## 1wauea2022007

## Source details

**Canonical URL:** [1wauea2022007](https://www.imf.org/-/media/files/publications/cr/2022/english/1wauea2022007.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2022/english/1wauea2022007.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2022/english/1wauea2022007.pdf.json)

---

### Structural changes and systemic implications
- Regional financial sector size and composition:
  - Total assets representing 69.2 percent of regional GDP in 2020.
  - Banks account for 75 percent of the financial system's assets.
  - Regional banking groups hold nearly 86 percent of banking assets.
  - Regional government securities outstanding expanded tenfold between 2010 and 2020.
- Bank sovereign exposures:
  - Bank sovereign exposures averaged 31 percent of banking assets at end-2020 (up from 7.1 percent at end-2004).
- Interconnectedness:
  - Predominance of regional banking groups and rising bank exposures to public debt instruments increased density and complexity of interconnectedness between banks, interbank counterparties, and governments.

### Cyclical and structural vulnerabilities
- Cyclical indicators:
  - Private sector credit growth: average growth decelerated to 6.5 percent (recent five years) after averaging 14.7 percent per year between 2006 and 2015.
  - Credit gap averaged 2.5 percent over 2006-2014, turned negative in 2015 at -0.8 percent, and averaged -2.2 percent per year over the past five years.
  - FSAP finds no evidence of excessive credit growth.
- Credit quality and NPLs:
  - Non-performing loans (NPLs) ratio to total loans declined to 11.4 percent in 2021.
  - NPLs in some jurisdictions (Benin, Guinea-Bissau, and Togo) remained at 15 to 21 percent of total loans at end-2020.
  - Region-wide loan loss provisioning coverage ratio: 67.1 percent in 2020 (improved from 63.7 percent in 2010); cross-country range 61.5 to 72.4 percent.
- Profitability and competition:
  - Return on assets stable at 1.2 percent between 2016 and 2020.
  - Return on equity declined to 13.3 percent in 2020 (below the 16.5 percent average for the previous four years).
  - Comparative 2020 examples: return on assets — Botswana 1.9 percent, Kenya 2.1 percent, Malawi 4.6 percent; return on equity — Botswana 15.9 percent, Kenya 14.5 percent, Malawi 35.9 percent.
- Bank solvency and heterogeneity:
  - Region-wide capital adequacy ratio: 12.4 percent at end-2020 (minimum regulatory requirement 8.25 percent); up 1.9 percentage points from 2018.
  - Country disparities: Guinea-Bissau average capital adequacy ratio -3.6 percent in 2020; Togo average 7.4 percent in 2020.
  - At end-2020, 19 banks (12.2 percent of total bank assets) had solvency ratios below the regulatory minimum; some (mainly government-owned) entities had negative capital.

### Recommendations on capital and concentration/contagion risk
- Additional capital buffers:
  - Introduce additional capital buffer requirements to cover concentration and contagion risks.
  - Additional capital surcharges should be structured as a nonlinear function of the degree of exposure and portfolio diversification.
  - Surcharges should be designed as a buffer requirement that can be relaxed in periods of stress (to reduce procyclicality).
  - Apply to both government securities holdings and private sector credit; complement with concentration limits for private credit.
  - Introduce gradually, possibly over a three-year period.
  - Buffer requirements for private credit concentration could factor in correlations or sectoral risks.
- Illustrative calibration:
  - On 82 banks (70 percent of total sectoral assets), additional capital needs of 0.4 percent of 2020 GDP at current sovereign exposure concentration levels (illustrative).

### Liquidity and interest-rate risk recommendations
- Liquidity risk management:
  - Encourage banks to manage liquidity risk internally.
  - Regulators should choose objective and verifiable liquidity indicators to determine HQLA eligible as Level 1 assets and the haircuts, including for government securities (numerator of the short-term liquidity ratio).
  - If the regulator does not wish to apply different haircuts for different sovereign issuers, at minimum apply a uniform haircut reflecting market liquidity constraints.
  - The LCR should take into account different risk profiles of bank funding under Pillar 2 of Basel III.
  - Support a gradual increase in the LCR requirement to 100 percent by 2028.
  - The FSAP team suggests counting bank required reserves toward Level 1 HQLA (without a haircut).
- Monitoring interest-rate and maturity mismatches:
  - Improve monitoring of interest rate risk given growth of securities portfolios with relatively long maturities.
  - Require regular reporting on residual maturities of assets and liabilities to establish a long-term liquidity ratio.
  - Introduce additional capital requirements based on Pillar II supervisory approach for interest-rate/maturity mismatches.

### Macroprudential framework, mandates, and governance
- Progress to date:
  - Core elements established: CSF-UMOA (est. 2010) chaired by BCEAO governor; BCEAO created CPMP in 2018.
  - New prudential framework since 2018 introduces CCyB, capital conservation buffers, systemic capital buffers, and borrower-based real estate measures.
- Gaps and recommended reforms:
  - Designate BCEAO as the macroprudential authority responsible for activating the CCyB, with a mechanism for national authorities to raise country-specific concerns.
  - Modify CSF-UMOA voting procedure: give each institutional member, including all state members collectively, a single vote to better balance power.
  - Introduce a "comply-or-explain" mechanism for implementation monitoring of CSF-UMOA recommendations; set clear timeframes for explanations.
  - Clarify and better define intermediate macroprudential policy objectives reflecting regional specifics (predominance of regional banking groups, structural constraints, financial sector development challenges).
  - Strengthen communication of systemic risk assessments and policy decisions to increase transparency; publish systemic risk assessment reports, minutes, and press releases more regularly.

### Data, monitoring, and surveillance improvements
- Data gaps and recommendations:
  - Enhance data collection on credit quality, profitability, debt characteristics (by maturity, rating, and sector), and corporate balance sheets for non-financial enterprises and households.
  - Draw on reforms of public credit registries to establish a permanent system for collecting data on non-financial enterprises and households.
  - Strengthen collection of granular data on residential real estate to identify and assess risks.
  - Calibrate thresholds of systemic risk indicators using peer-country experience and expert judgment.
- Dashboards and heatmaps:
  - Two dashboards developed: macrofinancial dashboard (six dimensions) and financial sector risk heatmap (analytically determined thresholds).
  - Upper 20th percentile values in the macrofinancial dashboard indicate high levels of risk; heatmap deviations classified as high (H), medium (M), or low (L).
  - Dashboards built using only banking sector data due to lack of long-term reliable series for DFSs, pension funds, and insurance companies—recommend close monitoring of DFS growth and performance.

### Concentration risks, sovereign exposures, and fiscal sustainability
- Large exposure limits and capital implications:
  - BCEAO reduced aggregate large exposure limit to 25 percent of bank Tier 1 capital under Basel II/III; to be implemented by 2023.
  - Banks currently struggle to meet existing 55 percent concentration limit and would need to further increase capital by almost 5 percent of regional GDP to meet the 25 percent limit by 2023.
  - A support plan for compliance will likely be required.
- Sovereign exposures and market structure:
  - Share of member state debt in bank portfolios: average 31 percent at end-2020 (from 7.1 percent at end-2004).
  - Banks’ purchases account for about 85 percent of outstanding government securities; banks account for around 90 percent of auction issuances and 75 percent of syndication placements.
  - High sovereign exposures concentrated in medium-sized banks with low capital buffers; banks with government securities holdings >20 percent of assets have on average the lowest solvency ratios.
- Interbank network and ownership:
  - At end-2020, 109 of the 149 banks operating in WAEMU (78 percent of total assets) were affiliated with 32 banking groups.
  - Bank group capital ownership by region: domestic shareholders 36.4 percent of assets; parent banks in North Africa 28 percent of assets; parent banks in Europe 13.1 percent of assets.
  - Direct interbank transactions dropped by 40.5 percent in 2020; intra-group share of cumulative interbank transactions increased to 56.2 percent in 2020 from 51 percent in 2019.
  - Most interbank lending/borrowing is unsecured; collateralized transactions accounted for, on average, 20 to 30 percent of cumulative interbank transactions in 2021.
  - Almost 41 percent of banks do not have repo framework agreements with counterparties.

