## 1prtea2026005

## Source details

**Canonical URL:** [1prtea2026005](https://www.imf.org/-/media/files/publications/cr/2026/english/1prtea2026005.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2026/english/1prtea2026005.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2026/english/1prtea2026005.pdf.json)

---

### Executive summary — macroprudential toolkit and recent measures
- Banco de Portugal (BdP) has developed a full‑fledged macroprudential toolkit.
- Activation, in January 2026, of a countercyclical capital buffer (CCyB) set at a positive cycle‑neutral (PCN) target rate of 0.75 percent.
- Since 2023, toolkit includes a sectoral systemic risk buffer (sSyRB) targeting residential real estate (RRE) risks.
- Borrower‑based measures (BBMs), including LTV and DSTI limits, have been in place since 2018 as recommendations for banks.
- The CCyB at 0.75 percent, when combined with the sSyRB, "totals about 1 percent of releasable capital."

### Institutional framework, governance, and decision making
- Three agencies responsible for systemic risk oversight: BdP, Securities Market Commission (CMVM), and Insurance and Pension Funds Supervisory Authority (ASF).
- BdP is the designated macroprudential authority since 2013; also microprudential supervisor under the SSM and national resolution authority under the Single Resolution Mechanism.
- BdP Board meets ordinarily at least once a week and approves staff proposals for macroprudential measures; Board is the sole decision‑making body.
- For EU‑harmonized instruments BdP must seek ECB approval; CNSF must be consulted for all instruments.
- Current BdP hard powers cover EU‑harmonized instruments under the Legal Framework of Credit Institutions and Financial Companies (LFCIFC): the CCoB, G‑SII, O‑SII buffers; the sSyRB; the CCyB; and certain risk weights for RRE and CRE lending.
- Non‑harmonized instruments (e.g., BBMs) are implemented via a Macroprudential Recommendation functioning as a comply‑or‑explain mechanism; compliance is high but may weaken with rising risk appetite or new lenders.
- Recommendation (Table 1): "Grant BdP binding powers to implement macroprudential policy instruments beyond those which exist in EU regulation." Timing: ST; Agency: MoF.

### Coordination, cooperation, and accountability
- Internal BdP coordination bodies:
  - CESEF (Specialized Commission for Supervision and Financial Stability) supports Board; meets at least annually.
  - SCRIP (Sub‑commission for Risks and Prudential Policy) meets at least twice a year; coordinates risk assessment between micro and macroprudential policies.
- National Council of Financial Supervisors (CNSF):
  - Meets at least quarterly; issues non‑binding opinions; permanent members include BdP Governor (chair), BdP Board member for prudential supervision, CMVM Chairperson, ASF Chairperson.
  - Secretariat has rotating Secretary of CNSF and staff from ASF, BdP, and CMVM.
  - Maintains a dashboard tracking four key high‑level risk areas: (i) macroeconomic and geopolitical environment; (ii) financial markets; (iii) sovereign financing conditions; and (iv) financing conditions for households and companies.
- International coordination: BdP participates in ECB/ESRB/EBA/European Commission fora; reciprocity example—Banco de España, Lietuvos Bankas, and National Bank of Belgium reciprocated the Portuguese sSyRB.
- Communication channels and publications:
  - Financial Stability Report (FSR), quarterly banking sector updates, annual report on macroprudential measures.
  - Public engagement products include "BdP Podcast", "90 Seconds of Financial Stability", and "Economics in a Picture" (each receiving over 400 page views and social media engagements).
- Recommendations:
  - "Enhance communication on macroprudential policy by launching a centralized webpage on the BdP website, collecting existing outputs together." Timing: ST; Agency: BdP.
  - "Review the breadth of external publications on macroprudential policy to streamline communications for efficiency." Timing: MT; Agency: BdP.

### Data access, operational capacity, and staffing
- Financial Stability Department is responsible for systemic risk assessment and macroprudential policy implementation.
- Division sizes and staffing:
  - Risks and Analysis Division and Regulatory Policy division: around 20 people each.
  - Macroprudential Policy Division: just over ten employees.
  - Staff vacancies close to 10 percent; internal mobility low; staff turnover is a significant challenge.
- Key datasets accessible to BdP:
  - Central Credit Register (CCR): monthly reporting by banks of all credit contracts.
  - Securities Holdings Statistics (SHS).
  - Centralized Securities Database (CSDB).
  - Central Balance Sheet Harmonized Panel (CBHP).
  - In‑house Credit Assessment System (SIAC).
  - Household Finance and Consumption Survey (HFCS) Microdata.
- Identified data gaps and mitigation:
  - Foreign subsidiaries’ securities holdings: limited granular access; mitigation plans include expanding SHS access and protocol with Spanish authorities.
  - Cyber risk transmission: insufficient ICT dependency data for non‑bank sector; mitigation via DORA and ad‑hoc exercises.
  - NBFIs: lack of micro‑level supervisory data (AIFMD, UCITS, MMFR, EMIR, SFTR); mitigation using aggregated Statistics Department data and CNSF joint analysis; bilateral protocol with CMVM being formalized.
  - NFCs: no bank‑firm level data on corporate deposits; limited firm‑level Scope 1–3 emissions data.
  - Household datasets: restricted access to energy efficiency certificates (ADENE); limited natural disaster insurance data.
- Recommendation: "Clarify BdP powers to collect data for ongoing monitoring of systemic risks and grant more powers where needed." Timing: MT; Agency: MoF.

### Modelling, monitoring, and stress‑testing toolkit
- BdP uses a large suite of models: network analysis, quantile regressions, machine learning for house prices, DSGE models, solvency and liquidity stress tests.
- Composite systemic risk indicator is used quarterly to assess cycle position and as key input for the CCyB.
- Housing and household models:
  - House price‑at‑risk model using quantile regression with local projections.
  - FSR features: FSR December 2019 (price‑to‑income, price‑to‑rent, trend deviation, error‑correction, inverted demand, quantile regression); FSR May 2024 (Housing Search Index using Google Trends); FSR November 2024 (regional determinants); Cascão & Neugebauer (2025) machine‑learning assessment.
  - Household distress model uses HFCS microdata; performs multi‑year adverse vs baseline scenarios and sensitivity analyses (income, interest rates, real estate prices, liquid assets, inflation, unemployment).
- NFCs model: projects firm balance sheets, income statements, cash flows; applied to pandemic, 2022 interest‑rate shock, CRE vulnerability.
- Banking sector solvency stress test:
  - Top‑down solvency stress test with inputs from Economics & Research and bank‑specific information; uses a dedicated annual data template for banks’ funding and capital plans.
  - No modeled second‑round transmission from corporate sector to real economy; BdP intends to address this.
  - Plans to enhance net interest income (NII) modeling for mixed‑rate loans and diversification effects.
  - Results used internally; not publicly disclosed.
- Liquidity stress tests and gaps:
  - Models: LCR, NSFR‑based stress test, cash‑flow‑based stress test; ad‑hoc exercises (e.g., digital euro).
  - NSFR‑based Stress Test migration rates: 50% from 6m–1y → <6m; 35% from ≥1y → 6m–1y; 15% from ≥1y → <6m. Threshold/benchmark: 100% regulatory minimum.
  - Cash‑flow‑based Stress Test: 4 scenarios; 4 horizons (5 days, 4 weeks, 3 months, 12 months). Output: Stressed net funding/contractual gap and Counterbalancing capacity (CBC). Threshold: CBC after net funding gap > 0.
  - LCR‑based Stress Test: 6 scenarios; Threshold/benchmark: 100% regulatory minimum.
  - Current liquidity stresses lack interbank interactions and system‑wide features (common asset sales, fire‑sale behavior).
  - Recommendation: "Further develop liquidity stress tests for macroprudential purposes, enhancing the modeling of systemwide interlinkages." Timing: ST; Agency: BdP.

