## 1. Positive neutral Countercyclical Capital Buffers (CCyB) have seen growing adoption in recent years

## Source details

**Canonical URL:** [1. Positive neutral Countercyclical Capital Buffers (CCyB) have seen growing adoption in recent years](https://www.imf.org/-/media/files/publications/selected-issues-papers/2025/english/sipea2025154.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/selected-issues-papers/2025/english/sipea2025154.pdf.md)
- [Structured JSON version](/-/media/files/publications/selected-issues-papers/2025/english/sipea2025154.pdf.json)

---

### Summary of key developments and international context
- Bank of England pioneered a CCyB framework with a positive neutral rate in 2016.
- A total of 23 countries have introduced CCyB regimes with explicit cycle-neutral rates by December 2025.
- Most adopters are in Central, Eastern and Southeastern Europe (CESEE) and other European countries.
- International evidence (including COVID-19) shows countries with active releasable buffers maintained stronger bank lending and stability during stress (BCBS, 2022b; Couaillier et al., 2022a, 2022b).

### Defining differences between positive neutral and zero-neutral CCyB
- Positive neutral CCyB features an ex-ante announced positive neutral rate that acts as a default anchor:
  - CCyB requirement expected to be set at the positive neutral rate for most of the time, or above it when cyclical systemic risks are heightened.
  - Exceptions: periods of financial stress (when CCyB is released) and brief recovery periods after severe stress (when the rate remains below the neutral rate).
  - Banks can expect the CCyB requirement to be re-raised to the positive neutral level soon after a stress episode (unlike zero-neutral regimes).
- Higher average CCyB rate through the cycle implies more room for bold releases:
  - Positive neutral rate is a higher buffer requirement (not a higher minimum capital requirement), so banks can expect the buffer to be released if cyclical systemic risk occurs.

### Albania-specific status and recent settings
- CCyB changes:
  - Raised to 0.25 percent effective from June 2025.
  - Raised to 0.5 percent effective from December 2025.
- Combined Buffer Requirement (CBR) applies on top of a Pillar 1 requirement of 12 percent (compared with 8 percent under the EU Single Rulebook).
- Leverage and MREL:
  - Albanian banks must maintain a 5.75 percent leverage ratio.
  - Some banks use CET1 capital to comply with MREL due to lack of eligible liabilities.
- System-wide capital and performance:
  - Total capital ratio exceeding 19 percent (system-wide).
  - Non-performing loan (NPL) ratio at 4.1 percent as of 2024.
  - RoA and RoE stood above historical averages in 2023 and 2024.
- Credit and macro-financial indicators:
  - Credit growth rates were in the double digits in 2023 and 2024, with particularly rapid increases in real estate lending.
  - Output gap estimated to be positive; economic sentiment and consumer confidence at or above historic averages.

### Benefits and costs of a positive neutral CCyB
- Benefits:
  - Reduces under-calibration from mismeasurement of cyclical risks.
  - Increases distance to the CBR, addressing buffer usability concerns.
  - Strengthens resilience against shocks not originating from domestic cyclical imbalances (including real-estate and sovereign risks).
- Costs:
  - Potential reduction in bank lending, impact on funding costs, and profitability.
  - Activation costs can be non-linear: more profitable banks can meet higher requirements via retained earnings, lowering deleveraging needs.
  - Earlier and more gradual activation (characteristic of positive neutral regimes) tends to lower overall activation costs.

### Modular six-part framework to explore positive neutral CCyB
- Building Block 1: Legal feasibility
  - Albania’s Regulation No. 41/2019 and the Bank of Albania’s 2018 CCyB methodology provide flexible inputs and a composite early warning indicator (scaled 0–1, mapped into CCyB benchmark rates via bucketing).
  - Introducing a positive neutral CCyB would require only minor regulatory changes to mandate the Bank of Albania to set the positive neutral target and operationalize build-up, release, and replenishment.
- Building Block 2: Macro-financial conditions
  - Evaluate system-wide solvency, stability, profitability; Albania’s conditions assessed as benign for introduction given robust profitability, capital headroom, and favorable growth.
- Building Block 3: Costs and benefits
  - Conduct cost-benefit analysis combining model outputs and qualitative assessment; for Albania a qualitative assessment indicates net benefits given current capital headroom and profitability.
- Building Block 4: Design considerations (four focal issues)
  - (i) Setting the country-specific positive neutral target rate.
  - (ii) Deciding build-up speed: gradual enough to minimize adverse lending effects.
  - (iii) Revisiting release methodology to ensure timely freeing of capital to prevent deleveraging.
  - (iv) Designing analytical framework for raising CCyB above neutral.
- Building Block 5: Integration in the macroprudential framework
  - Ensure no overlap with other micro- and macroprudential capital requirements.
  - Assess capital-accretive introduction versus capital neutrality (e.g., reducing Pillar 1).
- Building Block 6: Communication strategy
  - Foster understanding and acceptance by the banking sector and public; provide principles for CCyB increases, releases and replenishment to support transparency while preserving discretion.

