A. Macrofinancial Developments

7. WAEMU has experienced a decade of sustained economic growth. The growth—which averaged 6.4 percent during 2012–2019—was stimulated by private investment, spurred by public spending and robust credit growth (Figure 1; Table 2). External demand remains an important growth driver, but rising economic diversification has reduced regional economies’ dependence on raw materials exports and, thus, their exposures to terms-of-trade shocks.

8. The authorities took measures to mitigate the negative impact of the COVID-19 pandemic on economic growth (Table 3). Authorities increased fiscal spending to support domestic demand and public deficits increased by 3.5 percent of GDP in 2020, reversing the process of fiscal consolidation. Growth in 2020 slowed to 2 percent. The BCEAO reduced the policy rate by 50 basis points to ease monetary policy. It also fully met banks’ short-term refinancing needs and expanded the eligible collateral pool to accommodate the sharply raised demand for precautionary liquidity. Banks were permitted to reschedule 3.1 percent of private sector loans until end-2020, of which only 6 percent (less than a fifth of a percentage point of total loans) were reclassified as nonperforming by end-September 2021. Most measures, except the new bank refinancing mechanism, have been phased out.

9. The economic outlook appears favorable but is subject to significant risks.² The economy is on the rebound, with GDP growth expected to have bounced back to 5.7 percent in 2021. Bank credit also rebounded, growing by nearly 11 percent year-on-year at end-October 2021. The strong economic activity and supply disruptions have spurred inflationary pressures, with the region-wide headline inflation exceeding the BCEAO’s target band since April 2021. The region is altogether vulnerable to a rise in global commodity prices. Recent global price increases could lead to higher interest rates to counteract inflation and defend the currency peg to the euro, particularly in case of a significant fall of official FX reserves, and lead to higher funding costs for regional sovereigns. Other important risks include a deterioration in the regional security situation (as already seen in Mali and Burkina Faso) and/or a resurgence of the COVID-19 pandemic. Official reserves are at a comfortable level, but narrower fiscal space could limit the policy capacity to respond to external demand shocks.³

B. Financial Sector Structure

10. The WAEMU’s financial system has grown considerably since the 2008 FSAP but is still dominated by the banking sector (Figure 2). It has doubled in size since 2008, with total assets at 69 percent of regional GDP at end-2020. Banks remain the principal actor in the financial system (74 percent of total assets) and follow a traditional business model of taking deposits and providing mostly short to medium-term credit to nonfinancial corporations as well as financing public spending via the regional bond market.

11. Credit growth has been steady but uneven across the region. Private sector credit has increased by an average of 12 percent a year between 2010 and 2019 but still accounts for only 25 percent of region-wide GDP and is uneven across countries. Lending to the private sector grew faster than the economy until 2018 in line with the need for financial deepening, but the credit gap has since turned negative. Agricultural (mostly seasonal) lending is limited, at 2 percent of private sector loans, despite the importance of agricultural production in some member countries, notably Burkina Faso (cotton) and Côte d’Ivoire (cacao). The availability of financial data has improved with the development of a regional credit reporting bureau (BIC) and a corporate financial analysis unit, but still does not support banks’ ability to diversify their credit portfolios.

12. Important structural changes have redefined the financial sector since the 2008 FSAP. These include:

• Conglomeration of the banking sector: Banking groups—primarily pan-African and regional groups—now account for 86 percent of banking assets.

• Rise in sovereign risk exposures: The share of sovereign debt reached 35 percent of banks’ total assets at end-2021 (up from 7 percent in 2004). WAEMU’s member governments have overall become more dependent on the region’s banks to finance public spending, with banks’ share of public debt exceeding the average for emerging markets and low-income countries.

• Banks’ shift to a structural liquidity deficit: The share of BCEAO short-term financing increased to 13 percent of total bank assets in 2021 from 2 percent in 2008, reflecting the discontinuation of monetary financing of the government deficits in 2010. From the perspective of vulnerabilities, this does not pose a concern since liquidity deficits can prompt more robust bank liquidity management and interbank market development.

