = minimum of 0.25*Capex2019 and 0.5*depreciation2019 - Debt rollover ratio = 0.9 of maturing debt - Dividend payments = 0 ### Reporting Format for Results - Output presentation: - Distribution of ICR (median and interquartile range) - Debt-at-risk share by industry - Firm-at-risk share by industry - Cash ratio by industry --- ### Bank–NFC Liquidity Linkage ### Institutional Perimeter and Data - Institutions included: 151 NFCs and 40 UKBs. - Market share: - 44 percent of total NFC debt - 70 percent of total market capitalization - 46 percent of total bank loans - 90 percent of the banking system (UKBs) - Data and baseline date: Capital IQ, S&P Global Market Intelligence, the BSP (UKB data); consolidated NFC balance sheets and solo-based UKB balance sheets as of end-2019. - Domain Assumptions: Top-down by FSAP Team. ### Methodology and Linkage Mechanics - Based on 2020 IMF COVID-19 note in “system-wide FX liquidity stress test.” - Cash-flow based liquidity stress tests for banks and NFCs and link assumption parameters. - For banks, analysis considers shocks only to transactions with NFCs. - Aggregate NFC liquidity stress test results (e.g., deposit withdrawal rate and new financing need) are applied to bank-by-bank liquidity stress tests (uniform assumption across banks). - NFC cash flow calculation formula (cash balance at x months in 2020): - Cash balance (x months in 2020) = Cash balance 2019 + earnings shock × net cash flows from operations × (x/12) - stressed capital expenditure × (x/12) - debt repayment with moratorium (= original debt repayment × (1- moratorium utilization) × (x/12)) + rollover of repaid debt (= debt repayment with moratorium × rollover rate) + interest income × (x/12) – interest expense × (1-moratorium utilization) × (x/12) - dividend payment × (x/12) + new financing (to bring cash balance(x, 2020) to zero for each firm) - In case with 5-month moratorium, debt payments (principal and interest) are set at about 40 (=5/12) percent of the original amounts in the baseline. - Capex2020 = minimum of 0.25 × Capex2019 and 0.5 × depreciation2019 - Dividend payments = 0 - Bank cash flow (Counterbalancing capacity, CBC) at month x in 2020: - CBC at month x = CBC at end 2019 + net cash flows from operation within x (excl. loans and interests) + debt service receipt (= bank loan principal repayment in x months × (1-moratorium utilization)) – rollover rate × debt service receipt + interest income in x months × (1- moratorium utilization) – interest expense in x months - matured bank debt repayment within x + refinancing rate for banks (90 percent) × banks’ repaid borrowing - deposit runoff rate × stock of deposits (NFC deposit runoff rate = ∆cash by x month in 2020/cash at end 2019) - new financing to NFCs (to bring NFC cash position to 0 or above for each firm) - dividend payment within x + new financing inflows into banks + valuation change of CBC (i.e., haircut) - Note: Moratoria and loan rollovers affect all loans (including household credit) but only NFC deposit withdrawal is considered. ### Tail shocks and Parameters - Moratorium utilization rate = {0 (no moratorium), 50, 70, 90} percent. - NFC loans rollover rate = {50, 90} percent. - New financing to NFCs = {yes, no}. - Refinancing rate of repaid loans by banks = 90 percent. - Liquid asset haircuts: 0% for cash, 2% for high-quality sovereign securities, 8% for low-quality sovereign securities, 11% for corporate bond, 10% for covered bond. ### Reporting Format for Results - NFC: change in cash balance in month x, 2020 / cash balance at end 2019. - Bank: change in CBC by month x, 2020 / CBC at end 2019. --- ### Appendix II — NFC Stress Test Methodology (Hybrid Approach) - Purpose: project sample firms’ debt service capacity (ICR) and