### Calibration methods and recommended approach for Albania
- International practice:
  - Positive neutral rates range between 0.5 percent (Greece) and 2 percent (Netherlands, Poland, Sweden, United Kingdom).
  - Most countries set the target rate at 1 percent (examples: Australia, Czech Republic, Estonia, Georgia, Hungary, Latvia, Lithuania, Slovenia, Spain).
  - International practice tends to set positive neutral CCyB at 50 bps increments; actual settings done at 25 bps increments.
- Methods to set positive neutral CCyB:
  - Backward-looking methods: historical losses.
  - Forward-looking methods: stress tests, structural models, risk indicators.
  - Expert judgement.
  - Most European countries combine these approaches.
- Recommendation for Albania:
  - Calibrate using a combination of backward- and forward-looking methods, supported by expert judgement.
  - Rationale: Albania’s emerging market exposure to capital flows and FX shocks implies backward-looking history may be less informative; supplement with forward-looking assessments of (i) banks’ current risk exposures; (ii) available buffers; and (iii) projected changes under scenarios including capital flow reversals and adverse FX outcomes.
  - Analysis indicates a target range of around 100-200 basis points for the neutral rate.

### Operationalization preconditions and practical design choices
- Preconditions to assess:
  - Legal feasibility.
  - Macro-financial conditions: profitability, solvency, liquidity, and absence of current economic/financial stress.
  - Costs and benefits: benefits should exceed costs given macro-financial conditions.
- Practical choices:
  - Make the positive neutral rate explicit and stable to serve as an anchor.
  - Choose gradual build-up speed to avoid negative lending effects.
  - Define release thresholds and replenishment mechanics to ensure timely capital availability during stress.
  - Develop criteria for increasing CCyB above the neutral rate when warranted.

*Source: sipea2025154 - 1. Positive neutral Countercyclical Capital Buffers (CCyB) have seen growing adoption in recent years (PDF chapter).*

### 17. Backward-looking analysis assesses which CCyB size would have been useful in past stress episodes

### Backward-looking assessment: applicability to Albania
- Historical losses method limited for Albania: despite system-wide stress episodes (2002, 2008–10, 2013–14) the banking sector did not observe an overall loss-making year, and system-wide total capital ratio remained in the 15–18 percent range during 2008–14 (BoA 2014).
- Losses-to-buffer approach (De Nora et al. 2024) is more promising: calibrates target CCyB rate via a panel quantile model identifying portion of banks’ losses coverable by the positive neutral rate while controlling for bank-specific characteristics and macro developments.

### Losses-to-buffer methodology (Box 1 summary)
- Uses a location-scale quantile model (Machado and Santos Silva 2019) integrated into a local projection framework (Jordà 2005).
- Distinguishes average (location) and quantile-specific (scale) effects; includes bank- and time-fixed effects and prevents quantile crossing.
- Addresses small-sample bias in time dimension with a split-sample jackknife where needed (Dhaene and Jochmans, 2015).
- Calibration: average of statistically significant, non-positive time-fixed effects at selected quantiles and one year horizon, rescaled by banks’ average risk weights, translates unexplained system-wide losses into capital requirements.
- Rationale: focuses on systemic losses, minimizes overlap with other prudential tools, and allows policymakers to select quantiles reflecting tolerance for covering losses.

### Albania-specific estimation and results
- Data and sample:
  - 11 Albanian banks.
  - Period: 2006Q1 to 2019Q4 (quarterly).
  - Controls: pre-tax RoA, net interest margin, cost to income ratio, impairment over total assets, net loans over total assets, risk weighted assets over total assets, tangible equity over tangible assets, Tier 1 capital ratio, log of total assets, GDP growth, credit growth, and BoA policy rate.
  - Bank-specific and time fixed effects included.
  - Positive neutral CCyB calibrated based on sum of quarterly estimated coefficients over each year.
- Quantile estimation and calibrated neutral rates:
  - 50th percentile: 0.4 percent
  - 25th percentile: 1 percent
  - 10th percentile: 2.4 percent

### Robustness: CESEE panel results
- Panel: CESEE countries over 2006–2019 at annual frequency; sample includes 75 banks.
- Jackknife correction applied.
- Positive neutral rates for full CESEE sample including Albania:
  - 50th percentile: 0.53 percent
  - 25th percentile: 1.22 percent
  - 10th percentile: 2.3 percent
- Equivalent rates excluding Albania:
  - 50th percentile: 0.52 percent
  - 25th percentile: 1.16 percent
  - 10th percentile: 2.01 percent
- Conclusion: calibration results are stable across samples.