### Toolkit usage, capital buffers, and calibration
- BdP uses EU‑harmonized instruments and non‑harmonized instruments via comply‑or‑explain recommendations.
- Capital buffers and key figures:
  - Four capital buffers: capital conservation buffer (CCoB), O‑SII buffer, systemic systemic risk buffer (sSyRB), and countercyclical capital buffer (CCyB).
  - All buffers must be fully composed of Common Equity Tier 1 (CET1) capital and together form the combined buffer requirement (CBR).
  - At end‑2024, the CBR stood at 3.27 percent of risk‑weighted exposure amounts, equivalent to €5.9 billion.
- sSyRB details:
  - November 2023: BdP announced a 4 percent sSyRB on the risk‑weighted exposure amount of all retail exposures to individuals secured by RRE in Portugal; effective November 2024 for institutions using IRB models.
  - Market coverage: in 2023 Q2, four IRB banking groups’ market share for housing purchases was just over 60 percent; increased to 67 percent subsequently.
  - Overall impact limited because IRB models tend not to assign low risk weights and the Basel III output floor mitigates effect.
  - sSyRB calibrated to cover unexpected losses in RRE portfolios under a severely adverse scenario; expected to be released in a shock.
- CCyB details:
  - December 2024 decision: set CCyB rate at 0.75 percent for a neutral environment; effective January 2026.
  - Calibration based on a stress test using 2023 balance sheets and ECB BMPE December 2023 baseline and adverse scenarios; CCyB set in increments of 25 basis points.
  - Internal FSAP analysis: adjusting unemployment, interest rates, or real estate stress did not significantly affect credit loss outcomes for a given severity, indicating robustness of selected PCN rate.
  - Assessment: CCyB and sSyRB together represent about 1 percent of releasable capital.
  - Recommendation: BdP should regularly reassess appropriateness of the 0.75 percent PCN rate given balance sheet changes since 2023 and "evaluate releasable capital for the elevated environment, including determining the range of capital that is likely to be required and the strategy to build it." Timing: ST; Agency: BdP.
  - Banks maintain substantial voluntary capital buffers of around 6.5 percent; some banks anticipate significant reductions in these buffers per capital plans.

### Borrower‑based measures (BBMs), housing market, and policy implications
- BBMs in place (implemented as recommendations since Jul. 2018 and later):
  - Three LTV limits.
  - DSTI limit with three bank portfolio limits and detailed calculation adjustments.
  - Age‑based maturity limits (three bands).
  - Regular payment requirement.
  - Average maturity limit.
- 2025 H1 values and BBM specifics (from Table 8):
  - LTV Ratio: 2025 H1 value: 75% (average). Limits: Up to 90% for own and permanent residence; Up to 80% for other purposes; Up to 100% for immovable property held by institutions and for property financial leasing agreements. Implementation: Jul. 2018.
  - DSTI Ratio: 2025 H1 values: 94% of new loans (DSTI ratio ≤ 50%); 4% of new loans (50% < DSTI ratio ≤ 60%); 2% of new loans (DSTI ratio > 60%). Limits: Up to 50%, exceptions up to 60% for up to 10% of total credit amount; over 60% up to 5% of total credit amount. Implementation: Apr. 2020. Calculation adjustments: 0.5 p.p. for maturities ≤ 5 years; 1 p.p. for maturities between 5 and 10 years; 1.5 p.p. for maturities >10 years; income reduction for borrowers aged 70 and over upon expiry.
  - Maturity of Credit for House Purchase: 2025 H1 values: 32 years (average); 100% (borrowers aged ≤ 30); 94% (borrowers aged <30 and ≤ 35); 97% (borrowers aged > 35). Limits: 40 years for borrowers aged 30 or under; 37 years for borrowers aged over 30 and up to and including 35; 35 years for borrowers aged over 35; average maturity of new agreements up to and including 30 years. Implementation: Apr. 2022.
  - Maturity of Consumer Credit: 2025 H1 values: 6.7 years (personal credit average); 8.4 years (car credit average). Limits: 7 years for personal credit; 10 years for car credit and personal credit for education, healthcare and energy transition. Implementation: Apr. 2020.
  - Regular Payments Requirement: 2025 H1 value: 96% of new credit agreements have regular principal and interest payments.
- Observed housing and household trends:
  - House prices increased 18 percent in the year to end‑September 2025 and have climbed 169 percent since 2015 (EA average: 55 percent).
  - Growth in house and other household lending up 8.9 percent in 2025.
  - Household indebtedness increased to 56 percent of GDP by end‑September 2025, up from its low point at end‑2024.
  - About 60 percent of outstanding mortgage loans remain indexed to EURIBOR.
  - Banks’ loans for RRE account for nearly half of their loan portfolio.
  - BdP valuation models show house prices have been overvalued for several quarters.
  - Structural constraints (barriers to residential construction and building permits) make a short‑term price correction unlikely.
- Housing Guarantee Scheme (Box 1 summary):
  - Introduced in January 2025 (single contract recorded in December 2024); guarantees 15 percent of the house purchase price for buyers under 35.
  - First half of 2025: 39 percent of housing credit granted to borrowers under 35 (up from 30 percent in 2024).
  - First nine months of 2025: over 17,000 credit agreements valued at €3.4 billion under the scheme; these loans represented 23 percent of number of agreements and 26 percent of total credit amount for primary residence purchases.
  - Average loan amount under the scheme reached €205,000 by September 2025, higher than €184,000 for non‑guaranteed loans.
  - Recommendation: BdP should conduct a study to consider long‑term system‑wide impacts of government measures on financial stability; MoF participation could be useful.
- BBMs operational and design observations:
  - All BBMs are non‑binding "comply or explain" recommendations; BdP monitors compliance of 13 institutions via monthly reporting covering 91 percent of new lending and annual self‑assessments.
  - Operational difficulty in meeting quarterly average maturity limit of 30 years; BdP introduced age‑based maturity limits to assist compliance.
  - Exemptions exist for DSTI (wealth, guarantors, small loans below €8,000, etc.); some exemptions intended to meet social objectives.
  - Recommendations:
    - "Streamline and simplify the BBMs while continuing to uphold their objectives of enhancing lending standards and strengthening household and financial sector resilience." Timing: ST; Agency: BdP.
    - Replace excessive granularity and numerous exemptions with portfolio limits where appropriate; set clear benchmarks and continue monitoring.

### Financial sector structure, risks, and selected indicators
- Banking sector size and evolution:
  - Banking sector assets declined to 161 percent at end‑September 2025 from a peak of 293 percent of GDP at end‑2012.
  - Loans to customers declined to 90 percent of GDP at end‑September 2025 from 182 percent at end‑2012.
- NFCs:
  - Q3 2025 corporate profitability (EBITDA to total assets) was 9.2 percent year‑on‑year.
  - Corporate bond market capitalization approximately €128.5 billion, about €80 billion attributable to financial corporations.
- Sovereign debt holdings and composition:
  - By end‑September 2025, banks’ debt securities holdings increased to 29 percent of total assets (from 11 percent in 2008).
  - Share of government debt holdings allocated to the domestic sovereign decreased to 26 percent (from 73 percent in 2015).
  - Holdings of Portuguese government debt remain moderate at 5 percent of total assets.
- Nonbank financial sector (NBFIs):
  - As of September 2025, NBFIs accounted for about a third of financial sector assets.
  - Total NBFI assets approximately 80 percent of GDP (EA: 365 percent; end‑2010: 200 percent).
  - Captive financial institutions and money lenders: 33 percent of GDP; insurance companies: 20 percent of GDP; investment and pension funds combined: 19 percent of GDP.
- CRE:
  - CRE is not a major risk relative to many advanced economies; EA CRE prices decreased by 8.2 percent (after 1.5 percent growth in 2022); Portugal CRE prices remained broadly unchanged post‑pandemic.

### Emerging risks, innovation, and capacity building
- Climate change recognized as a systemic risk; BdP’s May 2025 FSR includes "A macroprudential approach to systemic climate‑related risk."
- Annual reporting mandated by the Climate Framework Law: "Annual Report on the Banking Sector's Exposure to Climate Risk."
- Climate analysis: BdP uses spatial modelling and adapted NGFS scenarios; expanded internal data to consider firms’ local units and real estate collateral locations.
- Cyber resilience testing (2024): assessed recovery from severe incidents, institution questionnaires on impact, response, second‑round effects, and coordination; systemic impact modules focused on common third‑party dependencies.
- BdP leverages international methodologies: ESRB framework for cyber exercises, ECB methodology for cyber tests, NGFS climate scenarios, private provider data for physical/transition exposures.
- Recommendation: prioritize resources for data, staff with technical skills, and forward‑looking models to monitor emerging systemic risks (climate, cyber, crypto).

