## 1surea2023002

## Source details

**Canonical URL:** [1surea2023002](https://www.imf.org/-/media/files/publications/cr/2023/english/1surea2023002.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2023/english/1surea2023002.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2023/english/1surea2023002.pdf.json)

---

### Summary of mission, objectives, and main findings
- TA mission: October 10 – 21, 2022, at the request of the Suriname authorities; conducted in coordination with the IMF’s Western Hemisphere Department.
- Main objectives:
  - reducing/explaining discrepancies between above- and below-the-line transactions;
  - updating the institutional table and sectorization of public units;
  - integrating stocks and flows of financial assets and liabilities;
  - checking consistency of GFS data and other macroeconomic statistics;
  - examining source data on other accounts payable and arrears;
  - providing hands-on assistance/training on related technical issues.
- Main counterparts: staff from the Ministry of Finance and Planning (MFP) and the Central Bank of Suriname (CBS) Statistics Department.
- Progress:
  - Partial follow-up of April 2022 recommendations: significant progress in below-the-line data; little progress in above-the-line data.
  - CBS compilers prepared a financial balance sheet of the Central Government (CG) to integrate flows with stocks and perform consistency checks of CG deposits and loans with banking system recordings (MFS and BoP).
  - Insufficient progress on institutional sector coverage and above-the-line compilation, mainly due to insufficient capacities of GFS compilers at the MFP and absence of an appropriate PFM system.
- Key data-quality constraint: no comprehensive data on government revenues and expenditures; budget execution data cover only transactions passing through a subset of government accounts (government’s 31-group-accounts in the CBS), excluding many operations through other accounts in group 31 and all operations through accounts 32 (projects), and many government accounts in commercial banks.
- Observed magnitudes:
  - government funds in commercial bank accounts represent more than 20 percent of all government deposits (1.2 bill SRD in Sep 2022).
  - mission found funds kept in government accounts in commercial banks present around 30 percent of all government deposits (1.2 bill SRD in Sep 2022 excluding taxes payable).
- Mission support: extended hands-on training for MFP GFS compilers; design of templates and data requests for the Tax Office; setup of a shared-folder database structure; introduction of coding of government transactions by economic categories.

### Priority recommendations (one-year action plan highlights)
- Immediate: Enhance institutional arrangements to improve collection and processing of source data as required by GFS. (Responsible: MFP)
- February 2023: Start using the “adjustment file” to compile GFS, starting from the CBS source data and showing explicitly all adjustments to the GFS revenue and expenditure categories, including adequate explanations on the source data used for each adjustment. (Responsible: MFP)
- March 2023: Set up a system to compile and maintain an up-to-date list of public units, focusing on central government entities. (Responsible: MFP)
- February 2023: The MFP should share the current institutional table with the CBS to sectorize government entities in MFS and thus expand coverage of central government below-the-line data. (Responsible: MFP and CBS)
- Selected milestones from TABLE 2 (selected entries):
  - High priority, Immediate: Sign Memorandum of Understanding (MoU) between the Tax Office and MFP, CBS and MFP, and CBS and SDMO specifying data provision requirements; introduce coding of transactions by economic categories; establish a GFS shared folder; ensure coordination and establish thematic technical working group(s). (Responsible: MFP, CBS and SDMO)
  - High priority, February 2023: Start using the “adjustment file” to compile GFS, starting from CBS source data and showing explicitly all adjustments to GFS categories and corresponding counterpart in below the line. (Responsible: MFP)
  - High priority, March 2023: Set up a system to compile and maintain an up-to-date list of public units, focusing on the central government entities. (Responsible: MFP)
  - Medium priority, January 2023: Agree on a uniform type of exchange rates for all transactions and stocks. (Responsible: MFP, CBS and SDMO)
  - Medium priority, gradually by June 2023: Follow-up on the April 2022 mission outstanding recommendations. (Responsible: MFP, CBS and SDMO)
  - Medium priority, January 2023: Set up a shared folder for GFS compilers at the MFP, including relevant source data, processing files, and final reports. (Responsible: MFP)

### A. Sector coverage — findings and recommended actions
Findings
- A draft list of central government units exists but comprehensiveness and coverage across source data and reports is uncertain.
- Existing institutional table shows differences between units covered in different reports/systems; neither above-the-line nor below-the-line data include all central government bodies.
- Large number of off-budget accounts (government accounts in commercial banks) are only covered below-the-line and not included in compilation of revenues and expenditures.
- Current coverage of central government units in MFS is incomplete because CBS lacks a comprehensive list of units.

Recommended actions (prioritized)
- MFP, with line ministries, should complete a list of all entities established by ministries and other government bodies; supplement with public units established by law or other legal acts; one source may be the government body responsible for registering legal entities.
- Centralize this information in a dedicated department/unit within MFP to serve as input for Treasury coverage and maintain the list on an ongoing basis (track newly established, restructured, or liquidated units).
- Coordinate the public sector institutional table with CBS and the national statistical institute; focus first on completing the central government list, then local government, social security funds, and public corporations.
- Clarify institutional coverage of the State Budget to understand what budget execution data refer to in terms of entity coverage.
- In the short term, provide the current draft institutional table to CBS to expand MFS coverage and derive a more comprehensive below-the-line indicator of government performance (e.g., net financing), while documenting that discrepancies with above-the-line may increase initially due to coverage differences.
- Responsibility: Ministry of Finance and Planning to complete and update the institutional table; coordinate among Budget, Treasury, Fiscal Affairs, and Public Accounting; share the existing institutional table with CBS in the meantime.