### Forward-looking stress-test implications for CCyB
- Forward-looking analysis should assess multiple severe yet plausible scenarios; absent multi-scenario approach use recent system-wide stress tests as starting point.
- Application to Albania (BoA stress test 2024):
  - System-wide total capital ratio projected to fall from 20.2 percent (Q3 2024) to 18.4 percent (Q4 2026) under the adverse scenario.
  - Aggregate resilience high, but some individual banks projected to breach Pillar 2 and Pillar 1 requirements.
- Stress-test implication:
  - An additional capital buffer requirement of 310 basis points would ensure all systemic banks remain compliant with combined Pillar 1 and 2 requirements.
  - Stress-test results provide a reference point and suggest the 50th-percentile losses-to-buffer estimate may be too low for severe shocks.

### Recommended positive neutral CCyB range and policymaker choices
- Combined evidence points to a positive neutral CCyB broadly between 100bps and 200 bps.
- Losses-to-buffer interpretation:
  - Policymakers select percentile based on risk tolerance.
  - 25th percentile or 10th percentile imply target rate in the 100bps–200bps dimension, similar to CESEE peers.
  - 50th percentile yields lower estimate (e.g., 0.4 percent for Albania) which may be insufficient for large unexpected losses.
- Stress-testing suggests a higher indication (around 300bps), closer to the 10th and 25th percentiles than the 50th percentile.

### Build-up speed for the positive neutral CCyB
- Transition should consider banks’ capital generation capacity; stepwise approach or forward guidance recommended.
- International practice: most adopters used a stepwise approach or introduction timeline of more than 12 months; high-rate countries (2 percent) often gave up to two years.
- Albania-specific capital generation capacity:
  - Estimated annual capital generation capacity of approx. 80-90 basis points in 2025 and 2026.
  - Return on assets in 2024: 1.9 percent, more than 50 percent above the previous 10-year average.
  - Under BoA stress test 2024 baseline, total capital ratio projected to increase by 89 bps in 2025 and by 87 bps in 2026, assuming full profit retention.
- Implication: healthy organic capital generation supports a swift build-up, especially if partially capital-neutral build-up options are used.

### Release methodology: recommendations and current limitations
- Higher buffer increases importance of full or partial releases; release methodology needs revisiting to provide flexibility and clearer signals while committing to replenishment.
- Current BoA release methodology (Bank of Albania 2018) focuses on evidence of loss materialization (NPL increases, capital erosion, losses, capital constraints), implying likely late releases after several months into stress.
- Recommendation: make release methodology more flexible to allow early response with partial or full CCyB releases from a higher CCyB level; add explicit reference to commitment to replenish positive neutral buffer and indicate envisaged duration of release and timeline for replenishment.

### Going beyond the neutral rate: differentiation and indicator thresholds
- Methodology should differentiate neutral environment and elevated cyclical systemic risk that warrants higher rate; guided discretion with clarifications supports transparency and predictability.
- For Albania: use existing Complementary Early Warning Indicator with recalibrated threshold to guide going beyond neutral rate.
  - Suggested threshold shift: from 0.5 to a higher threshold (subject to further analysis).
  - Ex-post performance indicates indicator values around 0.7 provided useful early warnings in the pre-GFC period.

### Integration in the macroprudential framework (Building Block 5)
- Positive neutral CCyB introduction options:
  - Capital-accretive introduction (higher capital requirements at adoption) or capital-neutral introduction (offsetting reductions elsewhere).
  - International experience: most EU activations followed capital-accretive default setting; capital neutrality considered in specific circumstances.
- No overlaps identified between positive neutral CCyB risk coverage and:
  - (i) D-SIB buffer (structural systemic risks).
  - (ii) CCoB: set at 2.5 percent under Basel III.
  - (iii) Minimum capital requirements.
- No overlap with Pillar 1 or Pillar 2 (Pillar 2 set via BoA SREP without explicit link to stress test results).

