## EXECUTIVE SUMMARY — 1cheea2025007-source-pdf

## Source details

**Canonical URL:** [EXECUTIVE SUMMARY — 1cheea2025007-source-pdf](https://www.imf.org/-/media/files/publications/cr/2025/english/1cheea2025007-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2025/english/1cheea2025007-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2025/english/1cheea2025007-source-pdf.pdf.json)

---

### Institutional framework for macroprudential policy
- Governance and mandates:
  - The SNB oversees macroprudential matters and makes Countercyclical Capital Buffer (CCyB) proposals but does not have voting power on CCyB decisions.
  - Neither the SNB, FINMA, nor any existing inter-agency committee hold a formal mandate to recommend adjustments to the Swiss macroprudential toolkit when current tools prove inadequate, creating an inaction bias.
  - Cooperation arrangements:
    - Bilateral MoU between SNB and FINMA establishes a “Standing Committee for Financial Stability” meeting at least four times per year; agendas jointly set; chair and secretariat rotate annually.
    - Trilateral MoU between SNB, FINMA and FDF establishes a “Trilateral Meeting” at least twice a year; forum has no decision-making competence and no joint communication after meetings.
- Capacity, information sharing, and accountability:
  - FINMA shares detailed data with the SNB, including data on systemic banks and real estate-related risk exposures.
  - No requirement for the four key stakeholders to evaluate regularly whether macroprudential objectives have been met or measures have been effective.
  - Communication: when the Federal Council (FC) decides on CCyB changes based on an SNB proposal, the FDF, FINMA, and SNB separately communicate decisions; no joint communication.
- Main institutional recommendation:
  - Establish a dedicated macroprudential committee (referred to as “Systemic Risk Council”), chaired by the SNB, with a clear mandate, procedures, and public accountability to reduce inaction risk and improve coordination (FDF, FINMA, SNB). Timing: MT; Priority: M.

### Macroprudential toolkit — current composition and limits
- Toolkit components and constraints:
  - Switzerland’s macroprudential toolkit consists mainly of the CCyB (broad-based and sectoral) and higher capital surcharges for systemically important banks; no legal basis for binding borrower-based measures.
  - Sectoral CCyB on mortgage-backed residential property:
    - Activated in 2013 at 1 percent, increased to 2 percent in 2014, fully released in March 2020, re-activated in January 2022 and raised to its legal maximum of 2.5 percent effective September 2022.
    - Legal hard maximum of the CCyB is 2.5 percent.
  - Broad-based CCyB on all domestic exposures has remained zero since establishment.
  - Swiss Banking Association self-regulation provides minimum requirements and qualitative guidelines; FINMA recognizes these as binding minimum standards and relies on external auditors to assess compliance.
- Limits and recommended legal changes:
  - As the sectoral CCyB is currently set at its legal maximum, it could not be increased further; the only option under current rules is to broaden its application beyond mortgage positions secured by residential property.
  - Recommendation: introduce a legal basis for a second time-varying capital-based instrument (similar to the sectoral Systemic Risk Buffer in the EU), allow both broad-based and sectoral capital tools to be used more flexibly, and consider removing the 2.5 percent upper limit on either instrument. (FDF, FINMA, SNB) — Timing: I; Priority: H.
  - Review CCyB framework toward a positive cycle-neutral CCyB regime and simplify the Capital Conservation Buffer (CCoB) framework. Suggested CCoB: set to 2.5% for all banks.

### Real estate market vulnerabilities and borrower risks
- Macroeconomic and market indicators:
  - Mortgage volume: after its peak in 2021, stands at around 145 percent to GDP.
  - Mortgages account for 86 percent of total loans as of end-2024.
  - Swiss household gross debt (loans and debt securities): more than 120 percent of GDP.
  - Owner-occupied mortgage debt concentrated in less than 40 percent of Swiss households.
  - House price overvaluation estimates:
    - Models: range between 20 and 35 percent; regression-based: 20 and 25 percent.
    - SNB estimates apartment prices overvaluation in a range between 15–40 percent.
  - Vacancy rate of dwellings: about 1 percent of the total housing stock.
  - Total number of dwellings grew by 10 percent over the last decade; building production and permits declined since 2018.
  - Monetary easing started March 2024.
- Affordability and lending-standard deviations:
  - Since 2017:
    - Around 50 percent of owner-occupied mortgage lending: stressed debt service and maintenance costs exceed the income-based affordability threshold.
    - Around 60 percent of investment-property related lending: stressed debt service and maintenance costs exceed the income-based affordability threshold.
  - Median net income for new mortgage loans in the owner-occupied segment: CHF 117 thousand as of 2024 (Hypo B Data).
  - Industry affordability benchmarks (per FINMA Guidance 2/2025):
    - Owner-occupied RRE: affordability threshold 33 percent of gross income (or 38 percent of net income); stressed costs use 5 percent imputed mortgage rate and 1 percent amortization; maintenance costs 0.8 percent of house value.
    - Investment-property real estate: 100 percent of net rental income used as affordability threshold.
    - Corresponding implied LTI multiples: 4.8 for owner-occupied RRE (80 percent LTV); 14.2 for investment-property RRE (75 percent LTV).
  - Compliance observations:
    - LTV cap compliance (owner-occupied): 1.5 percent of mortgage lending above 90 percent LTV from 2017 to 2024 (Hypo B sample of 29 banks).
    - Investment-property non-compliance higher: 5.5 percent (households) and 7.3 percent (commercial borrowers).
    - About half of new mortgage loans in the owner-occupied segment do not meet FINMA’s industry standards; around 60 percent in the investment property segment do not meet standards.
  - Basel III final risk weights calibration: assumed an estimated 15 percent deviation with prudent affordability and valuation in the mortgage portfolio (85 percent weight on preferential risk weights and 15 percent on penalty risk weights), which appears well below observed shares of non-compliance in new loans.

### Effectiveness of existing measures and proposed borrower-based tool (DSTI cap)
- Existing measures:
  - LTV caps (self-regulation) effective in limiting LGD; amortization practices: part below 66 percent LTV often interest-only; part above 66 percent LTV amortized linearly over 15 years; legal requirement to keep LTV unchanged for 5 years applies.
- Rationale for DSTI cap:
  - DSTI cap primarily targets borrower PD (probability of default) while LTV caps target LGD (loss given default).
  - Borrower-based measures better address incentives and borrower resilience than capital-based buffers alone.
- Counterfactual analysis setup:
  - Model: IDHBS+ (Integrated Dynamic Household Balance Sheet+).
  - Counterfactual DSTI cap evaluated: 30 percent of borrower’s income; alternatives considered: 30 percent, 25 percent, 20 percent.
  - Modeling assumptions: DSTI cap modeled on actual (not stressed) debt service; households above cap receive smaller loans to meet 30 percent cap while keeping resulting LTVs unchanged.
- Distributional and time-variation findings (2017–2024 lending flows):
  - Share of lending with DSTI >30 percent:
    - Owner-occupied: 5 percent.
    - Investment property (households): 39 percent.
    - Investment property (commercial borrowers): 35 percent.
  - A 30 percent DSTI cap would mainly affect high-LTI borrowers:
    - Owner-occupied: 91 percent of affected borrowers have debt exceeding 10 times sustainable net income.
    - Investment-property (scaled): 38 percent in the highest LTI bucket under scaled distribution.
  - DSTI cap binding by interest-rate fixation (owner-occupied):
    - Fixed-rate loans: 3 percent affected by 30 percent DSTI cap.
    - Variable-rate loans: 9 percent affected by 30 percent DSTI cap.
    - Mixed fixed/variable: 6 percent affected by 30 percent DSTI cap.
  - Time variation: affected shares rise during monetary tightening (owner-occupied from 5 percent average to close to 10 percent in 2023; household investment-property from 40 percent average to around 60 percent at peak policy rates in 2023).
- Quantitative impacts from IDHBS+ counterfactual (30 percent DSTI cap):
  - PD effects:
    - A 30 percent DSTI cap reduces the increase in mortgage PDs by about one third under the adverse supply-shock inflationary scenario.
  - LGD effects:
    - Very minor positive impact on LGDs based on the empirical relationship used.
  - Capital impact:
    - Capital ratios of Swiss banks gain 37 basis points of risk-weighted assets under the 30 percent DSTI counterfactual compared to the capital ratios resulting from the supply-shock adverse scenario.
  - Mortgage lending volume impact:
    - Mortgage lending volume would have been 1.6 percent lower for the entire period 2017 to 2024 under the 30 percent DSTI cap.
    - Segmental average reductions:
      - SORRE loans: 3 percent reduction on average.
      - IPRRE HH loans: 4 percent reduction on average.
      - IPRRE Commercial loans: 0.4 percent reduction on average.
      - Total SORRE and IPRRE: 1.6 percent reduction on average.
- Net assessment:
  - The 30 percent DSTI counterfactual shows limited costs on lending and access, while materially strengthening bank and household resilience and reducing expected losses.
  - International considerations: IMF Macroprudential Database reports 67 countries with active DSTI measures as of end-2023; DSTI/DTI designs vary (stressed vs actual service, inclusion of wealth, combination with LTV/maturity limits).
  - Data constraint: calibrating measures that incorporate financial wealth requires borrower-level wealth data which is not currently available in Switzerland.

