## _033105

## Source details

**Canonical URL:** [_033105](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2005/_033105.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2005/_033105.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2005/_033105.pdf.json)

---

### Executive Summary — objectives, process, and overall evaluation
- Objectives:
  - Introduced in March 2000 to provide reasonable assurance that central banks have adequate financial control systems to manage their resources and Fund disbursements.
  - Main objective: minimize the possibility of misreporting or misuse of Fund resources associated with the Fund’s lending activities.
  - Safeguards assessments apply to all member countries that have a Fund arrangement.
  - Assessments evaluate five areas (ELRIC): (i) the External audit mechanism, (ii) the Legal structure and independence, (iii) the financial Reporting framework, (iv) the Internal audit mechanism, and (v) the system of internal Controls.
- Process:
  - Diagnostic exercises carried out by Fund staff; apply to all new Fund arrangements, including augmentations.
  - Independent from Fund surveillance, program negotiation, and technical assistance.
  - Based on document review and discussions with authorities and external auditors; typically include off-site review and supplemental on-site visit.
  - Outcome: a safeguards assessment report with recommendations and associated timeframes; recommendations may be included in program conditionality.
  - Post-report monitoring: authorities may comment before finalization; Fund staff conducts off-site monitoring through periodic communications and review of annual financial statements and external auditors’ reports while Fund credit is outstanding.
  - Reports are confidential and transmitted to the central bank and the relevant Fund Executive Director; Executive Board receives summaries in staff reports and periodic updates.
- Overall evaluation:
  - Based on five years’ experience, the framework has been broadly appropriate.
  - Independent panel concluded the safeguards policy “has been highly successful in achieving its objectives.”
  - Panel recommended strengthening communication and broadening monitoring; staff concurs and will broaden channels of information and communication.
- Future direction:
  - Shift toward monitoring of previously assessed central bank frameworks and more focused, targeted follow-up assessments.
  - Safeguards assessments under Emergency Post-Conflict Assistance (EPCA) to be undertaken case-by-case reflecting re-establishment degree of central bank functionality.
  - Continued off-site monitoring for as long as Fund credit is outstanding.

### Implementation of safeguards recommendations — scope, rates, and key implementations
- Coverage and counts:
  - Completed safeguards assessments: 111.
  - Central banks covered: 69 (some central banks received transitional and full assessments or multiple assessments).
  - Total recommendations included in 111 safeguards assessments (excluding recommendations with suggested implementation date later than March 1, 2005 and recommendations not monitored): 569.
- Overall implementation performance (as of March 1, 2005):
  - Overall implementation rate: 80.7 percent.
  - Recommendations with formal commitment from the authorities: 124; Implementation Rate: 96.8 percent.
    - Under program conditionality: 57 recommendations; Implemented: 55; Implementation Rate: 96.5 percent.
    - LOI/MEFP commitments: 67 recommendations; Implemented: 65; Implementation Rate: 97.0 percent.
  - Recommendations not under program conditionality or LOI/MEFP commitments: 445; Implemented: 339; Implementation Rate: 76.2 percent.
  - Conditionality composition (included in the 57 conditionality recommendations): 11 prior actions (all implemented), 14 structural performance criteria (13 implemented), and 32 structural benchmarks (31 implemented) as conditionality in the program design for a total of 27 countries.
- Main findings and trends (2000–2002 versus 2003–present):
  - Non-existent or deficient external audit: 2000–2002: Number 34; Percent of Total 52. 2003–present: Number 11; Percent of Total 24.
  - No or delayed publication of financial statements: 2000–2002: Number 25; Percent of Total 38. 2003–present: Number 20; Percent of Total 43.
  - Inadequate controls over foreign reserves: 2000–2002: Number 18; Percent of Total 47. 2003–present: Number 15; Percent of Total 33.
  - Inadequate financial reporting standards: 2000–2002: Number 30; Percent of Total 79. 2003–present: Number 24; Percent of Total 52.
  - Deficient governance oversight: 2000–2002: Number 26; Percent of Total 68. 2003–present: Number 23; Percent of Total 50.
  - Deficient internal audit: 2000–2002: Number 29; Percent of Total 76. 2003–present: Number 26; Percent of Total 57.
  - Inadequate governing legislation: 2000–2002: Number 16; Percent of Total 42. 2003–present: Number 17; Percent of Total 37.
  - Inadequate accounting for Fund transactions: 2000–2002: Number 9; Percent of Total 24. 2003–present: Number 8; Percent of Total 17.
  - Memorandum: During 2000–2002, 15 transitional assessments identified a non-existent or deficient external audit mechanism and 10 noted issues with respect to the publication of financial statements.
- Selected institutional improvements (counts of central banks implementing measures):
  - First time appointment of an international auditor: implemented by 18 central banks.
  - First time reconciliation of key program data with accounting records: implemented by 34 central banks.
  - First time publication of full audited financial statements: implemented by 19 central banks.
  - Internationally recognized financial reporting framework introduced: implemented by 16 central banks.
  - Internal audit measures:
    - New internal audit function established: 2 central banks.
    - New reporting line to the governor or the board established: 3 central banks.
    - Audit charter adopted: 14 central banks.
  - Establishment of a new audit committee or supervisory board: implemented by 15 central banks.