### B. Above-the-line transactions — source data and institutional arrangements
Principal constraints
- Lack of a functional financial management information system (IFMIS).
- Inadequate capacities and absence of systematic procedures for source data collection, quality checks, processing, and automation.
- GFS compilers at MFP perform tasks that should be undertaken by other departments/agencies.

Recommended intermediate steps while PFM/IFMIS improvements proceed
- Sign a Memorandum of Understanding (MoU) with data providers specifying source data requirements, scope, periodicity, timeliness, and templates.
- Tax Office: provide aggregated summary data to the Economic Affairs Department (EAD) on indirect taxes in SRD and direct taxes in SRD, Euro, and USD (Euro and USD to be converted to SRD using daily ER). Data to be provided by month and by banks; crosschecks reconciling stocks and flows should show no discrepancies. Consider switching from daily to monthly data to reduce workload of GFS compilers.
- CBS: provide GFS compilers all data in electronic form (excel) on operations passing through government accounts (accounts 31, 32, 34, and 35); convert USD and Euro to SRD using daily exchange rate (ER); attribute transaction codes to all operations passing through all accounts (MFP currently does not receive data on transactions in 34 and 35).
- Introduce a coding system by economic categories for all payments and receipts passing through government accounts. Mission prepared a draft coding system (Appendix D). MFP should assign departments/divisions/bodies responsible for attributing codes. CBS bookkeeping would then adopt MFP coding. Ideally apply codes also to operations in commercial banks; if not technically possible, collect data on revenues and expenditures passing through commercial accounts by economic categories from line ministries (owners of the accounts).
- Set up a GFS shared folder accessible to all GFS compilers including input data, processing files, and final reports to track compilation flow and ensure updates are reflected in outputs (mission suggested an architecture in Appendix E).
- Ensure daily backup of GFS shared folders by IT department to prevent data loss.
- Exchange rates: agree on a uniform ER to be used for converting foreign-currency operations and stocks. Currently three ERs are published daily (at 10 am, 12:30 pm, and 3pm), distinguishing sales and purchases; divergent ERs are a source of discrepancies between above- and below-the-line data and between stocks and flows.

### Adjustment file for GFS compilation — setup and technical improvements
- Adjustment file was set-up during the mission in 2021 to reconcile the CBS source data with explicitly shown adjustments and the final GFS figures on revenues and expenditures by economic categories.
- Purpose: replace current compilation method which does not allow crosschecks and tracking of all input data and adjustments.
- Mission suggested running old and new systems in parallel and encouraged authorities to abandon the old system and start using the new adjustment file by the end 2022.

Technical improvements suggested
- Starting point:
  - Use raw unadjusted figures as classified in the CBS accounts as starting point.
  - Show all reclassifications in the summary sheet as adjustments rather than performing reclassifications outside the adjustment file.
  - Allow linking the master file with raw source data and verify consistency of subsequent adjustments.
- Links and crosschecks:
  - Master file should have a summary sheet showing transition from CBS source data to GFSM 2014 based revenue and expenditure categories and link to supplementary source-data sheets (e.g., tax revenues, CEBUMA, SDMO data).
  - Summary and relevant sheets should be linked with primary compilation files.
  - Automatic checks via formulas should ensure consistency between source data and adjustments, aggregates and underlying details, stocks and flows.
  - Master file should clearly identify adjustments in revenue and expenditure that should be done below-the-line and communicate these to CBS compilers to ensure consistency between above- and below-the-line data.

### Government accounts in commercial banks — outstanding issues and short-term proxies
Findings
- No comprehensive data on government revenues and expenditures exist in Suriname; budget execution data cover only a fraction of government spending by function and not by economic category.
- Budget execution data:
  - Comprise operational expenditures passing through one of the government accounts from the group 31 in the CBS.
  - Exclude operations going through other accounts in group 31 and all operations conducted through accounts 32 (projects).
  - A large number of government accounts in commercial banks are not captured at all.
- Institutional practice:
  - When an expense is executed, a transfer from the budget account 31 is made to the relevant ministry’s account in a commercial bank.
  - No evidence and control are in place on when and how government funds held in commercial bank accounts are used.
- Magnitudes:
  - Funds in government accounts in commercial banks present around 30 percent of all government deposits (1.2 bill SRD in Sep 2022 excluding taxes payable).
  - Government funds in commercial bank accounts represent more than 20 percent of all government deposits (1.2 bill SRD in Sep 2022).

Implications for above- and below-the-line data and sectorization
- Discrepancies arise because operations through commercial banks are not fully recorded above-the-line.
- Net impact of these operations (inflows less outflows) is reflected in the below-the-line data using MFS information on government accounts in commercial banks, but revenues and expenditures are missing above-the-line.
- Specific cases and implications:
  - Budgetary entities with accounts in commercial banks:
    - Under current system, expenditures are recorded when money is transferred from 31 accounts in CBS to the government account in a commercial bank (a transfer between government accounts and should not impact the deficit at that point).
    - Expenditure should be recorded only when funds are spent from the commercial bank account.
    - Central budgetary subsector expenditure is overestimated in budget reporting and GFS if funds in commercial banks are unspent; underestimated when money is spent.
    - Unspent amount of government funds in commercial banks accounted for 1.2 billion SRD in September 2022.
  - Transfers by BCG to commercial bank accounts owned by entities in other government subsectors:
    - Expense of the BCG subsector is correctly recorded at the transfer point and should be classified as a grant within the government sector.
    - Authorities should distinguish and sectorize units uniformly across subsectors.
  - Government units collecting own revenues via commercial bank accounts create additional discrepancies as these revenues are not reflected above-the-line but are covered below-the-line.

