A. Context

10. In 2007, a TA mission assisted the BCU in compiling MFS using the SRFs. The work focused on the compilation of central bank and ODC SRFs. The BCU currently use the SRFs to report monetary data for the central bank and ODC to the IMF. FFI (offshore banks), which only represented 0.01 percent of total ODC assets in December 2018, are reported as OFC to STA. These banks represented around 14 percent of total ODC assets back in 2014. However, due to legislative changes offshore banks have almost disappeared from the financial landscape of Uruguay, representing only 0.01 percent of total ODC assets in December 2019.

11. This mission, requested by the BCU and supported by WHD, focused on the compilation of OFC’s data, i.e., the non-deposit taker financial institutions. The OFC subsector is a significant part of the financial sector. According to the BCU’s 2018 Annual Report, the assets of the three main OFCs in Uruguay (IC, PF, and CAC) were of USD5,011 million, USD15,438 million, and USD1,116 million, respectively. These assets account for more than 36 percent of total financial sector assets; consequently, the compilation of the OFC survey is a high priority task.

B. The Financial System in Uruguay

12. The financial system of Uruguay is dominated by banks, but the size of non-bank financial intermediaries is significant. As shown in Table 2, in September 2019 ODC had 1,410,150.03 million pesos in total assets. The OFCs represented around 56 percent of total ODCs. Among OFCs, PF and IC were the largest components, with 39 percent and 13 percent respectively, followed by CAC, representing 3 percent. The ODCs include two public banks and 9 private banks. In addition, there is one cooperative and one financial house. The OFCs comprise 16 IC, 4 PF, 13 CAC, and 1 FFI.

^{Table 2.}

Structure of the Financial System (Summary) (September 2019)

(1) Relative to other depository corporations’ total assets.Source: Central Bank of Uruguay.

^{Table 2.}

Structure of the Financial System (Summary) (September 2019)

Type of Institution		Sep-19
		Total Assets (millions of pesos)	Relative Size (1) (in percent)
FINANCIAL CORPORATIONS		3,032,179.91
A. DEPOSITORY CORPORATIONS		2,249,535.92
1) Central Bank of Uruguay		839,384.98	60
2) Other Depository Corporations		1,410,150.93	100
	Public banks	698,952.21	50
	Private banks	708,023.92	50
	Cooperatives	810.40	0
	Financial houses	2,364.40	0
B. OTHER FINANCIAL CORPORATIONS		782,644.00	56
1) Pension Funds		551,533.00	39
2) Insurance Corporations		188,103.56	13
3) Credit administration companies		36,774.61	3
4) Foreign financial institutions		1,908.68	0
5) Financial services companies		3,051.03	0
6) Exchange houses		1,273.12	0

(1) Relative to other depository corporations’ total assets.Source: Central Bank of Uruguay.

View Table

^{Table 2.}

Structure of the Financial System (Summary) (September 2019)

Type of Institution		Sep-19
		Total Assets (millions of pesos)	Relative Size (1) (in percent)
FINANCIAL CORPORATIONS		3,032,179.91
A. DEPOSITORY CORPORATIONS		2,249,535.92
1) Central Bank of Uruguay		839,384.98	60
2) Other Depository Corporations		1,410,150.93	100
	Public banks	698,952.21	50
	Private banks	708,023.92	50
	Cooperatives	810.40	0
	Financial houses	2,364.40	0
B. OTHER FINANCIAL CORPORATIONS		782,644.00	56
1) Pension Funds		551,533.00	39
2) Insurance Corporations		188,103.56	13
3) Credit administration companies		36,774.61	3
4) Foreign financial institutions		1,908.68	0
5) Financial services companies		3,051.03	0
6) Exchange houses		1,273.12	0

(1) Relative to other depository corporations’ total assets.Source: Central Bank of Uruguay.

13. The OFC in Uruguay are supervised by BCU and source data are readily available to compile SRF 4SR. A description of the frequency of reported data, the time lag, and time series availability by main types of OFC is presented in Table 3 below.

