The contents of this report constitute technical advice provided by the staff of the IMF to the authorities of the Republic of Uzbekistan in response to their request for technical assistance. The main objective of the mission was to assist the Central Bank of the Republic of Uzbekistan (CBU) in improving the collection of statistical data on the balance of payments (BOP), including the adoption of an international transactions reporting system, and to assess progress in implementing the medium-term program for further development of External sector statistics. Along with the successes, the mission noted several shortages in the BOP compilation system. The transfer of the BOP compilation function from the Ministry of the Economy to the CBU, and the transition to Balance of Payments and International Investment Position Manual, sixth edition standards, have led to a data gap between the BOP compiled by two institutions. It is recommended to update the methodology for calculating the currency and deposits of households component per the mission’s recommendations, considering all possible inflows and outflows of foreign currency in cash by individuals.


The contents of this report constitute technical advice provided by the staff of the IMF to the authorities of the Republic of Uzbekistan in response to their request for technical assistance. The main objective of the mission was to assist the Central Bank of the Republic of Uzbekistan (CBU) in improving the collection of statistical data on the balance of payments (BOP), including the adoption of an international transactions reporting system, and to assess progress in implementing the medium-term program for further development of External sector statistics. Along with the successes, the mission noted several shortages in the BOP compilation system. The transfer of the BOP compilation function from the Ministry of the Economy to the CBU, and the transition to Balance of Payments and International Investment Position Manual, sixth edition standards, have led to a data gap between the BOP compiled by two institutions. It is recommended to update the methodology for calculating the currency and deposits of households component per the mission’s recommendations, considering all possible inflows and outflows of foreign currency in cash by individuals.

Summary of Mission Outcomes and Priority Recommendations

1. At the request of the Republic of Uzbekistan authorities for technical assistance (TA) on external sector statistics (ESS), and with the support of the Middle East and Central Asia Department (MCD) of the International Monetary Fund (IMF), a mission from the IMF Statistics Department (STA) visited Tashkent during October 15–26, 2018. This was the first TA mission under the auspices of the Data for Decision Fund and the second since the Republic of Uzbekistan Presidential Order of September 12, 2017, “On Measures to Ensure the Accessibility and Openness of Economic and Financial Data for the Republic of Uzbekistan” was issued.

2. The main objective of the mission was to assist the Central Bank of the Republic of Uzbekistan (CBU) in improving the collection of statistical data on the balance of payments (BP), including the adoption of an international transactions reporting system (ITRS), and to assess progress in implementing the medium-term program for further development of ESS. The main objectives of the mission were: (i) assist in compiling consistent historical balance of payments (BOP) time series and addressing structural gaps; (ii) for compilation of 2018 BOP data, ensure application of consistent methodology following international standards; (iii) assess BOP data reliability and facilitate improvements in specific components, including goods, services, remittances, and selected financial account components; (iv) advice on additional data sources, including administrative, and develop statistical methods for gap-filling missing observation; (v) assess feasibility of ITRS implementation; (vi) to follow up on the previous mission’s recommendations, including interagency data exchange and progress in the area of data dissemination; and (vii) conduct hands-on training on BOP compilation.

3. Over the past year, after the Republic of Uzbekistan authorities set a course toward opening the country, Uzbekistan has made progress in improving the compilation and dissemination of statistical data. The CBU has started the production of ESS in accordance with the Balance of Payments and International Investment Position (IIP) Manual, sixth edition (BPM6). Tables and analytical materials have been published on the CBU website on the BOP, IIP, international reserve assets (IRA), and external debt. In May 2018, Uzbekistan joined the enhanced general data dissemination system (e-GDDS).

4. Along with the successes, the mission noted a number of shortages in the BOP compilation system. A number of errors were made in compiling the BOP for the first half of 2018, primarily related to misclassification of selected BOP components, which would be relatively easy to rectify by reclassifying data from one component to another. A more serious problem is posed by undervalued components and components valued using the wrong methodology. Transactions between residents are also sometimes are recorded in the BOP. These departures from accepted methodology are described in detail in Part II of this report.

5. The transfer of the BOP compilation function from the Ministry of the Economy (ME) to the CBU, and the transition to BPM6 standards, have led to a data gap between the BOP compiled by two institutions. To improve the quality of external sector statistics and remove break in historic time-series, the mission has carried out the following work:

  • a) the mission assessed the quality of BOP data prepared for the first half of 2018. During their work, a number of errors involving data classification and scope were discovered and removed, and recommendations were made on improving the quality of specific BOP indicators (goods, transportation services, travel, wages, personal transfers, households’ holding of cash and deposits, reinvested earnings, reserve assets, and a number of other BOP components);

  • b) the mission adjusted a number of data collection forms for ESS, advised on new data sources and, together with the staff of the Balance of Payments Division, developed new reporting forms;

  • c) provided assistance in preparations for adopting the ITRS;

  • d) the mission analyzed the BOP for 2016–2017, prepared by the MOE. A tentative BOP for 2016–2017 was compiled together with Balance of Payments Division BOPD staff, based on existing data, and the BOP data for the first half of 2018 were validated;

  • e) the analytical BOP format used by the MCD mission was integrated into the working file for compiling the various BOP and IIP output tables;

  • f) an action plan was prepared and agreed on with the CBU management, setting forth the principal recommendations made based on the mission’s work.

6. To support progress in the above work areas, the mission identified the following priority recommendations carrying particular weight to make headway in improving ESS (Table 1).

Table 1.

Priority Recommendations

article image

7. Further details on the priority recommendations and the related actions/milestones can be found in the action plan under Detailed Technical Assessment and Recommendations.

Detailed Technical Assessment and Recommendations

A. Action Plan

8. The Action Plan presented below, which sets forth the key recommendations made upon the conclusion of the mission, was discussed and coordinated with CBU management. The mission believes that the timeframes and workloads will be acceptable to ESS compilers.

article image
article image
article image
Priority scale: H – High; M – Medium; O – Other.