### Data, monitoring, and CRE assessment
- Data improvements:
  - Hypo B dataset (since Q1 2017) covers flow of mortgage loans for 29 banks with segment-level metrics (LTV, LTI, DSTI, EtP flags).
  - From 2025 Q1: loan portfolio by LTV ratio (stock) becomes available for all Swiss banks due to Basel III final standard reporting.
  - From 2027: planned loan-by-loan reporting on the existing mortgage loan stock covering more than 90 percent of domestic credit market; first deliveries expected in 2027. The envisaged dataset does not cover all debts and debt service of each borrower.
- CRE data gaps and observations:
  - CRE appears broadly stable with less dynamic price developments than investment-property segment; signs of recovery in 2024.
  - New construction activity in CRE fell sharply in 2023 and 2024; no indication of systemwide oversupply.
  - Vacancy rates in office space in main cities remain elevated versus historical averages but lower than other European capitals.
  - Recommendation: ensure loan-by-loan data starting 2027 closes data gaps for CRE and consider closer monitoring of real estate investment funds.

### Main recommendations (selected, with timing and priority)
- Institutional Framework and Macroprudential Toolkit:
  1. Establish a formal Systemic Risk Committee (Systemic Risk Council), chaired by the SNB, with clear mandate and accountability. (FDF, FINMA, SNB) — Timing: MT; Priority: M.
  2. Introduce the legal basis for a sectoral capital-based instrument, separate from the CCyB. (FDF, FINMA, SNB) — Timing: I; Priority: H.
  3. Incorporate a more flexible and proactive CCyB approach (positive cycle-neutral CCyB) and remove the 2.5 percent upper limit. (FDF, SNB) — Timing: ST; Priority: M.
  4. Simplify the CCoB framework and separate it from buffers for systemically important banks. (FDF) — Timing: MT; Priority: L.
- Systemic Risks and Macroprudential Policy Settings:
  5. Ensure additional affordability-related credit risks in banks’ mortgage loan portfolios are adequately covered by capital. (FINMA, SNB) — Timing: I; Priority: H.
  6. Introduce a binding debt-service-to-income cap, if needed by establishing a legal basis for borrower-based measures. (FDF, FINMA, SNB) — Timing: I; Priority: H.
  7. Tighten the LTV cap in the investment property mortgage loan segment if mortgage loans with LTV ratios above 75 percent significantly increase. (FDF, FINMA, SNB) — Timing: ST; Priority: M.
  8. Remove tax deductibility of mortgage interest payments for households. (FDF) — Timing: ST; Priority: M.
  9. Close remaining data gaps on mortgage loan and borrower characteristics and enhance analysis of commercial real estate risks. (FINMA, SNB) — Timing: ST; Priority: M.
- Timing legend:
  - C: Continuous
  - I: Immediate (<1 year)
  - ST: Short Term (1–2 years)
  - MT: Medium Term (3–5 years)
- Priority legend:
  - H: High
  - M: Medium
  - L: Low

*Source: 1cheea2025007-source-pdf.*

### EXECUTIVE SUMMARY __________________________________________________________________________ 5

### EXECUTIVE SUMMARY

### Institutional framework for macroprudential policy
- The SNB oversees macroprudential matters and makes Countercyclical Capital Buffer (CCyB) proposals but does not have voting power on CCyB decisions.
- Neither the SNB, FINMA, nor any existing inter-agency committee hold a formal mandate to recommend adjustments to the Swiss macroprudential toolkit when current tools prove inadequate, creating an inaction bias.
- Recommendation: Establish a dedicated macroprudential committee, chaired by the SNB, to discuss and jointly decide on policies to improve macroprudential communication, public accountability, and mitigation of inaction risks.

### Macroprudential toolkit and limits
- The sectoral CCyB for mortgage loans:
  - Was set at 2 percent from 2014 until March 2020.
  - Was released during the Covid-19 pandemic.
  - Was raised again to its legal maximum of 2.5 percent in 2022.
- The sectoral CCyB has reached its legal limit and cannot be tightened further to counter mounting systemic risks.
- Authorities currently depend on the Swiss Banking Association self-regulation for borrower-based measures (minimum requirements and broad qualitative guidelines), which is insufficient given current risks.
- Recommendation: Establish a legal basis for a second time-varying capital-based instrument (similar to the sectoral Systemic Risk Buffer in the EU) to:
  - Enhance the ability to tackle risk build-up effectively.
  - Create space for more flexible use of both broad-based and sectoral capital tools.
  - Potentially introduce a positive cycle-neutral CCyB for the former and remove the 2.5 percent limit on either instrument.

### Real estate market vulnerabilities and borrower risks
- Housing market signals and borrower vulnerability:
  - Stretched housing market valuations, widespread loosening of underwriting standards, and record-high household debt to GDP indicate mounting systemic risks.
  - Monetary easing and cantonal initiatives easing affordability criteria for new borrowers could stimulate credit demand and reinforce a negative spiral of unsustainable lending standards, house price increases, and risk build-up.
- Affordability and prudential deviation:
  - Since 2017, in around 50 percent of owner-occupied mortgage lending, stressed debt service and maintenance costs exceed the income-based affordability threshold.
  - Since 2017, in around 60 percent of investment-property related lending, stressed debt service and maintenance costs exceed the income-based affordability threshold.
  - The recently introduced Swiss Basel III final risk weights for mortgage loans were calibrated based on an estimated 15 percent deviation with prudent affordability and valuation in the mortgage portfolio, which appears well below the observed share of new loans not meeting FINMA's affordability criteria.

### Effectiveness of existing measures and recommended borrower-based tool
- LTV caps (self-regulation) have been effective in limiting LGD, which is important given overvalued house prices.
- A debt-service-to-income (DSTI) cap would primarily aim at limiting borrower PD and address the riskiest mortgage lending types.
- Counterfactual analysis findings:
  - A DSTI cap would mainly affect the high-loan-to-income (LTI) portion of mortgage lending.
  - A DSTI cap would restrict investment-property lending more than owner-occupied mortgage lending.
  - A DSTI cap would be more binding for variable-rate mortgage loans, which have risen in popularity and could make borrowers vulnerable to interest rate increases.
  - Using the integrated dynamic household balance sheet model for a representative sample of household micro data, a counterfactual 30 percent DSTI cap would reduce the PD for mortgage loans by around one third under the adverse supply-shock inflationary scenario of the FSAP.