### Impact on data integrity, misreporting cases, and limits on preventing misuse
- Impact on program data and reporting:
  - Focus on controls over compilation of net international reserves (NIR) and net domestic assets (NDA) and consistency with TMU terms.
  - Improvements included independent oversight roles by external auditors and audit committees, periodic reconciliations of monetary program data with accounting records, strengthened timeliness of reporting, and enhanced interdepartmental collaboration.
  - Specific corrective actions: exclusion of restricted assets from NIR after anomalies; establishment of periodic reconciliations and internal audit reviews; adjustment where earmarked amounts were incorrectly included in NIR; transition to market-based valuation for foreign treasury instruments; movement toward full IFRS and appointment of international audit firms with IFRS experience; commissioning periodic treasury audits where applicable.
- Misreporting cases since policy introduction:
  - Three cases: Ghana, The Gambia, and Vietnam.
    - The Gambia: safeguards assessment key in detecting misreporting; revised program data and financial statements provided in October 2003; second on-site assessment in November 2003 confirmed governance issues, lack of auditor independence, and inadequate internal controls; remedial measures included change in external auditor, review of internal controls, reaudits of 2001 and 2002 financial statements, and a special audit of foreign reserves; two noncomplying disbursements from July and December 2001 were repaid.
    - Ghana: 2001 transitional assessment found serious concerns about audit quality and auditors’ qualifications and independence; an international audit firm was subsequently appointed.
    - Vietnam: safeguards assessment initiated in 2001 (not completed) found external audit not in accordance with ISA.
  - Example severity: an external auditor could obtain third-party confirmations for only about 15 percent of end-2002 foreign reserves; auditor issued a qualified opinion on currency in circulation; official foreign reserves reported as at end-2003 were subsequently revised downward by close to 50 percent.
- Limits and qualitative effects on misuse:
  - Policy cannot prevent misuse arising from willful override of controls or manipulation of data; it does not cover fiscal activities.
  - Qualitative deterrent effects: outside scrutiny; fraud allegations at two central banks led to replacement of senior management and strengthened controls.
  - Establishment of additional controls (audited financial statements publication, audit committees, supervisory boards) at many central banks.

### External audit quality, state auditors, and data-transparency context
- External audit assessment criteria:
  - Independence of auditor and quality of the audit (compliance with International Standards on Auditing—ISA).
- State auditors:
  - No prima facie case to exclude state auditors, but in all but one case state auditors did not meet minimum safeguards requirements.
  - The state auditor was the auditor of the central bank in 14 assessments.
  - Proposed remedies: joint-audits with a more experienced audit firm.
- Private audit firms:
  - Private firms, including local affiliates of international firms, sometimes found inadequate.
  - Proposed measures: add more experienced or international personnel to audit teams; conduct peer reviews by an international office of the same firm.
- Fiscal and data transparency milestones (as of December 31, 2004):
  - 73 fiscal ROSCs completed.
  - 51 data ROSCs completed.
  - Current work program projects an average of 12–15 fiscal ROSCs per year.
  - On average, 15 data ROSCs are conducted annually.
  - The Data Quality Assessment Framework (DQAF) developed; data ROSCs assess selected macroeconomic datasets based on DQAF.

### Obstacles, timing, confidentiality, and resources
- Main obstacles to implementation:
  1. Measures requiring changes to regulations or national law.
     - Legal-structure and independence recommendations represent about seven percent of total recommendations.
     - Legislative difficulties especially on operational independence matters; staff limited remedial measures requiring legislative amendments to operational aspects.
  2. Capacity constraints.
     - Lack of expertise to implement IFRS and cost of proper audits.
     - Staff responses: realistic deadlines for IFRS implementation; prioritizing TA requests; facilitating donor funding (e.g., for an ISA external audit).
  3. Political sensitivities.
     - Can delay implementation (examples: The Gambia special audit; proposed joint-audit with independent international firm).
     - Staff explored interim solutions (special audits, interim targeted NIR and reserve money reviews).
- Timing and confidentiality:
  - Operational deadline: assessments should be completed no later than the first review of the Fund-supported program; not met in nine cases.
  - Where completed after program approval, reports included review of quantitative performance criteria or TMU.
  - Implementation timeframes case-by-case; interim measures may include formal reconciliations and IFRS-compliant templates for interim financial statements.
  - Safeguards assessment reports are confidential and transmitted to the central bank and relevant Fund Executive Director.
  - Staff proposes that, with authorities’ consent, reports may be shared with other international organizations (World Bank, ADB, AfDB, ECB), subject to confidentiality.
- Resources and costs (FY2005):
  - Absorbed 15 full-time staff equivalents (FTEs).
  - Projected travel expenses about US$600,000.
  - Aggregate approximates US$2.9 million, within the overall Fund administrative budget of US$905.1 million.
  - Operations streamlined via methodology, specialized staff, collaboration with central banks, and IT; financed through redeployment within FIN and assistance from OIA and MFD in initial phases through mid-2004.
  - Staff view: resources believed adequate to meet expected demands.