cash positions end-2020 under different macro scenarios. - Rationale: COVID-19 crisis produces large supply disruptions and cross-industry impacts differing from past demand-led crises; regression-only approaches insufficient. - Hybrid approach: regression-based augmented with consensus earnings forecasts by market analysts to incorporate cross-industry differences. Step 1 — Apply shocks to ICR subcomponents: - Operating income shock: - Use consensus earnings forecasts for FY2020 from January 2 vintage and June 30 vintage. - For each firm, calculate percentage change of earnings forecasts between vintages; apply industry-median earning shock to individual firms’ 2019 earnings. - EBIT2020,j = (1+EBIT shocki/100) × EBIT2019,j - Interest payment shocks: - Interest payment shock and exchange rate shock applied separately. - Exchange rate shock set as percent change of bilateral exchange rate vis-à-vis the USD between end-2019 and projected end-2020 level in October 2020 IMF WEO. - Exchange rate shock applied to actual foreign currency-denominated portion of each firm’s outstanding debt as of end-2019. *Source: 2. Channels of Risk Propagation (chapter content).* ### 2019. The interest payment shock is set at 0, which is somewhat more conservative than the ### 1phlea2022002 - 2019. The interest payment shock is set at 0, which is somewhat more conservative than the ### Firm-level interest payment and ICR estimation (Steps 1–3) - Interest payment for firm j in 2020 defined as: - INTP2020,j = [FX debt share2019,j × (1+FX shock2020/100) + local currency debt share2019,j] × (1+INTP shock2020/100) × INTP2019,j - In the presented analysis INTP shock is set at 0. - Effective shock to interest payment is primarily from the exchange rate shock (currency depreciation raising FX debt interest payment in local currency). - Preliminary ICR estimate (Step 1): - ICR1 2020,j = EBIT2020,j / INTP2020,j - Regression-based alternative ICR forecasts (Step 2): - Sample: 29,161 non-financial firm-year observations (2002–2019) from ASEAN-6 (Indonesia, Malaysia, the Philippines, Thailand, Singapore, Vietnam). - Separate regressions for ROA, effective interest rate, and leverage (debt-to-asset ratio) with explanatory variables: - Macroeconomic Xk,t: real GDP growth, bilateral exchange rate against the US dollar, domestic interest rate. - Global Wt: world GDP growth, commodity price, LIBOR. - Firm-level Zi,t: lagged dependent variable, lagged total assets, lagged tangible assets-to-total assets ratio. - Macroeconomic and global 2020 values from October 2020 IMF WEO forecasts. - Predicted ROA, effective interest rate, and leverage used to back out predicted ICR: - ICR2020,j2 = (predicted ROA / EPPI) × LEE L? (formula preserved as in source) - Reconciliation of ICR estimates (Step 3): - Recalculate predicted earnings in Step 1 with a re-scaling factor θ: - EBIT2020,j(θ) = (1+ θ × EBIT shocki/100) × EBIT2019,j - θ chosen such that: - Median(ICR1 2020(θ)) = Median(ICR2 2020) - Re-scaling factor θ by scenario: - Baseline: θ = 3.2 - Upside: θ = 2.4 - Adverse: θ = 3.5 - Severe adverse: θ = 4.3 - Final ICR for firm j, ICR1 2020,j(θ), captures aggregate macroeconomic shock severity (sample median ICR) and relative industry earnings shock magnitudes. - Regression estimation results (selected coefficients and diagnostics): - Lagged Dep. Var.: ROA 0.453 (103.62)**, Effective Interest Rate 0.461 (94.45)**, Leverage (log) 0.834 (236.02)** - Real GDP Growth: ROA 0.002 (5.52)**, Effective Interest Rate 0.009 (3.12)**, Leverage 0.002 (0.570) - Exch. Rate (LCU/USD): 0.000 (5.87)**, 0.000 (6.45)**, 0.000 (0.000) - Lending Rate: ROA -0.001 (2.37)*, Effective Interest Rate 0.052 (12.38)**, Leverage 0.012 (2.91)** - LIBOR: ROA 0.000 (0.430), Effective