B. Goods and Services

9. The SSC is the official producer of international merchandise trade statistics (IMTS) in Uzbekistan. The SSC receives a database of customs declarations (DBCD) from the Republic of Uzbekistan State Customs Committee (RUSCC). The database is generated on the basis of freight customs declarations (FCD) and aggregated data are published without any adjustments, broken down by country and commodity group. IMTS recommends using a general trade system, but the SSC adheres to a special trade system that requires additional adjustments on the part of BOP compilers (see BPM6, component 10.25). In order to make the adjustments described in the previous TA mission’s report the decision was made to rely on the DBCD, which the CBU receives each month from the RUSCC, in calculating the Goods component.

10. The DBCD contains information about all goods crossing the border of Uzbekistan. Various customs regimes are assigned to them, allowing one to make a selection of the appropriate Goods component. In particular, excluded from the overall flow are the flows of goods associated with the repair or processing of goods, diplomatic goods, imports/exports of goods on a temporary basis (up to one year), and flows of goods from customs warehouses and temporary storage locations into free circulation are, while imports of goods to customs warehouses and temporary storage locations are added. Goods reexported from customs warehouses and temporary storage locations are also added. The MOE has made similar adjustments. The CBU sets apart imports/exports of nonmonetary gold, as required by the standard presentation on the BPM6.

11. Further adjustments consist of the following: (i) goods purchased by carriers in airports; (ii) estimated data on shuttle trade; (iii) conversion of CIF prices to FOB prices. Information on goods purchased by carriers (primarily aircraft fueling) is provided by the national aviation company (NAC). Shuttle trade is assessed following special studies at border posts. The conversion from CIF prices to FOB takes into account the goods’ weight, average cost to transport one ton of freight, depending on the mode of transport and the region of the trading country. Goods imports from neighboring countries by ground transport are not included in the calculation. The goods adjustments lack an assessment of unrecorded import/export transactions, which may occur with understated prices or volumes in customs declarations. This phenomenon can be detected and assessed using the mirror comparison method. This recommendation of the previous mission has not been implemented. Estimated data on import/export of migrants should also be removed from the goods component (BPM6, para. 10.22(b)); this recommendation of the previous mission has also not been implemented.

12. Some types of services were not properly evaluated. Among them, one may note debit for passenger transportation, which are estimated at a virtually zero value. The CBU used SSC data when estimating a transportation service by passengers, while the SSC for its part does not survey the offices of nonresident airlines that are represented in Uzbekistan, nor other nonresident carriers. As a result, the SSC data for credit entries are adequate, but zero values were obtained for debit entries on passenger transportation by air and rail transport. A rough estimate of this component yields about USD 100 million per quarter. To assess passenger transport, one could use border crossing data of the Committee for State Border Protection (CSBP) and the estimated cost of one trip by mode of transport. The CSBP furnishes data to the CBU on border crossings by residents and by mode of transport, including: air, rail, motor vehicle, inland water, and animal-drawn transport, as well as pedestrian movement. Each type of transportation must be assessed separately. The largest share of transportation costs falls to air transport; therefore, its assessment must be handled more rigorously.

13. In order to precisely determine the number of residents serviced by foreign airlines, a question must be added to the questionnaire for the NAC regarding the number of residents who have used the NAC’s services. For rail transport, assume that passenger transportation within Uzbekistan is performed by a resident (Uzbek Railways), and outside Uzbekistan by a nonresident. For passengers departing for Russia, assume that 90 percent of transportation services are provided by a nonresident and only 10 percent by the national railway. For Kazakhstan, use 80 percent as the nonresident share, and 50 percent for the remaining countries. In addition, one must estimate the number of residents by the principal destinations. This information may be obtained from the CSBP, by requesting information on border crossings by individuals along with a breakdown by countries and modes of transportation. At this time, the CSBP provides a separate breakdown by countries and by modes of transport. Detailed information can also be obtained by sending an inquiry to the Uzbek Railways for information on passenger flows (number and cost) by the principal destinations with a breakdown by residency. An expert assessment may be used until specific results are obtained.

14. For the remaining modes of transport (road transport and inland water transport – passengers), one may use an expert assessment until studies are conducted: 50 percent of residents use transportation services of nonresidents. A very large share of individuals cross the border by animal-drawn transport or on foot, with a subsequent transfer to a bus or other mode of motor vehicle transport. To include this phenomenon, the mission recommends that the approximately 20–25 percent of residents who have crossed the border on foot be included in the calculation of road passenger transportation (and, in specific cases, recognizing them under the “travel” component).

15. Another major services component in need of adjustment by the CBU is travel, primarily import. Travel services are assessed based on the results of a tourism survey conducted by the SSC. In that study, the SSC adheres strictly to the definition of tourism; thus, the result for tourism services needs to be adjusted when compiling the BOP component “travel”. Spending by excursionists (persons not overnighting in the host country), transit passengers, and short-term workers is not included in tourism services for either inbound or outbound tourists. On the other hand, tourism services do include transportation expenditures for round-trip travel from one’s own country to the visited country—that is, international transportation services. Besides spending on goods for personal use, tourism services also include spending to purchase valuables, which are included in goods in the BOP (BPM6, component 10.20).

16. In order to use SSC tourism data in compiling the BOP “travel” component, the CBU must make adjustments, which are described below in the recommendations. For purposes of correcting the SSC data on tourism imports, the CBU performs an assessment of spending by short-term workers abroad, but the CBU has overvalued this component. The main reason for the overvaluation is that the CBU considers all who cite work as their purpose of travel to be short-term workers who live and work abroad 90 days each quarter. The CBU has prepared a model for calculating the number of short-term workers located abroad. According to this model, of those departing to earn money during the current quarter 20 percent return in the next quarter, another 20 percent in the second quarter, and so on. This approach has yielded overstated data for both the number of short-term workers and their spending abroad, which amounted to about USD 1.3 billion for the first half of 2018. As a result, it appears that some of the short-term workers remain abroad more than a year.