^{Table 3.}

Types of OFC and Frequency of Data Reported to the BCU

^{Table 3.}

Types of OFC and Frequency of Data Reported to the BCU

OFC Type	Frequency	Time lag	Time series availability
Insurance Companies	Quarterly	30 days	September 2001
Pension Funds	Monthly	10 business days	October 2002
Credit Administration Companies	Monthly	7 business days	December 2012
Foreign Financial Institutions (offshore banks)	Monthly	7 business days	December 2001

View Table

^{Table 3.}

Types of OFC and Frequency of Data Reported to the BCU

OFC Type	Frequency	Time lag	Time series availability
Insurance Companies	Quarterly	30 days	September 2001
Pension Funds	Monthly	10 business days	October 2002
Credit Administration Companies	Monthly	7 business days	December 2012
Foreign Financial Institutions (offshore banks)	Monthly	7 business days	December 2001

A. Compilation of the Other Financial Corporations Survey: Overview

14. Source data for OFC are of good quality and allow for the compilation of quarterly SRF 4SR in line with MFSMCG. In general, information in primary data sources (accounting catalogs) are available with enough level of detail for the compilation of 4SR. Secondary data sources provide more granular information and were used to improve the data mapping to 4SR as needed; however, IC data mapping require further improvements (see IC subsection).

15. The mission compiled the SRF 4SR in coordination with compilers and in consultation with experts of the most relevant types of OFC. The mission reviewed the data sources for each type of OFC and prepared the compilation systems, conducting mapping exercises between available source data and the SRF 4SR for one reference period. The mapping was done in consultation with experts on IC, PF, and CAC and based on more granular information available in secondary data sources as needed. This allowed for improved classification of the accounts by currency, counterpart sector, and instrument.

16. Preliminary compilation of the SRF 4SR has been done according to the new SRFs, which follow the new MFSMCG. Further, the new SRFs have been mapped to the current SRFs; hence, BCU compilers would be able to report OFC data using the current and the new forms, when requested to do so. The expectation is that STA will start collecting MFS data in the new SRFs by end-2020.

17. It is essential for data analysis to compile historical OFC data. The mission recommended to compile and disseminate the SRF 4SR starting at least from December 2014 and report it to STA. Data should be reported on quarterly basis covering IC, PF, CAC and FFI.

18. Valuation of financial instruments, including recording of accrued interest, broadly follows international accounting standards. Provisions are separately identified in the asset side with negative sign; following the MFSMCG recommendation, in the SRF they are included as liabilities and recorded in Provisions for losses – Other – other accounts payable. Some negotiable instruments are valued at nominal value, but no adjustments to fair values are made.³

19. The BCU is moving towards the broad implementation of the International Financial Reporting Standards (IFRS). The BCU has recently introduced IFRS for banks and CAC. Their data are reported according to the IFRS since January 2019 and compilers are in the process of capturing information based on the IFRS. The BCU is to assess the impact of the IFRS on the data and document it in the metadata to inform users on any methodological changes that could affect time series analysis. Based on a preliminary assessment, compilers were of the view that the quantitative impact would be marginal and could affect asset valuation.

20. Recommendations: (i) Finalize the compilation of the SRF 4SR for one period, covering IC, PF, CAC and FFI, and using secondary available data sources to improve currency, counterpart sector, and instrument classification for IC (priority recommendation, deadline April 2020); and (ii) Compile and disseminate the SRF 4SR on quarterly basis covering IC, PF, CAC and FFI starting from December 2014, and report it to STA (priority recommendation, deadline March 2021).

21. The following sections summarize the findings on the types of OFC covered by the mission.

B. Insurance Companies

22. Source data are adequate to compile the SRF 4SR for IC. For some items, the information available in the primary data source is not detailed enough to obtain the breakdowns by currency, counterpart sector, and financial instrument; in these cases, the mapping was based on assumptions. However, more detailed information is available in secondary data sources, which will allow for improved classification. The mission, in collaboration with the compilers, identified the accounts for which secondary data sources are needed. Compilers should use the secondary data sources to improve the classification by currency, counterpart sector, and instrument.

23. Obtaining counterpart sector information from secondary data sources poses a challenge due to the high level of data granularity. The information in secondary data sources is organized in pivot tables. The high level of granularity by counterpart sector makes it difficult to extract data efficiently for compilation purposes. Sectorization ca n be based on a manual process (by accessing the detailed information through the currently available pivot tables), or on a more automatized fashion (by previously grouping institutional units into sectors). It is best practice among country’s compilers to create a list of institutional units and their correspondence with institutional sectors in a coordinated manner with other statistical domains, such as national accounts, external, and fiscal statistics. These lists identify units in the government sector (at central, state and local level), public enterprises, financial corporations, etc., assuring consistent allocation into institutional sectors. Units not specified in the list are treated residually as pertaining to the nonfinancial corporation’s sector or households .⁴ The BCU is to assess the method for sectorization that bes t fits its needs, based on available resources.

24. The bulk of the IC assets are debt securities (83 percent), of which 75 percent are issued by central government. The bulk of IC liabilities are insurance, pension, and standardized guarantee schemes (82 percent), of which more than a half correspond to nonlife insurance to households.