Interest Rate 0.034 (10.63)**, Leverage 0.004 (1.210) - World GDP Growth: ROA 0.002 (3.59)**, Effective Interest Rate -0.005 (1.030), Leverage -0.007 (1.640) - Commodity Price Index: ROA 0.000 (0.370), Effective Interest Rate 0.000 (1.98)*, Leverage 0.000 (0.640) - Lagged Assets: ROA 0.004 (9.16)**, Effective Interest Rate -0.046 (17.36)**, Leverage 0.032 (12.54)** - Lagged Tangibility: ROA -0.003 (1.040), Effective Interest Rate -0.043 (2.29)*, Leverage 0.072 (3.87)** - Constant: ROA 0.069 (4.43)**, Effective Interest Rate -2.132 (20.46)**, Leverage -0.625 (6.36)** - Adjusted R2: ROA 0.30, Effective Interest Rate 0.34, Leverage 0.67 - Observations: ROA 29,161; Effective Interest Rate 28,218; Leverage 30,803 - Note: All macroeconomic variables come from 2020 October IMF WEO forecasts. ### Projecting firms’ net cash positions at end-2020 - End-2020 cash position formula: - cash2020 = cash2019 + EBIT shock × (cash flow from operations2019) – capital expenditure2020 – debt amortization2020 + net interest payment2019 – dividend payment2020 - Assumptions and parameterization: - Capital expenditure2020 = min{0.5*depreciation2019, 0.25*capital expenditure2019} - Rationale: sample median CAPEX in GFC 2009 was ~80% of 2008; COVID-19 severity leads to CAPEX set at 25% of 2019 CAPEX or 50% of 2019 depreciation, whichever is lower. - Debt amortization2020 = 10 percent of maturing debt (debt rollover ratio = 90 percent) - Note: sample median rollover ratio in normal times ~110 percent (nominal debt growth ~10 percent); 2009 median ~100 percent. - Dividend payment2020 = 0 - Caveats: - Analysis does not consider effects of policy support measures. - Sample firms are a small share of non-financial firms in the Philippines and exclude micro-sized informal firms. - Consequently, estimated COVID-19 impact on non-financial corporate financial health is likely downward biased. ### Bank solvency stress test — proxy PDs (Appendix III.A) - BSP does not currently collect PDs and LGDs; proxy PDs calculated using NPL flow data. - When flow of new NPLs is available: - PD approximation: PD_it = 4 * NNI_NPL_t / NPL_stock_t-1 (formula preserved as in source) - When only stock of NPLs is available: - PD approximation: PD_it = 4 * [NPL_dt − (1 − α_t^) NPL_dt-1] / NPL_stock_t-1 (formula preserved) - α_t^ represents proportion of NPLs collected, written-off, cured, restructured, sold to SPVs, or transferred to ROPA. - α calculation and usage: - α derived using 2014–2019 information; change in NPLs between periods expressed and α computed (formulas preserved). - Assuming constant alpha (average over sample period), flow of new NPLs approximated as: - →NNI_NPL_t = NPL_dt − (1 − ᾱ) NPL_dt-1 - Average alpha by bank type: - UKBs: 16.8 percent - TBs: 25.9 percent - RCBs: 26.4 percent - Proxy PD evolution (system and bank types): - PD of UKBs reached one percent by December 2019. - PD of TBs reached 6.6 percent by end-2019. - PD of RCBs reached 9.7 percent by end-2019. - Relatively high PDs early in sample consistent with AFC credit risk increase. ### Credit risk models (Appendix III.B) - Credit risk satellite models estimated separately for UKBs, TBs, RCBs linking macro-financial scenarios to proxy PDs using quarterly 1999–2019 data. - Considered combinations of macro variables (real GDP growth, unemployment rate, short-term interest rates, term spread, stock prices, exchange rate) and different lags. - Model uncertainty: - Extremely severe COVID-19 shocks exceed historical shocks; estimated coefficients may not capture full PD sensitivity. - Projected PDs should be interpreted with caution due to high model uncertainty. - Alternative approaches and training samples yield a wider than usual range of PD paths. - Final models and