Banks’ Holdings of Public Debt

(In percent of total assets)

Citation: IMF Staff Country Reports 2022, 136; 10.5089/9798400207976.002.A001

Sources: BCEAO and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Banks’ Holdings of Public Debt

(In percent of total assets)

Citation: IMF Staff Country Reports 2022, 136; 10.5089/9798400207976.002.A001

Sources: BCEAO and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

Banks’ Holdings of Public Debt

(In percent of total assets)

Citation: IMF Staff Country Reports 2022, 136; 10.5089/9798400207976.002.A001

Sources: BCEAO and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

International comparisons of banks’ sovereign exposures

(In percent of total public debt; November 2021)

Citation: IMF Staff Country Reports 2022, 136; 10.5089/9798400207976.002.A001

Source: IMF’s Sovereign Debt Investor Base for Emerging Markets.Note: For regions with asterisk, the value is the regional average; LICs = low income countries and EMDEs = emerging markets and developing economies.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

International comparisons of banks’ sovereign exposures

(In percent of total public debt; November 2021)

Citation: IMF Staff Country Reports 2022, 136; 10.5089/9798400207976.002.A001

Source: IMF’s Sovereign Debt Investor Base for Emerging Markets.Note: For regions with asterisk, the value is the regional average; LICs = low income countries and EMDEs = emerging markets and developing economies.

• Download Figure
• Download figure as PowerPoint slide

International comparisons of banks’ sovereign exposures

(In percent of total public debt; November 2021)

Citation: IMF Staff Country Reports 2022, 136; 10.5089/9798400207976.002.A001

Source: IMF’s Sovereign Debt Investor Base for Emerging Markets.Note: For regions with asterisk, the value is the regional average; LICs = low income countries and EMDEs = emerging markets and developing economies.

• Download Figure
• Download figure as PowerPoint slide

C. Recommendations of the 2008 FSAP

13. The authorities have implemented most recommendations of the 2008 FSAP (Table 4). Key achievements include: (i) the increase in banks’ minimum capital requirement; (ii) the reform of the BCEAO and the CBU—an important step toward strengthening regulatory independence and enforcement capacity; (iii) the transition toward the Basel II/III regulatory framework; (iv) the introduction of risk-based supervision; and (v) introduction of a new AML/CFT regulatory framework via the adoption of the Uniform Law in 2015 and its implementing regulations in 2017.

A. Financial Sector Risks and Vulnerabilities

14. Credit risk is the most important risk faced by the system. Nonperforming loans (NPLs) are high, even though they have declined to 11.2 percent of total loans in November 2021 from 20 percent at end-2006. This improvement in asset quality is less than expected after a long period of strong GDP and credit growth, all else equal. The absence of effective mechanisms for NPL recovery and the uneven enforcement of regulatory write-off rules contribute to persistent asset quality issues. There are also considerable disparities across countries and institutions, with NPLs in some member countries as high as 20 percent of loans.

15. The high concentration of banks’ lending and sovereign exposures exacerbates credit risks. In their lending activities, banks are exposed to a limited set of private counterparties. Banks’ single largest exposure exceeds the limit of 55 percent of regulatory capital (the current WAEMU standard) for over a third of reporting banks, and the 25 percent large exposure limit under the Basel standard for two thirds of the banks (Figure 3). The sovereign concentration risk is even more pronounced, especially in Benin and Togo. Banks’ securities portfolios are partially diversified due to holdings of various member states’ debt. Yet, the diversification effects are limited, given the bias toward holding own-country and Côte d’Ivoire debt securities, and the possible correlation of member states’ vulnerabilities.