*Source: sipea2025154 - 17. Backward-looking analysis assesses which CCyB size would have been useful in past (PDF chapter).*

### 34. Integration options, trade-offs, and communication for a positive neutral CCyB

### Integration options for introduction
- Capital-neutral introduction
  - Would imply partial ‘swapping’ of part of the Pillar 1 requirement into a positive neutral CCyB (example: Armenia in 2023).
  - Drawback: would not leverage current favorable profitability to build buffers through earnings retention.
- Partially capital-accretive build-up
  - Example: build-up of a positive neutral CCyB in excess of 100 bps combined with a 100-bps reduction in the Pillar 1 requirement (to 11 percent).

### Regional peer experience and historical context
- Several CESEE countries had higher Pillar 1 requirements than 8 percent before joining the EU and introducing the EU Single Rulebook.
- Most CESEE EU peers have higher combined buffer requirements than Albania currently.
- Bulgaria introduced a SyRB simultaneously with Pillar 1 reduction in compliance with the EU Single Rulebook (capital-neutral).
- Transitions for most EU CESEE countries happened in 2014—before positive neutral CCyB regimes were common.
- EU member states that opted for capital-neutral build-up: Estonia and the Netherlands used offsets involving SyRB adjustments.

### Merits and trade-offs for Albania
- Merits of partially capital-accretive or capital-neutral introduction:
  - Albania has a gap vis-à-vis EU average in economic development and financial deepening; avoiding a large short-term spike in overall capital requirements may be warranted.
  - Significantly higher requirements could create competitiveness issues over time.
- High bank profitability could support fully capital-accretive transition, but merits exist for partial or full capital-neutral introduction.

### Communication strategy (Building Block 6)
- Albania’s current communication framework:
  - Quarterly CCyB decisions published on the BoA website with rationale including the credit-to-GDP gap and complementary early warning indicators.
  - Cyclical systemic risks and CCyB rate discussed in other BoA reports.
- Recommendation: transition to positive neutral CCyB should pay special attention to explaining its functioning to the general public, given the level of financial education and trust in financial institutions.

### Conclusion and calibration guidance
- Introducing a positive neutral CCyB regime would:
  - Strengthen BoA’s ability to react to stress episodes.
  - Improve usability of capital buffers without unduly restricting bank credit provision.
  - Maintain a higher CCyB through the cycle compared to the existing regime.
- Calibration guidance:
  - Use a suite of backward- and forward-looking methods combined with expert judgement.
  - Analysis points to a suitable range of around 100–200 bps for the positive neutral rate in Albania as a starting point.
  - Timely consultation of key stakeholders is essential to facilitate the process and build awareness.

*Source: sipea2025154 (IMF Selected Issues Paper).*

### 1.      Positive neutral Countercyclical Capital Buffers (CCyB) have seen growing adoption in

### Positive neutral Countercyclical Capital Buffers (CCyB) have seen growing adoption in recent years

### Summary of key developments and international context
- The Bank of England pioneered a CCyB framework with a positive neutral rate in 2016.
- A total of 23 countries have introduced CCyB regimes with explicit cycle-neutral rates by December 2025.
- Most adopters are in Central, Eastern and Southeastern Europe (CESEE) and other European countries.
- International evidence, including from the COVID-19 pandemic, shows that countries with active releasable buffers maintained stronger bank lending and stability during periods of financial stress (BCBS, 2022b; Couaillier et al., 2022a, 2022b).

### Two defining differences between positive neutral and zero-neutral CCyB
- Positive neutral CCyB provides an ex-ante announced positive neutral rate that acts as a default anchor for bank expectations:
  - CCyB requirement is expected to be set at the positive neutral rate for most of the time, or above the neutral rate in case of heightened cyclical systemic risks.
  - Exceptions: periods of financial stress (when CCyB is released) and brief recovery periods after severe stress (when the rate remains below the neutral rate).
  - Unlike a zero-neutral regime, banks can expect the CCyB requirement to be re-raised to the positive neutral level soon after a stress episode.
- Higher average CCyB rate through the cycle implies more room for bold releases if warranted:
  - The positive neutral rate is a higher buffer requirement (not a higher minimum capital requirement), so banks can expect the buffer to be released if cyclical systemic risk occurs.

### Albania-specific status and recent settings
- Albania raised the CCyB to 0.25 percent effective from June 2025, and 0.5 percent effective from December 2025.
- The Combined Buffer Requirement (CBR) applies on top of a Pillar 1 requirement of 12 percent (compared with 8 percent under the EU Single Rulebook).
- Albanian banks must maintain a 5.75 percent leverage ratio.
- Some banks use CET1 capital to comply with MREL due to lack of eligible liabilities.
- System-wide capital adequacy:
  - Total capital ratio exceeding 19 percent (system-wide).
  - Non-performing loan (NPL) ratio at 4.1 percent as of 2024.
  - Return on assets (RoA) and return on equity (RoE) stood above historical averages in 2023 and 2024.
- Credit and macro-financial signs:
  - Credit growth rates were in the double digits in 2023 and 2024, with particularly rapid increases in real estate lending.
  - Output gap is estimated to be positive; economic sentiment and consumer confidence indicators at or above historic averages.