### Key recommendations (selected and verbatim from Table 1 and body)
- "Grant BdP binding powers to implement macroprudential policy instruments beyond those which exist in EU regulation." Timing: ST; Agency: MoF.
- "Enhance communication on macroprudential policy by launching a centralized webpage on the BdP website, collecting existing outputs together." Timing: ST; Agency: BdP.
- "Review the breadth of external publications on macroprudential policy to streamline communications for efficiency." Timing: MT; Agency: BdP.
- "Clarify BdP powers to collect data for ongoing monitoring of systemic risks and grant more powers where needed." Timing: MT; Agency: MoF.
- "Further develop liquidity stress tests for macroprudential purposes, enhancing the modeling of systemwide interlinkages." Timing: ST; Agency: BdP.
- "Complete the strategy for the releasable capital in the elevated environment, including determining the range of capital that is likely to be required and the strategy to build it." Timing: ST; Agency: BdP.
- "Streamline and simplify the BBMs while continuing to uphold their objectives of enhancing lending standards and strengthening household and financial sector resilience." Timing: ST; Agency: BdP.
- Review the impact of government support measures for housing from a financial stability perspective. Timing: MT; Agency: BdP.

*Source: IMF staff technical note for the 2026 FSAP assessment of Portugal — Content unit: 1prtea2026005 (excerpts).*

### EXECUTIVE SUMMARY __________________________________________________________________________ 6

### EXECUTIVE SUMMARY

### Macroprudential toolkit and recent measures
- Banco de Portugal (BdP) has developed a full-fledged macroprudential toolkit.
- Activation, in January 2026, of a countercyclical capital buffer (CCyB) set at a positive cycle-neutral (PCN) target rate of 0.75 percent.
- Since 2023, toolkit includes a sectoral systemic risk buffer (sSyRB) targeting residential real estate (RRE) risks.
- Borrower-based measures, including LTV and DSTI limits, have been in place since 2018 as recommendations for banks.
- The CCyB at 0.75 percent, when combined with the sSyRB, "totals about 1 percent of releasable capital."

### Institutional framework and governance
- Portugal’s institutional framework for macroprudential policy is robust and largely aligned with best practices.
- Three agencies responsible for systemic risk oversight: BdP, Securities Market Commission (CMVM), and Insurance and Pension Funds Supervisory Authority (ASF).
- BdP is the designated macroprudential authority since 2013 and is also the microprudential supervisor under the SSM and national resolution authority under the Single Resolution Mechanism.
- The BdP Board meets weekly to make macroprudential decisions based on staff recommendations and regularly reviews policy.
- BdP has hard powers over macroprudential instruments harmonized in the EU (e.g., capital buffers and liquidity measures) but lacks powers to implement tools beyond the EU-harmonized toolkit, relying on recommendations and warnings for non-harmonized instruments.
- Compliance with non‑harmonized instruments is high in the current risk-off environment but may weaken if banks’ risk appetite rises or new lenders enter the market.
- Recommendation: Binding powers need to be granted by the legislator for BdP to enforce macroprudential instruments effectively. (Table 1 recommendation: "Grant BdP binding powers to implement macroprudential policy instruments beyond those which exist in EU regulation." Timing: ST; Agency: MoF.)

### Coordination and cooperation
- Macroprudential coordination is effective at all levels:
  - Internal BdP committee discusses macroprudential and related topics (microprudential issues, economics, markets, resolution, statistics, payments).
  - National Council of Financial Supervisors (CNSF) issues non-binding opinions on macroprudential actions.
  - Members maintain a shared dashboard tracking high-level risks (e.g., geopolitics) and discuss potential transmission and impact on the financial system.
  - BdP actively participates in multiple ECB/ESRB working groups and engages in early informal discussions to prevent delays in policymaking.

### Communication and accountability
- BdP communicates through multiple channels: Financial Stability Report (FSR), quarterly banking sector updates, and an annual report on implementation and effects of macroprudential measures.
- Accessible non-technical communications include a podcast, "90 Seconds of Financial Stability," and "Economics in a Picture," with strong engagement.
- Recommendation: Create a centralized "macroprudential hub" on the BdP website to collect existing outputs together. (Table 1 recommendation: "Enhance communication on macroprudential policy by launching a centralized webpage on the BdP website, collecting existing outputs together." Timing: ST; Agency: BdP.)
- Recommendation: Review the breadth of external publications to streamline communications for efficiency. (Table 1 recommendation: "Review the breadth of external publications on macroprudential policy to streamline communications for efficiency." Timing: MT; Agency: BdP.)

### Data, operational capacity, and modelling
- BdP has excellent data from banks and the credit registry, with minimal gaps (e.g., on foreign subsidiaries’ securities holdings).
- Although BdP has powers to collect data for financial stability purposes, in practice these have been used only for limited ad hoc analyses.
- Recommendation: Clarify BdP powers to collect data for ongoing monitoring of systemic risks and grant more powers where needed. (Table 1 recommendation: "Clarify BdP powers to collect data for ongoing monitoring of systemic risks and grant more powers where needed." Timing: MT; Agency: MoF.)
- BdP has developed an extensive suite of state‑of‑the‑art models supporting financial stability analysis, including models for emerging cyber, climate, and operational risks.
- Solvency stress‑testing model is used for regular system‑wide assessments and incorporates ad hoc scenarios.
- A liquidity stress‑testing framework has been recently developed; further enhancements are recommended to capture system‑wide dynamics (interbank interconnectedness and potential fire‑sale behavior).
- Recommendation: Further develop liquidity stress tests for macroprudential purposes, enhancing the modeling of systemwide interlinkages. (Table 1 recommendation: "Further develop liquidity stress tests for macroprudential purposes, enhancing the modeling of systemwide interlinkages." Timing: ST; Agency: BdP.)

### Toolkit usage and sectoral considerations
- BdP employs the EU‑harmonized toolkit and actively uses non‑harmonized instruments (e.g., borrower‑based measures) on a comply‑or‑explain basis.
- The non-bank sector is small and most subsectors are not growing; macroprudential instruments for non‑banks are under development at the EU level and will be applied by the relevant authority.
- After the 2020 pandemic, BdP shifted to building resilience earlier in the cycle.
- The CCyB rate of 0.75 percent is calibrated based on a 2023 adverse EBA scenario; FSAP sensitivity analysis shows credit losses remain relatively stable to variations in mild adverse scenarios, suggesting the current PCN rate may adequately guard against such shocks in the neutral cycle phase.
- The strategy for the elevated part of the financial cycle should be further developed:
  - The end-2024 policy statement on the PCN rate for the CCyB outlines calibration and indicators for aggregate risks but lacks guidance on capital build-up in the elevated-risk phase.
  - Studies suggest significantly more capital may be needed at cycle peaks, though this varies by country-specific factors.
  - Recommendation: Evaluate releasable capital for the elevated-risk phase and develop a strategy to ensure adequate resilience in time, particularly given ownership-related barriers to raising equity faced by many Portuguese banks. (Table 1 recommendation: "Complete the strategy for the releasable capital in the elevated environment, including determining the range of capital that is likely to be required and the strategy to build it." Timing: ST; Agency: BdP.)

### Borrower-based measures and housing risks
- BBMs have helped strengthen resilience to real estate risks, but rising house prices may weaken their effectiveness.
- BdP currently applies multiple BBMs:
  - Three LTV limits.
  - A DSTI limit (with three bank portfolio limits).
  - Three age-based maturity limits.
  - Regular payment requirements.
  - An average maturity limit.
- The BBM toolkit has become complex.
- Recommendation: Conduct a holistic review to streamline and simplify BBMs while maintaining objectives of enhancing lending standards and strengthening household and financial sector resilience. (Table 1 recommendation: "Streamline and simplify the BBMs be[?] while continuing to uphold their objectives of enhancing lending standards and strengthening household and financial sector resilience." Timing: ST; Agency: BdP.)

### Other recommendations captured in Table 1
- Review the impact of government support measures for housing from a financial stability perspective. (Timing: MT; Agency: BdP.)