### Main recommendations (Table 1)
- Institutional Framework and Macroprudential Toolkit
  1. Establish a formal Systemic Risk Committee, including the SNB, FINMA, and FDF, chaired by the SNB and with a clear mandate, procedures, and accountability to review and propose measures to mitigate systemic risks. (FDF, FINMA, SNB) — Timing: MT; Priority: M
  2. Introduce the legal basis for a sectoral capital-based instrument, separate from the CCyB. (FDF, FINMA, SNB) — Timing: I; Priority: H
  3. Incorporate a more flexible and proactive approach towards CCyB setting in Switzerland in the form of a positive cycle-neutral CCyB regime, and without the 2.5 percent upper limit. (FDF, SNB) — Timing: ST; Priority: M
  4. Simplify the CCoB framework and separate it from buffers for systemically important banks. (FDF). — Timing: MT; Priority: L

- Systemic Risks and Macroprudential Policy Settings
  5. Ensure that additional affordability-related credit risks in banks’ mortgage loan portfolios are adequately covered by capital. (FINMA, SNB) — Timing: I; Priority: H
  6. Introduce a binding debt-service-to-income cap, if needed by establishing a legal basis for borrower-based measures. (FDF, FINMA, SNB) — Timing: I; Priority: H
  7. Tighten the loan-to-value cap in the investment property mortgage loan segment in case of a significant increase in mortgage loans with loan-to-value ratios above 75 percent. (FDF, FINMA, SNB) — Timing: ST; Priority: M
  8. Remove the tax deductibility of mortgage interest payments for households. (FDF) — Timing: ST; Priority: M
  9. Close remaining data gaps on mortgage loan and borrower characteristics and enhance the analysis of commercial real estate risks. (FINMA, SNB) — Timing: ST; Priority: M

- Timing legend:
  - C: Continuous
  - I: Immediate (<1 year)
  - ST: Short Term (1–2 years)
  - MT: Medium Term (3–5 years)

- Priority legend:
  - H: High
  - M: Medium
  - L: Low

### Scope and purpose of the technical note
- The technical note examines the evolution and impact of Swiss macroprudential policy tools targeting real estate markets.
- Structure:
  - Review of institutional aspects of the macroprudential framework.
  - Analysis of real estate market dynamics since the 2019 FSAP and key financial stability risks.
  - Evaluation of the effectiveness of current macroprudential instruments and identification of vulnerabilities.
  - Counterfactual analysis assessing the introduction of additional borrower-based measures to the existing toolkit.

*Source: 1cheea2025007-source-pdf - EXECUTIVE SUMMARY.*

### 7.      The collaboration between the SNB, FINMA and FDF in the field of financial stability

### 7.      The collaboration between the SNB, FINMA and FDF in the field of financial stability

### Collaboration framework and governance
- Cooperation between the SNB and FINMA for financial stability is enshrined in a bilateral Memorandum of Understanding (MoU). The MoU provides the basis for an operational “Standing Committee for Financial Stability” that meets at least four times per year. The committee:
  - Discusses topical issues for financial stability in Switzerland and regulatory initiatives and their impact on Swiss banks, such as the recent Basel III finalization or the ongoing Too-Big-to -Fail (TBTF) reforms.
  - Neither has a formal advisory role in macroprudential measures nor fulfills this purpose in practice.
  - Has meeting agendas jointly set by the SNB and FINMA, with the chair and secretariat function rotating on an annual basis between the two institutions.
- The bilateral MoU also establishes a high-level “Steering Committee,” co-chaired by the SNB Chairman and the FINMA Chairman, which meets twice a year.
- A trilateral MoU between SNB, FINMA and FDF aims to support information exchange and cooperation in crisis times and continuously. Based on the trilateral MoU:
  - The FDF (SIF State Secretary), the FINMA CEO and the SNB Vice Chairman responsible for financial stability meet at least twice a year for a “Trilateral Meeting.”
  - The FDF is responsible for organizing the discussion and setting the agenda in consultation with the SNB and FINMA.
  - The forum has no decision-making competence and there is no joint communication after its meetings.
  - The trilateral MoU establishes two committees to facilitate inter-agency cooperation during financial crises: a high-level “Steering Committee” (broad mandate) and a “Committee on Financial Crisis.”
- The FDF and FINMA meet regularly to discuss FINMA’s strategic goals and regulatory issues, including in the field of financial stability (high-level “strategic dialogues” twice a year and “policy dialogues” at least three times a year).

### Information sharing and SNB capacity
- The SNB and FINMA have a well-established framework for information sharing in financial stability and macroprudential policy.
- In practice:
  - FINMA shares detailed data with the SNB, including data on systemic banks and real estate-related risk exposures.
  - The bilateral MoU requires FINMA to share important supervisory findings with the SNB, including banks’ compliance with macroprudential measures.
- Overall, the SNB has sufficient capacity to obtain information from FINMA to fulfill its financial stability function.

### Accountability and communication practices
- Accountability arrangements are very limited:
  - No requirement for the four key stakeholders to evaluate regularly whether macroprudential objectives have been met or measures have been effective.
  - No regular dedicated report on macroprudential policy.
- Communication practices on macroprudential policy have remained unchanged since the 2019 FSAP:
  - Whenever the FC decides on CCyB changes based on a SNB proposal, the FDF, FINMA, and SNB separately communicate the decision to the market; no joint communication is made.
  - The SNB, as leader on macroprudential oversight and CCyB proposals, communicates the underlying rationale of CCyB changes in more detail than the other institutions.
  - In quarters with no SNB proposal and thus no FC CCyB decision, no public communication takes place. At minimum, the SNB communicates the stance of the CCyB in Switzerland with a short note on the underlying rationale on an annual basis.

### Assessment: strengths and weaknesses of current setup
- Strength:
  - The current setup supports political buy-in for macroprudential actions: CCyB requirements have so far always followed the SNB proposals (four CCyB proposals since 2013), illustrating the SNB made convincing cases to the FDF before and after proposals.
- Weaknesses and risks:
  - Unclear accountability for macroprudential action and non-action: the SNB is accountable for macroprudential oversight and CCyB proposals but has no vote in CCyB decisions.
  - No formal mandate for SNB, FINMA, or existing inter-agency committees to advise on adjustments of the Swiss macroprudential toolkit if existing tools are insufficient or unusable for further action.
  - Consequence is a strong inaction bias, notably when the FDF (key player for legislative changes) may be reluctant to “lean against the wind.”
- Example of past policy action:
  - Reintroduction of the sectoral CCyB in 2022, increased from 0 percent to 2.5 percent with approx. 8 months lead time; context includes prior 2 percent sCCyB from 2014 to March 2020, release in March 2020 at onset of Covid-19 pandemic, and limited buffer utilization during the pandemic.