### Benefits and costs of a positive neutral CCyB
- Benefits:
  - Reduces potential under-calibration from mismeasurement of cyclical risks.
  - Increases distance to the CBR and therefore addresses buffer usability concerns.
  - Strengthens resilience against shocks not originating from domestic cyclical imbalances (including real-estate and sovereign risks).
- Costs:
  - Potential reduction in bank lending, impact on funding costs, and profitability.
  - Costs of activation can be non-linear: more profitable banks can meet higher requirements via retained earnings, lowering deleveraging needs.
  - Earlier and more gradual activation (characteristic of positive neutral regimes) tends to lower overall activation costs.

### Building blocks for exploring a positive neutral CCyB (modular six-part framework)
- Building Block 1: Legal feasibility
  - Assess whether the legal basis can accommodate a positive neutral CCyB and relevant indicators/timing.
  - Albania’s Regulation No. 41/2019 and the Bank of Albania’s 2018 CCyB methodology already provide flexible inputs and a composite early warning indicator (scaled 0–1, mapped into CCyB benchmark rates via bucketing).
  - Introducing a positive neutral CCyB would require only minor regulatory changes to mandate the Bank of Albania to set the positive neutral target and operationalize build-up, release, and replenishment.
- Building Block 2: Macro-financial conditions
  - Evaluate system-wide solvency, stability, profitability, and whether macro conditions are broadly stable.
  - Albania’s current macro-financial conditions are assessed as benign for introduction given robust bank profitability, capital headroom, and favorable growth.
- Building Block 3: Costs and benefits (cost-benefit analysis)
  - Conduct cost-benefit analysis combining model outputs and qualitative assessment.
  - For Albania, a qualitative assessment indicates net benefits given current capital headroom and profitability.
- Building Block 4: Design considerations (four focal issues)
  - (i) Setting the country-specific positive neutral target rate to ensure sufficient loss absorption capacity.
  - (ii) Deciding on build-up speed: gradual enough to minimize adverse lending effects.
  - (iii) Revisiting release methodology to ensure timely freeing of capital to prevent deleveraging.
  - (iv) Designing the analytical framework to determine conditions for raising the CCyB above neutral.
- Building Block 5: Integration in the macroprudential framework
  - Ensure no overlap in risk coverage with other micro- and macroprudential capital requirements.
  - Assess capital-accretive introduction (higher capital requirements at adoption) versus capital neutrality (introducing positive neutral CCyB while reducing another requirement, e.g., Pillar 1).
- Building Block 6: Communication strategy
  - Foster understanding and acceptance by the banking sector and public.
  - Provide principles for CCyB increases, releases and replenishment to support transparency and bank capital planning while preserving discretion.

### Calibration methods and recommended approach for Albania
- International practice:
  - Positive neutral rates internationally range broadly between 0.5 percent (Greece) and 2 percent (Netherlands, Poland, Sweden, United Kingdom).
  - Most countries set the target rate at 1 percent (examples: Australia, Czech Republic, Estonia, Georgia, Hungary, Latvia, Lithuania, Slovenia, Spain).
  - International practice tends to set positive neutral CCyB at 50 bps increments (e.g., 100 bps, 150 bps); actual settings done at 25 bps increments.
- Methods used to set positive neutral CCyB:
  - Backward-looking methods: historical losses.
  - Forward-looking methods: stress tests, structural models, risk indicators.
  - Other methods: expert judgement.
  - Most European countries combine these approaches.
- Recommendation for Albania:
  - Calibrate the buffer using a combination of backward- and forward-looking methods, supported by expert judgement.
  - Rationale: Albania’s status as an emerging market with exposure to capital flows and FX shocks implies backward-looking history may be less informative; supplement with forward-looking assessments of (i) banks’ current risk exposures; (ii) available buffers; and (iii) projected changes under scenarios including capital flow reversals and adverse FX outcomes.
  - The analysis indicates a target range of around 100-200 basis points for the neutral rate.

