Chapter

3. Data

Author(s):
International Monetary Fund
Published Date:
October 2003
Share
  • ShareShare
Show Summary Details

Data Availability

3.1 The BPM5 and the Benchmark recommend that FDI statistics be compiled as part of balance of payments statistics (the transactions data) and international investment position (IIP) statements (the position data). Countries are expected to compile and disseminate FDI statistics according to the standard components of BPM5: direct investment income, direct investment financial flows, and direct investment positions. The direct investment income component is broken down into (1) income on equity (dividends and distributed branch profits), (2) reinvested earnings and undistributed branch profits, and (3) income on debt (interest). Direct investment financial flows are broken down into (1) equity capital, (2) reinvested earnings (the counterpart of the income item for reinvested earnings), and (3) other capital (intercompany debt transactions). The data on direct investment positions are broken down into (1) equity capital and reinvested earnings, and (2) other capital.

Data Reported to the International Organizations

3.2 The 2001 SIMSDI update indicates that all 61 of the countries that participated in the update report at least one element of the FDI statistics to the international organizations. Table 3.1 shows the numbers of countries, broken down into OECD and other, that reported data for inward and outward direct investment income, direct investment financial flows, and direct investment positions, and also indicates the changes since the 1997 SIMSDI survey. (Tables 1 and 2 of Appendix I give the details by country for 2001, as well as further details for the elements of each component.)

Table 3.1.Data Reported to International Organizations
Countries Reporting FDI Statistics for:
Direct investment incomeDirect investment financial flowsDirect investment positions
Number of CountriesInwardOutwardInwardOutwardInwardOutward
Total 2001 (61)564859534745
Total 1997 (61)484250433431
Change+8+6+9+10+13+14
OECD 2001 (30)282729282726
OECD 1997 (29)282728272524
Other 2001 (31)282130252019
Other 1997 (32)2015221697

OECD countries

3.3 All 30 OECD countries report at least one of the standard components of inward FDI statistics—FDI income, financial flows, and positions—to the IMF or the OECD, or both. All 29 OECD countries that compile FDI transactions data1 now report inward data on at least one of the two elements of FDI financial flows, and all but Turkey report data for both equity capital and other capital. (Turkey reports data on equity capital only.) All but one of the 29 also report data on at least one of the three elements of inward FDI income. The exception is Denmark, although this country does compile data on all elements of FDI income and has reported in the past. Twenty-seven of the 30 OECD countries report inward FDI position data, the exceptions being Ireland, Korea, and Turkey. Ireland, however, compiles inward position data and indicated that it planned to disseminate data for the period from 1998 through 2001 by the end of October 2002.

3.4 Slightly fewer OECD countries report data for the outward FDI statistics. All but one of the 29 OECD countries that compile transactions data report at least one of the two elements of outward FDI financial flows. (Mexico does not compile outward FDI transactions data at present.) All of those 28 countries, except Turkey, report data for both equity capital and other capital. (As with the inward data, Turkey reports outward FDI transactions data on equity capital only.) All except two countries (Mexico and Denmark) also report data on at least one of the three elements of outward FDI income, and 26 report outward FDI position data—Ireland, Korea, Mexico, and Turkey being the exceptions. Ireland, however, indicated that it planned to disseminate outward position data for the period from 1998 through 2001 by the end of October 2002.

Other countries

3.5 There has been a marked improvement since the 1997 SIMSDI survey in the number of other IMF member countries that report inward FDI statistics to the IMF. With the exception of Malaysia, all of the 31 countries that participated in the 2001 update now report at least one of the two elements of inward FDI financial flows, an increase of 8 countries since 1997. Malaysia, however, does compile data on inward financial flows, and indicated that it planned to disseminate these data with effect from the 2001 reference period. Twenty-eight of the 31 countries now report at least one of the three elements of inward FDI income, an increase of 8 since 1997. The exceptions are Kuwait, Singapore, and Thailand, all of which compile, but do not disseminate, data on inward FDI income at present. Twenty of the 31 countries now report at least one component of inward FDI position data, an increase of 11 since 1997. The exceptions include Chile, Costa Rica, Croatia, Ecuador, Guatemala, Indonesia, Kuwait, Malaysia, Nigeria, and the Philippines. However, Croatia and Malaysia compile inward FDI position data and have indicated that they plan to disseminate the data in the future.

