A. Context

8. In 2007, a TA mission assisted the BCV in compiling MFS using the SRFs. The work focused on the compilation of central bank and ODC (SRF 1SR and 2SR) standardized reporting forms for monetary statistics. The mission worked in close collaboration with the staff of the BCV’s Statistics and Economic Research Department (DEE). The BCV currently use these SRFs to report monetary data to STA.

9. A number of differences in key monetary aggregates¹ are hampering IMF surveillance in Cabo Verde. In agreement with the BCV, AFR’s Cape Verde team requested TA to undertake a thorough review of the concepts and definitions used in the MFS published by the BCV and compare them with the concepts and definitions underlying the standardized reporting forms used to share data with the Fund (STA and AFR), to resolve inconsistencies identified in the data and provide recommendations to improve dissemination of consistent MFS following the MFSMCG.

B. The Financial Corporations Sector in Cabo Verde

10. Following the MFSMCG, the financial corporations sector in Cabo Verde should be divided in three subsectors. These subsectors are: (i) central bank; (ii) other depository corporations (ODCs), includes all financial corporations and quasi-corporations that have liabilities in the form of deposits or their equivalents included in the definition of broad money; and (iii) other financial corporations (OFCs), which includes the remaining financial intermediaries with liabilities that are not included in broad money, and financial auxiliaries. The entities belonging to the ODC and OFC subsectors, including its relative size measured by total assets, are shown in Table 2.

^{Table 2.}

Cabo Verde: Structure of the Financial Corporations Sector (Summary)

Source: Banco de Cabo Verde.

^{Table 2.}

Cabo Verde: Structure of the Financial Corporations Sector (Summary)

TYPE OF INSTITUTION	September 2020
	Total Assets (millions of escudos)	Relative Size (1) (in percent)
FINANCIAL CORPORATIONS	456,662	100
A. DEPOSITORY CORPORATIONS	357,909	78
1) Banco de Cabo Verde	84,674	19
2) Other Depository Corporations	273,235	60
Onshore Banks	273,235	60
B. OTHER FINANCIAL CORPORATIONS	98,753	22
1) Offshore banks	88,079	19
2) Insurance Companies	7,443	2
3) Other financial intermediaries and auxiliaries (2)	3,231	1
(1) Relative to financial corporations’ total assets. (2) Includes finance companies, microfinance institutions, venture capital companies, currency exchange agencies, insurance brokers, etc.

Source: Banco de Cabo Verde.

View Table

^{Table 2.}

Cabo Verde: Structure of the Financial Corporations Sector (Summary)

TYPE OF INSTITUTION	September 2020
	Total Assets (millions of escudos)	Relative Size (1) (in percent)
FINANCIAL CORPORATIONS	456,662	100
A. DEPOSITORY CORPORATIONS	357,909	78
1) Banco de Cabo Verde	84,674	19
2) Other Depository Corporations	273,235	60
Onshore Banks	273,235	60
B. OTHER FINANCIAL CORPORATIONS	98,753	22
1) Offshore banks	88,079	19
2) Insurance Companies	7,443	2
3) Other financial intermediaries and auxiliaries (2)	3,231	1
(1) Relative to financial corporations’ total assets. (2) Includes finance companies, microfinance institutions, venture capital companies, currency exchange agencies, insurance brokers, etc.

Source: Banco de Cabo Verde.

11. The financial corporations sector in Cabo Verde is dominated by banks. As shown in Table 2, in September 2020 onshore banks had 273,235 million escudos in total assets, representing 60 percent of the total financial corporations. The OFC comprise offshore banks, representing 19 percent of the total financial corporations sector; insurance companies, representing two percent; and remaining financial intermediaries and auxiliaries, for which total assets are very small.

12. All financial corporations are supervised by the BCV. The ODCs report aggregate monthly balance sheets for supervision and detailed monthly balance sheets for compiling the monetary statistics. The OFCs currently report aggregate monthly data only for supervisory purposes. OFC data are not included in the monetary statistics compiled and disseminated by the BCV.