Short-term proxy and data collection recommendations
- Use data on net transactions (change in stocks in the accounts) to adjust revenue/expenditures to ensure consistency with below-the-line fiscal balance impact; this will not reflect the full value of missing revenues and expenditures.
- Begin collecting data on inflows and outflows from commercial banks as a short-term measure and then focus on availability of data by economic categories.

### Below-the-line compilation, reconciliation and institutional coordination
- GFD in CBS Statistics Department compiles the Central Government financial balance sheet and financing, integrating flows with stocks of financial assets and liabilities.
- GFD monitors monthly consistency among macroeconomic statistics: GFS, MFS, ESS, and PSDS; conciliation started during previous TA mission and is now a regular task.
- Mission recommended creating a conciliation file between GFS and MFS, ESS, and PSDS using SDMO source data covering flows and stocks of deposits, credit, public securities holdings by CBS and commercial banks, and external financial assets and liabilities including interest in arrears.
- Share Public Sector Institutional table with MFS compilers even if still work in progress; MFS compilers regularly advise commercial banks on sectorization and should be consistent with GFS and debt sectorization.
- Cross-checking between CBS departments and SDMO has been beneficial and should continue to assure consistency.

Regular procedures and capacity recommendations
- Cross checking of stock and flow data between GFS below-the-line, PSDS, ESS and MFS should be part of regular procedures to enhance macroeconomic data quality.
- CBS is well placed to lead this task since most areas (except debt statistics) are supervised by the same director; CBS committed to devote appropriate resources to consistency control.
- Staffing and capacity:
  - CBS team compiled the CG financial balance sheet and its flows but currently only two resources work on the financial balance sheet and financing (one senior and one junior official).
  - It is important to strengthen the team responsible for compiling the CG primary balance from below-the-line, a main quantitative performance criterion (QPC) of the IMF program.

Monthly compilation practices (current)
- Team compiles monthly data on:
  - Stock of financial assets and liabilities by instrument and creditor when creditor can be identified; stocks in foreign currency are expressed in SRD using the ER at the end of each month.
  - Transactions in financial assets and liabilities; valuation in SRD depends on data availability:
    - Transactions in CG accounts in foreign currency are expressed in SRD using the ER of the day of the transactions. For disbursements or amortizations based on SDMO data, compilers check differences between SDMO and CBS accounts and make corresponding adjustments.
    - Transactions (disbursements, amortization, interest paid, and interest in arrears) in debt instruments in SDMO file not conducted through CG accounts in CBS in foreign currency are converted into SRD using the average (buy rate) ER of the month.
    - The rest of transactions in foreign currency are estimated by differences, in original currency, of stocks at the end of the month minus the end of the previous month and converted in SRD by the average ER of the month.

### Debt and arrears — SDMO data sharing, revisions, and integration
- SDMO system coverage and handling:
  - SDMO reported more than 90% of external debt liabilities and almost 80% of domestic debt liabilities are currently in the system.
  - Due to lack of coverage and difficulty downloading data, data shared with other departments is maintained in an Excel database.
- SDMO agreed to share with EAD and GFD additional monthly data; sharing will not be before November 2022. Additional monthly data to be shared:
  - CG financial assets (on lend) with the SOEs: Total stock, disbursements (from CG to the SOEs), principal, interest paid and arrears in principal and Interest; data will be shared one month after the closing date.
  - SDMO will share, in addition to the current file on External Debt in original currency and in USD, a file with the data converted in SRD. Stocks will be converted to SRD at the ER at the end of each month, and flows (disbursements, amortizations, interest, and costs) will be converted at the ER at the day of the transactions. This data will not be ready to share before end of November 2022.
- Historical revisions and file sharing:
  - SDMO will inform changes in historical data whenever there is a revision due to a correction or new loans/information added.
  - SDMO will revise historical monthly data for 2020 and 2021 by end of 2023; after revision is completed, SDMO will publish the revised data and share with MFP and CBS.
  - Files should be shared using formulas rather than hard numbers to make aggregates easier to comprehend by users; SDMO started sharing excel files with formulas since November 2022.
- Integration plans:
  - SDMO will include in debt liabilities data on other accounts payable resulting from transactions with suppliers, as well as arrears and overdrafts in the banking system.
  - Planned integration of these two data sets during 2023 after MFP shares with SDMO debt with suppliers, including arrears.
  - Currently, EAD compiles data on overdrafts in the CG accounts derived from MFS.
- Paymaster system and next steps:
  - Paymaster system contains domestic debt arrears with suppliers; further work required to determine how to download useful information in a format usable by GFS compilers.
  - Next TA mission will discuss how to better download useful information currently contained in the paymaster system with the paymaster and EAD.

### MOU structure and governance recommendations
MOU structure and core purpose
- MOU should distinguish between:
  - the Steering Committee (senior member Working Group) responsible for strategic decision making, and
  - operations Working Groups (expert level Working Group) responsible for executing the ongoing compilation process.
- MOU should establish core purpose: mutual commitment to cooperate to produce high-quality statistics.

Topics to consider in drafting an MOU (highlights)
- Partner organizations and preamble explaining need and usefulness of high-quality GFS.
- Purpose of compiling macroeconomic statistics/GFS; relation to other fiscal reporting.
- Legal context: explain legal remit of GFS compiling agencies and reference relevant Laws and Regulations.
- Timelines: project timelines/schedules for first dissemination; strategy to periodically review operational processes and adapt the MOU.
- Methodology/Compilation: reference to GFSM 2014, PSDSG 2011 and interlinkages requiring strict coordination.
- Governance structure and reporting: process for resolving disputes, meeting cadence, and operational liaison points for each GFS reporting period.