### The way forward — workload, assessment types, and EPCA modalities
- Workload and rationale for continuation:
  - Shift expected from new assessments to update assessments and monitoring; shift delayed in 2003 due to 22 new arrangements in FY2003 and 14 in FY2004.
  - Three reasons to continue safeguards:
    1. Assessments are a “snapshot”; internal control environments are dynamic and warrant periodic reassessment.
    2. Vigilance needed to protect Fund resources and manage reputational risk; monitor central bank changes in accounting/auditing.
    3. New arrangements may be requested by countries whose central banks have not had a safeguards assessment.
  - Evolution: greater reliance on previously implemented controls; move toward compliance-oriented assessments; resources allocated should decline over time as first-time assessments decrease, offset partially by ongoing monitoring.
  - Monitoring should remain while Fund credit is outstanding.
- Types of safeguards assessments (staff envisages three types):
  - New assessments: full safeguards assessments for countries not previously assessed; typically on-site after off-site evaluation.
  - Update assessments: expected to constitute the majority of future work; conducted when a previously assessed country requests a subsequent/successor arrangement; may be off-site or on-site depending on risk.
  - Monitoring assessments: normally conducted from headquarters unless on-site warranted by triggers (e.g., disruption in safeguards monitoring, de novo issues flagged by external auditor, safeguards issues identified in surveillance).
- Extension to EPCA cases:
  - Rationale: difficult to justify excluding EPCA cases when risks to Fund resources may be higher.
  - Practical constraints: full assessments may be infeasible in early post-conflict stages due to absence of a fully functional central bank and limited capacity.
  - Proposed modalities (case-by-case, based on central bank functionality and capacity):
    - If no functioning central bank: delay assessment until reconstruction establishes functional capability.
    - If central bank exists but lacks full functionality: evaluate functionality; if full assessment infeasible, undertake targeted safeguards assessment aimed at basic control functions.
    - If a functioning central bank exists (examples cited: Rwanda and the Democratic Republic of Congo), conduct a full safeguards assessment.

### Extension to emergency assistance, reporting periodicity, and independent panel observations
- Extension to emergency assistance:
  - Staff does not propose extension of safeguards policy to emergency assistance for natural disasters or outright purchases under the Compensatory Financing Facility (CFF).
  - Rationale: Fund lending under these facilities is required under very urgent circumstances, making full safeguards assessments unrealistic; unlike EPCA cases, no presumption member countries will subsequently move to an arrangement.
- Reporting periodicity:
  - Current practice: semi-annual update papers to the Board on safeguards assessment activities.
  - Staff proposal: prepare update papers on an annual basis going forward.
  - Periodic reviews of the safeguards policy expected every three years.
- Independent panel observations and staff response:
  - Panel: safeguards policy a success; process relatively efficient.
  - Panel recommended strengthening communications and monitoring procedures:
    - More regular contact between Fund staff and central bank staff on safeguards issues.
    - Enhanced safeguards information on the Fund’s web-site.
    - Possible coordination of regional and global forums on safeguards issues.
    - Monitoring modalities to include on-site visits when necessary and off-site updates on ELRIC developments.
  - Staff agrees and is operationalizing these recommendations.

### Issues for Board consideration and Annex I highlights
- Issues for Board discussion:
  - Whether the safeguards policy achieved desired objectives and whether changes should be made.
  - Views on the way forward, including modalities for the three assessment types, EPCA approach, periodicity of Board update papers (semi-annual to annual), and periodic review every three years.
  - Confidentiality: whether safeguards assessment reports could be shared with staff from other IFIs (World Bank, ADB, AfDB, ECB) with authorities’ consent and similar guidelines to technical assistance reports.
- Annex I — Completed Safeguards Assessments (as of March 1, 2005) — summary points:
  - Transitional Procedures - Report on the External Audit Mechanism: Completed in 2001.
  - Completed in 2002 (27), 2003 (16), 2004 (22), 2005 (24) and other tabulations as presented.
  - Country note: 19 central banks have undergone two full safeguards assessments (Albania, Argentina, Bangladesh, BEAC, Bolivia, Bosnia and Herzegovina, Bulgaria, Croatia, Georgia, Kenya, Lesotho, Malawi, Mongolia, Nepal, Nicaragua, Peru, Romania, Sri Lanka, and Tajikistan); Brazil has undergone three full safeguards assessments.
  - Footnotes: Vietnam excluded because arrangement expired before assessment completion; targeted assessment example: Burundi; data as of March 1, 2005; one assessment (Guatemala) in progress and near completion.