17. The following short-term worker return diagram is more suitable (Table 2). In addition, the estimated spending abroad of short-term workers should be reclassified from personal to business travel.

Table 2.

Short-term Worker Return Ratios

article image

18. The remaining services are not of such substantial proportions, and their assessment does not evoke serious doubts, and thus the mission did not review them at this stage. Going forward, compilers should study the methodology used by the SSC in assessing imports and exports of services, especially in such complex types of services as construction and insurance services, as well as pipeline transportation services.

Recommended Actions:

  • apply the mirror comparison method to estimate unrecorded transactions involving imports and exports of goods, and add the data obtained to the Goods component;

  • remove estimated data on imports/exports carried out by migrants from ‘goods’;

  • estimate debit of passenger transportation services for each mode of transport, and incorporate in the BOP;

  • prepare appropriate questionnaires and forward them to the NAC and Uzbekistan Railways in order to identify the number of residents who received passenger transport services from nonresident transportation companies;

  • correct the methodology for assessing the number of short-term workers located abroad and their spending, per the mission’s recommendations;

  • request information from the Ministry of Foreign Affairs concerning support for Uzbekistan exterritorial bodies and estimate debit side of the government services component. Include estimated data in the CBU database from the MOE BOP until the relevant information is received.

C. Primary and Secondary Income

19. The credit of compensation of employees component in the CBU balance sheet for the first half of 2018 is greatly overestimated at over USD 3.6 billion. The CBU assesses the number of short-term workers for the major countries hosting labor migrants from Uzbekistan (primarily Russia and Kazakhstan) and multiplies by the average wage in those countries. Information on average wages is received from the Ministry of Employment and Labor Relations (MZT). As noted above, the main reason for the wage overestimation lies in the overestimation of the number of short-term workers and their employment level. All those who cite work as their purpose of travel are assumed to be short-term workers and that they work 90 days each quarter. In reality, a fairly large proportion of labor migrants are long-term workers, and the average employment level of a short-term worker per quarter is generally no more than 2.0–2.5 months. The problem of dividing labor migrants into long-term and short-term was raised during the first mission as well. At that time, during a meeting at the SSC with representatives who were conducting a household budget survey, a proposal was made to include additional questions in the questionnaire for that survey. However, that recommendation has not yet been implemented. During the mission, this issue was again raised during a meeting with SSC staff, and firm promises were obtained as to its positive resolution. The proposed questions make it possible to divide the number of labor migrants into short-term (up to one year) and long-term (one year or more), and also to divide the funds incoming from short-term and long-term workers.

20. The mission visited the MZT’s Agency on External Labor Migration in order to address the above issue. During this meeting, a format was agreed on for the data on which this agency will report to the CBU. A reporting form was developed, and an official letter sent to the Agency during the mission. At this time a small number of workers are using the Agency’s services, but an increase in this number is anticipated, as Uzbekistan has introduced a system of benefits for labor migrants. Until usable results are obtained based on data from the above-mentioned surveys, the mission recommends the following expert estimates: define the share of short-term workers at 50–60 percent of all those citing work as their purpose of travel, take 2–2.5 months per quarter to be the average employment level, and use the diagram of return to Uzbekistan described in Table 2. Workers’ compensation calculated using this method will be 2–3 time less than in the current BOP. The CBU also estimates the workers’ compensations of Uzbekistan residents working in embassies and consulates of foreign states and international organizations located within Uzbekistan, adding the result to the compensation of employees component.

21. In parallel with the estimates of the compensation of employees component, the CBU also estimates so-called “net wages”—compensation of employees minus the worker’s expenses in the host country (’travel’ component) and taxes paid (secondary income). The methodology used to calculate workers’ expenses in the host country is described above, while taxes paid are calculated as follows: (i) the CBU estimates the number of workers and their wages by country, and an official income tax rate is identified for each country; and (ii) the amount of withheld tax is computed by multiplying wages by the income tax rate. The amount of net wages is used to calculate personal transfers as follows: money transfers made through electronic money transfer systems and imports of foreign currency in cash by residents are combined. The short-term workers’ net wages are then deducted from that amount, and the resulting value is considered a personal transfer. The mission believes that part of the inflow of foreign currency into Uzbekistan (either through electronic transfer systems or by importing cash) is not associated with either wages or personal transfers. It is very likely that a certain portion of the transferred currency is payment for exported goods or services. Part of the monetary flows into Uzbekistan are carried out among nonresidents. After publication of the ESS for the second quarter of 2018, the BOPD analyzed data sources among the electronic money transfer systems. It was discovered that it is possible to separate data according to the residency status of the Uzbekistan participant in a transaction. Going forward, when calculating wages, the BOPD will take into account money transfers to residents through electronic money transfer systems and the import/export of foreign currency in cash by residents and will revise data for previous periods.

Recommended Actions:

  • until suitable amounts are obtained identifying the proportion of short-term and long-term workers, use the following contingency estimate: define the share of short-term workers at 50–60 percent of the number of individuals who have left Uzbekistan for employment opportunities;

  • for countries that are located at a substantial distance or have complex procedures for obtaining a visa (such as the USA, Western European countries, Japan, and Australia)— consider all labor migrants to be long-term and exclude from COE;

  • define the average employment level per quarter at 2–2.5 months;

  • exclude 10–20 percent from the overall flow of foreign currency (money transfers to residents and imports of foreign currency in cash by residents), considering these flows unrelated to wages or personal transfers;

  • in countries where the income tax rate is 20 percent or more, adjust the income tax to 60–80 percent of the official amount.