Structure of Banks’ Public Securities Portfolio

(Percent of total assets)

Citation: IMF Staff Country Reports 2022, 136; 10.5089/9798400207976.002.A001

Source: BCEAO and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Structure of Banks’ Public Securities Portfolio

(Percent of total assets)

Citation: IMF Staff Country Reports 2022, 136; 10.5089/9798400207976.002.A001

Source: BCEAO and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

Structure of Banks’ Public Securities Portfolio

(Percent of total assets)

Citation: IMF Staff Country Reports 2022, 136; 10.5089/9798400207976.002.A001

Source: BCEAO and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

16. Contagion risks in the WAEMU are high, buoyed by the presence of multiple interconnectedness channels (Figure 4). These include: (i) mostly unsecured direct interbank exposures; and (ii) considerable common exposures to private and sovereign borrowers, including cross-border investments in other member states’ debt securities (e.g., Côte d’Ivoire).

17. The rise in banks’ sovereign exposures and greater reliance on BCEAO short-term funding have contributed to a rise in interest rate risk. Historically, the prevalence of short-term private credit claims in banks’ portfolios curbed this risk as maturity mismatches were minimal. The sharp rise in government debt holdings and expanded use of BCEAO short-term funding have raised banks’ asset-liability mismatches and the risk of a contraction in their interest rate margins should interest rates rise.

18. Banks’ capital buffers are limited and uneven across banks (Figure 5; Table 5).⁴ The system-wide capital adequacy ratio has risen, reaching 12.4 percent at end-June 2021, up from 10.5 percent at end-2018. This reflects an increase in the regulatory capital requirement in the context of the transition toward Basel II/III. Yet, banks’ capital positions are weaker than they seem. Full provisioning of NPLs would mostly absorb existing Tier 1 capital buffers, curbing banks’ capacity to cope with risk amplifiers (e.g., concentration and contagion) and emerging risks (e.g., interest rate risk).⁵ Disparities among institutions also persist, with 18 banks (10.2 percent of banks’ assets) not complying with the solvency norms at end-June 2021.

19. Liquidity risks are exacerbated by the underdevelopment of the regional bond markets (Figure 6). Banks’ funding, largely based on customer deposits, has been stable. Yet the heavy skew of banks’ funding base toward large deposits—with the five largest depositors on average accounting for 25 percent of deposits—amplifies the impact of liquidity shocks. The illiquidity of the secondary markets for government securities hampers banks’ ability to mitigate such shocks. It also stands in the way of developing the repo market, which offer a more robust form of interbank funding than the prevalent unsecured transactions. Finally, banks do not actively use debt issuances to curtail maturity mismatches and interest rate risks.

20. The quality of banks’ internal risk management systems is heterogeneous.⁶ Only 35 percent of banks systematically require approval by their risk management department before making a loan. Banks also do not uniformly develop formal audit plans or risk maps. Credit risks are monitored via analysis of borrowers’ financial statements and internal ratings, but credit risk analysis tools are rudimentary and suffer from the lack of borrower data. Few banks use external ratings and reporting bureaus. Only 50 of 152 banks report conducting regular solvency and liquidity stress tests, which are often limited in terms of scenarios and coverage.

21. Climate risk in the region is recognized as important but is inadequately monitored. The WAEMU countries, particularly those in the Sahel region, are highly exposed to the effects of climate change. At the sectoral level, the agriculture sector—particularly the production of cotton—is already experiencing its adverse impact. Banks’ exposures to the sector are limited and generally covered by government guarantees. Only less than a third of banks consider themselves capable of identifying assets that are directly exposed to climate risks.

B. Banking System Stress Tests

22. Solvency tests evaluated the resilience of the banking system under a severe adverse scenario (Figure 7; Appendices II and III). The adverse and baseline scenarios were estimated via a “growth-at-risk” model, which assumed a U-shaped post-COVID-19 recovery under the adverse scenario and a V-shaped recovery under the baseline scenario.^7,8 The adverse scenario entails an accelerated fiscal adjustment; a regionwide resurgence of the COVID-19 pandemic; a deterioration in the region’s security situation; an external shock due to surging commodity prices; and a climate shock that captures temperature variation, frequency of catastrophic events, and agricultural production. Regional real GDP growth cumulatively declines by 15 percentage points relative to the baseline over four years (2.3 standard deviations from historical mean in the peak year of stress).