A. Central Bank Survey: Analysis of Discrepancies

13. A number of discrepancies exist between the central bank survey disseminated by the BCV in its publications and the data sent to the IMF, through STA, for publication and surveillance purposes. These discrepancies were identified and discussed during the mission with the BCV’s MFS compilers. The MFS compilers agreed to eliminate these differences by adopting the methodology of the central bank survey compiled for the IMF, which is based on the MFSMCG. For example, this methodology includes the valuation of financial assets and liabilities including the accrued interest they generate, to better approximate the fair value of these instruments. Other discrepancies are described below.

14. Overall, discrepancies found are small, except for the net claims on central government, which are significant. A detailed analysis of the differences was made using December 2019 data as reference. A summary of the main analytical concepts and the differences are presented below.² Table 3 presents the specific types of assets and liabilities that explain the differences between IMF and BCV data for the central bank survey.

• Differences in net foreign assets are minor (71,869 vs 72,814 million escudos). See concepts 1 and 2 in Table 3.

• Differences in net claims on central government are significant (4,007 vs -3889 million escudos) See concept 3 in Table 3.

• Differences in claims on other sector are minor (1,000 vs 565 million escudos) See concept 4 in Table 3.

• Differences in monetary base are very minor (68,733 vs 68,710 million escudos).

^{Table 3.}

Central Bank Survey (1SG vs Consolidated Balance Sheet of BCV)

^{Table 3.}

Central Bank Survey (1SG vs Consolidated Balance Sheet of BCV)

Concept		IMF data (Central Bank Survey, 1SG)	BCV data (BCV Consolidated Balance Sheet)
1. External assets (claims on nonresidents
	Accrued interest on assets	Included	Not included
	SDR holdings	IMF record	BCV record
	Reserve position in the Fund	IMF record	BCV record
	Other foreign assets	Not included	Includes participation in international organization (Afrexibank)3
2. External liabilities (liabilities to nonresidents)
	Accrued interest on liabilities	Included	Not included
	SDR allocations	Included	Not included
3. Net claims on central government
	Securities and loans (assets)	Included	Excludes some securities (TÍT. DIV. PUB. ESTADOS DE C.V. DETIDOS PELO BCV) Excludes m/l credit to government and other minor items
	Deposits (liabilities)	Included	Not included
4. Claims on other sectors
Shares and other equity (nonfinancial corporations)		Included	Not included

View Table

^{Table 3.}

Central Bank Survey (1SG vs Consolidated Balance Sheet of BCV)

Concept		IMF data (Central Bank Survey, 1SG)	BCV data (BCV Consolidated Balance Sheet)
1. External assets (claims on nonresidents
	Accrued interest on assets	Included	Not included
	SDR holdings	IMF record	BCV record
	Reserve position in the Fund	IMF record	BCV record
	Other foreign assets	Not included	Includes participation in international organization (Afrexibank)3
2. External liabilities (liabilities to nonresidents)
	Accrued interest on liabilities	Included	Not included
	SDR allocations	Included	Not included
3. Net claims on central government
	Securities and loans (assets)	Included	Excludes some securities (TÍT. DIV. PUB. ESTADOS DE C.V. DETIDOS PELO BCV) Excludes m/l credit to government and other minor items
	Deposits (liabilities)	Included	Not included
4. Claims on other sectors
Shares and other equity (nonfinancial corporations)		Included	Not included

15. Recommendation: Eliminate the discrepancies between the central bank survey disseminated by the BCV in its publications and the data sent to the IMF by adopting the methodology of the central bank survey compiled for the IMF, which is based on the MFSMCG.

B. Other Depository Corporations Survey: Analysis of Discrepancies

16. Some discrepancies exist between the ODC survey disseminated by the BCV in its publications and the data sent to the IMF, through STA, for publication and surveillance purposes. These discrepancies were identified and discussed during the mission with the MFS compilers. The MFS compilers agreed to eliminate these differences following the implementation of a new compilation system for ODC. This compilation system was discussed during the mission and contain enough data to compile an ODC survey in line with the MFSMCG.

17. The new compilation system should produce data to be used consistently for national publications and to report these data to the IMF. The new source data are based on three elements: (1) detailed chart of accounts for banks, which is based on IFRS principles; (2) individual reporting of loans (loan-by-loan) and securities (security-by-security) held by banks; and (3) individual reporting of deposit liabilities (deposit-by-deposit). The combination of these three sources allow for the identification of the financial instruments, counterpart institutional sectors, and currency of denomination (national and foreign) recommended by the MFSMCG.