### Operationalization considerations and preconditions
- Preconditions to assess before introduction:
  - Legal feasibility.
  - Macro-financial conditions: profitability, solvency, liquidity, and absence of current economic/financial stress.
  - Costs and benefits: benefits should exceed costs given macro-financial conditions.
- Practical design choices:
  - Make the positive neutral rate explicit and stable over time to serve as an anchor.
  - Choose build-up speed that is gradual to avoid negative lending effects.
  - Define release thresholds and replenishment mechanics to ensure timely capital availability during stress.
  - Develop criteria for increasing CCyB above the neutral rate when warranted.

*Source: sipea2025154 - 1. Positive neutral Countercyclical Capital Buffers (CCyB) have seen growing adoption in recent years (PDF chapter).*

### 17.      Backward-looking analysis assesses which CCyB size would have been useful in past

### 17.      Backward-looking analysis assesses which CCyB size would have been useful in past stress episodes

### Backward-looking assessment: overview and applicability
- Historical losses method is of limited usefulness for Albania: despite three episodes of system-wide financial stress (2002, 2008–10, 2013–14) the banking sector did not observe an overall loss-making year, and the system-wide total capital ratio remained in the 15–18 percent range during 2008–14 (BoA 2014).
- Losses-to-buffer approach is more promising for Albania (De Nora et al. 2024). It calibrates the target CCyB rate based on a panel quantile model that identifies which portion of banks losses could be covered by the positive neutral rate while controlling for bank-specific characteristics and macroeconomic developments.

### Losses-to-buffer approach: methodology (Box 1 summary)
- Relies on a location-scale quantile model (Machado and Santos Silva 2019) integrated into a local projection framework (Jordà 2005).
- Distinguishes average (location) and quantile-specific (scale) effects of covariates; accommodates bank- and time-fixed effects and prevents quantile crossing.
- Conditional quantile of bank profitability at horizon h specified with bank-specific controls X, macro-financial variables Y, bank fixed effects, and time fixed effects; scale and location parameters capture quantile deviations and averages respectively.
- Small-sample bias in the time dimension addressed with a split-sample jackknife procedure where needed (Dhaene and Jochmans, 2015).
- Calibration: average of statistically significant, non-positive time-fixed effects at selected quantiles and one year horizon, rescaled by banks’ average risk weights, translates unexplained system-wide losses into capital requirements.
- Rationale: focuses on systemic (not idiosyncratic) losses, minimizes overlap with other prudential tools by controlling for bank and macro factors, and allows policymakers to select quantiles reflecting tolerance for covering losses.

### Losses-to-buffer: Albania-specific estimation and results
- Data and sample:
  - 11 Albanian banks.
  - Period: 2006Q1 to 2019Q4 (quarterly).
  - Controls included: pre-tax RoA, net interest margin, cost to income ratio, impairment over total assets, net loans over total assets, risk weighted assets over total assets, tangible equity over tangible assets, Tier 1 capital ratio, log of total assets, GDP growth, credit growth (proxy for domestic cyclical systemic risk), and BoA policy rate.
  - Bank-specific and time fixed effects included.
  - Positive neutral CCyB calibrated based on sum of quarterly estimated coefficients over each year.
- Quantile estimation:
  - Model estimated on the 10th, 25th and 50th percentiles of the conditional return on asset distribution.
  - For each percentile, model estimated one year ahead to account for CCyB implementation lag.
  - Calibrated positive neutral buffer rates for Albania:
    - 0.4 percent (50th percentile)
    - 1 percent (25th percentile)
    - 2.4 percent (10th percentile)

### Robustness check: CESEE panel results
- Panel: CESEE countries sample over 2006–2019 at annual frequency; data from S&P Capital IQ Pro and BoA for Albania; overall sample includes 75 banks.
- Jackknife correction applied due to larger N than T.
- Positive neutral rates for full CESEE sample including Albania:
  - 0.53 percent (50th percentile)
  - 1.22 percent (25th percentile)
  - 2.3 percent (10th percentile)
- Equivalent rates for CESEE sample excluding Albania:
  - 0.52 percent (50th percentile)
  - 1.16 percent (25th percentile)
  - 2.01 percent (10th percentile)
- Conclusion: calibration results are stable across the two estimation samples.

### Forward-looking analysis: stress-test modeling approach
- Forward-looking analysis should assess multiple severe yet plausible scenarios; in absence of multi-scenario approach, use recent system-wide bank solvency stress tests as starting point.
- Application to Albania: use Bank of Albania stress test 2024 results.
  - System-wide total capital ratio projected to fall from 20.2 percent (Q3 2024) to 18.4 percent (Q4 2026) under the adverse scenario.
  - Aggregate resilience high, but some individual banks projected to breach Pillar 2 and Pillar 1 requirements.
- Stress-test implication for CCyB:
  - An additional capital buffer requirement of 310 basis points would ensure that all systemic banks remain compliant with the combined Pillar 1 and 2 requirements.
  - Note: results need not be mechanically translated into CCyB target rates of corresponding size; stress-test results provide a reference point suggesting the 50th-percentile losses-to-buffer estimate may be too low for severe shocks.