training periods selected based on in-sample fit, significance of long-run multipliers among models complying with sign constraints, and expert judgment. - Selected model diagnostics (Appendix III, Table 1, summarized): - Sample Period: 2005–2019 - Observations: 60 (per reported columns) - R-squared examples: UKB 0.48; TB 0.40; adjusted R-squared values reported (e.g., 0.194 in table excerpt) - Coefficients and p-values reported for AR lags, Real GDP Growth, Short Term Rate, Unemployment Rate, Exchange Rate, Stock Price Growth. ### Sensitivity of bank solvency stress test results (Appendix III.C) - Key parameters with high sensitivity: LGD, cure rate, interest margin shock. - Parameters varied one at a time across system while holding others constant: - LGD: 35 percent and 85 percent - Interest margin shock: 0 percent during horizon and shock equivalent to 50 percent of paths observed during the AFC - Cure rate: 0 percent (as percent of NPLt-1) and 100 percent of average annualized cure rate observed for each bank type - Capital ratios (end of stress test horizon 2022) — selected figures from Appendix III. Table 2: - Latest actual: 15.6 (percent) - Baseline October: LGD 35% → 14.2; LGD 85% → 10.0; Interest margin shock 0% → 12.0; Interest margin shock 50% → 11.3; Cure rate 0% → 10.2; Cure rate 100% → 14.6 - Upside: 15.1; 12.5; 13.9; 13.2; 12.5; 15.7 (columns correspond to parameter variations as above) - Adverse: 12.9; 6.9; 9.8; 8.7; 7.1; 13.5 - Severe Adverse: 10.5; 1.0; 5.7; 4.1; 1.8; 11.5 - Capital shortfalls (in percent of GDP) — selected figures: - Latest actual: 0.0 - Baseline October: LGD 35% → 0.3; LGD 85% → 1.5; Interest margin shock 0% → 0.9; Interest margin shock 50% → 1.0; Cure rate 0% → 1.3; Cure rate 100% → 0.3 - Upside: 0.1; 0.8; 0.5; 0.6; 0.7; 0.1 - Adverse: 0.7; 2.9; 1.7; 2.1; 2.7; 0.5 - Severe Adverse: 1.3; 5.6; 3.5; 4.2; 5.1; 1.0 - Regulatory minima and hurdle rates (UKBs): - CAR 10 percent (Basel III 8 percent) - CET1 ratio 6 percent (Basel III 4.5 percent) - Tier 1 ratio 7.5 percent (Basel III 6 percent) - UKBs required to hold 2.5 percent capital conservation buffer and, if applicable, D-SIB buffer (1.5 or 2 percent) - TBs and RCBs minima: CAR 10 percent and Tier 1 ratio 6 percent (Basel I based); TBs and RCBs not subject to buffer and leverage ratio requirements ### Macro-financial linkage and credit growth (Appendix IV highlights) - Panel estimates of credit growth model (Appendix IV. Table 1) and SVAR identification (Appendix IV. Table 2) summarized. - Real credit growth long-run elasticities and diagnostics (selected ARDL results): - Real Credit Growth (lagged) sum of coefficients: -0.225; F-statistic 89.130; Marginal significance 0.000 - Real GDP Growth: sum of coefficients 3.235; F-statistic 7.023; Marginal significance 0.001; Long-run elasticity 2.641 - Change in Nominal Policy Rate: sum -3.029; F-statistic 2.811; Marginal significance 0.060; Long-run elasticity -2.473 - Change in Non-performing loan rate: sum -3.285; F-statistic 11.317; Marginal significance 0.000; Long-run elasticity -2.681 - Change in Loan loss reserve rate: sum -4.701; F-statistic 17.268; Marginal significance 0.000; Long-run elasticity -3.838 - Adjusted R2 0.134; Standard error of the regression 0.139; Number of observations 1701 - SVAR elasticities (Appendix IV. Table 4) — selected one-year responses: - Real Credit shock: One year 0.07; Two years 0.16 - Inflation shock: One year 0.08 - Real Policy Rate shock: One year -0.02 - Nominal Exchange Rate shock: One year -0.03 *Source: 1phlea2022002 - 2019. The interest payment shock is set at 0, which is somewhat more conservative than the (PDF chapter/section).* --- _Source: https://www.imf.org/-/media/files/publications/cr/2022/english/1phlea2022002.pdf_