*Source: IMF staff technical note for the 2026 FSAP assessment of Portugal.*

### 8. Decisions on policies are made by staff proposal and then approved by the BdP Board.

### 1prtea2026005 - 8. Decisions on policies are made by staff proposal and then approved by the BdP Board.

### Decision-making process and internal procedures
- BdP Board of Directors is the sole decision-making body and formally approves staff proposals for macroprudential measures.
- Board meeting frequency and planning:
  - Meets ordinarily at least once a week, and extraordinarily whenever convened by the Governor.
  - For existent and EU-harmonized capital-based macroprudential measures, a yearly work plan is established; the plan sets dates for submission to the Board and outlines timing of anticipated decisions, preceded by meetings with the Board member responsible for financial stability.
  - For other measures there is no preapproved plan; scheduling of discussions and approvals is the responsibility of the Board member overseeing the Financial Stability Department.
- For EU-harmonized instruments, BdP must also seek the approval of the ECB, and for all instruments, the CNSF must be consulted.

### Principle II — Ability to Act (powers and recommendations)
- Framework characterization:
  - Hard powers: provide direct control allowing prompt and decisive action.
  - Semi-hard powers: enable formal recommendations with a “comply or explain” mechanism.
  - Soft powers: limit the policymaker to advisory opinions.
- Current BdP powers:
  - BdP holds hard powers over EU-harmonized instruments under the Legal Framework of Credit Institutions and Financial Companies (LFCIFC).
  - Harmonized instruments listed: the CCoB, G-SII, and O-SII buffers; the sSyRB; the CCyB; and certain risk weights for RRE and commercial real estate (CRE) lending.
  - Other instruments, including borrower-based measures (BBMs), are left to national discretion; BdP lacks a defined binding regime for the non-harmonized instruments it has developed.
- Compliance and risks:
  - BdP’s Macroprudential Recommendation is a non-binding instrument functioning as a comply-or-explain mechanism; compliance has so far been high.
  - Banks face difficulties with meeting maturity limits (see section on BBMs).
  - As new institutions enter the market and competitive pressures increase, compliance may decrease, potentially undermining macroprudential effectiveness.
  - Non-harmonized instruments are likely to remain an important part of the framework, especially for new risks where international experience is limited (e.g., climate change related risks).
- Recommendation:
  - BdP should be given binding, hard powers over non-harmonized instruments, used as a backstop while preferably using recommendations first for flexibility.
  - Two legislative options to grant powers:
    - (i) Establish general powers in broad terms, giving BdP discretion to define design and conditions for implementation.
    - (ii) Grant powers in more specific terms.
  - Given learning needs for new risks and instruments, legislation should ideally grant powers in broad terms but explicitly include (and not be limited to) BBMs and sectoral/specific exposure requirements on banks to address systemic risks.

### Principle III — Coordination and cooperation
- Internal coordination within BdP:
  - Multiple formal and informal collaborations; internal commissions include:
    - Specialized Commission for Supervision and Financial Stability (CESEF):
      - A consultative commission within BdP composed of members of the Board (including the Governor and the Board member responsible for the Financial Stability Department), heads of several departments, and other relevant staff.
      - Aims to support the Board and promote broad discussion of financial stability issues; meets at least annually and additionally whenever a member requests a meeting.
    - Sub-commission for Risks and Prudential Policy (SCRIP):
      - Substructure of CESEF; coordinates risk assessment exercises to achieve consistency between microprudential and macroprudential policies.
      - Composition: Board member responsible for the Financial Stability Department and Heads of Financial Stability, Banking Prudential Supervision, Economics and Research, Markets, Banking Conduct Supervision, Statistics, and Resolution and Payments Systems Departments.
      - Meets at least twice a year.
  - Financial Stability Department collaborates closely with economics and research, statistics, and microprudential supervision departments; participates in Forum Monetário.
  - BdP is an integrated regulator with macroprudential, microprudential, resolution, and conduct mandates; coordination with the conduct department has been key to monitoring compliance (e.g., addressing product innovations like payment holidays or slow repayments).
- National coordination (CNSF):
  - National Council of Financial Supervisors (CNSF) advises BdP on macroprudential matters; set up in 2000, mandate amended in 2013 to include macroprudential topics.
  - CNSF meets at least quarterly; extraordinary meetings can be convened at any time.
  - In its advisory role, the Council:
    - Contributes to identification, monitoring, and assessment of risks to financial system stability.
    - Reviews specific macroprudential policy proposals submitted by BdP, issuing non-binding opinions.
  - CNSF membership and participation:
    - Permanent members: BdP Governor (chairs Council); the BdP Board member responsible for prudential supervision; Chairperson of the CMVM; Chairperson of the ASF.
    - A representative of the Minister of Finance (MoF) and the BdP Board member responsible for financial stability participate in macroprudential sessions as observers without voting rights.
  - Secretariat and groups:
    - Secretariat comprises the Secretary of the CNSF (two-year term, rotation principle) and staff from ASF, BdP, and CMVM.
    - Over 15 working and contact groups; the designated macroprudential contact group weighs on adoption of CNSF’s macroprudential opinions.
    - Working group on Risks and Financial Stability meets quarterly and produces risk reports covering banking, financial markets, and insurance and pension funds; maintains a dashboard tracking four key high-level risk areas: (i) macroeconomic and geopolitical environment; (ii) financial markets; (iii) sovereign financing conditions; and (iv) financing conditions for households and companies.
- International coordination:
  - BdP participates in numerous EU and international working groups and finds peer experiences helpful when considering macroprudential impacts.
  - Table 3 (summarized):
    - ECB (Top‑up Powers): may apply higher capital buffer requirements or stricter measures; national macroprudential authorities must notify the ECB before taking a final decision on applying macroprudential harmonized tools.
    - ESRB: responsible for macroprudential oversight in the EU; issues opinions, warnings and recommendations under a comply‑or‑explain mechanism; may issue recommendations on reciprocity to prevent cross‑border leakages.
    - EBA: assesses EU banking risks and vulnerabilities; promotes convergence; issues guidelines and recommendations under a comply‑or‑explain procedure.
    - European Commission: develops EU legal framework and proposes legislation; assesses notifications from Member States and issues opinions/authorizations before certain harmonized macroprudential measures are adopted.
  - Reciprocity usage example:
    - The voluntary reciprocity framework at EU level was used for the sSyRB; the ESRB recommended reciprocity due to Spanish banks’ significant role in the residential mortgage market.
    - Banco de España, Lietuvos Bankas, and the National Bank of Belgium decided to reciprocate the Portuguese measure.

### Communication and accountability
- Communication role:
  - Communication enhances effectiveness and accountability; in absence of hard powers, clear communication is an important “soft power.”
  - Accountability: BdP must demonstrate alignment with mandate by explaining decisions and publishing analysis and evidence.
- Disclosure and reporting:
  - BdP submits an annual report to parliament, including a dedicated section on macroprudential policy; the annual report is presented to parliament by the Governor in a specific hearing.
  - Macroprudential policy decisions are publicly disclosed on BdP’s website in Portuguese and English, and in the Official Bulletin; each measure implementation is accompanied by a press release. Regular press conferences and podcasts are used.
  - Publications:
    - Financial Stability Report (FSR) and a quarterly banking system update:
      - FSR is clearly written focusing on key risks and propagation channels.
      - Quarterly banking update is more data focused and has included household and corporate sector resilience; contains in-depth boxes (e.g., mortgage market).
      - FSR was typically published semiannually in past few years, but there was no FSR in H2 2025.
    - Annual report of the banking system’s exposures to climate change.
    - Annual report “Macroprudential measures in Portugal—progress report”: explains each measure’s purpose, rationale, market impact, and compliance; distinguishes structural or cyclical instruments and indicates whether BdP expects each capital element to be released during stress.
- Public engagement and digital outreach:
  - Communication products aimed at the general public: “BdP Podcast”; “90 seconds of financial stability” focused podcast; “Economics in a picture” charts (each receiving over 400 page views and social media engagements via BdP’s LinkedIn and Bluesky accounts).
  - Communications department notes these products are often picked up by broader media, increasing FSR impact.
- Recommendations for communication:
  - Streamline communications to ensure efficiency given limited resources; review and streamline macroprudential publication suite to maintain impactful, distinct outputs without overextending staff.
  - Launch a “Macroprudential hub” on BdP’s website to centralize and enhance access to analyses that are currently scattered across FSR, economics bulletin, quarterly update, and working papers; the hub should complement existing publications and consolidate mostly timeless analysis useful to financial stakeholders and international authorities.