*Source — _033105 (PDF, as of March 1, 2005).*

### Executive Summary ......................................................................................................

### Executive Summary

### Objectives of the Safeguards Policy
- Introduced in March 2000 to provide reasonable assurance that central banks have adequate financial control systems to manage their resources and Fund disbursements.
- Main objective: minimize the possibility of misreporting or misuse of Fund resources associated with the Fund’s lending activities.
- Safeguards assessments apply to all member countries that have a Fund arrangement.
- Assessments evaluate five areas (ELRIC): (i) the External audit mechanism, (ii) the Legal structure and independence, (iii) the financial Reporting framework, (iv) the Internal audit mechanism, and (v) the system of internal Controls.

### The Safeguards Process
- Safeguards assessments are diagnostic exercises carried out by Fund staff and are an integral part of the Fund’s lending activities; they apply to all new Fund arrangements, including augmentations to existing arrangements.
- Assessments are independent from Fund surveillance, program negotiation, and technical assistance activities.
- Assessments are based on review of documents and discussions with authorities and the central bank’s external auditors; typically include an off-site document review and a supplemental on-site visit.
- Outcome: a safeguards assessment report with recommendations and associated timeframes to address identified vulnerabilities; recommendations may be included in program conditionality when necessary.
- Post-report: country authorities may comment before finalization; Fund staff conducts off-site monitoring of recommendation implementation through periodic communications and review of annual financial statements and external auditors’ reports for as long as Fund credit is outstanding.
- Reports are confidential and transmitted to the central bank and the relevant Fund Executive Director; the Executive Board receives summaries in staff reports and periodic updates.

### Results and Impact to Date
- Completed safeguards assessments: 111.
- Central banks covered: 69 (some central banks received transitional and full assessments or multiple assessments due to subsequent arrangements or augmentations).
- Four broad categories used to evaluate results:
  - Impact on central banks’ operations and governance structures.
  - Enhanced awareness of the quality and reliability of data provided to the Fund.
  - Misreporting cases.
  - Possible misuse of Fund resources.
- Positive effects noted:
  - Considerable improvements in controls at central banks.
  - Heightened awareness of importance of proper audits, adherence to high-quality accounting standards, and appropriate control and oversight functions.
  - Policy helped focus central banks on the quality and robustness of data provided under Fund-supported programs.
- Limitations:
  - The policy cannot prevent misuse of resources resulting from willful override of controls or manipulation of data.
  - The policy does not cover fiscal data because it focuses on central banks.
- Implementation obstacles observed in some cases:
  - Specific legal restrictions in some countries.
  - Capacity constraints at central banks.
  - In a few cases, political sensitivities.
  - In most instances acceptable compromises were agreed without undermining the policy’s objectives.

### Overall Evaluation and Framework
- Based on five years’ experience, the framework used for safeguards assessments has been broadly appropriate.
- Reference benchmarks generally interpreted flexibly; country-specific circumstances taken into account where possible.
- The independent panel (appointed following the Board’s 2002 request) concluded that “the safeguards policy has been highly successful in achieving its objectives.”
- The panel recommended strengthening communication between the Fund and central banks on safeguards issues and broadening monitoring procedures for previously assessed central banks; staff concurs and will broaden available channels of information and communication.

### Future Directions (Way Forward)
- Staff expects a shift toward monitoring of previously assessed central bank frameworks and more focused, targeted follow-up assessments.
- Safeguards assessments under Emergency Post-Conflict Assistance (EPCA) cases: proposed to be undertaken on a case-by-case basis with the scope reflecting the degree of re-establishment of a fully functioning central bank.
- Continued application of off-site monitoring for as long as Fund credit is outstanding.

### Issues for Board Consideration
- Issues for Board consideration are set out in Section V of the source document.

*Source: Executive Summary and Sections I–II.C of the Safeguards Assessments review (pages excerpted).*

### 13.      The safeguards recommendations have strengthened the control frameworks in

### _033105 - 13.      The safeguards recommendations have strengthened the control frameworks in

### Implementation of safeguards recommendations
- Total recommendations included in 111 safeguards assessments (excluding recommendations with suggested implementation date later than March 1, 2005 and recommendations not monitored): 569.
- Overall implementation rate (as of March 1, 2005): 80.7 percent.
- Recommendations with formal commitment from the authorities: 124; Implementation Rate: 96.8 percent.
  - Under program conditionality: 57 recommendations; Implemented: 55; Implementation Rate: 96.5 percent.
  - LOI/MEFP commitments: 67 recommendations; Implemented: 65; Implementation Rate: 97.0 percent.
- Recommendations not under program conditionality or LOI/MEFP commitments: 445; Implemented: 339; Implementation Rate: 76.2 percent.
- Conditionality composition (included in the 57 conditionality recommendations): 11 prior actions (all implemented), 14 structural performance criteria (13 implemented), and 32 structural benchmarks (31 implemented) as conditionality in the program design for a total of 27 countries.