### Contagion simulations and quantitative outcomes
- Interbank credit and funding shock simulation assumptions:
  - Based on Espinosa-Vega and Sole (2010) combined credit and funding shock framework.
  - Assumed average discount on selling assets in the secondary market: 50 percent.
  - Assumed average of 42 percent of lost interbank funding lines are not replaceable.
  - Assumed loss given default of 80 percent.
  - Data coverage: interbank exposure data as of end-March 2021 covering 70 banks (nearly 86.5 percent of bank total assets).
  - Cumulative interbank transactions at end-March 2021: nearly CFAF 1,450 billion (pre-pandemic 2019 range CFAF 1,500 billion to CFAF 2,000 billion).
- Simulation results — interbank shocks:
  - Contagion due to combined credit and funding shocks could lead to capital losses up to 0.7 percent of 2021 GDP for the region’s banking system.
  - Further capital needs to cover losses from interbank exposures estimated at 0.3 percent of 2021 regional GDP.
  - Country-level maximum capital losses (average estimates): Burkina Faso 1.2 percent of 2021 GDP; Togo 1.1 percent of 2021 GDP; Senegal 0.6 percent of 2021 GDP; Côte d’Ivoire 0.2 percent of 2021 GDP (despite 12 banks failing in the simulation).
- Simulation results — common private exposures (credit + funding shock scenarios):
  - Capital losses from default of large common non-financial private exposures: 1.6 to 3.2 percent of 2021 regional GDP:
    - 1.6 percent of GDP given default of the 10 largest common exposures.
    - 2.7 percent of GDP given default of the 30 largest common exposures.
    - 3.2 percent of GDP given default of the 50 largest common exposures.
  - Capital needs between 0.8 and 2.1 percent of 2021 GDP:
    - Default of 10 largest common exposures: 38 banks fall below regulatory threshold of 8.25 percent; capital needs = 0.8 percent of regional GDP.
    - Default of 50 largest common exposures: 60 banks affected; capital needs = 2.1 percent of regional GDP.
  - Country-level vulnerability (default of the 10 largest regional common exposures): Togo capital losses 3.0 percent of 2021 national GDP; Senegal 1.6 percent; Burkina Faso 1 percent.
- Simulation results — sovereign default scenarios:
  - Scenario: sovereign issuers default on short-term domestic debt maturities outstanding and due by end-2022; outstanding debt maturing by 2022 represents on average 20 to 37 percent of total outstanding securities at end-September 2021.
  - Regional system vulnerability to Côte d'Ivoire default: almost 50 banks would not have sufficient capital buffers; additional capital needs up to 0.8 percent of 2021 regional GDP.
  - Impact of other issuer defaults: Senegal default causes failure of 7 banks; Togo default causes failure of 11 banks; Mali default causes failure of 13 banks.
  - Aggregate bank capital losses from sovereign default scenarios could reach 1.7 percent of 2021 regional GDP.
  - Maximum national GDP-equivalent capital losses under sovereign default scenarios: Burkina Faso 2.8 percent; Togo 2.4 percent; Côte d’Ivoire 1.9 percent; Mali around 1.7 percent; Senegal around 1.6 percent (all as percent of national GDP).
  - Capital needed to cover contagion risk from common sovereign exposures: ranges from 0.04 to 0.8 percent of 2021 regional GDP (0.8 percent for Côte d'Ivoire; 0.04 percent for Guinea-Bissau).
- Common exposures statistics:
  - The 50 largest common exposures accounted for 12.1 percent of bank loans at end-September 2021.
  - Common exposures concentrated in construction and public works; telecommunications; and water and energy sectors.
  - Government securities holdings display a national bias: between 8 and 16 percent of assets (depending on the country) are issued by the state in which banks reside.
  - Participation in syndicated loans for most large exposures: less than five banks.
  - Only the eight largest debtors have liabilities spread across more than 10 banks, representing on average 20 to 40 percent of the capital of the banks in question.

### Policy instruments, activation, and calibration
- Existing macroprudential instruments (since 2018):
  - CCyB with fixed ceiling of 2.5 percent of risk-weighted assets (not activated because macroprudential authority not designated).
  - Capital conservation buffer with fixed ceiling of 2.5 percent of risk-weighted assets.
  - Systemic buffer calibrated based on systemic importance of each SIBI; systemic buffer set at one percent of risk-weighted assets subject to transitional provisions.
  - Borrower-based measures: caps on loan-to-value and debt service-to-income ratios for real estate-backed loans.
- CCyB and calibration:
  - CCyB calibration range should be revised to ensure flexibility for heterogeneity of credit cycles across the Union.
  - Empirical finding: Except for Mali, credit cycles of WAEMU member countries between 1990 and 2020 are broadly synchronized with the regional cycle; regional-level CCyB calibration currently appropriate.
- Implementation modalities:
  - Additional capital surcharge for concentration and contagion risks could be implemented under Basel Pillar 2 or as a Pillar 1 macroprudential rule; as a buffer it should be relaxable in stress and introduced gradually (suggested three-year phase-in).
  - Non-linear calibration approach could be extended to private exposures using concentration factors (correlation of risks between debtors, sectoral exposures).