### Operational capacity and resources
- Effective systemic risk assessment and modeling require adequate data, staff with appropriate skills, and forward-looking models for the financial system.
- Resource needs highlighted: both data and people with the right skill set, and forward-looking models for the financial system.

*Source: 1prtea2026005 - 8. Decisions on policies are made by staff proposal and then approved by the BdP Board.*

### 35. BdP’s Financial Stability Department is responsible for carrying out BdP's financial

### 35. BdP’s Financial Stability Department is responsible for carrying out BdP's financial

### Organizational structure and staffing
- The Financial Stability Department is responsible for assessing systemic risks and implementing macroprudential policy.
- Divisions are small: the largest two—Risks and Analysis Division and Regulatory Policy division—comprising around 20 people each.
- The Macroprudential Policy Division has just over ten employees.
- Internal mobility within BdP is low; many staff in the financial stability unit have long tenures and deep expertise.
- Staff numbers are below headcount, with vacancies close to 10 percent.
- Staff turnover remains a significant challenge, driven by strong market demand for specialized technical skills; staff often leave for the private sector or other central banks and institutions in Europe.
- Vacant positions are difficult to fill due to scarce expertise in the labor market; emerging systemic risks (climate change, cyber) have exacerbated recruitment difficulties.

### Data access and key datasets
- BdP has good access to diverse datasets. Banks’ reporting generally follows EU standards, except for a data form on funding and capital plans implemented just before the Economic and Financial Adjustment Program, which both banks and BdP find beneficial.
- Collaboration with other authorities (e.g., tax authority for firms’ annual financial statements and monthly sales data) is valuable.
- Most firms’ data sources are mergeable; confidential information is anonymized when needed.
- Key data sources accessible to BdP (Table 4):
  - Central Credit Register (CCR): Managed by Banco de Portugal; monthly reporting by banks of all credit contracts. Information: Credit exposures, credit type (households & NFCs), loan amounts, detailed credit characteristics.
  - Securities Holdings Statistics (SHS): Information on securities held by resident sectors and banking groups. Information: Securities holdings at individual and group level.
  - Centralized Securities Database (CSDB): Data on securities issued in Portugal and worldwide. Information: Instruments, issuers, prices, ratings for debt, equity, investment fund shares.
  - Central Balance Sheet Harmonized Panel (CBHP): Dataset on economic & financial characteristics of Portuguese NFCs. Information: Firm-level balance sheet and financial information.
  - In-house Credit Assessment System (SIAC): Developed by Banco de Portugal for ECB monetary policy purposes. Information: Firm-level probability of default (PDs).
  - Household Finance and Consumption Survey (HFCS) Microdata: Survey data on Portuguese households. Information: Household assets, liabilities, income, consumption, savings, employment status.

### Macroprudential data gaps and mitigation measures
- Data gaps exist and may become material if emerging risks, NBFI and foreign subsidiaries’ role grow; some issues are operational or transitional as institutions adapt to new reporting templates.
- Summary of identified data gaps and implications (Table 5):
  - Foreign Subsidiaries’ Securities Holdings: Limited access to granular data on foreign subsidiaries; reliant on ECB SHS data with restricted access and delays. Implication: Hinders forward-looking analysis (e.g., interest rate risk on securities portfolios). Mitigation/plans: Increase number of staff with SHS access; establish protocol with Spanish authorities for timely data.
  - Cyber Risk Transmission Analysis: Insufficient data on ICT dependencies from the non-bank sector; ad-hoc data collection only. Implication: Limits ability to assess systemic cyber risk transmission channels. Mitigation/plans: Expect more harmonized data from non-bank financial sectors through DORA; continued exercises via Resolution Department.
  - Non‑Bank Financial Institutions (NBFIs): Lack of micro‑level supervisory data (AIFMD, UCITS, MMFR, EMIR, SFTR); no direct access to granular insurer/pension fund data. Implication: Weak visibility on leverage, exposures, solvency, interconnectedness. Mitigation/plans: Use aggregated Statistics Department data; participate in CNSF joint analysis.
  - Non‑Financial Corporations (NFCs): No bank‑firm level data on corporate deposits; lack of micro‑level climate data (Scope 1–3 emissions). Implication: Limits analysis of liquidity risks and precise climate risk assessments. Mitigation/plans: Regulatory developments expected; current reliance on sector‑level emission averages.
  - Household Sector (Income, Wealth, Climate & Risk Data): Restricted access to energy efficiency certificates (ADENE); limited natural disaster insurance data; inability to link credit data with timely income/wealth data. Implication: Reduces ability to assess vulnerability at borrower level and understand distributional risks. Mitigation/plans: Desire for more integrated borrower‑level datasets; reliance on surveys (HFCS).
  - Systemic Risk Identification (Financial Innovation & Digitalization): Data lags and gaps in fast‑changing areas (digitalization, AI, market innovations). Implication: Slower identification of emerging systemic risks. Mitigation/plans: Banco de Portugal expanding analytical tools, micro‑data use, and coordination with national/EU authorities.
- Note: the table is from BdP’s perspective; other institutions in Portugal may have access but be unable to fully share the data.

### Specific data-access limitations and planned remedies
- Securities holdings for foreign subsidiaries must be obtained from ECB’s SHS database; access is restricted to a small number of staff and subject to delays. Plans: (i) expand number of staff with direct SHS access; and/or (ii) establish a protocol with the Spanish authorities to obtain timely data (most significant foreign subsidiaries in Portugal have parent companies in Spain).
- Ad-hoc cyber exercise by BdP’s Resolution Department required banks to submit detailed information on internal and external relationships with ICT providers.
- For non-banks, BdP lacks direct access to data collected under AIFMD, UCITS, EMIR, SFTR, and granular insurer/pension fund data; BdP relies on aggregated Statistics Department data and CNSF joint analysis. A bilateral protocol between BdP and CMVM is being formalized for regular information exchange on UCITS and AIFMD; information related to EMIR and SFTR may be exchanged under the same protocol.

### Data collection powers and recommendations
- BdP’s statute includes a general provision granting powers to require any public or private entity to provide information deemed necessary; specific provisions exist for collecting data from the credit registry.
- To date, BdP has used these powers mainly for limited, ad hoc analyses.
- Identified needs:
  - Clarify and strengthen data collection powers for macroprudential purposes, particularly for regular reporting.
  - Ongoing monitoring is crucial for new and emerging risks (climate change, cyber, crypto), where data are not standardized.
  - Specific data gaps: missing firm-level emissions data; restricted access to household energy efficiency certificates managed by Agência Nacional de Energia (ADENE) due to data protection concerns; limited granular data on natural disaster insurance coverage at dwelling or company level.
- Recommendation: BdP should be given clarity on powers to request ongoing information or data for macroprudential purposes and granted more powers where needed.

### Systemic risk monitoring: models and indicators (overview)
- BdP employs a large suite of analytical models across multiple fields and techniques, despite small team size.
- Toolkit includes network analysis, quantile regressions for “at risk” metrics for key indicators, a machine learning model for house price analysis, and DSGE models for policy analysis currently being developed.
- These models are deeply integrated into policy analysis and often presented in Financial Stability Reports (FSRs).
- The composite systemic risk indicator is used quarterly to assess cycle position and serves as a key input for the CCyB.
- Three house price valuation models have been featured in FSRs.