### Main findings from safeguards assessments (scope and trends)
- Total safeguards assessments completed as of March 1, 2005: 111 assessments covering a total of 69 central banks (several central banks subject to multiple assessments).
- Transitional assessments (members with a Fund arrangement in place as at June 30, 2000): 27 transitional assessments covered only the central bank’s external audit function; full safeguards assessments cover all five ELRIC areas.
- Main findings, 2000–2002 versus 2003–present (percentages as reported):
  - Non-existent or deficient external audit: 2000–2002: Number 34; Percent of Total 52. 2003–present: Number 11; Percent of Total 24.
  - No or delayed publication of financial statements: 2000–2002: Number 25; Percent of Total 38. 2003–present: Number 20; Percent of Total 43.
  - Inadequate controls over foreign reserves: 2000–2002: Number 18; Percent of Total 47. 2003–present: Number 15; Percent of Total 33.
  - Inadequate financial reporting standards: 2000–2002: Number 30; Percent of Total 79. 2003–present: Number 24; Percent of Total 52.
  - Deficient governance oversight: 2000–2002: Number 26; Percent of Total 68. 2003–present: Number 23; Percent of Total 50.
  - Deficient internal audit: 2000–2002: Number 29; Percent of Total 76. 2003–present: Number 26; Percent of Total 57.
  - Inadequate governing legislation: 2000–2002: Number 16; Percent of Total 42. 2003–present: Number 17; Percent of Total 37.
  - Inadequate accounting for Fund transactions: 2000–2002: Number 9; Percent of Total 24. 2003–present: Number 8; Percent of Total 17.
- Memorandum: During 2000–2002, 15 transitional assessments identified a non-existent or deficient external audit mechanism and 10 noted issues with respect to the publication of financial statements.

### Key implementations and institutional improvements (selected measures and coverage)
- First time appointment of an international auditor: implemented by 18 central banks.
- First time reconciliation of key program data with accounting records: implemented by 34 central banks.
- First time publication of full audited financial statements: implemented by 19 central banks.
- Internationally recognized financial reporting framework introduced: implemented by 16 central banks.
- Internal audit measures:
  - New internal audit function established: 2 central banks.
  - New reporting line to the governor or the board established: 3 central banks.
  - Audit charter adopted: 14 central banks.
- Establishment of a new audit committee or supervisory board: implemented by 15 central banks.

### Impact on data integrity and program reporting
- Focus on controls over compilation of critical monetary data reported to the Fund, specifically net international reserves (NIR) and net domestic assets (NDA), and consistency with Technical Memorandum of Understanding (TMU) terms.
- Improvements instituted by central banks included establishment of independent oversight roles by external auditors and audit committees, periodic reconciliations of monetary program data with accounting records, strengthened timeliness of reporting, and enhanced collaboration among departments.
- Specific corrective actions identified:
  - Exclusion of restricted assets from NIR after anomalies; establishment of periodic reconciliations and internal audit reviews to ensure conformity with program requirements.
  - Adjustment of program data where earmarked amounts for specific imports were incorrectly included in NIR; steps to establish formal reporting procedures with Fund technical assistance.
  - Transition from historical cost valuation to market-based valuation for foreign treasury instruments; adoption or movement toward full IFRS and appointment of international audit firms with IFRS experience.
  - Commissioning of periodic audits by treasury where consolidated NIR targets included treasury operations not subject to central bank audit controls.

### Misreporting cases and role of safeguards
- Three misreporting cases of monetary program data since introduction of the safeguards policy: Ghana, The Gambia, and Vietnam.
  - The safeguards assessment was key in detecting misreporting in The Gambia; in The Gambia: revised program data and financial statements were provided in October 2003; a second on-site assessment in November 2003 confirmed governance issues, lack of auditor independence, and inadequate internal controls; remedial measures included change in external auditor, review of internal controls, reaudits of 2001 and 2002 financial statements, and a special audit of foreign reserves; two noncomplying disbursements from July and December 2001 were repaid.
  - Ghana: a 2001 transitional assessment found serious concerns about audit quality and auditors’ qualifications and independence; an international audit firm was subsequently appointed.
  - Vietnam: the safeguards assessment initiated in 2001 (not completed) found external audit not in accordance with ISA.
- Example of severe data issues: an external auditor could obtain third-party confirmations for only about 15 percent of end-2002 foreign reserves; auditor issued a qualified opinion on currency in circulation; official foreign reserves reported as at end-2003 were subsequently revised downward by close to 50 percent.