18. Overall, discrepancies found in current data are small, except for the reserve and other deposits of banks at the BCV. In the data reported to the IMF (2SR report form), these deposits include security deposits (depósitos de caução) that are excluded in the BCV data used to calculate the consolidated balance sheet of the deposit banks. A detailed analysis of the differences was made taking December 2019 data as reference. A summary of the main analytical concepts and the differences are presented below.⁴ Table 4 presents the specific types of assets and liabilities that explain the differences between IMF and BCV data for the ODC survey.

• Differences in net foreign assets are minor (-1,086 vs -2,100 million escudos). See concepts 1 and 2 in Table 4 below.

• Differences in net claims on central bank are significant (65,101 vs 60,362 million escudos). See concept 3 in Table 4.

• Differences in net claims on central government are minor (-19,715 vs -17,880 million escudos). See concept 4 in Table 4.

• Differences in claims on other sector are minor (123,506 vs 119,019 million escudos). See concept 5 in Table 4.

• Differences in broad money components are minor (189,106 vs 186,244 million escudos), although the composition (transferable vs other deposits) differ significantly. See concept 6 in Table 4.

^{Table 4.}

ODC Survey (2SG vs Consolidated Balance Sheet of ODC)

^{Table 4.}

ODC Survey (2SG vs Consolidated Balance Sheet of ODC)

Concept		IMF data (ODC Survey, 2SG)	BCV data (ODC Consolidated Balance Sheet)
1.External assets (claims on nonresidents
	Accrued interest on assets	Included	Not included
	Shares and other equity	Included	Not included
	Other accounts receivable	Included	Not included
2.External liabilities (liabilities to nonresidents)
	Accrued interest on liabilities	Included	Not included
	Other accounts payable	Included	Not included
3.Claims on central bank
	Reserve deposits - Security deposits (depósitos de caução)	Included	Not included
	Debt securities (Títulos do Banco Central (TIM e TRM)	Banks record (Valuation or time of recording issue)	BCV record
4.Net claims on central government
	Accrued interest on assets and liabilities	Included	Not included
	Deposits (liabilities)
5.Claims on other sectors
Classification of claims by institutional sector		Use of ponderations for banks not reporting accurate sectorization (not needed in the new compilation system)	Do not use ponderations
	Accrued interest on assets	Included	Not included
6.Broad Money components
	Transferable deposits	Higher than BCV data	Mostly a reclassification issue between with other deposits
	Other deposits	Lower than BCV data	Mostly a reclassification issue with transferable deposits

View Table

^{Table 4.}

ODC Survey (2SG vs Consolidated Balance Sheet of ODC)

Concept		IMF data (ODC Survey, 2SG)	BCV data (ODC Consolidated Balance Sheet)
1.External assets (claims on nonresidents
	Accrued interest on assets	Included	Not included
	Shares and other equity	Included	Not included
	Other accounts receivable	Included	Not included
2.External liabilities (liabilities to nonresidents)
	Accrued interest on liabilities	Included	Not included
	Other accounts payable	Included	Not included
3.Claims on central bank
	Reserve deposits - Security deposits (depósitos de caução)	Included	Not included
	Debt securities (Títulos do Banco Central (TIM e TRM)	Banks record (Valuation or time of recording issue)	BCV record
4.Net claims on central government
	Accrued interest on assets and liabilities	Included	Not included
	Deposits (liabilities)
5.Claims on other sectors
Classification of claims by institutional sector		Use of ponderations for banks not reporting accurate sectorization (not needed in the new compilation system)	Do not use ponderations
	Accrued interest on assets	Included	Not included
6.Broad Money components
	Transferable deposits	Higher than BCV data	Mostly a reclassification issue between with other deposits
	Other deposits	Lower than BCV data	Mostly a reclassification issue with transferable deposits

19. Recommendations: Use the new compilation system for ODC to eliminate the discrepancies between the ODC survey disseminated by the BCV in its publications and the data sent to the IMF. The new compilation system for ODC should follow the methodology of the current IMF report form 2SR, which is in line with the MFSMCG.

C. Compilation of the Depository Corporations Survey

20. The depository corporations (DC) survey should also be consistently compiled. This survey or analytical account (called Síntese Monetária in BCV terminology) result from the consolidation of the BCV survey and the ODC survey. The DC survey contain key aggregates for monetary analysis, such as broad money, external assets, and domestic credit by sector in the economy. To maintain consistency and transparency of official statistics, the mission recommended that these key aggregates are compiled with exact results both for national publications and international users, including the IMF.