### Recommendations for institutional reform
- Guiding principle: strengthen the role of the central bank in macroprudential policy.
- Consider creating a key macroprudential committee (“Systemic Risk Council”) to discuss and decide on macroprudential policy. Advantages of a decision-making macroprudential committee:
  - Improve transparency and accountability: issue joint press statements after each meeting and become the Swiss macroprudential authority accountable to the Swiss Parliament.
  - Reduce risk of inaction or delayed action; strengthen SNB’s ability to fulfill financial stability objective through stronger link between risk assessment and policy action.
  - Mandate the committee to advise executive and legislative branches on adequacy of macroprudential toolkit and suggest warranted changes.
  - Simplify decision-making: separate formal consultation of FINMA on CCyB proposals may be unnecessary due to FINMA’s participation in the committee.
  - Support expansion of the macroprudential toolkit beyond the CCyB.
- Suggested design characteristics for a Systemic Risk Council:
  - Clear mandate focused on system-wide perspective on financial stability risks and joint discussion and decision on macroprudential policy; define scope to distinguish from monetary policy, microprudential policy, and crisis management.
  - Assign committee’s secretariat function to the SNB and responsibility of introducing decision items; SNB to remain solely responsible for making CCyB proposals and retain powers regarding designation of systemically important banks.
  - SNB to play a leading role on additional macroprudential tools (e.g., binding borrower-based measures or recommendations regarding self-regulation amendments), engaging closely with FINMA.
  - Composition to involve senior representatives of SNB, FINMA and FDF, and potentially independent external experts; chaired by SNB given its legal mandate for financial stability, organized collegially with equal weight and influence among the three institutions.
  - Decision process: all macroprudential decisions preceded by well-informed discussion among members; chair leads discussion and decision phase; voting preferably based on simple majority, chair to aim for broad agreement without requiring unanimity.
  - Formal meetings at predefined frequency, preferably quarterly, followed by joint public communication agreed during each meeting; joint communication should not prevent SNB and FINMA from separate communications (e.g., SNB Financial Stability Report or FINMA Risk Monitor).
  - Primary legislation should assign sufficient powers to the committee—either hard powers to issue binding regulations or the ability to make recommendations to other agencies with hard powers.

### Strengthening communication practice (recommendation)
- SNB should communicate on macroprudential policy more often than the current annual basis if institutional setting remains unchanged.
- Good practice: communicate on a quarterly basis, including quarters with no CCyB proposal.
- Public communication should pay specific attention to policy implications from the SNB’s systemic risk analysis that may require introducing tools other than the CCyB (e.g., targeted borrower-based measures).

### Macroprudential toolkit and systemic risk analysis — current status
- The macroprudential toolkit in Switzerland is very narrow:
  - Consists of a CCyB (applicable on a sectoral and broad basis) and higher capital surcharges for systemically important banks.
  - No other broad-based or sectoral capital-based measures in place and no legal basis to enact binding borrower-based measures.
  - Swiss Banking Association sets self-regulation rules for mortgage lending, recognized by FINMA as binding minimum standards, enabling preferential risk weights under the Capital Adequacy Ordinance; FINMA relies on external auditors to assess compliance.
- Countercyclical Capital Buffer (CCyB) specifics:
  - Legal hard maximum of 2.5 percent, which can be broad-based or applied to a subset of exposures.
  - Sectoral CCyB on mortgage-backed positions secured by residential property:
    - Activated in 2013 at 1 percent of respective exposures.
    - Increased to 2 percent in 2014.
    - Fully released with immediate effect in March 2020 at the onset of the Covid-19 pandemic.
    - Re-activated in January 2022 and raised to its legal maximum of 2.5 percent, effective September 2022.
  - Broad-based CCyB on all domestic exposures has remained zero since establishment.
- SNB’s systemic risk assessment and CCyB decision process:
  - SNB performs a quarterly assessment separated into:
    - Residential real estate related exposures (mortgage loans cover approx. 90 percent of total loans in domestically oriented banks).
    - All domestic exposures (includes commercial real estate and other loans).
  - Guided discretion applied; proposals informed by reference indicators:
    - sCCyB proposal: mortgage-credit-to-GDP gap (similar to credit-to-GDP gap indicator (BCBS 2010)) and price-to-rent ratios (deviation from long-term average) for different real estate segments; additional indicators include LTI distributions, debt service ratio, real estate overvaluation estimates, and investment property yields.
    - CCyB proposal: bank-credit-to-GDP gap (which also includes corporate and consumer loans) and price-to-rent ratios in commercial real estate (deviations from long-term average). No composite indicator in place; interpretation of credit-to-GDP gaps follows BCBS 2010 guidance with activation signals for a positive gap above 2 percent.
- Frequency and process of CCyB decisions:
  - Actual CCyB decisions are taken infrequently: Federal Council took four CCyB decisions since 2013.
  - Decision-making steps:
    - SNB Governing Board decides whether to propose a CCyB change to the Federal Council (FC). Since the Covid-19-related release, SNB only submitted one proposal to reactivate and increase the CCyB (to 2.5 percent).
    - Formal consultation with FINMA follows the SNB Governing Board’s decision to propose.
    - SNB submits the proposal to the FC; the FC has followed SNB proposals to date.

*IMF staff summary based on the chapter text provided.*

### 20.      As the sectoral CCyB is    currently set at its legal maximum, it could not be increased

### 1cheea2025007-source-pdf - 20.      As the sectoral CCyB is    currently set at its legal maximum, it could not be increased

### Sectoral CCyB legal limit
- As the sectoral CCyB is currently set at its legal maximum, it could not be increased further if the SNB considered it being warranted.
- The only possibility to further tighten the CCyB within current Swiss regulation is broadening its application beyond mortgage positions secured by residential property.

### Capital Buffer Requirements for Systemically Important Banks (SIBs)
- SNB designates systemically important banks subject to Swiss G-SIB or D-SIB buffer requirements (going concern requirements) in consultation with FINMA.
- As of early 2025, the SNB classifies four Swiss banks as systemically important: UBS, PostFinance, Raiffeisen Group and ZKB.
- UBS has been classified as a global systemically important bank (G-SIB) by the Financial Stability Board (FSB) ever since the classification was introduced in 2011; following the state-assisted merger of Credit Suisse in 2023, UBS is the only remaining G-SIB in Switzerland.
- Main criteria for systemic relevance:
  - market share in systemically important functions (domestic credit, deposits, and payments);
  - size (measured in terms of total assets and covered deposits);
  - risk profile (based on foreign claims and liabilities, and replacement values of derivatives);
  - interconnectedness (claims to and liabilities from domestic banks).
- Structure of capital buffer requirements:
  - Banking category buffer requirements (Appendix 8 of the Capital Adequacy Ordinance):
    - category 5 (small banks): 2.5 percent
    - category 4 (medium-sized banks): 3.2 percent
    - category 3 (large complex banks): 4.0 percent
    - category 2 (very important complex market participants, including the three D-SIBs) and category 1 (extremely large, important, and complex market participants, which includes UBS): 4.8 percent
  - If a bank is designated as a SIB, the Swiss TBTF framework includes a capital buffer requirement of 4.86% of RWA.
  - Additional institution-specific capital surcharges (in percent of RWA) for the four systemic banks:
    - UBS: 1.44 percent
    - Raiffeisen Group: 0.36 percent
    - ZKB: 0 percent
    - PostFinance: 0 percent
  - These surcharges include additional capital buffer requirements for UBS and Raiffeisen Group that exceed certain thresholds for market size or total exposure as set out in Appendix 9 of the Capital Adequacy Ordinance.
- Ongoing TBTF reform: calculation of the capital surcharge linked to a systemically important bank’s size will be changed; authorities intend to raise the bucket thresholds of total exposure to take account of GDP growth in recent years.