Working Groups: commitments and expectations
- Members commit to attending scheduled meetings, championing partnership, sharing information, making timely decisions, and notifying members of issues.
- Members expect complete, accurate, timely information; reasonable time for decisions; alerts on potential risks; open and honest discussions.

Steering Committee (advisory and operations)
- Advisory functions: ongoing ‘health checks’, monitor and coordinate cooperation as defined in the MoU.
- Suggested meeting procedures: chair, quorum, decision rules (consensus or chair final decision), agenda and minutes preparation, meeting frequency and locations.
- Operational themes: collection of data; pre-compilation meeting; discuss policy changes and recording implications; integration outcomes and issues; pre-publication verification; post-compilation evaluation.

Division of responsibilities (areas to define)
- Expected format and level of detail from data suppliers; confidentiality considerations.
- Responsibilities for vertical and stock/flow reconciliation; horizontal reconciliation; time adjustments (above-and-below the line), imputations, methodological adjustments.

### Appendix highlights: templates, coding, and shared drive architecture
- Appendix C: Templates for collecting data on taxes (DIRECT TAXES and INDIRECT TAXES) with monthly bank breakdowns and reconciliation checks.
- Appendix D: Suggested coding of payments through accounts 31 and 32 (and 34, 35, 63) with example mappings for Credits (inflows) and Debits (outflows) across agencies (Tax Office, Paymaster, Treasury, SDMO, Line Ministries).
- Appendix E: Shared drive suggested architecture:
  - Folders by year:
    - I. Raw source data (subfolders by months Jan, Feb...)
    - II. Processed data (FILES by subject: 31 SRD, 31 Euro, 31 USD, 32 SRD, 32 Euro, 32 USD, 34, 35, 63, SDMO, Taxes, CEBUMA, Royalties, Settlements, Arrears)
    - III. Summary data (Weekly; Monthly)

*IMF | Technical Report – Suriname Government Finance Statistics*

### Section I. Detailed Technical Assessment and Recommendations ...................................................... 5

### Section I. Detailed Technical Assessment and Recommendations

### Summary of mission, objectives, and main findings
- A technical assistance (TA) mission took place during October 10 – 21, 2022, at the request of the Suriname authorities, conducted in coordination with the IMF’s Western Hemisphere Department.
- Main objectives included:
  - reducing/explaining discrepancies between above- and below-the-line transactions;
  - updating the institutional table and sectorization of public units;
  - integrating stocks and flows of financial assets and liabilities;
  - checking consistency of GFS data and other macroeconomic statistics;
  - examining source data on other accounts payable and arrears;
  - providing hands-on assistance/training on related technical issues.
- Main counterparts: staff from the Ministry of Finance and Planning (MFP) and the Central Bank of Suriname (CBS) Statistics Department.
- Progress summary:
  - Partial follow-up of April 2022 recommendations: significant progress in below-the-line data; little progress in above-the-line data.
  - CBS compilers have prepared a financial balance sheet of the Central Government (CG) to integrate flows with stocks and perform consistency checks of CG deposits and loans with banking system recordings (MFS and BoP).
  - Insufficient progress on institutional sector coverage and above-the-line compilation, mainly due to insufficient capacities of GFS compilers at the MFP and absence of an appropriate PFM system.
- Key data-quality constraint: no comprehensive data on government revenues and expenditures. Budget execution data cover only transactions passing through a subset of government accounts (government’s 31-group-accounts in the CBS), excluding many operations through other accounts in group 31 and all operations through accounts 32 (projects), and many government accounts in commercial banks.
- Observed magnitude: government funds in commercial bank accounts represent more than 20 percent of all government deposits (1.2 bill SRD in Sep 2022).
- Mission support: extended hands-on training for MFP GFS compilers; design of templates and data requests for the Tax Office; setup of a shared-folder database structure; introduction of coding of government transactions by economic categories.

### Priority recommendations (one-year action plan highlights)
- Immediate: Enhance institutional arrangements to improve collection and processing of source data as required by GFS. (Responsible: MFP)
- February 2023: Start using the “adjustment file” to compile GFS, starting from the CBS source data and showing explicitly all adjustments to the GFS revenue and expenditure categories, including adequate explanations on the source data used for each adjustment. (Responsible: MFP)
- March 2023: Set up a system to compile and maintain an up-to-date list of public units, focusing on central government entities. (Responsible: MFP)
- February 2023: The MFP should share the current institutional table with the CBS to sectorize government entities in MFS and thus expand coverage of central government below-the-line data. (Responsible: MFP and CBS)

Detailed priorities and milestones from TABLE 2 (selected entries)
- High priority, Immediate: Sign Memorandum of Understanding (MoU) between the Tax Office and MFP, CBS and MFP, and CBS and SDMO specifying data provision requirements; introduce coding of transactions by economic categories to be applied by MFP and other entities; establish a GFS shared folder; ensure coordination and establish thematic technical working group(s). (Responsible: MFP, CBS and SDMO)
- High priority, February 2023: Start using the “adjustment file” to compile GFS, starting from CBS source data and showing explicitly all adjustments to GFS categories and corresponding counterpart in below the line. (Responsible: MFP)
- High priority, March 2023: Set up a system to compile and maintain an up-to-date list of public units, focusing on the central government entities. (Responsible: MFP)
- High priority, February 2023: MFP should share the current draft institutional table with CBS to sectorize government entities in MFS and thus expand coverage of the central government below-the-line data. (Responsible: MFP)
- Medium priority, January 2023: Agree on a uniform type of exchange rates for all transactions and stocks. (Responsible: MFP, CBS and SDMO)
- Medium priority, gradually by June 2023: Follow-up on the April 2022 mission outstanding recommendations. (Responsible: MFP, CBS and SDMO)
- Medium priority, January 2023: Set up a shared folder for GFS compilers at the MFP, including relevant source data, processing files, and final reports. (Responsible: MFP)