### Misuse of Fund resources: limits and qualitative effects
- The safeguards policy cannot prevent misuse arising from willful override of controls or manipulation of data; it focuses on central bank operations and does not cover fiscal activities.
- Qualitative factors relevant to assessing effects on misuse:
  - Deterrent effect due to outside scrutiny.
  - Fraud allegations surfaced during assessments at two central banks leading to replacement of senior management and strengthening of controls.
  - Establishment of additional control and oversight processes (publication of audited financial statements, audit committees, supervisory boards) at many central banks.

### Overall evaluation and operational aspects
- Overall evaluation: The safeguards policy has had a positive impact on central banks’ governance and control frameworks; high implementation rates suggest broad acceptance of assessments’ usefulness; effectiveness regarding prevention of misuse of Fund resources is unknown but there has been a positive effect on the quality of data reported to the Fund.
- Scope of the policy: Focused on central banks and does not cover fiscal data; limitation accepted by the Board at policy introduction and during the 2002 review; staff does not propose extending assessments to fiscal area.
- Application and flexibility:
  - The ELRIC framework has been broadly appropriate for assessing central bank operations.
  - The uniformity of treatment policy has been critical to general acceptance and mandatory application for all users of Fund resources.
  - Flexibility examples: recognition of ECB accounting guidelines as an acceptable alternative to IFRS in some cases; not recommending an audit committee where the board provides adequate independent oversight.

*Italic: Source — _033105 - 13. The safeguards recommendations have strengthened the control frameworks in (PDF, as of March 1, 2005).*

### 27.      The quality of the external auditor has been a difficult issue in a number of

### 27.      The quality of the external auditor has been a difficult issue in a number of

### External audit quality and findings
- External audit mechanisms are judged on: (i) the independence of the auditor; and (ii) the quality of the audit (compliance with International Standards on Auditing—ISA).
- State auditors:
  - No prima facie case to exclude state auditors, but in all but one case state auditors did not meet the minimum safeguards requirements.
  - The state auditor was the auditor of the central bank in 14 assessments.
  - Proposed remedies for inadequate state audit performance included joint-audits with a more experienced audit firm.
- Private audit firms:
  - Private audit firms, including local affiliates of large international audit firms, have in some cases been found to be inadequate.
  - Proposed measures for inadequate private audit firms included adding more experienced or international personnel to the audit team and conducting peer reviews (by an international office of the same firm).

### Fiscal and data transparency developments (Box 3)
- As of December 31, 2004:
  - 73 fiscal ROSCs had been completed.
  - 51 data ROSCs had been completed.
- Fiscal ROSCs:
  - Examine clarity of fiscal management frameworks, publication of fiscal data, intergovernmental relations, and transparency/accountability of fiscal activities outside the budget.
  - A 2003 review noted fiscal ROSCs help systematize fiscal management information significant to Fund surveillance, program countries and TA activities.
  - The current work program projects an average of 12–15 fiscal ROSCs per year.
- Data ROSCs and DQAF:
  - The Data Quality Assessment Framework (DQAF) has been developed to assess data quality systematically.
  - Data ROSCs assess selected macroeconomic datasets based on the DQAF.
  - On average, 15 data ROSCs are conducted annually.

### Obstacles to implementing safeguards recommendations at central banks
- Three main obstacles:
  1. Measures requiring changes to regulations or national law.
     - Legislative difficulties often stem from recommendations concerning operational independence of the central bank.
     - Legal-structure and independence recommendations represent about seven percent of total recommendations.
     - Staff limited remedial measures requiring legislative amendments to operational aspects (e.g., clarification of provisions on credit to government, appointment/removal terms for board members and central bank head).
  2. Capacity constraints.
     - Central banks cited lack of available expertise to implement IFRS and the cost of conducting proper audits.
     - Staff responses included: (i) proposing realistic deadlines for major initiatives such as IFRS implementation; (ii) prioritizing technical assistance requests related to safeguards recommendations; and (iii) facilitating donor funding for improvements (e.g., for an ISA external audit of one central bank).
  3. Political sensitivities.
     - Political sensitivities can delay implementation (examples: special audit related to misreporting in The Gambia; proposed joint-audit involving an independent international firm).
     - Staff explored interim solutions (e.g., special audit underway in The Gambia; interim targeted review of NIR and reserve money in another case).
     - Close collaboration among staff and Fund country teams is necessary.