23. The solvency test results indicate moderate system-wide recapitalization needs but vulnerabilities for many small banks. Probabilities of default were estimated by bank type (with banks clustered via a statistical model) to account for heterogeneity within the sector and by degree of risk (estimated by quantile regression). In the adverse scenario, banks’ capital adequacy ratios could decline to 5.4 percent. Recapitalization needs are estimated to be limited, at up to 1.1 percent of regional GDP, reflecting the small size of banks’ risk weighted assets and the soundness of larger banks.⁹ This range is largely similar across the region (in terms of national GDP), except for Guinea Bissau and Togo where higher vulnerabilities and the larger size of the banking sector (in the case of Togo) put recapitalization needs at close to 3 percent of national GDP. Region-wide, the vulnerabilities center on small banks, 35 of which see their capital positions fall below the minimum requirement in the adverse scenario, joining another 20 banks that are already undercapitalized.

24. Concentration risks pose an additional challenge for banks in the region, accounting for significant recapitalization needs in some countries. A reverse stress test estimated a continuum of scenarios entailing a deterioration in the quality of banks’ largest exposures up to a “breaking point” of full depletion of the capital buffers (as defined for the solvency test). Region-wide, most banks’ capital buffers are found inadequate to cover the loss of even their single largest exposure. Recapitalization needs due to losses on the largest and up to the ten largest exposures are at 1 to 3 percent of regional GDP (Figure 3). At the country level, these needs range from 1.5 percent of national GDP for Côte d’Ivoire, the largest and most diversified economy in the Union, to 7 percent for Togo.

25. Many banks are exposed to contagion risks, but recapitalization needs should be modest. Contagion tests combine a credit shock and a funding shock propagated in the regional network via interbank and common exposures.¹⁰ The failure of WAEMU’s largest banking group would lead to the failure of 23 other banks. Yet, the recapitalization needs from a failure of all interbank positions would be small, at 0.3 percent of regional GDP. These needs are also small at the country level, at less than 0.2 percent of national GDP, except for Burkina Faso (close to 1 percent) and, to a lesser extent, Togo (Figure 8; Table 7). Recapitalization needs rise to about 0.8 percent of regional GDP in case of concomitant defaults of banks’ 10 largest common private exposures or defaults on debt payments due in 2021 and 2022 by the largest common sovereign exposure.

26. Interest rate stress tests suggest that inflation and, indirectly, higher interest rates have an impact on banks’ solvency. An “inflation-at-risk” model linked inflation to domestic and global macrofinancial conditions and evaluated the impact on banks’ profitability and solvency. An adverse scenario assumed an inflation uptick to 7 percent in the first year and a gradual reduction to the BCEAO’s 2-percent target (the baseline) over four years. The test results suggest that banks’ return on assets would decline by 2.5 percentage points. Capital adequacy ratios would decline to 3.6 percent, with recapitalization needs reaching 1.5 percent of regional GDP (Figure 9). Higher inflation could exacerbate the impact of a growth slowdown on banks’ solvency (as evaluated in the solvency test) and thus further increase recapitalization costs.

27. Liquidity test results show that banks’ liquidity gaps under stress are moderate, but the capacity to withstand liquidity pressures is heterogeneous. Stress tests of liquidity coverage ratios (LCR) assessed banks’ short-term resilience to sudden deposit outflows. In the event of a severe stress, most banks (58 of 91) are unable to withstand the cash outflows (Figure 10). Yet, the system-wide liquidity gap, even in the most adverse scenario, appears manageable at 2.1 percent of GDP. The inability to offset liquidity risks is most pronounced among small banks and the banks of certain member countries.