### A. Housing and households: models and analyses
- BdP has several models for assessing risks from housing markets and real estate overvaluation (Table 6).
- A house price-at-risk model using quantile regression combined with local projections was developed to model downside risks to house price growth.
- Selection of BdP real estate market models and publications (Table 6 highlights):
  - FSR December 2019—Special Feature: price‑to‑income, price‑to‑rent, trend deviation; error‑correction model; inverted demand equation; quantile regression. Provides multi‑method evaluation of house‑price misalignment; results updated regularly in FSRs.
  - FSR May 2024—Box 1: Housing Search Index using Google Trends (Google Trends search‑intensity modeling). Search behavior may act as a leading short‑term indicator of house‑price changes.
  - FSR November 2024—Box 2: Determinants of regional house‑price dynamics (municipality‑level econometrics). Confirms drivers: unemployment, interest rates, population dynamics, construction activity.
  - Cascão & Neugebauer (2025): Machine‑learning assessment of price determinants (Gradient tree boosting; Quantile regression). ML methods outperform alternatives; prediction intervals enhance uncertainty assessment.
- Household distress model:
  - Uses HFCS microdata to estimate household distress from various shocks, assessing aggregate and income‑quintile outcomes.
  - Considers a multi‑year adverse scenario vs baseline; performs sensitivity analyses to shocks: income, interest rates, real estate prices, liquid assets, inflation, unemployment.
  - Also used to assess banks’ exposure to vulnerable households and potential losses.

### B. Non‑financial corporates (NFCs)
- BdP has a model for evaluating NFCs’ financial soundness in the short to medium term using microdata.
- The model projects firms’ balance sheets, income statements, and cash flow statements; assesses impact of macro scenarios with variation by firm size and sector.
- Applied to scenarios: pandemic impact, 2022 interest rate hike, vulnerability of CRE firms; results published in BdP’s FSRs.

### C. Banking sector: solvency and liquidity stress testing
- Solvency stress test:
  - BdP uses a top‑down solvency stress test developed by the Financial Stability Department in collaboration with other departments.
  - Inputs: baseline scenario from Economics and Research Department quarterly projection; adverse scenarios from the same department; bank‑specific information from Banking Prudential Supervision Department.
  - Banks’ funding and capital plans are key inputs; a dedicated data template implemented just before the Economic and Financial Adjustment Program is produced once a year and is considered useful by banks.
  - Main transmission mechanisms: macroeconomy → firm/household PDs and loan/deposit interest rates; interest rates affect credit volumes and non‑interest income.
  - No modeled second‑round transmission from corporate sector to real economy via reduced bank lending; BdP intends to address this and prioritize it given large share of NFC loans in banks’ portfolios.
  - Plans to enhance modeling of net interest income (NII) to reflect diversification and “mixed” rate loans (rates fixed up to five years then variable).
  - Model is regularly used for BdP analysis and internal discussion; results are not publicly disclosed. Given EBA public stress tests for largest banks, publishing BdP model results is of limited marginal benefit and may be counterproductive.
- Liquidity stress tests:
  - Several liquidity models assess resilience: LCR, NFSR, and cash‑flow models covering nearly the entire banking sector; ad‑hoc sensitivity analyses on topics like the digital euro are also conducted.
  - Overview of liquidity models (Table 7 highlights):
    - NSFR‑based Stress Test: sources NSFR templates (C 81.00 & C 84.00). Migration rates: • 50% from 6m–1y → <6m. • 35% from ≥1y → 6m–1y. • 15% from ≥1y → <6m. Output: Stressed NSFR. Threshold/benchmark: 100% regulatory minimum.
    - Cash‑flow‑based Stress Test: sources ALMM maturity ladder templates (COREP 66.00). Scenarios/parameters: • 4 scenarios (roll‑off rates & haircuts). • 4 horizons: 5 days, 4 weeks, 3 months, 12 months. Outputs: Stressed net funding/contractual gap (cumulative inflows—outflows); Counterbalancing capacity (CBC—cumulative liquid assets). Threshold/benchmark: CBC after net funding gap > 0.
    - LCR‑based Stress Test: sources COREP LCR templates (C 72.00 / C 73.00 / C 74.00 / C 76.00). Scenarios: 6 scenarios (roll‑off rates & haircuts). Output: Stressed LCR per scenario. Threshold/benchmark: 100% regulatory minimum.
  - Current liquidity stresses apply simultaneously to all banks but lack interbank interactions and other features to enhance macroprudential perspective.
  - Recommendation: Prioritize development of interbank interactions and scenario analyses that capture feedback loops and interactions, focusing on large short‑term liquidity and funding shocks.
  - BdP should enhance toolkit for modeling short‑term liquidity stresses at system‑wide level, embedding common asset sales and interbank linkages.

### Emerging risks and analytical innovation
- New risk types require new techniques and broader institutional cooperation due to non‑traditional, complex systemic potential.
- Despite limited resourcing, BdP has produced analyses using new techniques:
  - Climate change risk analysis in BdP’s Annual Report on the Banking Sector to Climate Risk uses spatial modeling techniques to assess how climate change is likely to affect the financial system.
  - Systemic cyber risk analysis has involved complex network mapping and new models/approaches to cover operational risks.

_International Monetary Fund — Content unit: 1prtea2026005 (excerpts)_

### 58. Collaboration within BdP and with external stakeholders has been essential for

### 58. Collaboration within BdP and with external stakeholders has been essential for enabling this analysis

### Collaboration, capacity building, and data use
- BdP staff rapidly developed expertise through close collaboration across BdP departments and with external partners.
- Examples of frameworks, methodologies, and partnerships used:
  - Applied the ESRB framework in the cyber resilience exercise to address systemic risk aspects.
  - Drew on the ECB’s methodology for the cyber risk test.
  - Worked with the supervision team on operational arrangements and information on key service providers.
  - Adapted NGFS climate scenarios for climate risk analysis.
  - Leveraged data from established private providers on physical and transition risk exposures.
- Data gap status and improvements:
  - Some data gaps remain, but these are gradually narrowing.
  - For climate change-related physical risks, BdP expanded internal data sources and now considers firms’ local units and real estate collateral locations in its analysis.

### Climate risk recognition and reporting
- Climate change is recognized as a systemic risk.
- BdP strategy and reporting:
  - The BdP’s May 2025 FSR includes a special feature titled “A macroprudential approach to systemic climate-related risk,” outlining the high-level strategy for considering climate change-related systemic risks.
  - Key risks are published annually in the “Annual Report on the Banking Sector's Exposure to Climate Risk,” which also covers regulatory and supervisory initiatives.
  - This reporting is mandated by the Climate Framework Law.

### Cyber resilience testing
- 2024 operational cyber resilience test objectives and design:
  - Evaluated how institutions recover from a successful cyber incident.
  - Assessed institutions’ vulnerability and capacity to respond to a severe incident.
  - Each institution completed a questionnaire covering the incident’s impact, response and recovery measures, second-round effects, and coordination with other institutions and authorities.
  - Questionnaire modules addressing systemic impact focused on common reliance on third-party service providers, sector-wide coordination, and economic impact assuming multiple institutions were affected.

### A. Context and key risks for the Portuguese financial sector
- Banking sector size and evolution:
  - Banking sector assets declined to 161 percent at end-September 2025 from a peak of 293 percent of GDP at end-2012.
  - Loans to customers declined to 90 percent of GDP at end-September 2025 from 182 percent at end-2012.
- Drivers of balance-sheet changes:
  - Concerted effort by banks to address problem assets, bolster capitalization, and shift toward more stable funding sources to enable balance-sheet clean-up after the European debt crisis.

### NFCs (Non-Financial Corporations)
- Financial position and market structure:
  - Q3 2025 corporate profitability (EBITDA to total assets) was 9.2 percent year-on-year.
  - Corporate bond market capitalization approximately €128.5 billion, with most (about €80 billion) attributable to financial corporations.
  - Corporate debt significantly lower than pre-crisis levels and below the EA average since 2021.
  - Corporates hold substantial cash balances; NPLs remain low and did not worsen during the pandemic.

### Commercial Real Estate (CRE)
- CRE risk assessment:
  - CRE is not a major risk for Portugal relative to many advanced economies.
  - EA CRE prices decreased by 8.2 percent (after 1.5 percent growth in 2022); prices in Portugal remained broadly unchanged post-pandemic.