### Key numeric indicators (selected figures preserved exactly)
- Regional bank assets: 69.2 percent of regional GDP in 2020.
- Banks as share of financial system assets: 75 percent.
- Regional banking groups share of banking assets: nearly 86 percent.
- Bank sovereign exposures: averaged 31 percent of banking assets at end-2020.
- Capital adequacy ratio (region-wide): 12.4 percent at end-2020; regulatory minimum 8.25 percent.
- Number of banks below regulatory minimum at end-2020: 19 banks representing 12.2 percent of total bank assets.
- Large exposure limit: reduced to 25 percent of bank Tier 1 capital by 2023 (current limit 55 percent).
- Estimated additional capital to meet 25 percent limit: almost 5 percent of regional GDP.
- Banks’ share of outstanding government securities purchases: about 85 percent.
- Banks’ share of auction issuances: around 90 percent; syndication placements: 75 percent.
- Banks affiliated with groups at end-2020: 109 of 149 banks (78 percent of total assets).
- Ownership of banking group capital: domestic 36.4 percent of assets; North Africa 28 percent; Europe 13.1 percent.
- Decline in direct interbank transactions in 2020: 40.5 percent.
- Intra-group share of interbank transactions: 56.2 percent in 2020 vs 51 percent in 2019.
- Collateralized transactions share in 2021: 20 to 30 percent.
- Share of banks without repo agreements: almost 41 percent.
- Cumulative interbank transactions at end-March 2021: nearly CFAF 1,450 billion; 2019 pre-pandemic range CFAF 1,500 billion to CFAF 2,000 billion.
- Simulation assumptions and outcomes:
  - Secondary market selling discount assumed: 50 percent.
  - Share of lost interbank funding lines not replaceable: 42 percent.
  - Loss given default assumed: 80 percent.
  - Regional capital losses from combined credit and funding shocks: up to 0.7 percent of 2021 GDP.
  - Additional capital needs for interbank exposures: 0.3 percent of 2021 regional GDP.
  - Capital losses from default of 10/30/50 largest common private exposures: 1.6 ; 2.7 ; 3.2 percent of 2021 regional GDP respectively.
  - Corresponding additional capital needs: 0.8 ; 1.6 ; 2.1 percent of 2021 regional GDP respectively.
  - Aggregate bank capital losses from sovereign default scenarios: could reach 1.7 percent of 2021 regional GDP.
  - Country maximum capital loss examples (sovereign default scenarios): Burkina Faso 2.8 percent; Togo 2.4 percent; Côte d’Ivoire 1.9 percent; Mali ~1.7 percent; Senegal ~1.6 percent (all as percent of national GDP).
- Liquidity and LCR:
  - BCEAO Basel III target: move gradually toward 100 percent (LCR) by 2028.
- Credit and asset quality sample series (selected rows from tableau, values preserved exactly):
  - Change in credit/GDP ratio (ppt, annual): 0.9, 1.1, 0.3, 1.8, 0.8, -0.9, 0.7, 1.1, -0.4, 0.3, 0.6
  - Rate of growth in credit/GDP ratio (%, annual): 5.4, 6.3, 1.4, 9.6, 3.9, -4.1, 3.2, 5.0, -1.8, 1.2, 2.6
  - Credit gap (std. dev.): 0.5, 0.0, 0.2, -0.4, 0.8, 0.1, -1.0, 0.1, 0.4, -0.5, -0.1
  - Credit/deposit ratio: 142.4, 144.0, 141.1, 134.2, 140.4, 142.7, 143.2, 137.5, 144.2, 147.2, 138.4
  - Leverage ratio (%): 6.9, 7.3, 7.3, 7.2, 6.7, 5.7, 5.8, 6.3, 6.8, 6.8, 6.4
  - Return on assets: 1.1, 1.2, 0.9, 0.9, 1.1, 1.2, 1.3, 1.3, 1.2, 1.2, 1.2
  - Return on equity: 12.6, 13.7, 10.1, 11.5, 15.5, 16.4, 20.2, 17.6, 14.6, 13.4, 13.3
  - Non-performing loans ratio (% of total loans): 17.6, 15.9, 16.0, 15.3, 14.9, 14.4, 13.8, 13.9, 12.5, 11.8, 10.9

*Source: EXECUTIVE SUMMARY, WAEMU Financial Sector Assessment (excerpt).*

### EXECUTIVE SUMMARY __________________________________________________________________________ 5

### EXECUTIVE SUMMARY

### Structural changes and systemic implications
- Since the 2008 Financial Sector Assessment Program (FSAP), three structural changes have materially altered the WAEMU financial sector risk profile:
  - The regional financial sector has doubled in size since 2008, with total assets representing 69.2 percent of regional GDP in 2020.
  - Banks account for 75 percent of the financial system's assets.
  - Regional banking groups hold nearly 86 percent of banking assets.
  - The regional government securities market expanded tenfold in outstanding securities between 2010 and 2020, supporting a sharp increase in bank sovereign exposures.
- Bank sovereign exposures averaged 31 percent of banking assets at end-2020, almost triple the level observed in 2004.
- The predominance of regional banking groups and rising bank exposures to public debt instruments have increased the density and complexity of interconnectedness between banks, interbank counterparties, and governments.

### Cyclical and structural vulnerabilities
- Cyclical vulnerabilities:
  - The slowdown in credit growth in recent years points to weak cyclical pressures; the FSAP’s analysis shows no evidence of excessive credit growth.
- Structural vulnerabilities and credit risk:
  - Despite improvements since 2008, credit risk remains a structural concern because of:
    - Relatively high levels of nonperforming loans (NPLs).
    - High concentration of bank loan portfolios focused on a limited number of debtors, including sovereign borrowers, and on certain sectors of activity.
  - The structure of banks, and their large common exposures to private debtors and sovereign borrowers, are important vectors of contagion and amplification of credit risk in the union.
- Bank solvency and heterogeneity:
  - At end-2020, the region-wide capital adequacy ratio was 12.4 percent, above the minimum regulatory requirement of 8.25 percent and up 1.9 percentage points from 2018.
  - Disparities across countries are significant:
    - Guinea-Bissau: average capital adequacy ratio of -3.6 percent in 2020.
    - Togo: average capital adequacy ratio of 7.4 percent in 2020.
  - At end-2020, 19 banks, representing 12.2 percent of total bank assets, had solvency ratios below the regulatory minimum; some (mainly government-owned) entities continue to have negative capital.

### Recommendations on capital and concentration/contagion risk
- Introduce additional capital buffer requirements to cover concentration and contagion risks:
  - Additional capital surcharges should be structured as a nonlinear function of the degree of exposure and portfolio diversification to contain the highest risks.
  - These surcharges should be designed as a buffer requirement that can be relaxed in periods of stress (to reduce potential procyclicality).
  - They should apply to both government securities holdings and private sector credit, in addition to concentration limits for the latter.
  - Introduce gradually, possibly over a three-year period.
  - Additional capital buffer requirements for private credit concentration could factor in correlations or sectoral risks.

### Liquidity and interest-rate risk recommendations
- Liquidity:
  - Support for authorities’ efforts to encourage banks to manage liquidity risk internally.
  - Regulators should choose objective and verifiable liquidity indicators to determine HQLA eligible as Level 1 assets, and the haircuts, including those for government securities (numerator of the short-term liquidity ratio).
  - If the regulator does not wish to apply different haircuts for different sovereign issuers, a uniform haircut should at least be applied to reflect liquidity constraints across the market.
  - The LCR should take into account different risk profiles of bank funding under Pillar 2 of Basel III.
  - Support a gradual increase in the requirement to 100 percent by 2028 to take into account the difficulties some banks may have in complying.
- Interest-rate and maturity mismatch monitoring:
  - Improve monitoring of interest rate risk: growth of securities portfolios with relatively long maturities compared to bank resources could increase maturity and interest-rate mismatches.
  - Require regular reporting on the residual maturities of assets and liabilities to establish a long-term liquidity ratio.
  - Introduce additional capital requirements based on the Pillar II supervisory approach of Basel III.

### Macroprudential framework and institutional arrangements
- Progress to date:
  - The Central Bank of West African States (BCEAO) has put into place core elements of a macroprudential policy framework.
  - The Financial Stability Committee of the West African Monetary Union (CSF-UMOA), established in 2010 and chaired by the governor of the BCEAO, discusses facts relevant to financial stability and issues policy action recommendations and warnings.
  - The BCEAO created a Macroprudential Policy Committee (CPMP) in 2018 composed of senior staff from the BCEAO and the Banking Supervisor (CBU); the CPMP monitors systemic risks and can apply macroprudential instruments and implement CSF-UMOA recommendations on banking supervision.
  - The new prudential framework in force since 2018 introduces macroprudential instruments related to capital surcharges—countercyclical capital buffers (CCyB), capital conservation buffers, systemic capital buffers—and borrower-based measures for real estate lending.
- Institutional gaps and suggested reforms:
  - Introduce a comply-or-explain mechanism in the monitoring of the implementation of CSF-UMOA recommendations to increase accountability of institutional members.
  - Modify the CSF-UMOA voting procedure to give each institutional member, including all state members collectively, a single vote to ensure a better balance of power.
  - Clarify and better define intermediate macroprudential policy objectives, taking into account:
    - Predominance of regional banking groups.
    - Structural macroeconomic constraints (e.g., weak diversification of economic activities).
    - Regional financial sector development challenges (e.g., limited access to financial services).
  - Improve communication of systemic risk assessments and policy decisions to strengthen transparency.

