Bank of Korea

The ROSC process provided an opportunity to review the quality, including compliance with international standards, of Bank of Korea statistics and it will help us develop our own statistical reform plans.

We intend to implement the mission’s recommendations as soon as possible. A few of them have already been implemented and many of them will be in the near future; however, some of them ought to be delayed until the regular revision is over to cope with the work burden involved.

• In order to manage our implementation plan effectively, the BOK established an Advisory Committee for Data Quality Assessments as of July 16, 2001. The committee is comprised of 7 staff members who are not currently working in ESD but have statistical compilation experience.

• The Committee checks periodically on the performance of data quality improvements in the light of the mission’s recommendations and conducts periodic quality audits of individual programs.

There are wide differences in countries’ statistical settings such as culture, the legal environment, and the stage of statistical development. We, therefore, think it most desirable if such differences could be taken into account in application of the methodologies provided by DQAF.

Through the ROSC process to date, we feel that no one country is likely to meet all of the best practice criteria recommended by DQAF. In this regard, if only a few countries’ ROSC data modules are published, the credibility of those countries’ statistics could be adversely affected. From this standpoint, we would like publication to be postponed until the task of ROSC data quality assessment is completed. Furthermore, it needs to be borne in mind that the ROSC data modules should not be used to rank the quality of countries’ data when they are published.

Ministry of Finance and Economy

During the IMF Mission’s visit to Korea, we had the pleasure to be given the opportunity to review the quality of our government finance statistics by considering compliance with international standards, and to draw up statistical reform plans on the basis of the results from the review procedure.

The MOFE is committed to implementing the Mission’s recommendations at an earliest possible time. It is also hoped that the implementation will proceed in a smooth way and in a timely manner as well. However, there is also concern that a delay may occur in the process of implementation due to a couple of problems such as policy coordination among relevant government bodies and revision of relevant laws.

Statistics can vary from country to country due to wide differences in every respect of society of their own. Given this, it would be desirable if such difference should be taken into account in applying the methodologies as stipulated by DQAF.

In respect of the question of whether or not the ROSC should be made public, the position of the Korean government is that this time seems to be a bit too premature. However, we think that we will be able to publish it as soon as possible. One of the major reasons can be said that the public announcement, if happens at all, will adversely affect the confidence in GFS published by the Korean government. Furthermore, it would be greatly appreciated if you keep in mind that the ROSC data modules should not be used to rank the data quality of Member countries when they are published.

A. National Accounts

The BOK already put the pre-release access list on the Data Standards Bulletin Board (DSBB) in October 2001. “Data subject to an embargo are provided to the President, and the Minister of Finance and Economy in the late afternoon of the day previous to their release to the public.”

The BOK has drawn up the 1993 SNA implementation plan and has reviewed appropriate implementation approaches. Most of the items in the plan will be reflected in compiling national accounts by 2004 when the 2000 base year revision is finalized. The production and asset boundary will be extended to be in accordance with the 1993 SNA until the 2000 base year revision.

When converting import and export transactions to local currency terms from dollar terms, the BOK already began to use actual transaction exchange rates for goods instead of average exchange rates since the 2nd quarter of 2001. Whereas we will use average monthly exchange rates for other transactions such as services, incomes and transfers for which the daily transaction data are not available.

Regarding the GDP estimation in Korea, source data for production approach are well developed and comparatively easily available while those for expenditure approach are not. As a result, GDP estimation is centered on the production approach while the estimation of expenditure side is much relied on the results of production side estimation. The BOK will keep trying to develop expenditure approach estimation using survey data and ask the Korea National Statistical Office and other related institutions to make those data available for national accounts estimation. However, in the current Korean statistical environment, the expenditure approach might be no more reliable or accurate than the current production approach.

The BOK will update the document for sources and methods with a focus on the final annual data in Korean. The document in English will be published next year. The BOK will also make the summary methodology statement for the DSBB this year. The series of 18 quarters and 5 years were already added in the publication of national accounts for the second quarter of 2001. As a result, in all, series of 28 quarters and 16 years are now shown in the publication of national accounts.

B. Producer Price Index

Korea’s PPI includes domestic transactions at market prices in normal trade. But at present it does not include activities in free trade zones and bonded warehouses, and production for own use. In accord with the original purpose of PPI, the price survey should be based on actual transaction, and careful study should be given to including them in the scope of the Korea’s PPI.

BOK uses the carrying forward method for some seasonal products like agricultural and marine products. The imputation method using the price change of the similar items or the individual sub-period weight system could be a solution to the problems arising from the current BOK method. But the latter could overvalue the usual pattern of the price change of the seasonal products, because the total transaction amount of seasonal products is subject to wide fluctuations by season.