25. There are asset trade credit and advances related to insurance. A peculiarity of IC in Uruguay is that they provide coverage to policyholders prior to the payment of the premiums, hence IC finance policyholders through the provision of trade credits.

Table 4.

Simplified Balance Sheet of IC

Table 4.

Simplified Balance Sheet of IC

Concept	Percent
Currency and deposits	5
Debt securities	83
Insurance, pension, and standardized guarantee schemes (IPSGS) – Asset	2
Other accounts receivable (OAR)	8
Nonfinancial assets	3
TOTAL ASSETS	100
Insurance, pension, and standardized guarantee schemes (IPSGS) – Liability	82
Other accounts payable (OAP)	6
Equity Liability	12
TOTAL LIABILITIES	100

View Table

Table 4.

Simplified Balance Sheet of IC

Concept	Percent
Currency and deposits	5
Debt securities	83
Insurance, pension, and standardized guarantee schemes (IPSGS) – Asset	2
Other accounts receivable (OAR)	8
Nonfinancial assets	3
TOTAL ASSETS	100
Insurance, pension, and standardized guarantee schemes (IPSGS) – Liability	82
Other accounts payable (OAP)	6
Equity Liability	12
TOTAL LIABILITIES	100

26. Recommendation: Create a list of institutional units grouped into sectors in coordination with other statistical domains, such as national accounts, external, and fiscal statistics ⁵.

C. Pension Funds

27. In Uruguay, pension funds (PF) are the largest financial intermediary after banks, measured by total assets. In September 2019, total assets of these funds were 554 billion pesos, representing 39 percent of total bank assets. Pension funds are administered by four administradoras de fondos de ahorro previsional (AFAPs). In MFS, AFAPs are classified as financial auxiliaries and their balance sheets are negligible compared to the balance sheets of the PF under their administration. Therefore, compilation of data on PF focused on the balance sheets of the PF, not on the AFAP’s.

28. Aggregated PF data show they invest primarily in debt securities. 88 percent of total assets were invested in debt securities issued mostly by the central government and central bank. Liabilities are mostly in the form of pension entitlements (Table 5).

^{Table 5.}

Simplified Balance Sheet of Pension Funds

^{Table 5.}

Simplified Balance Sheet of Pension Funds

Concept	Percent
Currency and Deposits	10
Debt securities	88
Loans and other accounts receivable	2
TOTAL ASSETS	100
Pension Entitlements	99
Other accounts payable	1
TOTAL LIABILITIES	100

View Table

^{Table 5.}

Simplified Balance Sheet of Pension Funds

Concept	Percent
Currency and Deposits	10
Debt securities	88
Loans and other accounts receivable	2
TOTAL ASSETS	100
Pension Entitlements	99
Other accounts payable	1
TOTAL LIABILITIES	100

29. PF’s source data allow for the compilation of high quality SRF 4SR. The mission used available accounting data for PF to compile the 4SR for PF. These data provide enough information on the financial instrument, currency of denomination, and counterpart sector for most of the accounts. One outstanding issue under investigation is the nature of the investment of PF in financial trusts,⁶ which represent 14 percent of their assets. These investments were classified as debt securities, estimating that equity participations in financial trusts were small.

D. Credit Administration Companies

30. The CAC are natural or legal persons that regularly and professionally intervene in the financing of the sale of goods and services made by third parties, granting credit using cards, purchase orders, or other similar modalities. Other entities considered as CAC are consumer cooperatives, civil associations and other legal entities with a non-financial line of business that regularly issue purchase orders as long as such activity is significant within the set of activities that make up their line of business.

31. The core business of the CAC is to finance household consumption and, to a lesser extent, small and medium enterprises. The CAC facilitate the lending of money circumventing traditional financial intermediaries and charging a relatively high interest rate.

32. Source data for CAC are of good quality for 4SR compilation. Breakdown by resident and nonresident counter parts, national and foreign currency, and financial instrument are readily available; however, counterpart institutional sector information is available with some gaps. The mapping between source data (accounting catalog) and the sectoral balance sheet 4SR was done in consultation with experts on this type of entities and retrieving information from more detailed secondary data sources. For some items where the counterpart sector was not specifically identified, the breakdown was based on the type of credit provided (for example, credit for consumption was allocated to households, while commercial credit was allocated to nonfinancial corporations).

33. The mission focused on the mapping of all accounts with non-zero values. Because CAC use the same accounting catalog as banks, many accounts are not applicable to their business and, therefore, have zero values. Accounts with zero values that were not mapped were properly identified and flagged, so that compilers can easily monitor them. If any of these accounts show a non-zero value in the future (or if new accounts were created), compilers should map them into the 4SR.