### Data, monitoring, and surveillance improvements
- Systemic risk monitoring is well established but would benefit from closing data gaps and enhancing sectoral vulnerability monitoring.
- Recommended data and monitoring actions:
  - Enhance data collection on credit quality, profitability, debt characteristics (by maturity, rating, and sector), and corporate balance sheets for non-financial enterprises and households.
  - Draw on ongoing reforms of public credit registries to establish a permanent system for collecting data on non-financial enterprises and households.
  - Strengthen collection of granular data on residential real estate to identify and assess risks.
  - Calibrate thresholds of systemic risk indicators being monitored, drawing on the experience of instability in the banking sectors of peer countries and on expert judgment.

### Table of main recommendations (high-level)
- Monitoring of systemic risks:
  - Calibrate systemic risk thresholds for monitored macroprudential indicators, drawing on the experience of instability in the banking sectors of peer countries and on expert judgment. — BCEAO — ST
  - Improve the capacity to monitor vulnerabilities and risks in the non-financial corporate and household sectors. — BCEAO — MT
  - Fill in remaining data gaps, including those for the non-financial corporate, household, and real estate sectors. — CSF-UMOA/BCEAO — ST
- Operational framework for macroprudential policy:
  - Gradually introduce additional non-linear capital requirements to cover concentration, interest rate, and contagion risks arising from excessive concentration, including those for bank sovereign exposures. — BCEAO/CBU — ST
  - Introduce Basel-type liquidity ratios to strengthen resilience to liquidity shocks. Introduce a haircut for government securities, count banks’ required reserves toward Level 1 high-quality liquid assets (HQLA), and consider bank funding stability to impose higher LCR requirements under Pillar 2 of Basel III. — BCEAO/CBU — ST
  - Improve the monitoring of maturity mismatches and interest rate risk based on regular reporting of the residual maturities of bank assets and liabilities, as well as their interest rates. — BCEAO/CBU — ST
- Institutional framework for macroprudential policy:
  - Designate the BCEAO as the "macroprudential authority" responsible, inter alia, for activating the countercyclical capital buffer (CCyB), while introducing a suitable mechanism for national authorities to raise concerns if policies set by the BCEAO are ill-suited for their situation. — CSF-UMOA/BCEAO — I
  - Strengthen implementation of CSF-UMOA recommendations by introducing a "comply or explain" mechanism in follow-up and specify deadlines for explanations for failure to implement recommendations. — CSF-UMOA — I
  - Establish a comprehensive communication strategy for systemic risk assessments and macroprudential policy decisions. — CSF-UMOA/BCEAO — ST

*Source: EXECUTIVE SUMMARY, WAEMU Financial Sector Assessment (excerpt).*

### 6. The effective implementation of macroprudential policy requires continuous

### 6. The effective implementation of macroprudential policy requires continuous preventive monitoring of systemic risks

### Systemic risk dimensions and SIBI identification
- Systemic risks cover two main dimensions:
  - Cyclical dimension: accumulation of risks over time (e.g., phases of excessive credit growth leading to accumulation of credit risks).
  - Structural dimension: interdependence across institutions and distribution of risks within the financial system at a given time (e.g., strong direct interconnectedness transmitting shocks; indirect links due to large common exposures carrying contagion risks).
- Funding vulnerabilities highlighted:
  - Stronger reliance on less stable funding sources (e.g., wholesale or non-deposit funding) increases exposure to sudden liquidity outflows if not offset by liquid asset buffers.
- SIBI (systemically important bank/financial institution) framework by the BCEAO:
  - Multi-criteria approach based on Basel Committee recommendations.
  - Identification carried out by the CBU using a multi-indicator score considering: size, degree of interconnectedness, substitutability, and complexity.
  - A bank/financial institution is designated as a SIBI if its composite systemic importance score is above a certain threshold.
  - Published SIBI lists are broken into regional SIBIs and national SIBIs.
  - First WAEMU SIBI lists (published March 2020) include: six financial corporations in the regional SIBI list and 22 banks in the national list.
  - The CBU reviews the SIBIs list once a year.

### A. Cyclical vulnerabilities — key findings
- Private sector credit growth:
  - Average growth decelerated to 6.5 percent (recent five years) after averaging 14.7 percent per year between 2006 and 2015.
  - Credit gap (deviation of credit-to-GDP ratio from long-term trend) averaged 2.5 percent over 2006-2014, turned negative in 2015 at -0.8 percent, and has been steadily declining over the past five years, averaging -2.2 percent per year.
- Bank credit quality and NPLs:
  - With average real GDP growth of 5.6 percent and growth in private sector credit of 12 percent per year during 2010-2019, the ratio of non-performing loans (NPLs) to total loans declined by an average of 0.6 percentage point per year since 2010 to 11.4 percent in 2021.
  - NPLs in some jurisdictions (e.g., Benin, Guinea-Bissau, and Togo) remained high, at 15 to 21 percent of total loans at end-2020.
  - Bank loan loss provisioning: region-wide coverage ratio of 67.1 percent (improved from 63.7 percent in 2010 to 67.1 percent in 2020); cross-country range 61.5 to 72.4 percent.
- COVID-19 related measures and outcomes:
  - BCEAO temporary loan restructuring and relaxation of repayment rules allowed banks to reschedule 3.1 percent of their loans to the private sector.
  - Rules suspended in December 2020; at end-June 2021, 22 percent of rescheduled loans were downgraded to "doubtful."
- Bank profitability and competition:
  - Return on assets stable at 1.2 percent between 2016 and 2020.
  - Return on equity declined to 13.3 percent in 2020 (below the 16.5 percent average for the previous four years and below banks in other developing countries).
  - Specific country pressures: banks in Benin and Togo face significant profitability pressures.
  - Comparative examples (2020): return on assets — Botswana 1.9 percent, Kenya 2.1 percent, Malawi 4.6 percent; return on equity — Botswana 15.9 percent, Kenya 14.5 percent, Malawi 35.9 percent.
- Capital adequacy (Figure 2 summary):
  - Overall capital adequacy ratio rose from 10.5% in 2018 to 12.4% in 2020 under Basel II/III.
  - A few banks with negative capital continue to weigh on average capital adequacy in some countries (e.g., Guinea-Bissau and Togo).