3.6 There have been similar improvements since 1997 in the number of countries that report outward FDI statistics to the IMF. An additional 9 countries now report at least one of the two elements of outward FDI financial flows, with the result that 25 of the 31 countries now report these data. The exceptions are Bolivia, Ecuador, Guatemala, Indonesia, Malaysia, and Nigeria. However, Malaysia does compile outward data on FDI financial flows and has indicated that it plans to disseminate these data in the future. Twenty-one of the 31 countries now report data on at least one of the three elements of outward FDI income, an increase of 6 countries since 1997. The exceptions include Ecuador, Indonesia, Kazakhstan, Kuwait, Nigeria, Peru, Singapore, and Thailand. However, four countries (including Kuwait and Singapore) compile outward data on FDI income but do not disseminate them, and Singapore indicated that it planned to disseminate the data in the future. Nineteen countries now report at least one component of outward FDI position data, an additional 12 countries since 1997. The 12 countries that do not yet report outward FDI position data include Chile, Costa Rica, Croatia, Ecuador, Guatemala, Indonesia, Kuwait, Malaysia, Nigeria, and the Philippines. However, Croatia and Malaysia compile outward FDI position data and indicated that they planned to disseminate these data in the future.

Periodicity of the Disseminated FDI Transactions Data

3.7 Data “dissemination” refers to all the means by which data are made available to the public (including dissemination on the Internet). “Periodicity” for transactions (flow) data is specified in terms of the interval represented by a single data point, while “periodicity” for position (stock) data is specified in terms of the interval between two data points. The 2001 SIMSDI update indicates that more countries are compiling quarterly transactions data than in 1997, and that somewhat fewer are compiling monthly data and annual data. Table 3.2 indicates the periodicity of the reported data on inward equity capital transactions and indicates the changes since the 1997 SIMSDI survey. (Table 3 of Appendix I gives the details by country for 2001, as well as similar details for the inward FDI income data and the outward FDI transactions data.)

Table 3.2.Periodicity of the Most Timely and Most Comprehensive Equity Capital Transactions Data for Direct Investment in the Reporting Economy (Inward FDI)
Periodicity of the Most Timely Equity Capital Transactions DataPeriodicity of the Most Comprehensive Equity Capital Transactions Data
Number of CountriesMonthlyQuarterlyAnnualN.A.*MonthlyQuarterlyAnnualN.A.*
Total 2001 (61)22325**2151540
Total 1997 (61)24189107111330
Change–2+14–4–8–6–6+2+10
OECD 2001 (30)18110113818
OECD 1997 (29)1874056711
Other 2001 (31)4215**102722
Other 1997 (32)61151025619

Not applicable.

Includes one country that disseminates semiannual data.

Not applicable.

Includes one country that disseminates semiannual data.

OECD countries

3.8 The 2001 results indicate that, as in 1997, 18 (62 percent) of the 29 OECD countries that disseminate FDI transactions data, primarily European, compile monthly data as their “most timely”2 data on inward FDI equity capital transactions. The reporting of monthly balance of payments statistics is among the data requirements for the European Union (EU),3 which explains why a large number of European countries disseminate monthly FDI statistics. The results also show that no OECD country now compiles annual data as its most timely data on inward FDI equity capital flows, while four additional countries compile quarterly data. The 11 OECD countries that now compile quarterly data on inward equity capital transactions as their most timely data are Australia, Canada, the Czech Republic, Greece, Iceland, Ireland, Mexico, New Zealand, Switzerland, the United Kingdom, and the United States.

3.9 Six fewer OECD countries compile “most comprehensive”4 data on their inward equity capital transactions than in 1997. Of the 12 countries that still compile most comprehensive data, one (Italy) compiles monthly data, three (Australia, the Slovak Republic, and Spain) compile quarterly data, and eight (Austria, the Czech Republic, Finland, Iceland, Mexico, Poland, Switzerland, and the United Kingdom) compile annual data.