22. Direct investments reinvested earnings are to be adjusted. Reclassify dividends paid out by PSA enterprises (about USD 452 million over the first half of 2018) from ‘income from other investments’ to income from direct investments. Some PSA enterprises engaged in mineral production and sales pay rent to the state for use of natural resources; these payments amounted to about USD 152 million over the first half of the year. The CBU records these payments in income, as other primary income. Since PSA enterprises are residents of Uzbekistan, the rental payment is a payment between residents and should not be recorded in Uzbekistan’s BOP under the ‘primary income’ component, but rather in direct investments. There are certain problems in the calculations of reinvested earnings of direct investors (meaning investments into Uzbekistan). In the CBU balance of payments, reinvested earnings are calculated for the following economic entities:

  • a) Nonfinancial corporations, other than PSA enterprises; source of data—a SSC Survey of the Foreign Economic Activities of Nonfinancial Corporations (SFEANC).

  • b) Enterprises with PSA. Source of data—specialized statements of these enterprises.

  • c) Commercial banks; source of data—banking supervision information about a bank’s foreign shareholders and reports from banks.

  • d) Nonbanking financial corporations; source of data—accounting balance sheets of these enterprises and information about foreign shareholders.

23. The methodology used to calculate reinvested earnings is the same for all economic entities. For reinvested earnings, the CBU took an enterprise’s net profit for the reporting year without deducting declared dividends. The problem was compounded by the fact that some economic entities would declare their current year profit using their net income for the reporting quarter including dividends, while others would deduct dividends. The mission studied the chart of accounts for both enterprises and for commercial banks. Retained earnings in both cases are presented in two lines:

  • Prior years’ retained earnings

  • Current year retained earnings

24. Therefore, the mission proposed that the SSC amend the survey form for the SFEANC, in particular, by adding a line for current year retained earnings and, for the ‘current period net profits’ line, creating fields for data as of the beginning of the reporting quarter and the end of the reporting quarter. To complete these lines properly, the amount of retained earnings and declared dividends for the reporting quarter must be equal to the enterprise’s net profit. For financial corporations (bank and non-bank), additional information will be collected on net profits in the reporting quarter, making it possible to perform similar checks. The total value of transactions in the reporting quarter involving changes to prior year retained earnings and current year retained earnings will be reinvested earnings. In the case of PSA, the specialized survey provides for calculating reinvested earnings.

25. Income from other investments are calculated adequately. The bulk of these income are made up of interest on loans. Changes to the report form, made following the previous mission’s recommendations, allowed the CBU to produce a calculation of income on loans using the accrual method. For private debt, data are compiled through the banking system on each individual debt, in the original currency with allowance for all the requisite parameters and for both the principal and accrued interest.

26. Income on reserve assets is reported Department for Management of International Reserves (DMIR), which conveys prepared data to the BOPSO. During the meeting with DMIR representatives it became clear that they do not perform calculations of income from SDR allocations or SDR holdings. Together with BOPD staff, the mission generated a file for calculating income from SDR holdings and allocations. The latter will be included in interest income from other investments as a debit.

Recommended Actions:

  • Dividends paid out by enterprises to foreign investors under production-sharing agreements (PSA) are to be reclassified from ‘income from other investments’ to the ‘income on equity and investment fund share dividends and withdrawals from income of quasi-corporations’ component;

  • add rental payments from PSA to direct foreign investments. Remove existing data under the other primary income, rent’ component;

  • calculate reinvested earnings according to the mission’s updated recommendations; and

  • use the file developed by the mission to calculate income on SDR allocations and SDR holdings.

D. Direct and Portfolio Investments

27. The CBU made definite progress in improving the system for collecting data and compiling direct investments (DI). The SSC completed work on the transition to a comprehensive reporting form for foreign economic activities of enterprises and has already obtained the first results. Information has been gathered from the private sector for the first and second quarters of 2018. Using the new reporting forms, the CBU has collected information on direct investments from deposit corporations other than the central bank (DTC) and from nonbank financial corporations (microfinance organizations, insurance companies, and leasing companies). In addition, the CBU is using the existing form to compile data on PSA, which is used to calculate DI in Uzbekistan. Investments abroad (assets) are also compiled, but their volume is negligible. The format of the collected data allows for statistics to be compiled on both positions and transactions.

28. The mission also noted a number of shortcomings in the data collection system for DI and the calculations for several components. The existing form for collecting data from PSA fails to provide for data collection the expenditure of such enterprises before mineral extraction begins. These are generally borne by the parent company and they are direct investments in the form of equity. Instead of direct investments/ debt instruments, the PSA’s debt obligations to foreign investors are recorded as other accounts payable/receivable in other investments component. The preferred (non-voting) shares of a DTC are recorded in share capital instead of debt instruments. There were problems in data collection and in the method used to calculate reinvested earnings. The SSC data on direct investments are not reconciled; change in stocks are not reconciled with transactions during the reporting period. Trade credits between direct investment enterprises and direct investors were recorded along with trade credits in other investments, instead of direct investments/ debt instruments.

29. The mission made a recommendation to remove the above shortcomings. In the PSA survey form, questions were added to collect information on the parent company’s costs before commencing mineral extraction and production. It is recommended that preferred shares in a DTC be withdrawn from share capital and included in either portfolio investments long-term debt securities if their owner is a financial corporation or added to debt instruments in the form of direct investments if the owner of these preferred shares is a nonfinancial corporation. The mission visited the SSC and studied the process of collecting and processing data pursuant to the SFEANC. The meeting included a discussion of inconsistencies between transactions and positions during the reporting period; promises were obtained that these inconsistencies would be eliminated. The issues related to the calculation of reinvested earnings are described above.

30. The CBU performs estimates investment in real estate. Information about real estate purchases and sales by residents of Uzbekistan abroad, and about real estate purchases by nonresidents in Uzbekistan, is collected in the context of exchange regulation reports, which is quite sufficient for recording real estate transactions. Information about real estate purchases/sales can also be obtained using the ITRS, whose introduction is planned in the second half of 2019.