### Deadlines, confidentiality, and assessment timing
- Operational deadline:
  - Safeguards assessments should be completed no later than the first review of the member’s Fund-supported program.
  - This requirement was not met in nine cases.
  - Where assessments were completed after program approval, they included a review of quantitative performance criteria or the TMU agreed under the program.
- Implementation timeframes:
  - Timeframes are case-by-case and depend on nature of vulnerability.
  - Examples of interim/supplementary remedial actions when long implementation periods are expected:
    - Formal reconciliation of account balances with key monetary data reported to the Fund at specified dates.
    - Use of IFRS-compliant templates, developed with external auditors, for interim financial statements prior to full IFRS adoption.
- Confidentiality:
  - Safeguards assessment reports are confidential and transmitted to the central bank and relevant Fund Executive Director.
  - The Executive Board receives a summary of findings/recommendations in staff reports and periodic update reports.
  - Staff proposes that, with authorities’ consent, safeguards assessment reports may be shared with other international organizations (e.g., World Bank, ADB, AfDB, ECB), mirroring guidance on technical assistance reports, on the understanding that reports will be kept confidential.

### Resources, costs, and staffing
- FY2005 safeguards operations:
  - Absorbed 15 full-time staff equivalents (FTEs).
  - Projected travel expenses of about US$600,000.
  - Aggregate approximates US$2.9 million, within the overall Fund administrative budget of US$905.1 million.
- Implementation approach:
  - Streamlined operation achieved via efficient methodology, specialized staff, collaboration with central banks, and IT to manage data.
  - Safeguards operations financed through redeployment within FIN and assistance from specialized staff in other departments (OIA and MFD) in initial phases through mid-2004.
- Staff view:
  - Resources believed adequate to meet expected demands going forward.

### The way forward — workload and rationale for continued safeguards
- Transition in workload:
  - Expected shift from new assessments to update assessments and monitoring did not fully materialize in 2003 due to more new Fund arrangements than anticipated: 22 new arrangements in FY2003 and 14 in FY2004.
  - Focus is shifting toward update assessments of previously assessed central banks and monitoring of prior recommendations.
- Three reasons to continue consistent safeguards application:
  1. Assessments are a “snapshot”; institutional internal control environments are dynamic and warrant periodic reassessment.
  2. Vigilance needed to protect Fund resources and manage reputational risk; monitoring central bank changes (e.g., accounting or auditing practices) is important.
  3. New arrangements may be requested by countries whose central banks have not yet had a safeguards assessment.
- Evolution of assessments:
  - Greater reliance on previously implemented controls; high-quality external audits should make it easier to identify internal control issues.
  - Intensity will move toward compliance-oriented assessments.
  - Resources allocated to safeguards assessments should decline over time as first-time assessments decrease, partially offset by ongoing monitoring.
  - Monitoring should remain in place for as long as Fund credit remains outstanding.

### Types of safeguards assessments (staff envisages three types)
- New assessments:
  - Full safeguards assessments for countries not previously assessed; typically on-site after off-site document evaluation.
- Update assessments:
  - Expected to constitute the majority of future safeguards work.
  - Conducted when a country previously assessed requests a subsequent or successor arrangement.
  - Take into account prior assessment, monitoring of recommendations, and new information.
  - May be off-site or on-site depending on perceived risk.
- Monitoring assessments:
  - Normally conducted from headquarters unless on-site is warranted.
  - Triggers for on-site monitoring include: (i) disruption in safeguards monitoring (e.g., significant delays in priority recommendations, program off-track); (ii) need to assess de novo safeguards issues due to control breakdowns (e.g., external auditor raises critical concerns or flags issues in audit opinion); (iii) safeguards issues identified in Fund surveillance requiring follow-up.

### Extension of safeguards policy to EPCA cases
- Rationale:
  - Executive Board discussed extension in March 2004.
  - Hard to justify excluding EPCA cases when risks to Fund resources may be higher.
- Practical constraints:
  - Full assessments may be infeasible in early post-conflict stages due to absence of a fully functional central bank and limited capacity to implement recommendations.
  - A full assessment may not be meaningful until fundamental controls are in place.
  - Targeted assessments may be more appropriate to monitor progress toward necessary systems before moving to upper-credit tranche or PRGF arrangements.
- Proposed modalities (case-by-case, based on central bank functionality and administrative capacity):
  - If no functioning central bank: delay assessment until reconstruction establishes sufficient functional capability.
  - If a central bank exists but lacks full functionality: evaluate functional capability; if full assessment infeasible, undertake targeted safeguards assessment aimed at basic control functions.
  - If a functioning central bank exists (as in some post-conflict cases such as Rwanda and the Democratic Republic of Congo), conduct a full safeguards assessment.