### Self-Regulation for Mortgage Lending
- The Swiss Banking Association (SBA) self-regulation on mortgage loans has been partly in place since 1994 and covers mortgage lending to all domestic counterparties (private borrowers and companies), including owner-occupied RRE, investment-property RRE, SMEs and CRE.
- Changes to self-regulation are discussed by a working group led by the SBA and involve the FDF, SNB and FINMA; the process aims for consensus among the four stakeholders and can be lengthy (e.g., the 2019 revision took around two years).
- The self-regulation has two parts:
  - Guidelines on minimum requirements for mortgage loans (minimum requirements):
    - (i) a minimum downpayment of 10 percent, which corresponds to a uniform 90 percent LTV cap, for most mortgage loans as of early 2025;
    - (ii) a linear amortization requirement of the loan amount exceeding 66 percent LTV over a period of 15 years.
  - Guidelines on assessing, valuing, and processing loans secured by property (qualitative guidelines):
    - principle-based; relate to lending policies including affordability assessment, valuation, and loan monitoring;
    - require banks to define an ‘affordability threshold’ in relation to borrower’s income and expenses (owner-occupied RRE) or investment income (income-producing real estate);
    - require individual banks to define an appropriate ‘imputed mortgage interest rate’ for each segment without giving indication of appropriate levels or ranges.
- Stricter minimum requirements for investment property-related lending were put in place at end-2019 and lifted at end-2024:
  - 2019 tighter requirements included a minimum 25 percent downpayment and 10-year linear amortization of the loan amount exceeding 66 percent LTV.
  - With the revised Capital Adequacy Ordinance (CAO), in force from January 2025, the same minimum requirements apply again to new mortgages in all segments.
  - FINMA recommends that banks continue to maintain the 25 percent minimum down payment for investment property going forward.
- FINMA considered issuing supervisory guidance related to real estate and mortgage risks to clarify supervisory expectations regarding affordability assessment and valuation.

### Box 1 — Industry Standards in the Swiss Mortgage Market (key points)
- Industry standards (non-binding) relate to LTV ratios, amortization, and affordability.
- Loan-to-value and down payments:
  - Industry standards for maximum LTV ratios:
    - owner-occupied RRE: 80 percent
    - investment-property RRE: 75 percent
    - CRE: 50 – 70 percent (depending on property type)
  - Down payments need to be made in hard equity for the portion required by the SR minimum standard (90 percent LTV); for owner-occupied first homes, the remainder could be covered by withdrawing or pledging pension fund savings or third pillar.
- Amortization:
  - Industry standard: (i) the part of the loan below 66 percent LTV is often not amortized (first mortgage) and thus constitutes an interest-only loan, and (ii) the part above 66 percent LTV is amortized (second mortgage) linearly over 15 years.
  - Indirect amortization is common for second mortgages (payments into a Pillar 3 pension scheme pledged against the mortgage); popularity driven by tax treatment and higher return potential.
- Affordability:
  - Industry standard for owner-occupied RRE (FINMA considers sustainable based on FINMA Guidance 2/2025):
    - an affordability threshold of 33 percent of gross income (or 38 percent of net income);
    - stressed costs calculated by a 5 percent imputed mortgage rate and 1 percent amortization;
    - build-related maintenance costs of 0.8 percent of the house value.
  - For investment-property real estate, 100 percent of the net rental income is used as the affordability threshold.
  - If income-based affordability is calculated per above, it corresponds to:
    - a LTI multiple of 4.8 in the owner-occupied RRE segment (for 80 percent LTV, gross income * 0.33 ≥ loan*(0.05+0.01)+value*0.008);
    - a LTI multiple of 14.2 for the investment-property real estate segment (same calculation with 100 percent net income and LtV 75%).
- Notes from Box 1:
  - Amortization practice may differ across banks and customers; some banks may require some amortization depending on property type or counterparty rating.
  - Payments into pension schemes can be deducted from taxable income; pension assets, capital gains and income are exempt from taxation until payout and not subject to wealth taxes; interest income is exempt from income and withholding tax; payouts are taxed at a lower rate.
  - For a standard owner-occupied RRE mortgage loan with an LTV of 80% at origination, yearly amortization payments of 1.1% are required to be in line with the self-regulation.

### Additional Pillar 2 Requirement Related to Mortgage Risk
- FINMA imposes Pillar 2 measures for selected banks with elevated mortgage lending risk (microprudential measure related to mortgage lending risks).
- Applied based on FINMA stress-test findings (using an 8 percent hurdle rate in terms of total capital) where available capital buffers for the mortgage book are insufficient to absorb stress losses or based on on-site inspection findings of unsustainable mortgage growth or inadequate controls.
- As of end-2024, FINMA imposed 15 such Pillar 2 add-ons to Swiss banks in relation to mortgage market risks.

### Comparison of the Swiss Macroprudential Toolkit with European Peers
- Two main differences between Switzerland and most EEA countries:
  - Most EEA countries (27 out of 30) have borrower-based measures in place; Switzerland relies on self-regulation by the banking association rather than statutory borrower-based measures.
  - Most EEA countries have at least one additional capital-based tool available which can be activated jointly with the CCyB and SIB buffers; 21 out of 30 EEA countries have activated a broad-based or sectoral Systemic Risk Buffer (SyRB) as of early 2025.
- Table summary highlights (as of early April 2025):
  - Examples of activated CCyB rates in peers include: 1.0% (broad), 2.5% (broad), 0.75% (broad), 2.0% (broad), 2.5% (broad), 2.5% (sectoral), 2.0% (broad).
  - SyRB (or equivalent) activated sizes cited include 0.5-1% (broad), 6% (sectoral), 7% (sectoral), 2% (sectoral), 4.5% (broad), 3% (broad).
  - SIB buffer ranges in peers include 0.45-1.75%, 0.75-1.5%, 1-3%, 0.25-2%, 0.25-2%, 1-2%, 0-1.44%, 1%.
  - CCoB (equivalent) in place commonly at 2.50%; Switzerland’s CCoB transposition ranges 2.5-4.8% depending on category.
  - Borrower-based measures: most peers have LTV, DSTI and loan maturity caps; Switzerland uses self-regulation with LTV cap and amortization rule.
- Commentary:
  - Banks in some peer countries (Denmark, Norway, Sweden) have combined buffer requirements more than double those applied to Swiss banks as a percent of risk-weighted assets.
  - All peers in the table have adopted the Basel III final standards.
  - Except for Germany, all peer countries have activated borrower-based measures as of early 2025.

*Italic: Source: 1cheea2025007-source-pdf (excerpt).*

### 31.      The SNB publishes a financial stability report on an annual basis and communicates

### The SNB publishes a financial stability report on an annual basis and communicates risks to key stakeholders and the general public.

### Risk analysis framework and reporting
- The FSAP Technical Note on Systemic Risk Analysis provides a detailed assessment of the SNB’s operational framework for financial stability risk analysis in the banking sector and non-bank financial institutions.
- The SNB publishes a financial stability report on an annual basis and communicates risks to key stakeholders and the general public.

### Data sources for mortgage-lending risk assessment
- FINMA and the SNB mainly rely on two data sources to assess mortgage lending risks:
  - Mortgage lending data (since 2017): the ‘Hypo B’ dataset includes the flow of mortgage loans since Q1 2017 for 29 banks48 and is differentiated by mortgage segment49, lending metrics (such as LTV, LTI, DSTI), whether the loan is considered an ”Exception-to-Policy“ (EtP loan), and other characteristics. FINMA uses it to identify outlier banks (e.g., displaying high EtP ratios) and inform supervisory action.
  - Survey data on mortgage lending criteria (biannually): FINMA collects quantitative survey data on mortgage lending criteria from 45 banks every two years; the latest analysis was performed in 2024 with a focus on mortgage-related lending criteria and valuation methods.

### Improvements to mortgage loan data coverage (timeline and scope)
- Two-step improvements since the 2019 FSAP:
  - Loan portfolio by LTV ratio (from 2025): More granular data on the stock of mortgage loans, differentiated by LTV and property type (e.g., owner-occupied, investment property), becomes available from 2025 Q1 onwards due to the introduction of the Basel III final standard. This database allows portfolio-level analysis for the stock of mortgage loans for all Swiss banks.50
  - Loan-by-loan data on the existing mortgage loan stock (from 2027): Following a 2019 IMF FSAP recommendation, the SNB and FINMA decided in 2022 to introduce a reporting requirement covering more than 90 percent of the domestic credit market. It will cover both existing and new loans and shed further light on LTV and LTI distributions, include more granular data on the income of private borrowers, firm characteristics, and outstanding debt of commercial borrowers. First data deliveries expected in 2027.51 52
- Current limitation noted: the envisaged dataset does not cover information about all debts and debt service of each respective borrower, which would be necessary to calculate and analyze DTI and DSTI ratios at the single loan level.