21. Depending on user needs, the DC survey main contain different disaggregation. For example, in national publications the BCV may decide to provide more granularity or identify especial operations, such specific monetary policy instruments. At the IMF, the structure of the DC survey is also different depending on its intended use: publication in the IMF’s International Financial Statistics bulletin for general audience, or analysis by internal users, such as AFR. However, the different breakdowns of key analytical concepts should add up to the same figures for all users.

D. Compilation of the Other Financial Corporations Survey

22. The other financial corporations (OFC) are not required to report data for monetary statistics purposes, except for insurance companies. The OFCs report quarterly data (analytical balance sheet) following the instructions from the banking supervision department (BSD), and the annual report of the BCSTP already includes annual data on insurance companies. Although the analytical balance sheets contain more aggregated data, they are useful as a starting point for collecting and compiling statistics on the OFCs.

23. As presented in Table 2, the two main types of OFC in Cabo Verde are offshore banks and insurance companies. MFS compilers compile data on insurance companies only. However, historically important offshore banks are expected to become onshore by the end of 2021. Due to a change in legislation, Cabo Verde will no longer allow offshore banks to operate from its economic territory. Therefore, it is expected that the assets remaining in offshore banks will become onshore and their assets and liabilities will be added to the ODC survey.

24. Offshore banks should be included in the OFC survey. The mission explained that these banks are not money issuers in Cabo Verde because they are not allowed by law to accept deposits from Cabo Verdean residents. Therefore, they are not included in the category of ODC (i.e. in the SRF 2SR). Consequently, offshore banks should be included in the SRF 4SR, together with other types of financial intermediaries such as insurance companies.⁵

25. Growth rates for external assets and liabilities of ODC need to be properly calculated. Because of the likely impact of the transformation of the offshore banks into onshore entities, the ODC survey for end of 2021 will contain new assets and liabilities from the offshore banks than did not appear in the ODC survey of the previous period (month or year). For the correct calculation of growth rates between periods, these assets/liabilities should not be considered, as they represent a reclassification rather than new acquisition of assets or incurrence in liabilities.

26. Insurance corporations are also not included in the data shared with the IMF. However, this limitation is not important because of their relative size is small (only two percent of the financial system). Nonetheless, MFS compilers are very advanced in the incorporation of these corporations into the OFC survey (report form 4SR).

27. Recommendations: (i) Take into consideration the reclassification of offshore bank and incorporation into the ODC survey for the calculation of growth rates; and (ii) Disseminate data on insurance companies following the SRF 4SR, to be included in the OFC survey.

E. Resources and Technical Assistance

28. The BCV has adequate staff for compiling MFS. The staff are comprised of a balanced combination of very experienced and relative new MFS compilers. Most staff have attended MFS courses provided by the IMF or other institutions, such as the Bank of Portugal.

29. Hands-on training was well received but limited. More comprehensive training is recommended to familiarize staff with the new MFSMCG-based reporting forms, which are expected to be required by the end of 2021. The mission recommends the compiling staff members who have not yet attended any MFS courses to do so in the coming opportunities, which may be online.

F. DC Compilation System and Integrated Monetary Database

30. Since 2007, STA and AFR have shared an integrated monetary database (IMD) used for publication and operational needs. The mission discussed the importance of maintaining the IMD for Cabo Verde. The current IMD contains the SRFs for the central bank and ODC, and the corresponding surveys. The IMD contains data for December 2001 onward.

G. Compilation of Credit by Economic Activity

31. At AFR’s request, the mission discussed plans to improve the compilation and dissemination of data on credit by economic activity. MFS compilers are implementing a data collection system for credit by economic activity using the 2008 Classification by Economic Activities of Cabo Verde of the National Institute of Statistics (INE). According to INE, this classification was designed to preserve international statistical comparability, in a harmonized manner with the International Standard Classification of Economic Activities, Revision 4 (CITA-Rev.4) published by the United Nations Statistical Services (ST/ESA/STAT/SER.M/4/Rev.4).

32. Recommendation: Improve data on credit by economic activity using the Classification by Economic Activity of the INE.