34. The bulk of the CAC assets are loans (81 percent), of which 97 percent are extended to households. Most of the loans (96 percent) are granted in national currency. On the liability side, 36 percent correspond to equity liabilities, 40 percent to other accounts payable (of which 68 percent are provisions for losses), and 24 percent to loan liabilities (of which 34 percent are to ODC and 61 percent to nonfinancial corporations – mostly businesses adhered to the credit cards and purchase orders system).

Table 6.

Simplified Balance Sheet of Credit Administration Companies

Table 6.

Simplified Balance Sheet of Credit Administration Companies

Description	Percent
Currency and deposits	6
Debt securities	3
Loans – Asset	81
Other accounts receivable (OAR)	6
Nonfinancial assets	3
TOTAL ASSETS	100
Loans – Liability	24
Other accounts payable (OAP)	40
Equity Liability	36
TOTAL LIABILITIES	100

View Table

Table 6.

Simplified Balance Sheet of Credit Administration Companies

Description	Percent
Currency and deposits	6
Debt securities	3
Loans – Asset	81
Other accounts receivable (OAR)	6
Nonfinancial assets	3
TOTAL ASSETS	100
Loans – Liability	24
Other accounts payable (OAP)	40
Equity Liability	36
TOTAL LIABILITIES	100

E . Mutual Funds and Financial Trusts

35. Other financial intermediaries for which data were not readily available during the mission include investment funds and financial trusts. The mission requested information on the asset size of investment funds and financial trust, but this information is not available. Financial trusts do not follow a common accounting catalog, which makes it very costly to compile data on them.

36. Recommendation: Investigate the size and available data on investment funds and financial trusts. If their size is significant, include them in the SRF 4SR to be reported to STA.

F. Balance Sheet Approach

37. The BSA is a systematic analytical framework for exploring how balance sheet weaknesses can contribute to the origin and propagation of financial crises. BSA departs from the traditional analysis based on the examination of flow variables such as current account and fiscal balance. It focuses on the examination of stock variables in a country’s sectoral balance sheets and its aggregate balance sheet (assets and liabilities); and thus, facilitates the identification and analysis of vulnerabilities (imbalances).

38. In Uruguay, the improved MFS coverage including OFC will allow for BSA analysis to be conducted. The most important building block for BSA analysis are the data on the entire financial system, including the most relevant OFCs, which will be available as soon as the 4SR historical series are compiled. The other important data sources are quarterly international investment position and general government financial balance sheet, which are also available in Uruguay.

39. The mission conducted a workshop on the BSA, comprising a presentation and a case study. The workshop was attended by compilers and analysts of national accounts, external, monetary, and fiscal statistics. The workshop provided a historical perspective of the evolution of flow-of-funds compilation in OECD countries and the links with BSA. The key messages of the workshop were: (i) the interest by the international community since the 1990s to devote resources to compile intersectoral positions data, in addition to the flow-of-funds data currently compiled by many countries; (ii) the need for more frequent (quarterly vs annual) and timely data for macro-financial stability analysis; and (iii) the focus on a from-whom-to -whom approach, i.e., allowing for the identification of creditor and debtor sectors and the financial dependence among them.⁷

40. The mission encouraged the compilation of BSA matrices on a regular basis. Ideally, BSA matrices should be compiled by specialists in intersectoral financial analysis, such as those responsible for the compilation of sectoral accounts, flow-of-funds accounts, or financial statistics. In Uruguay, the compilation of sectoral accounts by national accounts compilers is in an early stage. In addition, the Financial Analysis Department is responsible for intersectoral financial analysis in connection with the financial programming exercise. This department uses flow data mostly, whereas the BSA focuses on positions. Given their experience and the need for very timely statistics, this department could be a candidate for compiling BSA data. Other possible candidate is the department responsible for macro-financial stability, which was not represented in the workshop.

41. The authorities noted that the BCU should be informed of any plans to compile and analyze BSA matrices by IMF users, such as Uruguay’s WHD team. This is so because the BCU has more granular data than those available in STA, which can be used to replace STA estimates in the BSA matrix. For example, for certain items in the matrix for which international data sources does not provide counterpart sectors, such as direct investment, the BSA data entries are estimated using data system of predefined allocations.

42. Recommendation: Using more granular, national data than those available in STA, compile quarterly BSA matrices once the OFC historical series are available, and share these data with WHD on request.