### Recommended positive neutral CCyB range and policymaker choices
- Combined evidence from backward- and forward-looking methods points to a positive neutral CCyB broadly between 100bps and 200 bps.
- Losses-to-buffer interpretation:
  - Policymakers can select percentile based on risk tolerance.
  - 25th percentile or 10th percentile imply target rate in the dimension of 100bps to 200bps, similar to CESEE peers.
  - 50th percentile yields lower estimate (e.g., 0.4 percent for Albania) which may be insufficient for large unexpected losses.
- Stress-testing approach suggests a somewhat higher indication (around 300bps), closer to the 10th and 25th percentiles than the 50th percentile.

### Build-up speed for the positive neutral CCyB
- Transition timeline should consider banks’ capital generation capacity to avoid strong impact on lending; stepwise approach or forward guidance recommended.
- International practice: most countries that adopted positive neutral CCyB used a stepwise approach or introduction timeline of more than 12 months; high-rate countries (2 percent) often gave up to two years.
- Albania-specific capital generation capacity:
  - Estimated annual capital generation capacity of approx. 80-90 basis points in 2025 and 2026.
  - Return on assets in 2024: 1.9 percent, more than 50 percent above the previous 10-year average.
  - Under the BoA stress test 2024 baseline scenario, total capital ratio projected to increase by 89 bps in 2025 and by 87 bps in 2026, assuming full profit retention.
- Implication: healthy organic capital generation supports a swift build-up of the positive neutral CCyB, especially if partially capital-neutral build-up options are used.

### Release methodology: recommendations and current limitations
- Higher buffer through the cycle increases importance of full or partial releases; release methodology needs revisiting to provide flexibility and indications to markets while committing to swift replenishment after release.
- Current BoA CCyB release methodology (Bank of Albania 2018) focuses on evidence of loss materialization (NPL increases, capital erosion, losses, capital constraints), implying likely late releases after several months into a stress episode.
- Recommendation: make release methodology more flexible to allow early response to system-wide financial stress with partial or full CCyB releases from a higher CCyB level; add explicit reference to commitment to replenish positive neutral buffer and indicate envisaged duration of release and timeline for replenishment.

### Going beyond the neutral rate: differentiation and indicator thresholds
- Methodology should differentiate standard risk environment (neutral CCyB) and elevated cyclical systemic risk that warrants higher rate; guided discretion with clarifications supports transparency and predictability.
- For Albania: use existing Complementary Early Warning Indicator with recalibrated threshold to guide going beyond neutral rate.
  - Suggested threshold shift: from 0.5 to a higher threshold (subject to further analysis).
  - Ex-post performance indicates indicator values around 0.7 provided useful early warnings in the pre-GFC period.

### Integration in the macroprudential framework (Building Block 5)
- Positive neutral CCyB can be introduced capital-accretively or partially/fully capital-neutrally. International experience: most EU activations followed capital-accretive default setting; capital neutrality considered in specific circumstances.
- No overlaps identified between positive neutral CCyB risk coverage and other requirements:
  - (i) D-SIB buffer: addresses structural systemic risks.
  - (ii) CCoB: set at 2.5 percent under Basel III and supports resilience once other buffers used/released.
  - (iii) Minimum capital requirements.
- No overlap with Pillar 1 (covers idiosyncratic credit, market, operational risks standardized) nor with Pillar 2 (set via BoA SREP without explicit link to stress test results).

*Source: sipea2025154 - 17.      Backward-looking analysis assesses which CCyB size would have been useful in past*

### 34.      However, the high Pillar 1 requirement in Albania raises questions about the medium-

### 34.      However, the high Pillar 1 requirement in Albania raises questions about the medium-

### Integration options for a positive neutral CCyB
- Capital-neutral introduction
  - Would imply a de facto partial ‘swapping’ of part of the Pillar 1 requirement into a positive neutral CCyB.
  - A similar approach was followed by Armenia in 2023 (see Central Bank of Armenia 2024).
  - Drawback: would not use the current favorable profitability of Albanian banks to build capital buffers through earnings retention in response to CBR changes.
- Partially capital-accretive build-up
  - Example: build-up of a positive neutral CCyB in excess of 100 bps combined with a 100-bps reduction in the Pillar 1 requirement (to 11 percent).