31. To estimate other capital in the form of direct investments, the CBU uses data from the SFEANC survey and the banking system for collecting data on foreign loans. Per a recommendation of the previous mission, the CBU updated the information collection form for foreign loans through the banking system. The new system makes it possible to separate inter-company lending and ordinary loans. Information is collected both on principal and accrued interest. For banking system loans, the data collection system furnishes more suitable data, upon which the CBU fully relies. For trade credits between related enterprises, the mission recommends using information from the SFEANC survey conducted by the SSC. The mission also recommends recording PSA debts to foreign investors in ‘direct investments/debt instruments’.

32. Portfolio investments in Uzbekistan are negligible for both assets and liabilities. There are virtually no debt securities; the one component that exists in negligible amounts is capital participation instruments. At this stage, no financial derivatives have been discovered, either.

The mission recommends:

  • updating the DI reporting forms and the methodology for calculating reinvested earnings per the mission’s recommendations;

  • adding questions to the PSA reporting form about the foreign investor’s costs before beginning mineral production and sales, recording the data obtained in equity;

  • updating the ‘direct investments/debt instruments’ component per the mission’s recommendations;

  • withdrawing preferred shares in a DTC from equity and recording them in either portfolio investments, long-term debt securities or adding them to debt instruments in the form of direct investments, depending on the type of owner.

E. Other Investments and Reserve Assets

33. Despite certain successes in the collection and methods of processing data for the ‘other investments’ component, the mission discovered a number of deficiencies and deviations from accepted standards. Some of them were rectified during the mission. The most problematic component is ‘assets, other sectors, currency and deposits’. The CBU recorded central bank deposits placed in commercial banks of Uzbekistan in the ‘central bank, currency and deposits’ component, assets. At the mission’s recommendation these data were removed from that component but added the external assets in the form of cash and deposits denominated in non-reserve currencies or currencies not meeting other criteria for reserve assets.

34. Currency and deposits of DTC are calculated based on banking information; these are balances of commercial banks and information collected by the CBU in the context of exchange regulation. When calculating transactions under the ‘currency and deposits’ components for both liabilities and assets, all net changes are counted regardless of whether a commercial bank performed a transaction with a resident or nonresident. This approach yields adequate estimates of asset and liability positions, but partially distorts transaction volume.

35. The greatest problem with the ‘currency and deposits’ component concerns the estimation of households’ position and transactions’. It is brought about by the lack of direct data, and by the complexity inherent in conducting proper sampling surveys for volume of cash kept at home or deposits placed in overseas banks. The CBU methodology for calculating foreign currency cash holding and households’ deposits fails to encompass all possible flows for this instrument.

36. The existing system for classifying transactions involving currency and deposits in BOP transactions includes some transactions among residents; this does not alter the overall transaction for all institutional sectors but distorts the transactions of specific institutional sectors. According to CBU methodology, households’ currency and deposits consist of the following components:

  • net transfer of foreign currency into Uzbekistan (inflow/outflow, through electronic money transfer systems);

  • net import of foreign currency in cash into Uzbekistan (import/export of foreign currency in cash);

  • sales of foreign currency by individuals through exchange offices;

  • wages received by residents working in foreign embassies and international organizations in Uzbekistan;

  • shuttle imports/exports;

37. Both residents and nonresidents participate in most of these components: (i) not all money transfer recipients are residents of Uzbekistan; (ii) not all senders of foreign currency are residents of Uzbekistan; (iii) both residents and nonresidents carry foreign cash across the border; and (iv) exchange office services are used by nonresidents as well as residents. The CBU analyzed the data sources. It emerged that it is possible to separate the above components by residency. The CBU intends to recalculate data for previous periods on currency and deposits, factoring in the residency of the recipient of funds. The mission recommends identifying households’ currency and deposits based on the following components:

  • Net transfers of foreign currency to Uzbekistan by a recipient/sender who is a resident;

  • Net imports of foreign cash into Uzbekistan by residents;

  • Purchase/sale of foreign cash by resident individuals through exchange offices;

  • Wages received by residents working in foreign embassies and international organizations within Uzbekistan (residents receive wages in a foreign currency);

  • Other flows of foreign cash between banks and individual residents. This category includes: flows related to a deposit placement, withdrawal from a deposit account, obtaining/repaying a bank loan, down payment in foreign cash to exercise non-cash payment from the bank account, etc.

  • Other flows of foreign cash from individuals (illegal export of foreign cash by individuals, other potential expenses of individuals with nonresidents). It is impossible to estimate this component with any precision, but an estimate of 5 to 15 percent of the net inflow of money transfers would not be an exaggeration.

38. The mission does not recommend including expenditures on shuttle imports/exports when calculating the households’ currency and deposits component, since these transactions are carried out either by exported cash or by remitting funds through money transfer systems. Part of these costs are considered in the components listed above. The CBU’s current method and the method proposed by the mission make allowance for some inter-sectoral transactions (primarily between banks and individuals), but this will not affect the overall transaction or the balances of the respective institutional sectors.

39. The principal sources for calculating the loans component are: reporting by the Ministry of Finance (MOF) on government and government-guaranteed debt, the SFEANC conducted by the SSC, and CBU reporting in the context of exchange regulation. CBU reporting provides for data collection on each individual loan, collection of general information on such loan positions at the end of the reporting period, and actual payments. Asset side for the loans component is negligible. Loan liabilities in the government sector are calculated on the basis of MOF reports. These reports offer data presented in the original currency, separately for principal and for accrued interest. In addition, the MOF presents data separately for direct government debt and for government-guaranteed debt that is serviced by the creditor itself. This allows the CBU to delineate government-sector debt and debt of other institutional sectors. However, despite the separate presentation of data, the CBU aggregate them and records in the BOP under the government sector.