G. Financial Soundness Indicators

43. The mission met the Financial Services Superintendent to share information on the status and use of FSI within the IMF. The mission explained that more than 140 countries now report FSI to STA, which share these data with other departments, in particular with WHD and the Monetary and Capitals Markets (MCM) Department, for their analysis and publications. FSI are widely used in Article IV consultation reports and in the MCM’s flagship publication, the Global Financial Stability Report.

44. The mission stressed the need to improve the frequency of the FSI data reported by BCU to STA. While the great majority of countries report monthly or quarterly FSI data to STA, the BCU only submit annual data, which prevents timely macro-financial stability analysis. The authorities committed to start FSI data reporting on a quarterly basis. Monthly FSI reporting should be considered as a medium-term goal to better support the analytical needs of WHD.

45. In addition, the FSI coverage could be expanded. Current FSI data reported to STA focuses on deposit-taking corporations only. However, the new FSI guide published in December 2019⁸ significantly expand the number of FSI of other sectors /subsectors of the economy. In particular, the new guide recommends the compilation of sectoral financial statements and memorandum items for three types of non-bank financial intermediaries (IC, PF, and money market funds).

46. STA plans on launching the request for the new FSI in September 2020. STA will request countries to report sectoral financial statements, memorandum items, and underlying series only. With these data, the new list of FSI will be compiled by STA. The mission alerted the authorities of this new approach, as the BCU stopped reporting sectoral financial statements in 2012. Therefore, the mission recommended the BCU to resume reporting of sectoral financial statements, memorandum items, and underlying series based on the new list of FSI.

47. The authorities showed interest in getting support in the form of training or TA. The mission acknowledges that the BCU may have capacity constrains to meet the recommendations for higher frequency, reporting of underlying statements, and implementation of the new FSI guide. The mission reminded the authorities of available courses on FSI and of the possibility of requesting a TA mission to help implementing the requirements of the new guide.

48. Recommendations: (i) Increase periodicity of FSI by compiling quarterly data and report them with a lag of one quarter to STA, including the underlying statements (priority recommendation, by September 2020⁹); and (ii) Start reviewing the new FSI guide and its requirements, including the new list of FSI (expected to be requested by STA starting in September 2020).

H. Resources and Technical Assistance

49. The BCU should allocate new staff for compiling MFS. The mission learned that the main MFS compiler is retiring at the end of the mission. With more than 30 years’ experience, the absence of this compiler will be surely felt. To avoid disruptions or delays in the production of data, new staff should be allocated to compiling MFS without delay. This is particularly important for the implementation of the recommendations of this report.

50. Implementation of IFRS based reporting by banks and other financial intermediaries, as well as implementation of the MFSMCG for the central bank and ODCs may require further TA. This mission focused on the compilation of new standardized data for OFCs, but it did not review data on the central bank or ODCs. However, in 2019 the BCU required banks to adopt IFRS, which involved significant changes in their plan of accounts and catalog. The impact of these changes in the compilation of MFS was not analyzed in this mission, but it could be part of a follow-up mission, together with the implementation of the MFSMCG.

I. OFC Compilation Systems and Integrated Monetary Database

51. Since 2007, the BCU, STA, and WHD have shared an integrated monetary database (IMD) used for publication and operational needs. The mission discussed the importance of maintaining and improving the IMD for Uruguay. The current IMD contains the SRFs for the central bank and ODC ¹⁰ and the corresponding surveys. The IMD is consistent with the recommendations of the Monetary and Financial Statistics Manual of 2000. It contains data for December 2001 onward.

52. The IMD should be upgraded to include the available data for OFCs, with historical series starting in December 2014. The upgraded IMD should contain data from the three financial subsectors: central bank, ODC, and OFC. With these data as source, it generates surveys for: central bank (1SG), ODC (2SG), depository corporations (3SG), OFC (4SG), and financial corporations (5SG). The financial corporations survey consolidates all financial subsectors into one single survey.

53. The mission encouraged the dissemination of detailed monetary data from IMD on the IMF external website. Detailed monetary data from IMD for over 40 countries are disseminated on the IMF’s website.¹¹

54. The mission completed several compilation files to be used in the IMD. These files and their use are:

• 1. IC-Compilation system: it compiles the new 4SR for IC for one quarter.

• 2. PF-Compilation system: it compiles the new 4SR for PF for one quarter.

• 3. CAC-Compilation system: it compiles the new 4SR for CAC for one quarter.

• 4. New to old 4SR: it converts the new 4SR into the old (currently used to report data to STA) 4SR.

• 5. IMD file: it contains data for the entire financial sector, including CB, ODC, and OFC, as explained above.

55. Recommendation: Consider authorizing STA to disseminate the IMD on the IMF external website.