### A. Sector coverage — findings and recommended actions
Findings
- A draft list of central government units exists but its comprehensiveness and coverage across source data and reports is uncertain.
- Existing institutional table shows differences between units covered in different reports/systems and indicates neither above-the-line nor below-the-line data include all central government bodies.
- Large number of off-budget accounts (government accounts in commercial banks) are only covered below-the-line and not included in compilation of revenues and expenditures.
- The current coverage of central government units in MFS is incomplete because CBS lacks a comprehensive list of units.

Recommended actions (prioritized)
- MFP, in cooperation with line ministries, should complete a list of all entities established by ministries and other government bodies, supplemented by public units established by law or other legal acts; one source may be the government body responsible for registering legal entities.
- Centralize this information in a dedicated department/unit within MFP to serve as input for Treasury coverage and maintain the list on an on-going basis (tracking newly established, restructured, or liquidated units).
- Coordinate the public sector institutional table with CBS and the national statistical institute; focus first on completing the central government list, then local government, social security funds, and public corporations.
- Clarify institutional coverage of the State Budget to understand what budget execution data refer to in terms of entity coverage.
- In the short term, provide the current draft institutional table to CBS to expand MFS coverage and derive a more comprehensive below-the-line indicator of government performance (e.g., net financing), while documenting that discrepancies with above-the-line may increase initially due to coverage differences.
- Responsibility: Ministry of Finance and Planning to complete and update the institutional table; coordinate among Budget, Treasury, Fiscal Affairs, and Public Accounting; share the existing institutional table with CBS in the meantime.

### B. Above-the-line transactions — improving the source data and institutional arrangements
Principal constraints
- Lack of a functional financial management information system (IFMIS).
- Inadequate capacities and absence of systematic procedures for source data collection, quality checks, processing, and automation.
- GFS compilers at MFP currently perform tasks that should be undertaken by other departments/agencies.

Recommended intermediate steps while PFM/IFMIS improvements proceed
- Sign a Memorandum of Understanding (MoU) with data providers specifying source data requirements, scope, periodicity, timeliness, and templates (Appendix B provides an outline).
- Tax Office: provide aggregated summary data to the Economic Affairs Department (EAD) on indirect taxes in SRD and direct taxes in SRD, Euro, and USD (Euro and USD to be converted to SRD using daily ER). Data to be provided by month and by banks; crosschecks reconciling stocks and flows should show no discrepancies. Consider switching from daily to monthly data to reduce workload of GFS compilers.
- CBS: provide GFS compilers all data in electronic form (excel) on operations passing through government accounts (accounts 31, 32, 34, and 35); convert USD and Euro to SRD using daily exchange rate (ER); attribute transaction codes to all operations passing through all accounts (MFP currently does not receive data on transactions in 34 and 35).
- Introduce a coding system by economic categories for all payments and receipts passing through government accounts. Mission prepared a draft coding system (Appendix D). MFP should assign departments/divisions/bodies responsible for attributing codes. CBS bookkeeping would then adopt MFP coding. Ideally apply codes also to operations in commercial banks; if not technically possible, collect data on revenues and expenditures passing through commercial accounts by economic categories from line ministries (owners of the accounts).
- Set up a GFS shared folder accessible to all GFS compilers including input data, processing files, and final reports to track compilation flow and ensure updates are reflected in outputs (mission suggested an architecture in Appendix E).
- Ensure daily backup of GFS shared folders by IT department to prevent data loss.
- Exchange rates: agree on a uniform ER to be used for converting foreign-currency operations and stocks to SRD. Currently three ERs are published daily (at 10 am, 12:30 pm, and 3pm), distinguishing sales and purchases; divergent ERs are a source of discrepancies between above- and below-the-line data and between stocks and flows.

### Improving compilation procedures (introductory points identified by the mission)
- Use of an “adjustment file” to compile GFS beginning from CBS source data and explicitly showing all adjustments to GFS revenue and expenditure categories, with explanations on source data used for each adjustment and corresponding counterparts below the line.
- Introduce uniform transaction coding and consistent exchange-rate application across data providers (MFP, CBS, SDMO).
- Build capacity at MFP GFS compilers to reduce substitution of other departments’ responsibilities and to improve above-the-line data quality.

*Section I. Detailed Technical Assessment and Recommendations — 1surea2023002*

### 13. The mission reviewed the so called “adjustment” file to compile GFS. The file was set-up

### 1surea2023002 - 13. The mission reviewed the so called “adjustment” file to compile GFS. The file was set-up

### Adjustment file for GFS compilation
- The adjustment file was set-up during the mission in 2021 to reconcile the CBS source data with the relevant explicitly shown adjustments and the final GFS figures on revenues and expenditures by economic categories.
- The adjustment file is intended to replace the current way of GFS compilation which does not allow crosschecks and tracking of all input data and adjustments.
- The mission suggested running the old and new systems in parallel for a certain period to examine correctness of the new adjustment file.
- The mission encouraged the authorities to abandon the old system and start using the new adjustment file by the end 2022.