40. Liabilities on private-sector loans are estimated on the basis of information collected through the banking system. Following the previous mission’s recommendation, the CBU updated the format of collected information, making it possible to prepare statistics for both the BOP and the IIP. Within the framework of these reports, information was collected on 759 borrowings as of June 30. In particular, the CBU calculates transactions and balances by principal, accrued interest, past-due principal, and past-due interest. Interest income is calculated separately for the primary income component. The only shortcoming is that debt forgiveness was recorded in the BOP sheet only in capital transfers and, instead of a BOP transaction reducing the loan obligation, the loan was reduced in the IIP by other changes. This was corrected during the mission.

41. The CBU calculates trade credits of other investments using its own system for tracking accounts payable/receivable related to foreign trade transactions. This system uses banking and customs statistics to identify transactions involving trade credits, but it cannot determine whether a specific trade credit falls under direct or other investments. For this reason, the CBU includes everything furnished by this system in trade credits of other investments. The mission recommends deducting the amount of trade credits recorded under direct investments from trade credits received using the banking system and recording only the remaining amount under the ‘trade credits – other investments’ component.

42. The mission met with representatives of the Department for Management of International Reserves (DMIR) and discussed the methods used to calculate data on IRAs and to complete the forms agreed on with the previous mission. IRA calculations are done first in the original currency in accordance with the BPM6, then converted into USD, and exchange-rate changes are calculated. When converting to USD, the respective exchange rates are applied to balances as of the beginning and the end of the reporting period, while the rate as of the end of the reporting period is also applied for the transaction. For the transaction, the mission recommended either applying the rate as of the transaction date or the average rate for the reporting period. After consulting with management, the DMIR made the decision to apply average rates to the conversion of reserve asset transactions from the original currency to US dollars. Before the mission’s conclusion, the DMIR provided updated data on IRAs starting with December 31, 2015, which were used in compiling the 2016–2017 BOP.

The mission recommends that:

  • initial data from the MOF be processed using the methodology proposed by the mission, using the working files developed with the mission;

  • guaranteed debts on which the government has assumed repayment obligations be recorded among government-sector debts, and that debts serviced by the debtor itself be recorded in the appropriate sector depending on the borrower; memorandum items should clearly state the amount of government guaranteed debt and debt service;

  • a list of enterprises that had foreign economic transactions during the reporting period be provided to the SSC, for inclusion in the list of respondents for the new survey;

  • data on trade credits of other investments component be adjusted per the mission’s recommendations.

F. Program to Develop and Introduce an ITRS

43. One of the actions recommended by the November 2017 mission was to study the possibility of adopting an ITRS for the external sector of the economy of Uzbekistan. The information collected in the course of exchange regulation partially overlaps with the scope of ITRS but does not record receipts and payments by type of transaction or by receipt of funds. For this reason, it provides the ability to calculate only certain BOP components. The CBU expressed interest in the ITRS, studied the experiences of several countries including Kazakhstan and Georgia, and made the decision to create a system using the Georgian ITRS as a prototype.

44. Even before the current mission began, the CBU had taken the following first steps toward introducing an ITRS:

  • developing a draft ITRS reporting form for banks and instructions for completing it;

  • developing a specialized system of ITRS codes for banks;

  • holding consultations with commercial banks, familiarizing them with the ITRS system itself, and coordinating the form and code system with them;

  • defining a report presentation format that does not allow for consolidated BOP transactions to be performed;

  • developing a program for processing, analyzing, and assessing the intermediate data base.

45. After the above work was completed, the CBU launched pilot data collection in commercial banks at the beginning of October 2018. The Uzbekistan ITRS system has a number of its own special features, in particular: (i) banks report on all transactions, in both internal and external correspondent accounts; for internal transactions a different code system is used; (ii) only those transactions that pass through external correspondent accounts are sent to the BOPSO; (iii) banks do not report cash transactions; and (iv) threshold values are not used, all transactions are coded, and a resident transaction participant is identified for all transactions. The file sent to the CBU includes the name of the client/transaction initiator and their fiscal code.

46. The due dates for submitting reports to the CBU are fairly tight. Express information is submitted immediately following the closing of the business day. A special software program was developed to ensure timely submission of data. To ensure such rapid data submission, the ITRS codes are identified by the commercial bank operators. The final and revised data are forwarded to the CBU one month later. Commercial banks must send revised data to the CBU for the reporting month.

47. The mission familiarized themselves with the code system and reporting form and provided recommendations on their further improvement. The ITRS form does not contain a field for the institutional sector; the absence of such a field restricts the ability to apply the collected information. The mission prepared codes for the institutional sectors cited in Table 3.

Table 3.

Codes for Institutional Sectors

article image

48. The ITRS code system is in need of revision. The existing system was created in accordance with the criteria of the fifth edition of the BPM5 and does not take into account the changes occurring with the transition to the BPM6. The updated code system is cited in Appendix I. One the one hand, the adoption of ITRS codes by operators makes it possible to provide timely data to the CBU, but on the other it increases the risk of erroneous identification of transaction codes. It is difficult to train a large number of operators on the nuances of the BOP and the finer points of identifying a BOP transaction code.

49. To further improve the quality of bank reporting under the ITRS, the mission met with representatives of commercial banks. The meeting included discussions of matters concerning training bank specialists (back office personnel) to determine transaction codes and institutional sector codes and establishing communications between commercial banks and the BOPD in order to hold prompt consultations on determining the transaction code and institutional sector code. The mission proposed that the CBU arrange for regular seminars on issues of ITRS reporting. The commercial banks expressed their readiness to collaborate in the area of introducing the ITRS and promised to implement the CBU’s requirements.