*Source: Excerpt from safeguards policy review chapter/section.*

### 44.      Staff does not propose extension of the safeguards policy to emergency assistance

### _033105 - 44.      Staff does not propose extension of the safeguards policy to emergency assistance

### Extension of the safeguards policy to emergency assistance
- Staff does not propose extension of the safeguards policy to:
  - emergency assistance for natural disasters cases; or
  - outright purchases under the Compensatory Financing Facility (CFF).
- Rationale:
  - Emergency assistance for natural disasters is available where a member country cannot meet its immediate financing needs arising from a natural disaster (e.g., earthquakes, floods, or hurricanes) without serious depletion of its foreign reserves.
  - The CFF was established to ensure timely external financing for member countries experiencing balance of payment difficulties resulting from a temporary decline in export earnings.
  - In both cases:
    - Fund lending is required under very urgent circumstances which would not present realistic conditions for the conduct of a full safeguards assessment.
    - Unlike EPCA cases, there is no presumption that member countries would subsequently move to an arrangement.

### Periodicity of Safeguards Update Reports to the Board
- Current practice:
  - Semi-annual update papers are presented to the Board on safeguards assessment activities.
  - These reports summarize assessments conducted to date, including the number of safeguards recommendations made and their implementation rates.
- Staff proposal:
  - Update papers should be prepared on an annual basis going forward.
  - Periodic reviews of the safeguards policy are expected to be performed every three years.
- Supporting detail:
  - In past years, about twenty safeguards assessments were completed annually, but since fewer new assessments are expected going forward, it may be more meaningful to provide an annual rather than a semi-annual update.

### Independent Panel’s observations
- Panel conclusions:
  - The safeguards policy has been a success and the safeguards process has been relatively efficient.
- Panel observations and recommendations in two areas: communications and monitoring procedures:
  - Communications: need to strengthen communication in three main areas:
    - more regular contact between Fund staff and the central bank staff on safeguards issues;
    - enhancement of safeguards information on the Fund’s web-site;
    - possible coordination of regional and global forums on safeguards issues.
  - Monitoring procedures:
    - Modalities for monitoring safeguards at central banks previously subject to an assessment should include on-site visits to central banks when deemed necessary.
    - Off-site monitoring should include updates on pertinent developments in a central bank’s ELRIC areas, in addition to the current monitoring of the implementation of safeguards assessment recommendations.
- Staff response:
  - Staff agrees with the panel’s recommendations and is taking steps to operationalize them.

### Issues for Board discussion (paragraph 47)
- Directors are invited to comment on:
  - Whether the safeguards policy achieved its desired objectives and whether any changes should be made (Section II).
  - Views on the way forward for the safeguards policy (Section V), specifically:
    - whether the modalities for the three types of assessment going forward and the proposed approach for EPCA cases are broadly appropriate; and
    - whether (i) the periodicity of update Board papers on safeguards assessments could be changed from a semi-annual to an annual basis, and (ii) the policy should continue to be periodically reviewed, say, every three years.
  - Confidentiality of safeguards reports:
    - Whether the safeguards assessment reports could be shared with staff from other IFIs (World Bank, ADB, AfDB, ECB), with the authorities’ consent and in accordance with guidelines similar to those for technical assistance reports.

### Annex I — Completed Safeguards Assessments (as of March 1, 2005)
- Transitional Procedures - Report on the External Audit Mechanism: Completed in 2001
- Completed in 2002 (27)
- Completed in 2003 (16)
- Completed in 2004 (22)
- Completed in 2005 (24)
- Completed in 2005 (19) and (3) [table entries as presented]
- Note entries and example country listings included in the annex (excerpted):
  - Countries appearing across years include: Argentina, Albania, Bolivia, BEAC, Bosnia and Herzegovina, Bulgaria, Croatia, Georgia, Kenya, Lesotho, Malawi, Mongolia, Nepal, Nicaragua, Peru, Romania, Sri Lanka, Tajikistan, Brazil (Brazil has undergone three full safeguards assessments).
- Footnotes and explanatory notes in Annex I:
  - 1 Vietnam is excluded because the arrangement expired before the assessment was completed.
  - 2 In light of the constrained administrative and technical capacity in Burundi, a targeted assessment focusing on the external audit, financial reporting, and internal controls was conducted.
  - 3 19 central banks have undergone two full safeguards assessments: Albania, Argentina, Bangladesh, BEAC, Bolivia, Bosnia and Herzegovina, Bulgaria, Croatia, Georgia, Kenya, Lesotho, Malawi, Mongolia, Nepal, Nicaragua, Peru, Romania, Sri Lanka, and Tajikistan; and one central bank, Brazil, has undergone three full safeguards assessments. Subsequent full assessments at central banks constitute update assessments as described in paragraph 40 of the paper.
  - 4 Data as of March 1, 2005. There was one other assessment, Guatemala, that was in progress and near completion.

*Source: IMF staff paper excerpt (paragraphs 44–47 and Annex I).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2005/_033105.pdf_