### Macrofinancial indicators, dashboards, and heatmaps
- Rationale:
  - Simple indicators may not capture cyclical characteristics and macrofinancial interrelationships; systemic vulnerability dashboards provide a holistic view essential for macroprudential policymaking.
- Two complementary dashboards developed by the FSAP team:
  - Macrofinancial dashboard:
    - Groups WAEMU macrofinancial vulnerabilities into six dimensions: (i) macroeconomic risk; (ii) vulnerabilities to external shocks; (iii) credit risks; (iv) liquidity risks; (v) monetary and financial conditions; and (vi) risk appetite.
    - Thresholds are endogenous, based on empirical distribution of individual indicators; upper 20th percentile values indicate high levels of risk.
  - Financial sector risk heatmap:
    - Provides financial stability indicators with exogenous, analytically determined thresholds.
    - Combines early warning indicators (e.g., indicator of credit cycle) and measures of financial system soundness (e.g., balance sheet strength and adequacy of capital buffers).
    - Thresholds determined via empirical cross-country work and expert judgment (not based on individual indicator distributions).
    - Deviations from analytical thresholds classify vulnerability as high (H), medium (M), or low (L).
- Dashboard findings:
  - Accumulation of macrofinancial vulnerabilities after the COVID-19 outbreak in 2020 reflected in spidergrams and heatmaps.
  - Specific vulnerability changes since 2020:
    - Increased macroeconomic risk driven by deterioration in domestic and global demand and rising uncertainty related to security problems in the Sahel (Burkina Faso, Mali, Niger).
    - Easing monetary and financial conditions due to policy rate decline and accommodative measures by BCEAO; prolonged accommodation could encourage excessive risk-taking amid stronger WAEMU-wide bank competition and lower lending rates.
    - Credit risks stabilized at 2019 levels, but continued high NPLs and heavy concentration of credit portfolios in a small number of borrowers sustain uncertainty.
    - Mitigated risks from external shocks due to strong external buffers as measured by external reserve import coverage, but structural vulnerabilities from limited export diversification and competitiveness constraints persist.
- Heatmap conclusion:
  - Despite recent improvements, low asset quality of banks remains a significant structural risk; the heatmap indicates structurally high credit risks with delinquency levels remaining high despite moderate credit growth.
- Data coverage limitation and recommendation for monitoring:
  - Dashboards built using only banking sector data due to lack of long-term reliable series for decentralized financial systems (DFSs), and absence of data on pension funds and insurance companies.
  - Recommendation: closely monitor growth and financial performance of DFSs to prevent transfer of risk from banking to non-banking institutions.
- Illustrative thresholds note:
  - The thresholds presented are illustrative and can be fine-tuned using complete data, other analytical work, and expert opinions; the 20th percentile threshold may be varied.

### B. Concentration risks — key findings and statistics
- Sovereign and private debtor concentration:
  - WAEMU bank balance sheets show high concentration of assets to a limited number of private debtors and to sovereign borrowers.
  - The regional market provides opportunity to diversify sovereign exposures, which banks have begun to utilize.
- Concentration of private sector loan portfolios:
  - Concentration of exposures to individual borrowers declined since 2017 but remains high: average concentration ratio of 78.1 percent of capital in 2020.
  - For more than half of WAEMU banks, the ten largest exposures on average represent 20 percent of assets.
  - High sectoral concentration in Guinea-Bissau (and to some extent Burkina Faso, Niger, Togo) increases vulnerability; Côte d'Ivoire and Senegal show more diversified sectoral exposures.
- Capital buffers vs large exposures:
  - The ten largest exposures are larger than total capital for 80 percent of WAEMU banks.
  - More than 60 percent of banks do not have sufficient surplus capital to address credit risk associated with a potential failure of their largest debtors.

### Key tableau of selected numeric indicators (as presented)
- Credit and growth metrics (selected values):
  - Change in credit/GDP ratio (ppt, annual): 0.9, 1.1, 0.3, 1.8, 0.8, -0.9, 0.7, 1.1, -0.4, 0.3, 0.6
  - Rate of growth in credit/GDP ratio (%, annual): 5.4, 6.3, 1.4, 9.6, 3.9, -4.1, 3.2, 5.0, -1.8, 1.2, 2.6
  - Credit gap (std. dev.): 0.5, 0.0, 0.2, -0.4, 0.8, 0.1, -1.0, 0.1, 0.4, -0.5, -0.1
- Soundness and balance sheet metrics:
  - Credit/deposit ratio: 142.4, 144.0, 141.1, 134.2, 140.4, 142.7, 143.2, 137.5, 144.2, 147.2, 138.4
  - Leverage ratio (%): 6.9, 7.3, 7.3, 7.2, 6.7, 5.7, 5.8, 6.3, 6.8, 6.8, 6.4
- Profitability:
  - Return on assets: 1.1, 1.2, 0.9, 0.9, 1.1, 1.2, 1.3, 1.3, 1.2, 1.2, 1.2
  - Return on equity: 12.6, 13.7, 10.1, 11.5, 15.5, 16.4, 20.2, 17.6, 14.6, 13.4, 13.3
- Asset quality:
  - Non-performing loans ratio (% of total loans): 17.6, 15.9, 16.0, 15.3, 14.9, 14.4, 13.8, 13.9, 12.5, 11.8, 10.9
  - Change in non-performing loans ratio (%, annuel): 2.2, -9.4, 0.3, -3.9, -2.7, -3.9, -3.6, 0.5, -10.1, -5.6, -7.6
- Macrofinancial dashboard heatmap sample values (2017–2020 averages and items; selected entries):
  - Macroeconomic risks: 2017 5.2; 2018 4.8; 2019 4.8; 2020 5.4
  - Fiscal space: 2017 5.0; 2018 4.5; 2019 4.0; 2020 7.5
  - Credit to the economy: 2017 6.0; 2018 4.0; 2019 5.0; 2020 5.0
  - Liquidity risks — Bank exposure to liquidity shocks: 2017 6.0; 2018 6.0; 2019 6.3; 2020 6.5
  - Risk appetite — Portfolio investment flows: 2017 9.0; 2018 9.0; 2019 5.0; 2020 4.0

*Source: IMF staff analysis and BCEAO data as presented in the WAEMU FSAP chapter.*

### 19. The BCEAO has adopted measures to limit the concentration risk, but their application

### 19. The BCEAO has adopted measures to limit the concentration risk, but their application

### Reduction in large exposure limit and capital implications
- The BCEAO reduced the aggregate large exposure limit to 25 percent of bank Tier 1 capital under Basel II/III; this should be implemented by 2023.
- Banks currently struggle to meet the existing 55 percent concentration limit and would need to further increase their capital by almost 5 percent of regional GDP to meet the 25 percent limit required by 2023.
- A support plan for compliance with the new limit will likely be required.

### Concentration of sovereign exposures and fiscal sustainability risks
- The share of member state debt in bank portfolios more than quadrupled in 15 years, reaching an average of 31 percent at end-2020, up from 7.1 percent at end-2004.
- Bank securities purchases account for about 85 percent of outstanding government securities; WAEMU banks account for around 90 percent of issuances via auctions and 75 percent of the syndication placements.
- High sovereign exposures are concentrated in medium-sized banks with low capital buffers; banks whose government securities holdings account for more than 20 percent of their assets have on average the lowest solvency ratios.
- WAEMU member countries are at medium or high risk of debt distress according to published IMF debt sustainability analyses; sovereign ratings are below BB. Some countries (Benin, Côte d’Ivoire, and Senegal) have access to international capital markets, with average spreads in the performing bond category.

### Interbank network structure and contagion channels
- At end-2020, 109 of the 149 banks operating in the WAEMU (78 percent of total assets) were affiliated with 32 banking groups.
- Bank group capital ownership by region: domestic shareholders 36.4 percent of assets; parent banks in North Africa 28 percent of assets; parent banks in Europe 13.1 percent of assets.
- Three identified contagion channels: (i) direct interbank exposures (volumes increased over the past five years), (ii) regional and cross-border interconnections via bank capital structures, and (iii) large common exposures to member governments and large private debtors.
- Intra-group interbank flows provided stabilizing liquidity during COVID-19: direct interbank transactions dropped by 40.5 percent in 2020, while the share of cumulative interbank transactions between banks belonging to the same group increased to 56.2 percent from 51 percent in 2019.
- Most interbank lending and borrowing is unsecured; collateralized transactions accounted for, on average, 20 to 30 percent of the cumulative volume of interbank transactions in 2021.
- Almost 41 percent of banks do not have repo framework agreements with counterparties; few banks have signed such agreements with more than four counterparties.
- The absence of a BCEAO framework for providing emergency liquidity and low secondary market liquidity in government securities limit banks’ ability to sell assets in a liquidity shock.