C. Policies to Reduce Vulnerabilities

28. An additional capital requirement should be calibrated to discourage banks’ excessive concentration of sovereign exposures. The calibration should be nonlinear, with the requirement increasing gradually beyond a minimum concentration threshold based on the level of a bank’s exposure to a specific sovereign relative to its risk-weighted assets (Figure 11). The additional capital requirement should be based on Basel Pillar 2 or, alternatively, on Pillar 1 with an accompanying macroprudential rule. The difference in the two approaches relates to the institution in charge (the CBU for Pillar 2 and the BCEAO for Pillar 1) and the extent of public disclosure (the Pillar 1 approach entails publication of regulations).

29. The supplementary capital requirement could be extended to cover other risk types, including:

• Large exposures to private borrowers. The additional requirement (Pillar 2) has the advantage of accounting for more specific concentration factors, such as the correlation of risks between borrowers and sectoral exposures. It is an important complement to the Basel large exposure limit—which the authorities are strongly encouraged to implement as planned by 2023—to offset concentration risks in the banking sector.

• Interest rate risk. The additional requirement (Pillar 2) should be commensurate with the maturities and interest rate mismatches in banks’ balance sheets, which necessitates measuring these on a regular basis.

30. The mission supports the authorities’ efforts to encourage banks to internalize liquidity risks. An LCR requirement is expected to be introduced soon and then gradually increased to 100 percent by 2028. It is important that the requirement abides by the following:

• The haircuts on high-quality liquid assets (HQLA) should reflect the assets’ market liquidity, as revealed by liquidity indicators, to achieve risk equivalence. If the regulator opts not to apply different haircuts across sovereign issuers, a uniform haircut should be introduced at a minimum.

• Required reserves could also be counted toward HQLA without applying a discount since they are available to absorb temporary liquidity needs.¹¹

• Over time, LCR requirements should be differentiated based on banks’ relative liquidity risk profiles under the Basel Pillar 2 approach.

31. The introduction of Basel liquidity ratios will require a systematic data collection process by the supervisor. To permit the introduction of higher Pillar 2 requirements on banks more prone to larger outflows, the regulator should develop capacity to monitor the distribution of monthly changes in banks’ various funding sources. Introducing the long-term liquidity ratio should entail regular reporting requirements on the residual maturities of banks’ assets and liabilities to enable ongoing monitoring of maturity mismatches.

A. System-level Oversight Institutional Architecture

32. WAEMU’s institutional framework for financial stability and the legal mechanisms underpinning its functioning have seen substantial progress since the 2008 FSAP. Institutional reforms have clarified the mandates of the BCEAO and the CBU and strengthened the CBU’s legal autonomy and enforcement powers. In addition to the objective of preserving price stability, the BCEAO has explicitly been charged with the core task of ensuring financial stability. The law on the banking regulatory framework has established an overarching framework for the supervision of banks’ activities, which has been made more proactive and risk based with the gradual implementation of the Basel II/III framework since 2016. A macroprudential policy framework, a bank resolution regime, and a deposit guarantee fund—whose mandate was extended to bank resolution funding—have been introduced.

33. The CBU’s independence from member states should be strengthened. The principle of independence, which prohibits CBU members from receiving instructions from external entities, including member states, should be explicitly enunciated in the CBU’s governing documents (specifically the Annex to the CBU’s governing Convention). The CBU’s governing documents should also stipulate that commissioners serving on the Supervisory Board because of their status or position in a WAEMU member state’s administration (eight of the 16 members) should do so in a non-voting capacity. As an alternative, the CBU’s composition could be amended to increase representation of CM-nominated commissioners based on professional qualifications. Finally, maintaining the CM’s appellate jurisdiction over CBU decisions raises institutional independence concerns. As per good practices, such decisions should only be appealed before the WAEMU’s Court of Justice, which also has jurisdiction on these matters.