50. Analysis of the first results from data received from commercial banks revealed a high number of errors in the identification of transaction codes and the unsuitability of accumulated data for subsequent use. Prior to the official (full-fledged) introduction of the ITRS, the CBU must develop procedures to ensure that the accumulated data are of suitable quality. A program of regular seminars for commercial banks must be prepared, to explain the goals of data collection and provide training in compiling reports under the ITRS system; develop procedures for detecting indicators that are outliers from the general series and other atypical deviations in banks’ regular reports (logical checks); develop methods and procedures for regular reconciliation of individual ITRS indicators with banks’ balance sheets (reconciliation of balance-sheet positions and flows, broken down by certain financial instruments).

51. The priority order for introducing the 4 major ITRS segments must also be determined: (i) CBU segment; (ii) commercial bank segment (transactions in cash and non-cash); and (iii) non-banking financial institution segment; and (iv) segment of nonfinancial corporations with current accounts abroad.

Recommended Actions:

  • insert an additional field for the institutional sector in the reporting form;

  • update the system of transaction codes and the instruction on their application per the mission’s recommendations;

  • identify a date for the official introduction of the ITRS and provide legal support;

  • provide for regular seminars for commercial banks, to explain the goals of data collection and train staff in compiling reports under the system;

  • improve data collection for the CBU under the ITRS and study the advisability of extending the ITRS to the remaining segments.

G. Compilation of Balance of Payments, 2016–2017

52. One of the current mission’s objectives was to assist the CBU in analyzing the causes of the gap between the new BOP data and historic time-series, which emerged when the production of ESS data was moved from the MOE to the CBU, and to assist in eliminating this gap. There are both objective and subjective reasons for the appearance of a gap between historical data and the new data. These two BOP datasets are constructed using different formats and data were estimated based on different assumptions that were sometimes not entirely suitable. There were deviations from accepted standards, some indicators were estimated on different source data, and there were also errors, which all together resulted in a rather serious data gap. The lack of a harmonized data series diminishes the ability to make forecasts and to provide proper support for decision-making regarding the development of the external sector of Uzbekistan’s economy.

53. The mission studied the MOE and CBU balances, analyzed the causes of the discrepancies between the data in those balances, and assisted the CBU in generating the BOP for 2016–2017. When generating historical data, all available historical data were used in the new (BPM6) format, including data from the BOP prepared by the MOE. The CBU sent requests to various ministries, departments, and private corporations (MOF, RUSCC, SSC, commercial banks, and others) to provide data for 2016–2017. Prior to receiving these data, the mission and the CBU performed a rough estimate of the missing data. Upon receipt of the requested data, the estimated data will be replaced by revised estimates, thus the BOP for 2016–2017 prepared during the mission should be viewed as a trial version that could change significantly after real data are obtained from the various sources. The mission also assisted the CBU in integrating the analytical BOP format into the working file. This means that when generating new balances or reviewing time series, the analytical presentation of the BOP will be updated automatically, which in turn will facilitate cooperation between the BOPD and the MCD IMF mission. A comparative analysis of historical time-series produced by the MOE and the CBU’s for 2016–2017 is given below.

54. Goods. For goods, the difference between the balances of the MOE and the CBU is minimal; both use the general trade system but arrive at those results differently. The MOE takes SSC data and requests data from the RUSCC on flows to and from customs warehouses in free circulation. They then remove from imports goods leaving customs warehouses, and add goods entering warehouses. The MOE then removes from goods those flows that are associated with processing of goods. Similar approach is applied at the CBU but based entirely on RUSCC data. The CBU requests data from the RUSCC on trading systems and selects those that conform to the general trade system, excluding goods for processing and repair. Since the RUSCC does not perform any data adjustments, the result of these two methods should in principle be identical. The difference in goods is caused by subsequent adjustments. The MOE adds to goods component the unrecorded import/export transactions conducted by individuals. These goods are estimated at 20 percent of money transfers.

55. The CBU estimates shuttle trade based on surveys at border crossing points. These estimates are more modest for exports—about 40 percent of the MOE’s estimation. The CBU additionally asks the National Air Company (NAC) about goods purchased by carriers in ports and airports and adds them to goods. These goods are recorded by the SSC, and accordingly in the BOP of the MOE, among services. There is a difference in the conversion from the CIF price to the FOB price. The MOE assumes the cost of freight and insurance at 10 percent of the CIF price but does not take energy imports into account, which lowers the cost of freight and insurance to 9.5 percent. The CBU estimates freight based on geography and the physical volume (weight) of the goods. Calculations are performed using the average transportation cost of one ton of cargo; imports from neighboring countries with shared border are not taken into account, resulting in higher figures—up to 13–14 percent. In the trial BOP for 2016–2017, data for goods were taken from the MOE balance. They are presented in two lines—nonmonetary gold and other goods.

56. Services. In the CBU estimates there was no debit on passenger transportation services for any mode of transport. These services were estimated based on information about border crossings by individuals for all modes of transport. Data for the years 2016–2017 were taken from the BOP of the MOE; the overall figure for all modes of transport was distributed among the individual modes of transport according to passenger flows. During the last days of the mission, the CBU obtained data on international trade in services from the SSC for 2016–2017. The MOE did not adjust services exports, and a decision was made to derive services exports for the trial BOP from the newly obtained data. In imports of services, a portion of insignificant components that are not adjusted in the BOP of the MOE are calculated based on SSC information.

57. The debit entry on the government services component, which is absent from the SSC data, was taken from the MOE balance. Freight is estimated using the CBU methodology as the more suitable method, compared to 10 percent from imports. The results turned out to be slightly greater than in the MOE’s BOP. As for the travel component, the MOE’s BOP lacked the travel services incurred by short-term workers. This component is also absent from compensation of employees estimates. In the 2016–2018 trial balance, unlike the MOE balance, expenses of short-term workers have been added to business travel. All the remaining minor service components are recorded in a single component—”other business services”—and the data have been harmonized with data from the MOE BOP.