The arithmetic average of price relatives formula is discouraged because it may result in an upward bias. So, BOK plans to compile an index calculated using the geometric mean at the basic level as a reference index during the next PPI revision (2000 year base).

Generally the Laspeyres formula tends to overstate inflation with the lapse of time. In solving this problem, the Chain Laspeyres index, Paasche index and Fisher index for the PPI could be helpful. But it is hard to collect the source data (reference year volume data) for these formulas in time. BOK plans to conduct a pilot compilation of the Chain Laspeyres index for import/export commodities during the next revision (2000 year base) to see if it helps to mitigate the effects of the substitution bias.

C. Government Finance Statistics

The MOFE is committed to implementing the Mission’s recommendations at an earliest possible time. It is also hoped that the implementation will proceed in a smooth way and in a timely manner as well.

• The detailed GFS of the central government, including local governments, will be compiled and disseminated starting 2003. In line with the 1986 Manual on Government Finance Statistics, the Ministry of Government Administration & Home Affairs and the Ministry of Education are in preparation for GFS while, at the same time, in consultation with the Ministry of Finance & Economy and the Ministry of Planning & Budget.

• The problems that may arise in reclassifying extra-budgetary funds will be addressed in the process of reforming the government accounting system, which will include the rearrangement of government accounts such as consolidation, abolishment or reclassification. Also, under the current plan, a draft will be prepared by the Korean government to rearrange government accounts at the end of this year or early next year. The anticipated year of implementation will be around 2003, when the consultation among relevant ministries and other government agencies, and revision of relevant laws will emerge as a prerequisite.

• To improve metadata on government finance statistics, “The Government Finance Statistics in Korea (2001),” which will be published at the end of this year, will cover the concepts, definitions, and classification systems used in compiling GFS in more detail to help users to get a better understanding of how data are put together in comparison with those used in national budget and budget execution data.

D. Monetary Statistics

BOK is well on the way toward revision of monetary and financial statistics system according to the Monetary and Financial Statistics Manual (MFSM) and will release the revised data by the end of 2002. Thus, the monetary and financial statistics of Korea will meet almost all of the criteria of MFSM after that time.

Therefore, after revising our monetary and financial statistics system, inconsistencies with MFSM, concerning liquidity based aggregates, classification/sectorization, a common definition of the money holding and money issuing sector etc., will varnish. In addition, this revision will improve consistency between monetary and financial statistics and other macroeconomic statistics such as Flow of Funds, BOP and National Accounts.

Also there was some revision in May 2001, adopting the recommendations of IMF mission staff. These are set out below.

• The offshore accounts and inter-office accounts of both deposit money banks and other financial institutions are recorded on a gross basis.

• SDR allocations have been reclassified as part of share and other equity (capital).

• Positions in financial derivatives have been separately identified in the surveys and the contracts with non-residents have been identified as part of foreign assets and liabilities.

• Compensation (receipt) for oversea branches’ losses (incomes) is included in foreign assets (liabilities) in order to prevent their being understated.

E. Balance of Payments Statistics

We plan to endeavor to implement the mission’s recommendations progressively starting from those most applicable. Though some recommendations are to be implemented in the near future, others may take time for implementation.

• We already enhanced the metadata in August 2001 on the DSBB to provide sufficient information for users specifically about internal governmental access to statistics on the day prior to the official release date, deviations from the Balance of Payments Manual (fifth edition) (BPM5) guidelines, and change in the foreign direct investment threshold in April 1999 (from 20% to 10%). And we reclassified securities issues by the Foreign Exchange Equalization Fund from General Government to the Monetary Authority Sector and sent revised quarterly BOP data for 1998 to the IMF in August 2001.

• Regarding the residency definition in the Foreign Exchange Transactions Act and accounting treatment for losses of foreign branches of domestic banks, it is necessary to amend the Act and to revise the Accounting Procedure. However, it will take time to complete consultation with the authorities concerned, as well as the procedures for amendment.

• In order to implement some recommendations, e.g., reinvested earnings, trade credits or short-term loans with affiliated entities, and non-cash transactions such as share exchanges and stock dividends, it is necessary to carry out direct surveys of the transactors based on a new survey method.

• For the other recommendations, such as disaggregation of other transportation services by mode of transport and recording valuation adjustments under the specific instruments to which the valuation adjustments apply, it is necessary to improve source data such as the foreign exchange transactions data (KFX) and trade statistics. As improvement of the source data involves an increase in the reporting burden of raw data providers and source data compilers, we need to come up with an appropriate method.