### Housing and households
- Deleveraging and indebtedness:
  - Household indebtedness as a share of disposable income has remained below the euro-area average since 2021.
  - Nominal household credit bottomed out in 2017 and has only recently started rising.
- Recent credit growth and exposures:
  - Growth in house and other household lending up 8.9 percent in 2025.
  - Household indebtedness increased to 56 percent of GDP by end-September 2025, up from its low point at end-2024.
  - About 60 percent of outstanding mortgage loans remain indexed to EURIBOR.
- House price dynamics:
  - House prices increased 18 percent in the year to end-September 2025 and have climbed 169 percent since 2015, compared to an EA average of 55 percent.
  - About half of house purchases are made without credit.
  - Banks’ loans for RRE account for nearly half of their loan portfolio.
  - BdP’s valuation models show house prices have been overvalued for several quarters.
  - Structural constraints—especially barriers to residential construction and building permits—make a short-term price correction unlikely.
- Policy guidance:
  - BdP should continue to closely monitor household indebtedness and take appropriate actions.
  - BBMs are unlikely to significantly cool the housing market given the limited role of credit in total house purchase transactions, but they can strengthen household resilience and help reduce procyclicality.

### Box 1 — Housing Guarantee Scheme (summary)
- Scheme design and uptake:
  - Introduced in January 2025 (single contract recorded in December 2024); guarantees 15 percent of the house purchase price for buyers under 35.
  - In the first half of 2025, 39 percent of housing credit was granted to borrowers under 35, up from 30 percent in 2024.
  - In the first nine months of 2025, over 17,000 credit agreements valued at €3.4 billion were made under the scheme.
  - These loans represented 23 percent of the total number of agreements and 26 percent of the total credit amount for primary residence purchases.
  - Average loan amount under the scheme reached €205,000 by September 2025, higher than €184,000 for non-guaranteed loans.
- Distributional and systemic considerations:
  - Guarantees most notable in Lisbon and Porto.
  - LTV ratios have increased and loan maturities have lengthened; DSTI are higher overall but stressed DSTI are lower.
  - Direct risks are fiscal; indirect risks to financial stability exist if household leverage increases.
  - Current statistics suggest guarantees are being taken more by people in the third and fourth quintile of the income distribution.
- Recommendation:
  - BdP should conduct a study to consider the long-term system-wide impacts of the government measures on financial stability. The MoF could also usefully participate in the study.

### Sovereign debt
- Shift in sovereign exposure composition:
  - By end-September 2025, banks’ debt securities holdings increased to 29 percent of total assets, up from 11 percent in 2008.
  - Share of government debt holdings allocated to the domestic sovereign decreased to 26 percent, down from 73 percent in 2015.
  - Holdings of Portuguese government debt remain moderate (5 percent of total assets).

### Nonbank financial sector (NBFIs)
- Size and composition:
  - As of September 2025, NBFIs accounted for about a third of the financial sector’s assets.
  - Total NBFI assets approximately 80 percent of GDP, compared to 365 percent in the EA and 200 percent at end-2010.
  - Portugal ranks 14th out of 20 EA NBFI sectors.
  - Captive financial institutions and money lenders dominate with assets at 33 percent of GDP.
  - Insurance companies assets at 20 percent of GDP.
  - Investment funds and pension funds combined assets at 19 percent of GDP.

### Box 2 — Actions during the COVID-19 pandemic (summary)
- Temporary relaxations in 2020:
  - Banks allowed to operate below Pillar 2 Guidance requirements, the Combined Buffer Requirement (CBR), and the LCR.
  - Implementation of the O-SII buffer was postponed by a year.
  - Banks received state-guaranteed loans and applied moratoria to avoid credit disruptions.
- Evidence on effectiveness:
  - Avezum et al. (2023) found banks with greater headroom above overall capital requirement extended more credit after the COVID-19 shock.
- Other measures:
  - Short-term personal credit was excluded from the DSTI limit; take-up was limited and measure discontinued.

### B. Sector-specific measures: Housing and households
- Borrower-based measures (summary of limits and 2025 H1 values from Table 8):
  - LTV Ratio:
    - Limits: Up to 90% for own and permanent residence; Up to 80% for other purposes; Up to 100% for immovable property held by institutions and for property financial leasing agreements.
    - Implementation: Jul. 2018
    - 2025 H1 value: 75% (average)
  - DSTI Ratio:
    - Definition includes interest rate rise and/or income reduction shock with adjustments: 0.5 p.p. for maturities ≤ 5 years; 1 p.p. for maturities between 5 and 10 years; 1.5 p.p. for maturities >10 years; income reduction for borrowers aged 70 and over upon expiry.
    - Limits: Up to 50%, with exceptions up to 60% for up to 10% of the total credit amount; Over 60% up to 5% of the total credit amount.
    - Implementation: Apr. 2020
    - 2025 H1 values: 94% of new loans (DSTI ratio ≤ 50%); 4% of new loans (50% < DSTI ratio ≤ 60%); 2% of new loans (DSTI ratio > 60%)
  - Maturity of Credit for House Purchase:
    - Limits by age and average maturity: 40 years for borrowers aged 30 or under; 37 years for borrowers aged over 30 and up to and including 35; 35 years for borrowers aged over 35; Average maturity of new agreements up to and including 30 years.
    - Implementation: Apr. 2022
    - 2025 H1 values: 32 years (average); 100% (borrowers aged ≤ 30); 94% (borrowers aged <30 and ≤ 35); 97% (borrowers aged > 35)
  - Maturity of Consumer Credit:
    - Limits: 7 years for personal credit; 10 years for car credit and personal credit for education, healthcare and energy transition.
    - Implementation: Apr. 2020
    - 2025 H1 values: 6.7 years (personal credit average); 8.4 years (car credit average)
  - Regular Payments Requirement:
    - New credit agreements should have regular principal and interest payments.
    - 2025 H1 value: 96%
- Notes on DSTI calculation and renewable energy credit:
  - DSTI calculation assumptions include constant instalments, impact of interest rate rise according to maturity for variable or mixed rate agreements, and income reduction at age 70 unless retired at credit assessment.
  - In April 2025, the concept of renewable energy credit was extended to “energy transition” credit; eligible uses include purchase and installation of renewable energy equipment, or other equipment or operations, including works to improve buildings’ energy efficiency.

### Effectiveness and bank behavior
- BBMs appear effective in containing household leverage and shaping lending standards.
- Observed outcomes:
  - Number of households with both high LTV and high DSTI ratios remains low.
  - Banks report DSTI levels well below BdP limits.
  - Measures helped establish expectations on lending standards, reducing pressure to compete through looser lending standards.

### Consumer credit and climate considerations
- Consumer credit composition and limits:
  - Consumer credit primarily consists of car purchase loans (70 percent of the total).
  - Little evidence households use consumer credit to circumvent LTV limits.
  - Consumer credit agreements above €8,000 are included in the DSTI limit calculation.
- Green lending maturity extension:
  - In 2020, maximum maturity for loans financing renewable energy extended from 7 years to 10 years.
  - Since April 2025, “renewable energy” credit broadened to “energy transition” purposes.

*International Monetary Fund — 1prtea2026005*

### 73. All the BBMs are implemented as “comply or explain” recommendations rather than

### All the BBMs are implemented as “comply or explain” recommendations rather than mandatory requirements.

### Implementation and compliance
- All borrower-based measures (BBMs) are non-binding “comply or explain” recommendations, yet compliance is high.
- BdP monitors compliance of 13 institutions through:
  - monthly reporting covering 91 percent of new lending operations; and
  - annual self-assessment reports approved by institutions’ Boards of Directors.
- For other institutions, BdP relies on individual data from the Central Credit Register (CRC) to monitor compliance with limits.
- The limits apply to all entities authorized to make loans in Portugal, minimizing opportunities for regulatory arbitrage.

### Operational challenges with BBMs
- Banks report operational difficulties meeting the recommendation that average loan maturities remain at 30 years (or less) each quarter due to:
  - operational difficulties in managing maturities on a quarterly basis given various time lags in the mortgage process; and
  - recent policies aimed at increasing lending to younger borrowers, which tend to involve longer maturities and shift the borrower age profile.
- BdP introduced maturity limits based on borrower age to help banks comply more effectively.
- Recent shifts to younger borrowers, driven by government initiatives, have made compliance more difficult; some banks attempt to increase shorter maturity “senior” loans to older borrowers to balance the average loan maturity at 30 years.