### Interbank contagion simulations: assumptions and results
- Simulation setup details:
  - Combined credit and funding shock framework based on Espinosa-Vega and Sole (2010).
  - Assumed average discount on selling assets in the secondary market: 50 percent.
  - Assumed average of 42 percent of lost interbank funding lines are not replaceable.
  - Assumed loss given default of 80 percent.
- Data coverage: interbank exposure data as of end-March 2021 covering 70 banks (nearly 86.5 percent of bank total assets); cumulative interbank transactions at end-March 2021 stood at nearly CFAF 1,450 billion (pre-pandemic 2019 range CFAF 1,500 billion to CFAF 2,000 billion).
- Simulation outcomes:
  - Contagion due to combined credit and funding shocks would lead to capital losses of up to 0.7 percent of 2021 GDP for the region’s banking system.
  - Further capital needs to cover losses from interbank exposures were estimated at 0.3 percent of 2021 regional GDP.
  - Country-level maximum capital losses (average estimates):
    - Burkina Faso: 1.2 percent of 2021 GDP.
    - Togo: 1.1 percent of 2021 GDP.
    - Senegal: 0.6 percent of 2021 GDP.
    - Côte d’Ivoire: 0.2 percent of 2021 GDP (despite 12 banks failing in the simulation).

### Contagion from large common exposures (sovereign and private)
- The 50 largest common exposures of WAEMU banks accounted for 12.1 percent of bank loans at end-September 2021.
- Common exposures are concentrated in construction and public works; telecommunications; and water and energy sectors.
- Government securities holdings display a national bias: between 8 and 16 percent of assets (depending on the country) are issued by the state in which banks reside.
- Côte d’Ivoire securities account for a prominent share of securities portfolios across the Union; Ivorian banks hold almost exclusively Côte d’Ivoire securities.
- Syndicated loans exhibit limited risk sharing: on average syndicated loans are issued by only a few banks; for most large common exposures, participation is limited to less than five banks.
- Only the eight largest debtors have liabilities spread across more than 10 banks, representing on average 20 to 40 percent of the capital of the banks in question.

### Key quantitative metrics (preserved exactly as reported)
- Aggregate large exposure limit to be reduced to 25 percent of bank Tier 1 capital by 2023.
- Current concentration limit: 55 percent.
- Required additional capital to meet 25 percent limit: almost 5 percent of regional GDP.
- Share of sovereign exposures in bank portfolios: average 31 percent at end-2020; 7.1 percent at end-2004.
- Banks’ share of outstanding government securities purchases: about 85 percent.
- Banks’ share of auction issuances: around 90 percent; syndication placements: 75 percent.
- Banks affiliated with groups at end-2020: 109 of 149 banks (78 percent of total assets) affiliated with 32 banking groups.
- Ownership of banking group capital: domestic 36.4 percent of assets; North Africa 28 percent; Europe 13.1 percent.
- Decline in direct interbank transactions in 2020: 40.5 percent.
- Intra-group share of interbank transactions: 56.2 percent in 2020 vs 51 percent in 2019.
- Collateralized transactions share in 2021: 20 to 30 percent.
- Share of banks without repo agreements: almost 41 percent.
- Secondary market selling discount assumed: 50 percent.
- Share of lost interbank funding lines not replaceable: 42 percent.
- Cumulative interbank transactions at end-March 2021: nearly CFAF 1,450 billion; 2019 pre-pandemic range CFAF 1,500 billion to CFAF 2,000 billion.
- Regional capital losses from combined shocks: up to 0.7 percent of 2021 GDP.
- Additional capital needs for interbank exposures: 0.3 percent of 2021 regional GDP.
- Country maximum capital loss averages: Burkina Faso 1.2 percent of 2021 GDP; Togo 1.1 percent of 2021 GDP; Senegal 0.6 percent of 2021 GDP; Côte d’Ivoire 0.2 percent of 2021 GDP.
- 50 largest common exposures share of bank loans at end-September 2021: 12.1 percent.
- Participation in syndicated loans for most large exposures: less than five banks.
- Eight largest debtors’ liabilities spread across more than 10 banks represent on average 20 to 40 percent of the banks’ capital.

*International Monetary Fund — WAEMU Financial Sector Assessment (chapter text).*

### 37. An analytical approach was used to assess the potential contagion effects associated

### 1wauea2022007 - 37. An analytical approach was used to assess the potential contagion effects associated

### Analytical approach to contagion from common exposures
- Methodology based on the network analysis framework of Espinosa-Vega and Sole (2010).
- Expands interbank exposure matrix to include common exposures of banks to non-financial entities (governments or private debtors).
- Captures amplifying effect of interbank market: default of common exposures leads to
  - direct capital losses for banks with outstanding loans to the common exposures, and
  - indirect losses for other banks via the interbank exposure channel.

### Results: Common non-financial private exposures (credit + funding shock scenarios)
- Capital losses from default of large common non-financial private exposures: range from 1.6 to 3.2 percent of 2021 regional GDP.
  - 1.6 percent of GDP given default of the 10 largest common exposures.
  - 2.7 percent of GDP given default of the 30 largest common exposures.
  - 3.2 percent of GDP given default of the 50 largest common exposures.
- Capital needs: between 0.8 and 2.1 percent of 2021 GDP.
  - Default of the 10 largest common exposures: 38 banks fall below regulatory threshold of 8.25 percent; capital needs = 0.8 percent of regional GDP.
  - Default of the 50 largest common exposures: 60 banks affected; capital needs = 2.1 percent of regional GDP.
- Country-level vulnerability (default of the 10 largest regional common exposures):
  - Togo: capital losses of 3.0 percent of 2021 national GDP.
  - Senegal: capital losses of 1.6 percent of 2021 national GDP.
  - Burkina Faso: capital losses of 1 percent of 2021 national GDP.

### Results: Common sovereign exposures (sovereign default scenarios)
- Scenario: sovereign issuers default on short-term domestic debt maturities outstanding and due by end-2022.
- Outstanding debt maturing by 2022 represents on average 20 to 37 percent of total outstanding securities at end-September 2021.
- Regional system vulnerability to Côte d'Ivoire default:
  - Almost 50 banks would not have sufficient capital buffers to cope with a default by Côte d'Ivoire on its short-term maturities.
  - Additional capital needs would be up to 0.8 percent of 2021 regional GDP in the event of a Côte d'Ivoire default.
- Impact of other issuer defaults:
  - Senegal default: causes failure of 7 banks.
  - Togo default: causes failure of 11 banks.
  - Mali default (more concentrated short-term repayment profile): causes failure of 13 banks.
- Aggregate bank capital losses from sovereign default scenarios: could reach 1.7 percent of 2021 regional GDP.
- Maximum national GDP-equivalent capital losses under sovereign default scenarios:
  - Burkina Faso: 2.8 percent of national GDP.
  - Togo: 2.4 percent of national GDP.
  - Côte d'Ivoire: 1.9 percent of national GDP.
  - Mali: around 1.7 percent of national GDP.
  - Senegal: around 1.6 percent of national GDP.
- Capital needed to cover contagion risk from common sovereign exposures: ranges from 0.04 to 0.8 percent of 2021 regional GDP.
  - 0.8 percent of regional GDP in event of default by Côte d'Ivoire.
  - 0.04 percent of regional GDP in event of default by Guinea-Bissau.