Other countries

3.10 The results of the 2001 update summarized in Table 3.2 show that 21 (70 percent) of the 30 countries that disseminate data on inward FDI equity capital transactions compile quarterly data as their most timely data, compared with 50 percent in 1997—an increase of 10 countries. This shift to quarterly periodicity may, in part, have been driven by the requirement for quarterly balance of payments statistics by the SDDS, to which many of these countries subscribe. Of the remaining nine countries, four (Brazil, the Philippines, Slovenia, and Thailand) disseminate monthly data, one (China) disseminates semiannual data, and four (Botswana, Kuwait, Nigeria, and Tunisia) disseminate annual data.

3.11 Four fewer countries now compile most comprehensive data on equity capital transactions than did in 1997. Of the nine countries that compile these data, two countries (Argentina and Colombia) disseminate quarterly data, and seven (Botswana, Chile, Ecuador, India, Kuwait, Peru, and Singapore) disseminate annual data.

Periodicity of the Disseminated FDI Position Data

3.12 The 2001 SIMSDI update shows that a higher proportion of the countries that disseminate inward data on equity capital and reinvested earnings positions compile quarterly data as their most timely data than in 1997 (38 percent in 2001, compared with 27 percent in 1997) and that most of the remainder disseminate annual data for their most timely data. Table 3.3 indicates the periodicity of the reported data on inward FDI equity capital and reinvested earnings positions and indicates the changes since the 1997 SIMSDI survey. (Table 4 of Appendix I gives the details by country for 2001, as well as similar data for the inward FDI other capital position data and the components of the outward FDI position data.)

OECD countries

3.13 For the most timely data on inward equity capital and reinvested earnings positions, the 2001 results show that only one OECD country, Hungary, disseminates monthly data. (Hungary disseminated annual data in 1997.) In 1997, another OECD country (Finland) disseminated monthly data as its most timely data but has since moved to quarterly data. Eight OECD countries (Australia, the Czech Republic, Finland, Iceland, New Zealand, Portugal, Spain, and the United Kingdom) now disseminate quarterly data as their most timely data, compared with four countries in 1997. The remaining 19 countries that disseminate data on inward equity capital and reinvested earnings positions have annual data as their most timely data, the same number of countries as in 1997.

3.14Table 3.3 indicates that seven fewer OECD countries now compile most comprehensive inward data on equity capital and reinvested earnings positions than in 1997. Of the seven countries that still compile these data, one country (Australia) compiles quarterly data, and the remaining six (the Czech Republic, Finland, France, Iceland, Sweden, and the United Kingdom) compile annual data.

Table 3.3.Periodicity of the Most Timely and Most Comprehensive Equity Capital and Reinvested Earnings Position Data for Direct Investment in the Reporting Economy (Inward FDI)
Periodicity of the Most Timely Equity Capital and Reinvested Earnings Position DataPeriodicity of the Most Comprehensive Equity Capital and Reinvested Earnings Position Data
Number of CountriesMonthlyQuarterlyAnnualN.A.*MonthlyQuarterlyAnnualN.A.*
Total 2001 (61)1**17**28**16**02851
Total 1997 (61)192328031345
Change0**+8**+5**–12**0–1–5+6
OECD 2001 (30)1**8**19**3**01623
OECD 1997 (29)14195021215
Other 2001 (31)099***1301228
Other 1997 (32)0542301130

Not applicable.

Will not add to the totals as one OECD country has more than one periodicity for its most timely position data.

Includes one country that disseminates semiannual data.

Not applicable.

Will not add to the totals as one OECD country has more than one periodicity for its most timely position data.

Includes one country that disseminates semiannual data.

Other countries

3.15 Of the 18 countries that disseminate most timely inward FDI statistics on equity capital and reinvested earnings positions, 9 compile quarterly data, 1 (Colombia) compiles semiannual data, and 8 compile annual data. None of the countries compiles monthly data as its most timely inward FDI statistics on equity capital and reinvested earnings positions.

3.16 Only three countries disseminate most comprehensive inward FDI statistics on equity capital and reinvested earnings positions, one more than in 1997. Of these, one country (Peru) disseminates quarterly data and two (Argentina and Botswana) disseminate annual data.