### Commercial real estate (CRE) data gaps and market assessment
- Existing data gaps hinder thorough CRE market analysis from a financial stability perspective.
- Market observations:
  - CRE appears broadly stable based on price developments, with less dynamic price developments than the investment-property segment in recent years and signs of recovery in 2024.53
  - New construction activity in the CRE segment fell sharply in 2023 and 2024, yet there is no indication of systemwide oversupply.
  - Price indices based on asking rents have recovered.
  - Structural challenges: increasing use of e-commerce and working from home.
  - Demand indicators suggest a stable market, including office and retail segments.
  - Vacancy rates in office space in main cities remain elevated compared to historical averages but are lower than in other European capitals.54
- Recommendation: FINMA and the SNB should ensure the envisaged loan-by-loan data starting from 2027 closes relevant current data gaps to support enhanced micro- and macroprudential analyses for the CRE segment of the mortgage loan book. Closer monitoring of real estate investment funds should also be considered.55

### Macroprudential policy settings and Basel III impact
- Evolution since 2019 FSAP:
  - Tightened in 2019 (stricter self-regulation for investment properties).
  - Loosened in 2020 at the onset of the Covid-19 pandemic (full release of the 2 percent sCCyB).
  - Tightened again in 2022 (sCCyB at 2.5 percent as of September 2022).
  - Somewhat loosened as of January 2025 (lifting stricter self-regulation for investment properties, while introducing final Basel III risk weights).35
- Impact of Basel III final standard on mortgage loan capital requirements:
  - For banks under the standardized approach (STA), differentiated risk weights for mortgage loans were introduced depending on LTV and mortgage market segment.
  - Raised risk weights for high-LTV loans in the RRE investment property segment (above 60 percent LTV).
  - Reduced risk weights for owner-occupied RRE mortgage loans below 80 percent LTV (shift from loan splitting to whole loan approach and lower penalty risk weights for loans non-compliant with self-regulation).
  - FINMA estimates, based on the Quantitative Impact Study performed on end-2020 data, that the change in Basel III risk weights for mortgage loans will result in an overall reduction in risk-weighted assets for credit risk by 5-10 percentage points for domestically oriented banks, depending on their portfolio composition.56

### Assessment of current toolkit limits and recommended additions
- Current limitations:
  - The macroprudential toolkit is too narrow and has reached its limits for further tightening possibility: the CCyB as the only time-varying capital-based measure cannot be practically further increased in its current sectoral application, and there is no legal basis for borrower-based measures.57 58 38
- Recommended enhancements:
  - Introduce an additional capital-based tool separate from the CCyB, not subject to an upper limit, designed similarly to the sectoral SyRB used by European peers and currently applied to RRE exposures in 7 EEA countries.59
    - The same decision-making process as for the sCCyB could apply.
    - Allow the SNB to propose the broad-based CCyB as well as the new sectoral buffer requirement to apply in parallel.
    - Neither the CCyB nor the second capital-based measure should be subject to an upper limit; remove the existing legal maximum of the CCyB at 2.5% so it is possible to set the CCyB in excess of 2.5% as in the European Union.60
  - Review the existing CCyB framework to make it more flexible and proactive, possibly as a positive cycle-neutral broad-based CCyB to reduce under-calibration from mismeasurement of cyclical risks and to build resilience against exogenous shocks.
  - Simplify the Capital Conservation Buffer (CCoB) framework and more clearly separate it from buffers for SIBs:
    - Set the CCoB requirement to 2.5% for all banks, as prescribed by Basel Risk-Based Capital Requirements (RBC 30 – Buffers above the regulatory minimum).
    - Improve the D-SIB framework by imposing higher buffers commensurate with banks' size and business model to support compliance with Basel RBC 30 and international comparability/transparency.

### Self-regulation: deficiencies and proposed changes
- Identified deficiencies of principle-based self-regulation:
  - Heterogeneous and partly aggressive bank interpretations of principles have given rise to unsustainable lending practices; it is unclear whether higher expected and unexpected losses are adequately covered by provisions and risk weights.
  - Minimum requirements set by the Swiss Bankers Association sometimes fall short of supervisory expectations (current example: investment-property lending). FINMA encouraged banks to use tighter standards than the minimum, but banks have limited incentives to comply since FINMA has formally accepted the looser requirements as a binding minimum standard.
  - Changes to self-regulation involve lengthy consensus processes, which may hamper timely action.
- Recommendations to address self-regulation deficiencies:
  - Set quantitative affordability criteria by regulation or minimum requirements in self-regulation, which banks should incorporate into lending policies, with appropriate supervisory follow-up for non-prudent lending standards.
  - Grant FINMA legal powers to enforce tighter minimum standards (for instance, via a FINMA Ordinance) where self-regulation falls short; preparation of such measures would benefit from triparty discussions with the SNB and FDF in the context of the recommended Systemic Risk Council.61
  - Consider rule-based regulation with industry consultation and public authority decision-making to improve timeliness and leverage in negotiations with the Swiss Bankers Association.

*Source: IMF staff summary of IMF Financial Sector Assessment content provided in the source PDF.*

### 43.      The need for more rule-based regulation is evident, especially regarding the

### The need for more rule-based regulation is evident, especially regarding the assessment of affordability

### Rule-based regulation and affordability limits
- Recommended substance: define quantitative limits for affordability in the form of LTI or DSTI caps, or for each element of the affordability calculation (affordability threshold, imputed interest rate, maintenance costs, and amortization).
- Application: limits should be applied on new and existing loans to identify mortgage loans with the highest affordability risks and thus the highest probabilities of default in relative terms.
- Legal/design options:
  - For new loans, limits could be set as a macroprudential borrower-based measure, for which a legal basis would need to be created.
  - For new and existing loans, limits could be set as a minimum requirement beyond which loans would be subject to penalty risk weights for loans that are non-compliant with the self-regulation.
- Regulatory process note: FINMA Ordinances also require a public consultation before they can be finalized and issued.

### Data collection and granularity
- SNB and FINMA have expanded efforts to collect more granular data on mortgage loans, but the extent of collection should be extended to cover all relevant borrower and loan characteristics.
- Essential data needs:
  - Data on all debts and all debt service of a single borrower to assess elevated affordability risks and perform counterfactual analyses of alternative macroprudential policies.
  - Reporting should include all relevant information on loan amortization modalities, which is currently lacking.
  - More granular data regarding the CRE segment, including information on the frequency and mode of appraisals, to assess build-up of risks for individual banks and the banking sector as a whole.
- Reference: See the companion FSAP Technical Note on Systemic Risk Analysis for further details.