### Liquidity risks and deposit concentration
- Market liquidity limitations: apart from bank reserves at the BCEAO, many bank assets have zero or limited market liquidity (e.g., non-marketable debt, government securities).
- Deposit concentration:
  - The five largest depositors, including non-financial corporations, account for an average of 26 percent of total deposits.
  - The largest depositor accounts for an average of 10 percent of total deposits.
- Maturity and interest rate mismatches: emerging risk due to expansion of government securities holdings with long maturities relative to bank resources.
- Data constraint: lack of granular residual maturity and rate data prevented full measurement of this risk.

### Systemic risk monitoring: current practices and gaps
- BCEAO role:
  - Financial stability department responsible for analysis and monitoring of systemic risks.
  - Supports CSF-UMOA, which meets at least twice a year and is chaired by the BCEAO’s governor.
- Tools and practices:
  - Use of a broad set of macroprudential indicators presented as a financial stability dashboard.
  - Use of simulation and stress testing with satellite models linking macroeconomic factors (e.g., GDP and inflation) to bank-specific variables (e.g., profitability, credit growth, asset quality deterioration).
  - Recommendation to publish the financial stability dashboard or stress testing results to improve transparency.
- Monitoring gaps and recommendations:
  - No identified systemic risk thresholds or early warning system in the BCEAO framework; reliance on expert judgment recommended.
  - Need for enhanced monitoring of corporate, household, and real estate sector risks:
    - Corporate debt accounts for about two-thirds of bank loan portfolios.
    - Need more disaggregated indicators on corporate and household balance sheets, profitability, debt repayment capacity, and property prices.
  - Data initiatives noted: corporate balance sheets database and credit and payment registers, regional credit bureau; recommendation to accelerate overhaul of credit register and reorient it to focus on macroprudential surveillance.

### Macroprudential instruments and policy recommendations
- Existing instruments introduced since 2018:
  - CCyB (countercyclical capital buffer) with a fixed ceiling of 2.5 percent of risk-weighted assets.
  - Capital conservation buffer with a fixed ceiling of 2.5 percent of risk-weighted assets.
  - Systemic buffer calibrated based on systemic importance of each SIBI.
  - Credit containment measures: caps on loan-to-value and debt service-to-income ratios for real estate-backed loans.
- CCyB activation status:
  - CCyB has not been activated because the macroprudential authority responsible for activation has not been designated.
  - Given moderate cyclical vulnerabilities and a negative credit-to-GDP gap over the last five years, non-activation has not been problematic to date.
  - Recommendation: designate the macroprudential authority and finalize procedures, triggers, and implementation scope (regional or national).
- SIBI requirements:
  - Systemic buffer set at one percent of risk-weighted assets, subject to transitional provisions; CBU can adjust based on systemic importance.
  - SIBIs subject to more intensive CBU supervision.
- Measures to mitigate concentration-related systemic vulnerabilities:
  - Promote diversification of securities portfolios across issuing member states, noting limits due to similar debt sustainability profiles and national bias in holdings.
  - Warning that rapid expansion of bank holdings of government securities could crowd out private sector credit without an appropriate framework.
- Additional capital buffer for concentration and contagion risks:
  - Recommend capital surcharge calibrated as a non-linear function of portfolio concentration above a certain exposure threshold for each debtor.
  - Could be implemented under Basel Pillar 2 or as a Pillar 1 macroprudential rule; choice affects implementing authority and transparency.
  - As a buffer, surcharge should be relaxable in periods of stress to reduce procyclicality.
  - Suggested gradual introduction over a three-year period.
  - Illustrative calibration on 82 banks (70 percent of total sectoral assets) suggests additional capital needs of 0.4 percent of 2020 GDP at current levels of sovereign exposure concentration.
- Extension to private exposures:
  - Non-linear calibration approach could be extended to bank loan portfolios to account for concentrations in private exposures, using concentration factors such as correlation of risks between debtors and sectoral exposures.
  - Could complement risk sharing where many banks, particularly smaller ones, cannot meet the 25 percent limit by 2024.

*Source: IMF staff report excerpt (WAEMU FSAP).*

### 57. Additional capital buffer requirements should also be imposed on banks to cover

### 1wauea2022007 - 57. Additional capital buffer requirements should also be imposed on banks to cover

### Capital buffers and CCyB calibration
- Additional capital buffer requirements (Pillar 2 capital surcharge) should be imposed on banks to cover interest rate risk; the surcharge should be proportional to each bank's specific interest rate and maturity mismatches, measured regularly.
- The calibration of the Countercyclical Capital Buffer (CCyB) range should be revised to ensure flexibility for potential heterogeneity of credit cycles across the Union.
- The CCyB calibration should at least reflect the estimated average loss of capital experienced by the banking system following adverse changes in macroeconomic and financial conditions (typically reversals in the credit cycle).
- In the context of a currency union like WAEMU, differences in magnitude and duration of credit cycles pose a challenge for a uniform CCyB; some flexibility to account for member-state differences is necessary.
- Empirical finding: Except for Mali, an analysis of the synchronization of credit cycles between 1990 and 2020 shows that the credit cycles of WAEMU member countries are broadly synchronized with the regional cycle; therefore, a regional-level CCyB calibration is currently appropriate without risk of bias.

### Liquidity risk management and Basel III implementation
- The mission supports authorities’ efforts to enhance bank internal liquidity risk management.
- The BCEAO plans to introduce the short-term liquidity coverage ratio (LCR) under Basel III and move gradually toward 100 percent by 2028.
- Short-term LCR recommendations:
  - Numerator: The regulator must choose objective and verifiable liquidity indicators to determine Level 1 eligible HQLA and appropriate haircuts, including for government securities. If the regulator does not want to differentiate haircuts for individual sovereign issuers in the region, at minimum a uniform haircut should be applied to account for limited liquidity in the regional government securities market.
  - The FSAP team suggests counting bank required reserves toward Level 1 HQLA, albeit without a haircut, since they are available to absorb temporary liquidity needs.
  - Denominator: Differences in deposit stability should be considered when determining LCR requirements (Pillar 2 of Basel III). To impose a higher requirement on banks with significantly larger withdrawals, the regulator should improve tracking of the historical distribution of monthly changes in each bank’s funding sources.
- Long-term liquidity ratio:
  - The supervisor needs to strengthen the requirement for regular reporting of the residual maturities of bank assets and liabilities to monitor maturity mismatches and be able to enforce the ratio.
- Note: For detailed analysis of liquidity-related issues, see the technical note “Analysis of Systemic Liquidity” produced by this FSAP.