### Regional peer experience and historical context
- Several CESEE countries had historically higher Pillar 1 requirements than 8 percent before joining the EU and introducing the EU Single Rulebook.
- Most CESEE peers in the EU have higher combined buffer requirements than Albania at this point.
- Bulgaria introduced a SyRB simultaneously with the reduction of the Pillar 1 requirement in compliance with the EU Single Rulebook, i.e., in a capital-neutral way.
- These transitions happened in 2014 for most EU CESEE countries—before the “invention” of a positive neutral CCyB regime.
- Note: Among the EU member states that opted for a capital-neutral build-up, Estonia offset the introduction of a positive neutral CCyB by discontinuing a previously applicable SyRB and the Netherlands partially offset it by not reintroducing a previously lowered SyRB for three systemic banks during the COVID-19 pandemic.

### Merits and trade-offs for Albania
- Merits of partially capital-accretive or fully capital-neutral introduction:
  - Albania has a significant gap vis-à-vis the EU average in economic development and financial deepening; avoiding a considerable short-term spike in overall capital requirements may be warranted.
  - Significantly higher requirements for Albanian banks, compared to EU peers, might create competitiveness issues over time.
- High bank profitability could support a fully capital-accretive transition, but there are merits to a partially capital-accretive or capital-neutral introduction.

### Communication strategy (Building Block 6)
- Albania has a well-established macroprudential communication framework with active communication of CCyB decisions.
  - Quarterly CCyB decisions are published on the BoA website, supported by a decision rationale that includes the credit-to -GDP gap and complementary early warning indicators.
  - Cyclical systemic risks and the CCyB rate are discussed in other BoA reports.
- Recommendation: While the communication setup appears well suited for transitioning to a positive neutral CCyB, special attention should be paid to explaining the functioning of the positive neutral CCyB framework to the general public, taking into account the level of financial education and trust in financial institutions.

### Conclusion and calibration
- Introducing a positive neutral CCyB regime would:
  - Strengthen the BoA’s ability to react to stress episodes.
  - Improve the usability of capital buffers without unduly restricting bank credit provision.
  - Maintain a higher CCyB through the cycle compared to the existing buffer regime.
- Calibration guidance:
  - The calibration of the positive neutral rate could be based on a suite of backward- and forward-looking methods, combined with expert judgement.
  - The analysis points to a suitable range of around 100–200 bps for the positive neutral rate in Albania as a starting point for more refined analysis.
  - Timely consultation of key stakeholders will be essential to facilitate the process and to build awareness in advance.

### Annex I — Losses-to-Buffer Approach: Horizon 1 Statistics (Location–Scale Quantile Panel, h=1, q=50th)
- Return on Assets
  - Sample 1 (ALB only): 0.426 ***  (0.051)
  - Sample 2 (CESEE + ALB): .155 ***  (0.002)
- Net interest margin
  - Sample 1: 0.181 ***  (0.064)
  - Sample 2: .156 *  (0.011)
- Cost-to-income
  - Sample 1: -0.000  (0.000)
  - Sample 2: -0.000  (0.991)
- Impairments / assets
  - Sample 1: -0.053  (0.070)
  - Sample 2: -.0264  (0.788)
- Net loans / assets
  - Sample 1: 0.000  (0.010)
  - Sample 2: -.026 *  (0.012)
- RWA / assets
  - Sample 1: -0.019 *  (0.011)
  - Sample 2: .009  (0.321)
- Total equity / assets
  - Sample 1: 0.006  (0.034)
  - Sample 2: -.099 ***  (0.004)
- Tier 1 ratio
  - Sample 1: -0.019  (0.033)
  - Sample 2: .004  (0.837)
- Log assets
  - Sample 1: -0.000  (0.003)
  - Sample 2: -.728 ***  (0.005)
- Yield slope
  - Sample 1: -0.021 **  (0.009)
  - Sample 2: -.069  (0.206)
- Real GDP growth
  - Sample 1: 0.002 **  (0.001)
  - Sample 2: .1155 **  (0.001)
- Domestic Systemic Risk Indicator (d-SRI)
  - Sample 1: 0.202 **  (0.087)
  - Sample 2: -.324  (0.441)
- Tier 1 × d-SRI
  - Sample 1: -0.001  (0.040)
  - Sample 2: .000  (0.739)
- Observations: 616 (Sample 1), 1,050 (Sample 2)
- Notes: Bootstrap s.e., clustered by bank; bank and time fixed effects included in both models.

*Source: sipea2025154 (IMF Selected Issues Paper).*

---


_Source: https://www.imf.org/-/media/files/publications/selected-issues-papers/2025/english/sipea2025154.pdf_