Timeliness of the Disseminated FDI Transactions Data

3.17 “Timeliness” refers to the number of weeks, months, or years between the end of the reference period (or the reference date for position data) and the dissemination of the data. The results of the 2001 SIMSDI update indicate that, as in 1997, a significantly higher proportion of the OECD countries have shorter time lags in disseminating their data than the non-OECD countries. This reflects, in part, the reporting requirements of the EU, which require balance of payments statistics within six weeks after the end of the reporting period. However, none of the 61 countries that participated in the 2001 SIMSDI update now has a timeliness of more than six months for its most timely data on FDI equity capital transactions. (See Table 3.4; Table 5 of Appendix I gives details by individual country for 2001, including the timeliness of the components of the inward and outward FDI transactions data.) Information on the revision practices of the countries is available in the individual country metadata on the IMF website and also in the completed questionnaires of the OECD member countries on the OECD website. (See paragraph 1.8 in Chapter 1 for Internet addresses.)

OECD countries

3.18 Twenty-one (72 percent) of the 29 OECD countries that disseminate FDI transactions data have a timeliness for their most timely inward equity capital transactions data of less than three months. Of the remaining eight OECD countries, seven have a timeliness of three months, and one, Belgium, has a timeliness of between 12 and 20 weeks. Since the 1997 SIMSDI survey, Ireland and New Zealand have improved the timeliness of their most timely FDI transactions data.

3.19 As might be expected, the most comprehensive data tend to be less timely, and no OECD country has a timeliness of less than three months. Of the 12 OECD countries that disseminate most comprehensive data for equity capital transactions, 5 have a timeliness of 3 to 6 months, 6 have a timeliness of 6 to 12 months, and 1 (the Czech Republic) has a timeliness of 14 months.

Table 3.4.Timeliness of the Most Timely and Most Comprehensive Equity Capital Transactions Data for Direct Investment in the Reporting Economy (Inward FDI)
Number of CountriesTimeliness of the Most Timely Equity Capital Transactions DataTimeliness of the Most Comprehensive Equity Capital Transactions Data
–3M3–6M6–12M12+M–3M3–6M6–12M12+M
Total 2001 (61)3326001695
OECD 2001 (30)218000561
Other 2001 (31)1218001134
–3M = Timeliness of less than 3 months, i.e., less than 12 weeks.3–6M = Timeliness of 3 to 6 months.6–12M = Timeliness of more than 6 months to 12 months.12+M = Timeliness of more than 12 months.
–3M = Timeliness of less than 3 months, i.e., less than 12 weeks.3–6M = Timeliness of 3 to 6 months.6–12M = Timeliness of more than 6 months to 12 months.12+M = Timeliness of more than 12 months.

Other countries

3.20 Although lower than the proportion of OECD countries, a sizable percentage of the 30 countries that disseminate inward data on FDI equity capital transactions now have a timeliness of less than three months for their most timely data—12 countries or 40 percent. Of the remaining 18 countries, 15 have a timeliness of three months, 1 (Kuwait) has a timeliness of four months, and 2 (Indonesia and Tunisia) have a timeliness of six months. Since 1997, there have been significant improvements in the timeliness of the most timely data for a number of countries, including Botswana, Costa Rica, Croatia, Hong Kong SAR, Indonesia, Israel, Peru, Singapore, South Africa, and Thailand. These improvements, in part, may have been driven by the timeliness requirements for the balance of payments statistics of the SDDS, to which almost all of these countries subscribe.

Table 3.5.Timeliness of the Most Timely and Most Comprehensive Equity Capital and Reinvested Earnings Position Data for Direct Investment in the Reporting Economy (Inward FDI)
Number of CountriesTimeliness of the Most Timely Equity Capital and Reinvested Earnings Position DataTimeliness of the Most Comprehensive Equity Capital and Reinvested Earnings Position Data
–3M3–6M6–12M12+M–3M3–6M6–12M12+M
Total 2001 (61)8231040154
OECD 2001 (30)610830142
Other 2001 (31)213210012
–3M = Timeliness of less than 3 months, i.e., less than 12 weeks.3–6M = Timeliness of 3 to 6 months.6–12M = Timeliness of more than 6 months to 12 months.12+M = Timeliness of more than 12 months.
–3M = Timeliness of less than 3 months, i.e., less than 12 weeks.3–6M = Timeliness of 3 to 6 months.6–12M = Timeliness of more than 6 months to 12 months.12+M = Timeliness of more than 12 months.

3.21 Of the nine countries that disseminate most comprehensive inward FDI statistics on equity capital transactions, only Ecuador has a timeliness of less than three months. Of the remaining eight, one country has a timeliness of 6 months, three countries have a timeliness of 6 to 12 months, and four countries have a timeliness of more than 12 months.