### Institutional framework for macroprudential policy — mandates and governance
- A strong, functional institutional framework is essential for effective macroprudential policymaking; it should clarify missions and objectives, divide responsibilities, enable quick responses, ensure access to information, and provide a broad set of tools and reporting mechanisms.
- Since the 2008 FSAP, core elements of a macroprudential framework have been put in place:
  - CSF-UMOA created in 2010 and chaired by the BCEAO governor; charged with safeguarding financial stability, strengthening cooperation in macroprudential oversight, conducting regular discussions, issuing alerts, and making macroprudential recommendations.
  - The BCEAO’s financial stability mandate is defined in its statute; financial stability is defined as a state in which “the financial system, which includes financial intermediaries, markets, and market infrastructures, is able to withstand shocks and correct financial imbalances.”
- Intermediate objectives:
  - The current framework lacks guidance on intermediate macroprudential objectives; the BCEAO planned to define such objectives as part of its 2019-2021 strategic action plan for financial stability.
  - Recommended intermediate objectives include: (i) increasing the system's resilience to shocks and supporting continued availability of credit across the cycle; (ii) containing accumulation of systemic vulnerabilities and limiting excessive growth in leverage and volatility of systemic liquidity; (iii) containing structural risks from financial interconnectedness.
  - Definitions should consider regional specificities (predominance of regional banking groups, limited production diversification) and financial sector development challenges (limited private-sector access to bank financing).

### Decision-making powers, representation, and reforms
- CSF-UMOA composition and influence:
  - Government representatives hold eight out of the thirteen seats on the CSF-UMOA and thus have a dominant position; government representatives account for more than half of CSF-UMOA membership and can significantly influence macroprudential measures.
  - Risk: Overrepresentation of state members may affect the Committee’s ability to take necessary macroprudential measures that incur costs for states.
- Recommended reforms:
  - Revise CSF-UMOA decision-making to better balance powers; if CSF-UMOA is to serve as a macroprudential authority for all financial sectors in the long run, voting procedures should be reformed to ensure a more balanced decision-making approach among regulators.
  - Specifically, give each institutional member of CSF-UMOA, including all member states, a single vote to reduce influence of states.
- Ability to act:
  - CSF-UMOA can recommend policy actions but lacks adequate powers to ensure implementation; the framework does not specify whether regulators are required to implement recommendations nor a timeframe for explaining delays.
  - Recommendation: Strengthen monitoring of CSF-UMOA recommendations by introducing a “comply-or-explain” basis and clear timeframes for regulatory members to explain any failure to implement proposed measures.
- Role of BCEAO and CPMP:
  - BCEAO should assume the role of macroprudential authority responsible for activating the CCyB, given its statutory mandate, expertise, and dominant role of banks.
  - BCEAO holds Chairmanship and Secretariat of CSF-UMOA and leads analysis and monitoring of systemic risks.
  - In 2018 BCEAO created CPMP (composed of senior staff from BCEAO and the Banking Supervisor, CBU) to monitor and analyze systemic risks and to apply macroprudential instruments and implement CSF-UMOA recommendations related to banking.
  - Given heterogeneities across WAEMU economies, the macroprudential framework should incorporate a mechanism for national authorities to raise country-specific concerns related to regional policies.
- Activity record:
  - CSF-UMOA has not always met at the minimum frequency prescribed and has not often produced recommendations on macroprudential instruments since its establishment in 2010.
  - CPMP, established in 2018, did not hold its first meeting until March 2020.

### Information sharing, coordination, and cross-border supervision
- Cooperation and coordination:
  - The institutional framework encourages cooperation; risks monitoring and analysis are centered on preparations for CSF-UMOA and CPMP meetings, led by BCEAO in close collaboration with other institutions and regulators.
  - Cooperation and data-sharing arrangements among regulators are working well but remain focused on preparation for CSF-UMOA. The CSF-UMOA provides a forum for exchanges; the "Note on Recent Developments in the WAMU Macrofinancial Environment" is the main document for discussion.
  - Since December 2014, CSF-UMOA regulatory members have a cooperation charter providing a framework for information exchange and mutual assistance; in practice information exchange remains limited and mostly geared to CSF-UMOA work preparation.
- Cross-border contagion arrangements:
  - Significant progress in coordinating monitoring of cross-border contagion risks via supervisory boards for certain WAEMU-based banking groups.
  - Over the past ten years, the CBU has entered into about 10 agreements with supervisors in several jurisdictions to establish information-sharing arrangements, supervisory boards, and crisis committees; these arrangements are generally working well with organized joint supervisory reviews.

### Transparency and communication
- Current state:
  - Implementation of macroprudential policy in WAEMU lacks transparency; systemic risk assessments and macroprudential policy actions adopted by authorities are not published.
  - The latest annual financial stability reports and reviews published by the BCEAO date back to 2008; press releases from CSF-UMOA and CPMP meetings are not published regularly.
- Recommendations:
  - Improve communication of systemic assessments and policy decisions in macroprudential publications to strengthen transparency.
  - Produce and publish systemic risk assessment reports regularly; publish minutes and press releases of CSF-UMOA meetings systematically.
  - CSF-UMOA and CPMP publications (press releases and annual reports) should be more risk focused and provide prescriptive comments on developments of key risks since the last assessment.
  - Use a core set of indicators more frequently in financial stability communications to better inform the public about CSF-UMOA decisions.

### Key statistics and simulation results (selected figures preserved exactly)
- BCEAO Basel III target:
  - Move gradually toward 100 percent (LCR) by 2028.
- Credit cycle synchronization:
  - Analysis period: between 1990 and 2020; exception: Mali not broadly synchronized with regional cycle.
- CSF-UMOA composition:
  - Government representatives: eight out of the thirteen seats on the CSF-UMOA.
- Table 2: Summary of contagion risk simulations — selected results (all values reported exactly as in source)
  - Contagion due to direct interbank exposures — Maximum capital losses (as a % of national GDP in 2021) and Additional capital needs (as a % of national GDP in 2021):
    - Benin: 1.2 ; 0.9
    - Burkina Faso: 0.3 ; 0.17
    - Côte d’Ivoire: 0.2 ; 0.12
    - Mali: 0.1 ; 0.08
    - Niger: 0.1 ; 0.04
    - Senegal: 0.6 ; 0.09
    - Togo: 1.1 ; 0.5
    - WAEMU: 0.7 ; 0.3
  - Contagion due to large common private exposures — Cumulative large exposures; Capital losses (% of regional GDP in 2021); Additional capital needs (% of regional GDP in 2021):
    - 10 largest common exposures: 1.62 ; 0.85
    - 20 largest common exposures: 2.21 ; 1.22
    - 30 largest common exposures: 2.66 ; 1.57
    - 40 largest common exposures: 2.99 ; 1.88
    - 50 largest common exposures: 3.25 ; 2.07
  - Contagion due to sovereign exposures — Issuing country defaulting in the regional securities market; Capital losses (% of regional GDP in 2021); Additional capital needs (% of regional GDP in 2021):
    - Benin: 0.2 ; 0.06
    - Burkina Faso: 0.4 ; 0.1
    - Côte d’Ivoire: 1.7 ; 0.8
    - Guinea-Bissau: 0.1 ; 0.04
    - Mali: 0.4 ; 0.15
    - Niger: 0.4 ; 0.13
    - Senegal: 0.2 ; 0.05
    - Togo: 0.2 ; 0.06
- Table 1: Cross-correlation matrix of credit cycles (selected exact values)
  - WAEMU–Benin: 0.7545*
  - WAEMU–Burkina Faso: 0.5645*
  - WAEMU–Côte d’Ivoire: 0.8647*
  - WAEMU–Guinea-Bissau: 0.5054*
  - WAEMU–Mali: 0.1000
  - WAEMU–Niger: 0.6503*
  - WAEMU–Senegal: 0.9157*
  - WAEMU–Togo: 0.5325*
  - Note: (*) Significant at 5 percent.

*Source: IMF staff and BCEAO material contained in the FSAP chapter.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2022/english/1wauea2022007.pdf_