### Technical improvements suggested for the adjustment file
- Starting point
  - Use as starting point the raw unadjusted figures as classified in the CBS accounts.
  - Show all reclassifications in the summary sheet as adjustments rather than performing reclassifications outside the adjustment file.
  - This approach would allow linking the master file with raw source data and verify consistency of subsequent adjustments.
- Links and crosschecks
  - The master file should consist of a summary sheet showing the transition from CBS source data to GFSM 2014 based revenue and expenditure categories, and clearly link to supplementary source-data sheets (e.g., tax revenues, CEBUMA, SDMO data, etc.).
  - Summary sheet and relevant sheets should be linked with primary compilation files.
  - Checks should be performed automatically in the master file via formulas to ensure consistency between source data and adjustments, aggregates and underlying details, stocks, and flows.
  - The master file should clearly identify adjustments in revenue and expenditure that should be done below-the-line and communicate these to CBS compilers to ensure consistency between above- and below-the-line data.

### Outstanding issues — Government accounts in commercial banks
- No comprehensive data on government revenues and expenditures exist in Suriname; budget execution data cover only a fraction of government spending by function and not by economic category.
- Budget execution data:
  - Comprise operational expenditures passing through one of the government accounts from the group 31 in the CBS.
  - Exclude operations going through other accounts in group 31 and all operations conducted through accounts 32 (projects).
  - A large number of government accounts in commercial banks are not captured at all.
- Institutional practice and implications:
  - National rules: when an expense is executed, a transfer from the budget account 31 is made to the relevant ministry’s account in a commercial bank.
  - No evidence and control are in place on when and how government funds held in commercial bank accounts are used.
  - Mission found funds kept in government accounts in commercial banks present around 30 percent of all government deposits (1.2 bill SRD in Sep 2022 excluding taxes payable).
  - The mission and technical staff identify absence of a properly functioning public financial management (PFM) system in Suriname as the main issue.

### Implications for above- and below-the-line data and sectorization
- Discrepancies between above- and below-the-line data arise because operations through commercial banks are not fully recorded above-the-line.
- Net impact of these operations (inflows less outflows) is reflected in the below-the-line data using MFS information on government accounts in commercial banks, but revenues and expenditures are missing above-the-line.
- Specific cases:
  - Budgetary entities with accounts in commercial banks:
    - Under current system, expenditures are recorded when money is transferred from 31 accounts in CBS to the government account in a commercial bank, but this is a transfer between government accounts and should not impact the deficit at that point.
    - Expenditure should be recorded only when funds are spent from the commercial bank account.
    - Central budgetary subsector expenditure is overestimated in budget reporting and GFS if funds in commercial banks are unspent; underestimated when money is spent.
    - Unspent amount of government funds in commercial banks accounted for 1.2 billion SRD in September 2022.
  - Transfers by BCG to commercial bank accounts owned by entities in other government subsectors:
    - The expense of the BCG subsector is correctly recorded at the transfer point and should be classified as a grant within the government sector.
    - Authorities should distinguish and sectorize units uniformly across subsectors for compilation of revenues/expenditure and financing data.
  - Government units collecting own revenues via commercial bank accounts create additional discrepancies as these revenues are not reflected above-the-line but are covered below-the-line.
- Short-term proxy and data collection recommendations:
  - Use data on net transactions (change in stocks in the accounts) to adjust revenue/expenditures to ensure consistency with below-the-line fiscal balance impact; this will not reflect the full value of missing revenues and expenditures.
  - Authorities should begin collecting data on inflows and outflows from commercial banks as a short-term measure and then focus on availability of data by economic categories.

### Below-the-line compilation, reconciliation and institutional coordination
- GFD (Government Finance Division) in CBS Statistics Department compiles the Central Government financial balance sheet and financing, integrating flows with stocks of financial assets and liabilities.
- GFD monitors monthly consistency among macroeconomic statistics: GFS, MFS, ESS, and PSDS; conciliation started during previous TA mission and is now a regular task.
- Mission recommended creating a conciliation file between GFS and MFS, ESS, and PSDS using SDMO source data covering flows and stocks of deposits, credit, public securities holdings by CBS and commercial banks, and external financial assets and liabilities including interest in arrears.
- Share Public Sector Institutional table with MFS compilers even if still work in progress; MFS compilers regularly advise commercial banks on sectorization and should be consistent with GFS and debt sectorization.
- Cross-checking of data between CBS departments and SDMO has been beneficial and should continue to assure consistency.

### Differences across source data and treatment of interest and debt stocks
- Main differences between GFS and ESS: amounts of interest payments and stock of public securities issued abroad.
  - ESS records public securities issued abroad as debt with nonresidents only for amounts not held by domestic sector, included in IIP at market value including interest in arrears at the due date.
  - SDMO debt stock includes all public securities issued abroad in external debt at face value; interest in arrears is included in the stock of outstanding debt.
  - Interest in arrears considered by external sector is included in the BOP when interest payment is due; in GFS the interest in arrears has not been recorded above-the-line at the due date.
  - Neither ESS nor SDMO calculate accrued interest monthly; interest is calculated only when due, each three months in the case of public securities issued abroad.
- Cross-checks between GFS and MFS:
  - Main sources of differences are sectorization of institutions and accrued interest in MFS data.
  - GFS uses data from SDMO which does not calculate accrued interest.

