IMF国际货币组织全球-Djibouti_Technical-Assistance-Report_14页_703kb

Djibouti Technical Assistance Report: Financial Soundness Indicators Mission Summary

Core Content Overview

This report outlines the findings and recommendations of a technical assistance mission conducted by the International Monetary Fund (IMF) in Djibouti from March 4 to 11, 2018. The mission aimed to support the Central Bank of Djibouti (CBD) in compiling financial soundness indicators (FSIs) in accordance with the IMF’s Financial Soundness Indicators Compilation Guide (FSI Guide). The report includes an assessment of the financial system, source data, and technical recommendations for improving the FSI process.

Main Objectives of the Mission

1. Ensure that source data were adequate for FSI compilation.
2. Assist the CBD in compiling FSIs based on international standards.
3. Guide the CBD in preparing FSI metadata in line with IMF forms.
4. Agree on an action plan for the production and regular reporting of FSIs to the IMF’s Statistics Department (STA).

Key Findings

• Financial System Structure: The banking sector dominates Djibouti’s financial system, accounting for 94% of total assets and 80% of total loans. As of end-2017, there were 12 active commercial banks (1 domestically-controlled, 11 foreign-controlled) and two banks under liquidation.
• Islamic Banks: Represented 20% of total banking assets in 2017, up from 1% in 2006. A national Sharia committee was established to oversee Islamic financial instruments.
• Microfinance Institutions (MFIs): Although growing, MFIs have a modest share of total assets and loans (less than 1%), and were not included in the FSI calculation due to time constraints.
• Source Data Adequacy: Overall, the source data is sufficient for FSI compilation, but inconsistencies in reporting by banks complicate data manipulation and aggregation.
• Accounting Principles: Djibouti commercial banks follow national accounting principles based on the French chart of accounts, using the accrual basis for recording, with historical cost for domestic assets and market price revaluation for foreign assets.

Key Recommendations

Priority Recommendations

Technical Recommendations

• Automation of Data Submission: Implement a web-based tool to automate and harmonize banks' data submissions.
• Improvements to Call Report Forms:
  • Deactivate unnecessary cells.
  • Include automatic calculation of subtotals.
  • Develop automated validation controls.
  • Add automatic verification mechanisms to alert banks of inconsistencies before submission.
• Training: The CBD should provide regular training for its staff, especially the Banking Supervision Unit (BSU), on the FSI methodology and report forms.

FSI Compilation and Methodology

• Core FSIs (12): Focus on deposit takers (DTs) and include indicators such as regulatory capital to risk-weighted assets, nonperforming loans to total gross loans, return on assets, and others.
• Additional FSIs (13): Cover various aspects including geographic and sectoral distribution of loans, market liquidity, and real estate loans.
• OFC Indicators (2): Relate to other financial corporations, including their share of total financial system assets and GDP.
• Nonfinancial Corporations (5): Include total debt to equity, return on equity, earnings to interest and principal expenses, and net foreign exchange exposure.
• Households (2): Cover household debt to GDP and debt service and principal payments to income.
• Market Liquidity (2): Reflect the bid-ask spread and daily turnover ratio in the securities market.
• Real Estate Markets (4): Include residential and commercial real estate prices and loan ratios.

Action Plan

Conclusion

The mission confirmed that the source data is broadly adequate for compiling FSIs, but improvements in standardization, automation, and training are essential to enhance the efficiency and accuracy of the process. The CBD is encouraged to implement the recommended action plan to ensure consistent and reliable reporting of financial soundness indicators in line with international standards.