B. Banking Supervision and Regulation

38. The BCEAO has undertaken an ambitious regulatory reform. The 2017 publication of four circulars on governance, risk management, internal control, and compliance has aligned the regulatory environment more closely with international best practices and have imposed stricter regulatory requirements on credit institutions. Substantial progress has been made on aligning regulatory capital requirements with Basel standards. The regulatory framework on liquidity risk management standards, published in 2017, appears complete and consistent with the Basel Core Principles. This regulatory effort has consolidated the prudential base and established the conditions for strengthened supervision.

39. The offsite supervision capacity for assessing banking sector risks should be enhanced. Specifically, off-site supervision should conduct cross-cutting analyses more frequently and its systemic risk map should be further developed. The stress testing expertise of the Secretariat General of the CBU (SGCB) should be strengthened to ensure adequate risk oversight. The methodology for rating credit institutions should be amended to: (i) account for market and interest rate risks; and (ii) weigh risk factors by their relative importance, with a higher weight for concentration risk, given its importance in banks’ risk profiles. The robustness and performance of the rating system should be assessed on a regular basis.

40. The resources of the CBU and its secretariat should be strengthened. The SGCB teams are experienced and qualified, but their capacity should be strengthened further to respond adequately to the changing banking sector landscape, including the rising complexity and stronger presence of cross-border groups, which give rise to new risk types. Continued IT investments will be needed to optimize and automate the off-site supervision’s toolkit.

41. The supervisory authority’s preventive efforts should be accompanied by vigorous deterrent actions. The CBU should strengthen the frequency and intensity of onsite supervision, particularly the inspection of banks’ governance and risk management. The CBU has various types of sanctions at its disposal, which it can apply separately or jointly. Yet, it has not been sufficiently strict toward entities that violate prudential regulations over extended time periods, particularly in enforcing the minimum capital requirements. The CBU tends to rely on unpublished written censures and on stays of proceedings—at times used repeatedly. Censures and sanctions should be published to enhance their effectiveness, with monetary sanctions used more frequently. Repeated stays of proceedings should be avoided to the extent possible.

C. Supervision of Money Laundering and Terrorism Financing Risks

42. Despite important reforms to enhance the supervisory framework for anti-money laundering/combating the financing of terrorism (AML/CFT), significant weaknesses persist. The CBU has made important reforms to its AML/CFT supervisory program following the adoption of a new Uniform Law in 2015 and the issuance of implementing regulations in 2017. These include the introduction of a new overall supervisory risk assessment model that assigns risk ratings to each bank, the administration of an AML/CFT-specific bank questionnaire, and the issuance of updated onsite inspection checklists.¹³ Yet, significant weaknesses persist: (i) the offsite supervision program for AML-CFT is underdeveloped and not fully risk based; (ii) the AML/CFT-related inputs to the risk assessment model are unduly limited; (iii) the AML/CFT onsite inspection program requires methodological improvements in certain key areas, including to distinguish between lower-and higher-risk domestic politically exposed persons (PEPs); and (iv) the CBU does not cooperate effectively with key partners in and outside the region.

43. The authorities should undertake further steps to fully implement the AML/CFT supervisory program based on a risk-based approach and enhance supervisory capacity. These include: (i) expanding the role of offsite supervision and moving to a risk-based model; (ii) requiring banks to submit, at least annually, basic clientele statistics to serve as additional ML/TF risk indicators, including on the number of foreign customers and high-risk domestic PEPs, and residents of higher-risk regions or countries; (iii) strengthening the onsite inspection program by enhancing the methodology with respect to politically exposed persons, suspicious transaction reporting, and targeted financial sanctions; (iv) improving domestic and international cooperation by eliminating obstacles to information exchange between the CBU and national financial intelligence units, and making better use of existing powers and tools for information exchange with foreign counterparts; and (v) considering the establishment of an internal pool of AML/CFT expertise to further develop AML/CFT supervisory capacity.