### DSTI, LTV, maturity limits and exemptions
- Exemptions exist to make the DSTI limit less restrictive:
  - Banks may consider additional factors in assessing borrowers’ creditworthiness, such as overall wealth, the presence of a guarantor, and other regular expenses.
  - Certain credit agreements are excluded from the limit, such as short-term loans below €8000; these smaller loans are included in the DSTI calculation if the borrower subsequently takes out a larger loan.
- Some exemptions are intended to address social objectives despite banks already having discretion to apply similar measures (e.g., discretion to hold a certain share of loans exceeding the 50 percent limit).
- Mortgage repayment by retirement is not required in some cases, partly due to the existence of other credit agreements without age limits used by older borrowers for health-related expenses; insurance would be a more appropriate product in such cases.

### Effectiveness, trade-offs, and design guidance
- Multiple measures serve complementary roles; key functions include:
  - LTV limits: reduce lender losses in default and incentivize prudent borrowing by requiring down payments; particularly important given the high level of recourse of mortgages in Portugal.
  - DSTI limits: contain household leverage and indebtedness and limit the aggregate demand channel.
  - Maturity limits: prevent circumvention of DSTI limits via long-term loans.
- Most EU macroprudential frameworks include either LTV, DSTI, and maturity limits, or combinations of LTV and DTI limits. Granular age-related limits, regular repayment requirements, and average maturity caps are uncommon.
- Recommendations:
  - Streamline and simplify BBMs while upholding objectives of enhancing lending standards and strengthening household and financial sector resilience.
  - Replace excessive granularity and numerous exemptions with portfolio limits where appropriate.
  - Establish clear expectations and communication; set clear benchmarks and continue ongoing monitoring of compliance and risk evolution over time.
- Example of redundancy: the 100 percent LTV limit was introduced after the European debt crisis to reduce banks’ on-balance-sheet exposures, but after nearly a decade such loans are likely scarce.

### Capital measures: overview and key figures
- Four capital buffers apply to the Portuguese banking sector: the capital conservation buffer (CCoB), the other systemically important institutions (O-SII) buffer, the systemic systemic risk buffer (sSyRB), and the countercyclical capital buffer (CCyB).
- All buffers must be fully composed of Common Equity Tier 1 (CET1) capital and together form the combined buffer requirement (CBR).
- At end-2024, the CBR stood at 3.27 percent of the risk-weighted exposure amounts, equivalent to €5.9 billion.
- The CCoB and O-SII buffers are structural; the sSyRB and CCyB are cyclical and intended to be released during periods of stress.
- Capital requirements in Portugal are broadly aligned with the EU average.

### O-SII framework, methodology changes, and scope for differentiation
- O-SII buffer: harmonized EU capital instrument to address structural too-big-to-fail risks; introduced in 2015 and phased in over time.
- BdP assesses systemic risk annually based on size, importance for the Portuguese economy or the EU, complexity, interconnectedness, and potential contagion.
- Methodology updates adopted in April 2025 reflect CRDV changes, including raising the maximum O-SII buffer limit from 2 to 3 percent and incorporate the ECB application of a floor methodology lowering the floor to maintain a stable group of O-SII banks.
- The increasing share of branches of EU banks—excluded from the O-SII calculation since the buffer is institution-specific—has contributed to a decline in Portuguese institutions’ scores: branches accounted for 4.1 percent of the composite systemic importance indicator in 2016 and 9.3 percent in 2023.
- Portugal’s mapping of O-SII scores to buffer requirements yields limited variation in capital requirements despite heterogeneity in O-SII scores across banks; there is scope to differentiate more and impose higher O-SII capital requirements on more systemic institutions.

### Countercyclical Capital Buffer (CCyB)
- International context: growing recognition of building releasable capital earlier; BCBS endorses maintaining a “positive cycle-neutral rate” for the CCyB; several countries including 17 in the EU have moved toward this approach.
- In December 2024, authorities set the CCyB rate at 0.75 percent for a neutral environment, effective from January 2026.
- Calibration:
  - The neutral CCyB rate of 0.75 percent was calibrated using a stress test based on 2023 balance sheets and the December 2023 ECB BMPE baseline and adverse macrofinancial scenarios.
  - The chosen rate is slightly higher than the projected loss rate in the scenario; CCyB can only be set in increments of 25 basis points.
  - Internal FSAP analysis found that adjusting dimensions such as unemployment, interest rates, or real estate stress did not significantly affect credit loss outcomes for a given severity, suggesting robustness of the selected PCN rate.
- Assessment and recommendations:
  - The banking system is currently resilient considering Portugal’s position in the credit cycle, but BdP should regularly reassess the appropriateness of the 0.75 percent rate as the neutral CCyB given changes in balance sheets since 2023.
  - Both the CCyB and the sSyRB are fully releasable buffers and together represent about 1 percent of releasable capital.
  - Although the PCN CCyB rate is low compared to EA peers, when combined with the sSyRB, total releasable capital at the neutral point aligns with the EA average.
  - If authorities wished to build additional releasable capital now, costs and disruptions would likely be minimal given current bank profitability, but delaying buildup could lead to higher costs later.
  - Banks maintain substantial voluntary capital buffers of around 6.5 percent; some banks have capital plans anticipating significant reductions in these buffers.
  - BdP should evaluate the amount of releasable capital required for an elevated risk environment and assess whether banks can realistically raise required capital within necessary timeframes; publishing this capital range might be beneficial.

### sSyRB (sectoral systemic risk buffer) specific design and impact
- In November 2023, BdP announced a four percent sSyRB on the risk-weighted exposure amount of all retail exposures to individuals secured by RRE in Portugal; the buffer took effect in November 2024 and applies to institutions using internal rating models (IRB) for credit risk.
- Market coverage: in 2023 Q2, the four IRB banking groups’ market share for housing purchases was just over 60 percent and has since increased to 67 percent.
- The overall impact on banks’ capital requirements is limited because IRB models tend not to assign low risk weights to these exposures and the Basel III output floor further mitigates the effect.
- The sSyRB rate was calibrated to cover unexpected losses in RRE portfolios under a severely adverse scenario, assuming banks will not draw on the CCoB or O-SII buffer to absorb losses; authorities expect the sSyRB to remain in place during normal times but to be released in the event of a shock.

### Monitoring policy effectiveness and modeling toolkit
- BdP employs various models to assess macroprudential policy impacts prior to and after implementation; the DSGE toolkit was used in the 2024 CCyB decision.
- Selected BdP research findings (methodology and key findings):
  - Basto et al. (2019): Structural multi‑country model; BVAR; EAGLE‑FLI. Findings: LTV curbs excessive credit growth; lower mortgage supply & GDP; higher bank solvency; slight reduction in house-price growth.
  - Abreu & Passinhas (2021): BVAR. Findings: Household loan growth reduced after 4 months; no short‑term effect on house prices or economic activity.
  - Abreu et al. (2024): Credit registry data; Diff in Diff. Findings: Lower leverage & LTIs; higher spreads & service‑to‑income; purchase of cheaper houses; improved household risk profile.
  - Neugebauer et al. (2021): Integrated micro‑macro model. Findings: LTV limits reduce PDs & LGDs; capital ratio rises by 0.74 p.p.
  - Mateus & Ramos (2023): LGD modeling. Findings: LTV strongly predicts LGD; non‑linear effect: higher LGD only above ~80–100% LTV.
- BdP has developed Portugal-specific DSGE models jointly with the University of Surrey to examine:
  - Interaction of borrower-based measures (LTV, DSTI, average loan maturities) mirroring BBMs in Portugal.
  - Effectiveness of macroprudential policies in a small economy within a monetary union facing inflationary pressures and transmission of non-financial shocks.
  - Interplay between BBMs and capital instruments under trade shocks.
- BdP publishes an annual report “Macroprudential measures in Portugal—progress report” showing measures in place and banks’ and borrowers’ responses; this report shows a high level of banking system‑level compliance.
- BdP’s ongoing, systematic ex-post reporting and research represent strong accountability and best practice in monitoring macroprudential policy effectiveness.

*Source: Excerpt from IMF chapter on Portugal macroprudential and capital measures.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2026/english/1prtea2026005.pdf_