### Regular procedures and capacity recommendations
- Cross checking of stock and flow data between GFS below-the-line, PSDS, ESS and MFS should be part of regular procedures to enhance macroeconomic data quality.
- CBS is well placed to lead this task since most areas (except debt statistics) are supervised by the same director; CBS committed to devote appropriate resources to consistency control.
- Staffing and capacity:
  - CBS team has made significant progress compiling the CG financial balance sheet and its flows.
  - There are currently only two resources working on the financial balance sheet and financing (one senior and one junior official).
  - It is particularly important to strengthen the team responsible for compiling the CG primary balance from below-the-line, which is one of the main quantitative performance criteria (QPC) of the IMF program.

### Monthly compilation practices (current)
- Team compiles monthly data on:
  - Stock of financial assets and liabilities by instrument and creditor when creditor can be identified; stocks in foreign currency are expressed in SRD using the ER at the end of each month.
  - Transactions in financial assets and liabilities; valuation in SRD depends on data availability:
    - Transactions in CG accounts in foreign currency are expressed in SRD using the ER of the day of the transactions. For disbursements or amortizations based on SDMO data, compilers check differences between SDMO and CBS accounts and make corresponding adjustments.
    - Transactions (disbursements, amortization, interest paid, and interest in arrears) in debt instruments in SDMO file not conducted through CG accounts in CBS in foreign currency are converted into SRD using the average (buy rate) ER of the month.
    - The rest of transactions in foreign currency are estimated by differences, in original currency, of stocks at the end of the month minus the end of the previous month and converted in SRD by the average ER of the month.

### Debt and arrears — SDMO data sharing and revisions
- SDMO system coverage and data handling:
  - SDMO reported more than 90% of external debt liabilities and almost 80% of domestic debt liabilities are currently in the system.
  - Due to lack of coverage and difficulty downloading data, data shared with other departments is maintained in an Excel database.
- SDMO agreed to share with EAD and GFD additional monthly data; sharing will not be before November 2022. Additional monthly data to be shared:
  - CG financial assets (on lend) with the SOEs: Total stock, disbursements (from CG to the SOEs), principal, interest paid and arrears in principal and Interest; data will be shared one month after the closing date.
  - SDMO will share, in addition to the current file on External Debt in original currency and in USD, a file with the data converted in SRD. Stocks will be converted to SRD at the ER at the end of each month, and flows (disbursements, amortizations, interest, and costs) will be converted at the ER at the day of the transactions. This data will not be ready to share before end of November 2022.
- Historical revisions and file sharing:
  - SDMO will inform changes in historical data whenever there is a revision due to a correction or new loans/information added.
  - SDMO will revise historical monthly data for 2020 and 2021 by end of 2023; after revision is completed, SDMO will publish the revised data and share with MFP and CBS.
  - Files should be shared using formulas rather than hard numbers to make aggregates easier to comprehend by users; SDMO started sharing excel files with formulas since November 2022.
- Integration plans:
  - SDMO will include in debt liabilities data on other accounts payable resulting from transactions with suppliers, as well as arrears and overdrafts in the banking system.
  - Planned integration of these two data sets during 2023 after MFP shares with SDMO debt with suppliers, including arrears.
  - Currently, EAD compiles data on overdrafts in the CG accounts derived from MFS.
- Paymaster system data on domestic debt arrears with suppliers:
  - Mission met a paymaster technician who presented the system, but due to time constraints could not follow up; further work is required to determine how to download useful information in a format usable by GFS compilers.
  - The next TA mission will discuss how to better download useful information currently contained in the paymaster system with the paymaster and EAD.

*Italic: IMF | Technical Report – Suriname Government Finance Statistics*

### 4. It is useful if the MOU makes a distinction between the Steering Committee

### 4. It is useful if the MOU makes a distinction between the Steering Committee

### MOU structure and core purpose
- The MOU should distinguish between:
  - the Steering Committee (senior member Working Group) who are responsible for strategic decision making, and
  - operations Working Groups (expert level Working Group) who are responsible for executing the ongoing compilation process.
- A MOU can start out by establishing the core purpose of the document: the will between parties to commit to reach a common gold by cooperating on a number of actions delivered at the right time.

### Topics to Consider in Drafting a MOU
1. Partner organizations - the MOU is between:
- Name, Title, Organization
- Name, Title, Organization
- Name, Title, Organization

2. Preamble
- Why we need this agreement?
- Why is important to make use of existing synergy to produce high-quality statistics?
- Who needs high quality GFS (and other macroeconomic statistics) and why they are important?
- What does this MoU provision set out (e.g., fundamental cooperation between xy in the field of statistics related to the Albanian economy)?

3. Purpose of compiling macroeconomic statistics/GFS
- To provide context, it is useful to explain the usefulness GFS and how this differs from or relates to other fiscal reporting in the country – including their legal status
- Maybe also with a brief description of outputs

4. Legal context
- Explain the legal remit of GFS compiling agencies – including the relevant Laws, Regulations that explain roles and responsibilities
- Include reference numbers and dates of laws and their amendments

5. Timelines
- Considering the complete GFS compilation universe is still in development, it is useful to incorporate projected timelines / schedules for first dissemination of various sets of data.
- Include a strategy to periodically review operational processes and identify issues of concern and adapt the MOU where needed.
- Either party may request changes to this MOU. Any changes, modifications, revisions, or amendments to this MOU which are mutually agreed upon by and between the parties to this MOU shall be incorporated by written instrument, and effective when executed and signed by all parties to this MOU.