### Real estate market vulnerabilities and recent trends
- Non-financial private sector debt has grown rapidly over the past two decades, particularly in the mortgage sector.
- Mortgage volume:
  - After its peak in 2021, mortgage volume stands at around 145 percent to GDP.
- Mortgage loans share of bank portfolios:
  - Mortgages account for 86 percent of total loans as of end-2024.
- Household debt and wealth:
  - Swiss household gross debt (loans and debt securities) is at more than 120 percent of GDP.
  - Household debt is more than double the level in Germany or France, and remains well above Norway.
  - Swiss households have on aggregate high and liquid net worth, but the distribution of household wealth is not well known.
- Home ownership and debt concentration:
  - Owner-occupied mortgage debt is concentrated in less than 40 percent of Swiss households.
- House price overvaluation estimates:
  - Models produce a range of overvaluation between 20 and 35 percent; regression-based models produce estimates between 20 and 25 percent.
  - SNB estimates apartment prices overvaluation in a range between 15–40 percent (see SNB 2024).
- Housing supply and vacancy:
  - Rate of vacant dwellings is about 1 percent of the total housing stock.
  - Total number of dwellings has grown by 10 percent over the last decade, but new housing supply has recently stagnated; building production and building permits have started to decline since 2018.
- Monetary stance and housing market:
  - Recent monetary loosening started in March 2024 and increased attractiveness of real estate investment and renewed demand for mortgages.
  - Sustained low or negative interest rates could put further upward pressures on house prices.
- Housing cost burdens and distribution:
  - Swiss households allocate 20 percent of their disposable income to housing-related costs (mortgage or rent burden).
  - Housing costs vary by tenure: 20 percent for renter vs 9 percent for homeowners.
  - The burden is higher in the lower quintile of the income distribution (40 percent) vs the top quintile (14 percent).
  - Median net income for new mortgage loans in the owner-occupied segment was CHF 117 thousand as of 2024 (Hypo B Data), which is above the 90th percentile of the household income distribution.
  - Some cantons have promoted initiatives to ease affordability for new borrowers through loans or cantonal guarantees (examples noted for Canton of Geneva and Canton of Zurich).

### Household survey and debt characteristics (Box 5 highlights)
- Data source: experimental module on Wealth of the Statistics on Income and Living Conditions, collected in 2020 by the Federal Statistics Office; representative sample of 8,900 households.
- Key statistics:
  - 53 percent of households in Switzerland are indebted.
  - 33 percent hold mortgage loans.
  - 19 percent have other (mainly consumer) debt.
  - Mortgages (on both first and second residence) account for 70 percent of total household debt.
  - The median debt to disposable income (DTI) value is 3.5 across all debt holders.
  - Households who hold only a non-amortizing mortgage have a median DTI of 5.
  - That group spends approximately 7 percent of their gross income on debt service.
  - 2.6 percent of indebted households allocate more than 30 of their disposable income for total debt servicing.

### Effectiveness of macroprudential self-regulation and compliance
- LTV cap compliance (owner-occupied segment):
  - Based on a representative sample of lenders (Hypo B dataset covering 29 banks with more than CHF 6 billion mortgage stock outstanding), only 1.5 percent of mortgage lending was provided with LTV ratios above 90 percent from 2017 and 2024.
  - Penalty risk weights apply to these loans.
- Investment-property residential (IPRRE) segment non-compliance:
  - Non-compliance higher at 5.5 percent (households) and 7.3 percent (commercial borrowers).
  - After self-regulation tightened effective Q1 2020 (lower LTV cap of 75 percent), compliance improved over time.
- Caveats on LTV effectiveness:
  - Data does not provide a full picture of the form of minimum down payments (pledged assets or pledged savings from second or third pillar observable; source of cash down payments not known).
  - Reported property values used to calculate LTVs might be based on optimistic valuations (e.g., low capitalization rates for investment-property lending or upper ends of valuations based on hedonic models).
  - Median maturity of an owner-occupied mortgage loan is rather short (6 years) and non-amortization of first mortgages is common; mortgages are frequently rolled over and new loan LTV is based on appraisal value, which may be higher than original purchase value. A legal requirement to keep the LTV unchanged for 5 years applies.
- Affordability compliance (Figure 10 summary):
  - More than half of mortgages in the owner-occupied segment do not meet prudent affordability criteria.
  - Two thirds of investment property mortgages do not meet prudent affordability criteria.
  - Affordability criteria used for owner-occupied segment: (i) affordability threshold of 38 percent of the sustainable net income; (ii) stressed costs calculated by a 5 percent imputed mortgage rate and 1 percent amortization; and building-related maintenance costs of 0.8 percent of the house value.
  - For investment property, the affordability threshold is 100 percent of the net rental income.
  - These criteria are set out in a recent FINMA Guidance, published in May 2025 after the main FSAP mission.

### Systemic risk context and policy implications
- Recent relaxation of the macroprudential stance contrasts with currently heightened real estate related risks and, together with the recent reduction of policy rates, has the potential to further increase mortgage market related vulnerabilities.
- Counterfactual analysis: the report performs counterfactual analysis of additional macroprudential measures, based on an assumed broadening of the macroprudential toolkit along the lines of rule-based affordability limits and improved data collection.
- Suggested policy priorities (grounded in the analysis above):
  - Establish quantitative affordability limits (LTI, DSTI, or explicit component caps) applicable to new and existing loans, with clear treatment for non-compliance (e.g., penalty risk weights).
  - Enhance borrower-level data collection to include all debts and debt service, loan amortization modalities, and CRE appraisal practices.
  - Consider legal and institutional arrangements to enable borrower-based macroprudential measures for new loans.
  - Monitor and reassess self-regulation effectiveness and upgrade qualitative guidelines to minimum requirements where necessary.

*SWITZERLAND  INTERNATIONAL MONETARY FUND*

### 54.      Available data does not allow assessing bank compliance with the linear amortization

### 54.      Available data does not allow assessing bank compliance with the linear amortization requirement over 10 and 15 years, respectively. However, evidence from household micro data (discussed in the next section) suggests that all non-amortizing loans in the representative sample have LTV ratios below 66 percent LTV.

### Deviations from affordability benchmarks and LTV caps — key findings
- About half of new mortgage loans in the owner-occupied segment do not meet the respective “industry standards” used by FINMA as supervisory benchmarks.
- Around 60 percent of new mortgage loans in the investment property segment do not meet the respective “industry standards” used by FINMA as supervisory benchmarks.
- The share of new mortgage loans that deviate from prudent affordability benchmarks is high and has further increased over time.
- All non-amortizing loans in the representative household micro-data sample have LTV ratios below 66 percent LTV.
- The Basel III final risk weights for mortgage loans were calibrated assuming an average 15 percent non-compliance at the single loan level regarding affordability and prudent valuation (85 percent weight on preferential risk weights and 15 percent weight on penalty risk weights).

### Distribution across banks and evolution of lending standards
- Looser lending practices with respect to affordability benchmarks are widespread across banks, not confined to a few institutions.
- The distribution of the ratio of loans that deviate from prudent lending practices with respect to affordability criteria (to total mortgage loans granted in the respective year) shows elevated median, 25th/75th percentiles, and wide ranges across banks.
- Results point to somewhat looser lending standards applied over time despite qualitative self-regulation guidelines.

### Assessment and macroprudential recommendations
- The widespread deviation from industry standards for affordability and supervisory benchmarks warrants macroprudential action.
- Authorities responsible for financial stability need to:
  - Address deviation from prudent lending standards.
  - Ensure that additional affordability-related credit risks incurred by banks are adequately covered by capital.
- Policy options discussed:
  - Instead of adjusting risk weights or FINMA applying Pillar 2 add-ons, the SNB should explore increasing the sectoral CCyB to a corresponding extent to ensure risks are adequately covered by capital.
- Monitoring and borrower-based measures:
  - Authorities should closely monitor loan-to-value ratios in view of the recent loosening of the LTV cap in the investment-property segment.
  - Should the share of mortgage loans with LTV ratios above 75 percent significantly increase, authorities should tighten the LTV cap either via amendment of the minimum requirements of the self-regulation or by resorting to binding borrower-based measures.
- Tax recommendation:
  - Reiterate the 2019 FSAP recommendation that the tax deductibility of mortgage interest payments should be abolished to remove incentives for higher mortgage debt.
  - Rationale: Current tax regime incentivizes households to take on more mortgage debt, contributing to high mortgage debt and household vulnerability to adverse macroeconomic and interest rate shocks.
  - Policy window: Renewed low interest environment represents a window of opportunity for reform.