58. Primary and secondary income. Compensation of employees was overestimated in the CBU BOP and was adjusted during the mission. However, in the MOE balance this e component was underestimated. The MOE data reflect only net funds transfers, not all compensation of employees, some of which are spent within the host country (primarily on housing and taxes). In the CBU BOP for 2016–2017, compensation of employees component is estimated according to the BPM6 criteria (it includes total earnings, not only the portion that was transferred to Uzbekistan). Data for the first half of 2018 were adjusted. The new data provide an adequate time-series and are comparable with the MOE’s BOP data, with the provision that the latter do not reflect all earnings but only the portion transferred to Uzbekistan.

59. The MOE lacks reinvested earnings within direct investment income component. They are in the working files, but the data are rather understated. In the final line where all direct investment income is calculated, existing reinvested earnings are multiplied by zero—that is, they are excluded. Reinvested earnings obtained using the new reporting forms are fairly large, amounting to USD 400 million over the first half of the year. In the trial balance, dividends and interest income were taken from the MOE balance, while reinvested earnings were calculated as follows: an estimated position was taken as of the beginning of 2016, the average corporate yield was defined at 5 percent, and the enterprise’s profit was calculated on this basis, deducting paid-out dividends from it. The data obtained were recorded both in DI reinvested and under DI. For interest income from other investments, only data on interest payments were available, and they were recorded in the experimental balance. As soon as relevant data are available (from the MOF and from commercial banks), paid interest will be replaced by accrued interest. Updated DIMR data were reflected in income from reserve assets. Personal transfers were identified at 54 percent of money transfers; this component will be reviewed later.

60. Transfers of migrants are recorded in capital transfers in the BOP of the MOE, which is not allowed according to the BPM6. Naturally, these transfers were removed in the experimental balance; in addition, debt forgiveness appeared in debt of other sectors in the first half of 2018 and is recorded as a capital transfer. It is not inconceivable that, when data are obtained on 2016–2017 debt, instances of debt forgiveness will turn up there too, but for now there is no record in capital transfer components.

61. The mission performed an adjustment of DI. To calculate DI into Uzbekistan, the MOE uses only data on equity capital and debt capital. Even though there are data on reinvested earnings in the MOE working file, they are not used in the calculation of DI (see above). The CBU receives its data on equity capital and debt capital for 2016–2017 from the MOE balance, except for debt capital for the fourth quarter of 2017. Compiling the BOP for the fourth quarter of 2017 at the request of the MCD IMF mission, the MOE experienced a shortage of data. In particular, it lacked information about loans attracted from direct investors and in the fourth quarter of 2017, the CBU repeated data from the fourth quarter of 2016, in the amount of USD+326 million. As of the time the CBU trial BOP was compiled, the real size of DI debt capital was known; it comprised USD -175 million. As for reinvested earnings, the methods used to calculate them are described above. Direct investments abroad are negligible. For the trial BOP for 2016–2017, these data were taken from the MOE working file.

62. As has already been noted, portfolio investments and financial derivatives in Uzbekistan are quite negligible. They are not set apart in the MOE balance, and they did not bother to make any estimates. Other investments are presented in the MOE balance in a rather limited form. Loans are represented only in liabilities and are broken down into government and government-guaranteed on one side and unsecured loans on the other. The mission and the CBU jointly decided to record data from the MOE balance in public sector borrowing, despite the fact that they differ slightly from the MOF data. Loans of DTC and other sectors must be taken from commercial bank reports.

63. Data on currency and deposits in the MOE balance have little credibility; imports/exports of Uzbek Sum are recorded in liabilities for this instrument. First, it is doubtful that nonresidents are holding Uzbekistan Sum in such considerable amount and, if this is so, it is a CBU liability. In assets, rather sizable amounts are in the central bank sector in all quarters of 2016–2017. But, on the other hand, DMIR representatives have asserted that the CBU has no external assets in non-reserve currencies. All deposits in non-reserve currencies are in resident banks; based on the above, deposits of the DTC sector were estimated based on bank statements. For deposits of other sectors where the large share falls to households, expert assessments were done based on imports/exports of foreign currency into Uzbekistan. Additionally, estimates of inflows and outflows of foreign currency by households were accounted for.

64. To compile the 2016–2017 BOP, the mission obtained data on reserve assets and external assets in non-reserve currencies from the DIMR. Major transactions are performed under the currency and deposits subcomponent. Transactions involving monetary gold are not carried out; all changes in monetary gold are assigned to other changes, and these data were recorded in the trial BOP.

65. A table for the trial BOP beginning with 2016 is provided below in Appendix II. Many components are roughly estimated for lack of reliable source data. The BOPD will continue efforts to collect additional data for 2016–2017 in order to update the 2016–2017 trial balance, for which reason the latter could change greatly.

H. Other Issues (Human Resources and Training)

66. At the beginning of 2018, a BOPD was created within the Department of Foreign Exchange Regulation and Control. This Department is responsible for generating the ESS of Uzbekistan. Following the reorganization, the department’s name was changed—it became the Department of Foreign Exchange Regulation and the Balance of Payments (DERBP). At present the BOPD staff consists of 8 employees; in reality only 4 have been hired, including the head of the office. At this time the CBU is selecting candidates and plans to fill the remaining open positions in the very near future. As of October 2018, only 2 staff members of the 4 had completed special training courses. One BOPD staff member was transferred to another CBU unit. Three employees were nominated for a seminar on ESS, to take place in the Joint Vienna Institute in February of 2019. The mission recommends referring personnel responsible for compiling BOP and IIP for training in ESS organized by the IMF and other international organizations or central banks of other countries.

The mission recommends:

  • bringing the BOPD to full strength within the shortest possible time;

  • provide one position within the new office’s staff for a programmer, or assign one programmer to the BOPSO;

  • nominate personnel for relevant IMF seminars to further study the BPM6 and external debt statistics.

I. Officials with Whom the Mission Met

article image
article image