6. Methodology / Compilation
- It is important to make reference to the suite of methodology that the various statistical outputs need to adhere to (e.g., GFSM 2014, PSDSG 2011).
- It is also useful to explain that these methodologies are strongly interlinked and thus require a strict coordination effort to ensure consistency – also with other macroeconomic statistics

7. Governance structure and reporting
- Include a description of the governance structure and reporting responsibilities may be useful
- Describe the process for resolving disputes that may arise amongst the partners to the agreement

### Working Groups: commitments and expectations
8. The members of the Working Groups will commit to:
- attending all scheduled Working Groups meetings
- champion the partnership within and outside of work areas
- share all communications and information across all Working Groups members
- make timely decisions and take actions so as to not hold up the compilation process
- notifying members of the Working Groups, as soon as practical, if any matter arises which may be deemed to affect the development of the partnership
- attendance at all meetings and if necessary, nominate a proxy.

9. Members of the Working Groups expect:
- that each member will be provided with complete, accurate and meaningful information in a timely manner
- to be given reasonable time to make key decisions
- to be alerted to potential risks and issues that could impact the project, as they arise
- open and honest discussions

### Steering Committee (advisory function and operations)
- ongoing ‘health checks’ to verify the overall status and ‘health’ of the partnership.
- To monitor and coordinate the cooperation as defined in MoU, it would be useful to establish a Steering Committee with an advisory function. E.g., two members from each institution.
- Meetings
- Meetings will be chaired by (insert name and organization)
- A meeting quorum will be (insert number) members of the advisory group
- Decisions made by consensus (i.e., members are satisfied with the decision even though it may not be their first choice). If not possible, advisory group chair makes final decision
- Meeting agendas and minutes will be provided by (Insert name and organization), who will:
  - prepare agendas and supporting papers
  - prepare meeting notes and information.
- Meetings will be held (how often) for (specify time) at (specify location).
- If required, subgroup meetings will be arranged outside of these times at a time convenient to subgroup members
- On an operational level. What are the various liaison points in time for each GFS reporting period? Themes to consider:
  - Collection of data
  - Pre-compilation meeting
  - Discuss important policy changes / economic transactions and the impact on recording
  - Integration outcomes and issues
  - Pre-publication final verification
  - Post-compilation evaluation (with action points as outcome)

### Division of responsibilities
- All agencies have a collective responsibility in compilation – while each agency has a particular assignment that is either individual or shared according to the compilation process. A description of responsibilities can be based on the following issues (not exhaustive):
  - What is expected from data suppliers – the expected format of supplied data, the required level of attributes, the required metadata, the required level of detail, whether data should be aggregated to a certain level or not, whether data should be bridged to GFSM
  - How are information and data administered by all parties to the agreement – include confidentiality considerations
  - Who is responsible for vertical and stock/flow reconciliation – unit-by-unit / sector-by-sector
  - Who is responsible for horizontal reconciliation – transaction-by-transaction / balance sheet position-by-position
  - Who is responsible for time adjustments (above-and-below the line in parallel), reconciling accounting time of recording with statistics time of recording, imputations, methodological adjustments etc.

### Appendix highlights: templates, coding, and shared drive architecture
- Appendix C: Templates for collecting data on taxes
  - DIRECT TAXES: templates labelled 2022 and Jan-22 with fields such as:
    - TOAL in SRD (SRD+Euro+USD)JanFebrMarAprJunJulAugSepOctNovDecTotal
    - Rows including Opening stock; Total credit tax collection by banks; Interest (credit); Debit service charge; Transfer to central bank; Closing stock; Check 1+2+3-4-5-6=0; Chceck closing stock t= opening stock t+1
  - INDIRECT TAXES: templates labelled 2022 and Jan-22 with similar row structure and bank breakdowns.
- Appendix D: Suggested coding of payments through accounts 31 and 32 (To be also applied for 34, 35 and 63 accounts)
  - Credits (inflows) — example mappings:
    - Direct tax revenue — Commercial banks under the supervision of the Tax Office
    - Indirect tax revenue — Commercial banks under the supervision of the Tax Office
    - Non-tax revenue — Paymaster
    - Dividends — Treasury
    - Interest — SDMO
    - Line Ministries
    - Internal transfers 31&32 — Paymaster; Treasury; SDMO
    - Borrowing (debt issuance) — Loans SDMO; Debt securities SDMO; Financial Leasing SDMO
    - Repayments of on-lending and of other receivables — SDMO
    - Sale of shares and equities — Treasury
  - Debits (outflows) — example mappings:
    - Wages and pensions — Paymaster
    - Goods &Services &Capex — G&S Paymaster
    - Business trips — Paymaster; Treasury
    - Interest — Domestic SDMO; SDMO
    - Internal transfers 31&32 — Paymaster; Treasury; SDMO
    - Transfer to gov acc in commercial banks — Paymaster
    - Transfers distinguish current and capital
    - Subsidies — Paymaster
    - Other transfers — Treasury
    - Reversals — Tax office (tax returns)
    - Borrowing (debt repayment) — Loans SDMO; Debt securities SDMO; Financial Leasing SDMO
    - Loans granted (incl. on—lending) — SDMO
    - Acquisition of shares and equities — Treasury

- Appendix E: Shared drive - suggested architecture
  - A. Folders by year
    - I. Raw source data
      - i. Subfolders by months (Jan, Feb...)
    - II. Processed data
      - i. FILES by subject, would include monthly and weekly/daily data (as relevant):
        - 31 SRD
        - 31 Euro
        - 31 USD
        - 32 SRD
        - 32 Euro
        - 32 USD
        - 34
        - 35
        - 63
        - SDMO
        - Taxes
        - CEBUMA
        - Royalties
        - Settlements
        - Arrears
    - III. Summary data
      - i. Weekly
      - ii. Monthly

*IMF | Technical Report – Suriname Government Finance Statistics*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2023/english/1surea2023002.pdf_