### Rationale for borrower-based measures and role of DSTI caps
- Capital-based measures, even sectoral CCyB, are not well suited to reduce bank incentives to engage in risky mortgage lending practices; borrower-based measures can address imbalances in real estate markets and improve borrower resilience.
- A DSTI cap would primarily be targeted at limiting PDs, while the existing complementary LTV cap focuses on limiting LGD.
- A DSTI cap is conceptually aligned with the way affordability is measured in the qualitative guidelines of the Swiss Banking Association, though by design it does not capture property maintenance costs.

### Counterfactual analysis design and model
- Analysis tool: IDHBS+ (Integrated Dynamic Household Balance Sheet+) developed by Gross and Poblacion (2017) and Gross et al. (2022).
- Counterfactual DSTI cap evaluated: 30 percent of borrower’s income, assessed under both baseline and adverse scenarios.
- DSTI cap modeled on actual (not stressed) debt service; households with DSTI ratio above 30 percent are assumed to receive smaller loans to meet the 30 percent cap while keeping resulting LTV ratios unchanged compared to observed LTVs.
- The model generates individual households’ simulated PDs, LGDs, and loss rates, linked to bank mortgage portfolios to assess impact on bank capital; also investigates impact on mortgage lending volumes and composition based on Hypo B data.
- Alternative DSTI caps considered in analysis: 30 percent, 25 percent, and 20 percent (presentation focuses on 30 percent).

### Distributional context and which borrowers would be affected
- DSTI distribution (2017–2024 lending flows):
  - Owner-occupied segment: 5 percent of mortgage lending had DSTI ratios above 30 percent.
  - Investment property segment (households): 39 percent of mortgage lending had DSTI ratios above 30 percent.
  - Investment property segment (commercial borrowers): 35 percent of mortgage lending had DSTI ratios above 30 percent.
- A 30 percent DSTI cap would mainly affect borrowers with high LTI multiples:
  - Owner-occupied segment: 91 percent of the affected borrowers have debt exceeding 10 times their sustainable net income per year (highest LTI bucket shown).
  - Investment-property segment: when scaled to be comparable to owner-occupied assumptions, the DSTI cap would broadly have similar effects; 38 percent in the highest LTI bucket under the scaled distribution.
- The analysis notes it is partial as it does not consider borrowers’ overall debt situation (unlike microdata-based counterfactuals).

### Time variation and interest-rate fixation effects
- The DSTI cap would have been slightly more binding during recent monetary tightening:
  - Owner-occupied segment: affected share would have increased from 5 percent average over 2017–24 to levels close to 10 percent in 2023.
  - Household investment-property segment: affected share would have increased from 40 percent average over 2017–24 to around 60 percent at peak policy rates in 2023.
- DSTI cap binding by interest-rate fixation (owner-occupied segment, 2017–2024 lending flows):
  - Fixed-rate mortgage loans: 3 percent would have been affected by a 30 percent DSTI cap (3 percent on average).
  - Variable-rate mortgage loans: 9 percent would have been affected by a 30 percent DSTI cap (9 percent on average).
  - Mixed fixed/variable-rate loans: 6 percent would have been affected by a 30 percent DSTI cap (6 percent on average).
- Implication: DSTI cap tends to be more binding for variable-rate loans and favors fixed-rate loans, reducing newly borrowing households’ interest rate exposures.

### Role of household wealth
- Inclusion of financial wealth is important for PD analysis given high household wealth in Switzerland; wealth is a possible mitigating factor for low default rates.
- No representative-sample analysis including household financial wealth is noted to be available so far.

### Quantitative impact from IDHBS+ counterfactual (30 percent DSTI cap)
- Probability of default (PD):
  - The DSTI cap leads to a lower starting-point PD for the mortgage portfolio and a lower PD increase during the adverse scenario.
  - Specifically, a DSTI cap at 30 percent reduces the increase in mortgage PDs by about a third under the adverse supply-shock inflationary scenario.
- Loss given default (LGD):
  - The DSTI cap has a very minor positive impact on LGDs based on the empirical relationship used.
- Expected loss and capital impact:
  - Capital ratios of Swiss banks gain 37 basis points of risk-weighted assets under the counterfactual setup with a 30 percent DSTI cap in place compared to the capital ratios resulting from the supply-shock adverse scenario (comparison of peak capital reduction under the adverse scenario).
- Impact on mortgage lending volume:
  - Mortgage lending volume would have been 1.6 percent lower for the entire period 2017 to 2024 under the counterfactual 30 percent DSTI cap.
  - Segmental average reductions noted:
    - SORRE loans: 3 percent reduction on average.
    - IPRRE HH loans: 4 percent reduction on average.
    - IPRRE Commercial loans: 0.4 percent reduction on average.
    - Total SORRE and IPRRE: 1.6 percent reduction on average.

### Net assessment and policy recommendation on DSTI
- The counterfactual analysis for a 30 percent DSTI cap illustrates net benefits:
  - Limited costs in terms of impact on lending and access compared to the status quo.
  - Strengthens resilience of banks and households by reducing expected losses and improving lending composition from an aggregate credit risk perspective.
- Further design considerations:
  - Specific policy design could benefit from international experience; IMF Macroprudential Database includes 67 countries with active DSTI measures as of end-2023.
  - European peers with DSTI caps often combine them with LTV caps and maturity limits (examples cited in the analysis include Austria, Estonia, Finland, Latvia, Lithuania, Malta, Slovenia, and Slovakia).
  - DSTI definitions vary (numerator: some add interest rate shock for stressed debt service; denominator: some deduct minimum living expenses from income).
  - Experience with DTI or LTI in countries such as Denmark, Ireland, and the United Kingdom can provide additional insights.

*Italic: Content summarized from the supplied IMF chapter/section PDF content.*

### 73.      Stressed DSTI caps and DTIs can be seen as close substitutes. International experience

### Stressed DSTI caps and DTIs can be seen as close substitutes.

### Substitutability and international experience
- Macroprudential authorities typically apply only one income-based measure: either a DSTI cap based on stressed or actual interest rates or a DTI/LTI cap (see CGFS 2023).
- Both types of income-based measures strengthen borrower resilience and may have partly overlapping effects, as noted by the Bank of England on the UK experience with LTI and stressed DSTI caps.
- One exception is Slovakia that has both DSTI and DTI caps in place.
- Based on the IMF Macroprudential Database Dec. 2024 update, reporting measures were in place in 161 member countries as of end-2023.

### Design options and data constraints
- Stressed DSTI or DTI caps can be designed to account not only for income but also for financial wealth (for example, the DTI/LTI cap in Denmark).
- Calibrating borrower-based measures to incorporate financial wealth requires that the respective data be available to banks and supervisors; this data is not available in Switzerland and thus limits this design option in practice.
- DSTI caps are often combined with maturity limits to avoid circumvention via maturity extensions.

### Mitigations for potential exclusion of suitable borrowers
- Potential concerns that DSTI caps could exclude suitable borrowers (for example, young or low-income borrowers) could be mitigated via careful design of speed limits that grant exceptions.
- High-net-worth individuals for whom DSTI caps bind may be limited in number but could in principle be included within speed limits as well.
- Conversely, policy can be tightened for more speculative borrowers (for example, the investment property RRE segment), which the partial Hypo B data analysis shows to be riskier than the owner-occupied RRE segment in terms of affordability.

### Current Swiss market assessment and policy implications
- The current situation in the Swiss real estate and mortgage market calls for adoption of additional measures.
- Monetary easing and initiatives aimed at easing affordability criteria of new borrowers through cantonal guarantees may further stimulate credit demand and reinforce a spiral of unsustainable lending standards, house price increases, and systemic risk build-up.
- While housing affordability requires a holistic policy approach, including tax and land reforms, borrower-based measures could be an additional tool to enforce prudent lending practices in Switzerland.

*SWITZERLAND INTERNATIONAL MONETARY FUND*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2025/english/1cheea2025007-source-pdf.pdf_