Timeliness of the Disseminated FDI Position Data

3.22Table 3.5 indicates that the most timely inward FDI statistics on equity capital and reinvested earnings positions tend to be less timely than the most timely inward FDI statistics on equity capital transactions. Table 6 of Appendix I gives details of the countries’ practices, as well as information regarding the timeliness of the other capital component of the inward FDI position data and the components of the outward position data for each country.

OECD countries

3.23 In contrast to the transactions data, less than one quarter (6) of the 27 OECD countries that disseminate inward FDI statistics on equity capital and reinvested earnings positions have a timeliness of less than three months for their most timely data, and only 10 have a timeliness of three to six months. The remaining 11 countries (41 percent of the total) have a timeliness of either 6 to 12 months (8 countries) or more than 12 months (3 countries) for their most timely position data. Of the seven OECD countries that disseminate most comprehensive position data, the timeliness ranges from 22 weeks for Australia to 18 months for France, with the rest having a timeliness of between 9 and 16 months.

Other countries

3.24 A higher proportion than the OECD countries have a timeliness of six months or less for their most timely inward FDI statistics on equity capital and reinvested earnings positions (83 percent of the 18 countries that disseminate inward FDI statistics on equity capital and reinvested earnings positions, compared with 59 percent of the OECD countries). Table 6 of Appendix I provides the details for each country. For the three countries that disseminate most comprehensive position data, the timeliness ranges from 9 months for Botswana, to 18 months for Peru, and 18–24 months for Argentina.

Data Sources

3.25 The sources used by compilers have an impact on their ability to implement the international recommendations for the compilation of FDI statistics. There are three major data sources: enterprise surveys, international transactions reporting systems (ITRS),5 and “administrative” data from exchange control or investment control authorities. The advantages and disadvantages of these three main data sources are reviewed in the IMF’s Balance of Payments Compilation Guide.6

Table 3.6.Primary Data Sources for the Most Timely FDI Transactions Data
Number of CountriesEnterprise SurveysInternational Transactions Reporting System (ITRS)Exchange Control or Investment Approval AuthoritiesOther (Published Sources, Press Reports, Bilateral Sources, etc.)
InwardOutwardInwardOutwardInwardOutwardInwardOutward
Total 2001 (61)36342525981510
Total 1997 (61)2823242411833
Change+8+11+1+1–20+12+7
OECD 2001 (30)161616163244
OECD 1997 (29)121118182000
Other 2001 (31)20189966116
Other 1997 (32)1612669833

Sources for the FDI Transactions Data

Most timely data

3.26Table 3.6 shows the primary data sources for the most timely FDI transactions data and indicates that since 1997 there has been a move toward the use of enterprise surveys as a primary source for compiling these data. The table also indicates that there has been a move toward the use of published sources, such as company accounts, which are often used to supplement data obtained from a cash-based ITRS data source that cannot, for example, provide the information needed for the compilation of data on reinvested earnings or data on the noncash acquisition of equity. Table 7 of Appendix I gives the details by country for 2001.

OECD countries

3.27 The 2001 update indicates that enterprise surveys and an ITRS continue to be the most commonly used data sources for the most timely FDI transactions data and that the same number of the OECD countries (16) now use enterprise surveys as a primary data source as use an ITRS. However, 4 of those 16 countries (Austria, Denmark, Italy, and the Netherlands) use enterprise surveys as a primary source for their reinvested earnings data only, and Norway uses enterprise surveys only for the data on FDI income. Slightly fewer countries now use an ITRS as their primary data source than in 1997, while an additional four countries use an enterprise survey for their inward FDI transactions data and five for their outward FDI transactions data. With the exception of Japan, all countries that use an ITRS are European. The continued use of the ITRS by these countries can be explained, in part, by the requirement of the EU for monthly data with a timeliness of six weeks—periodicity and timeliness that are difficult to achieve with enterprise surveys. The table also shows that four OECD countries now use other data sources, such as company accounts or press reports—Iceland uses these as its sole primary data source, while Australia, the Slovak Republic, and Sweden use them as an additional primary data source. Of the remaining three OECD countries, the Czech Republic uses reports on individual transactions submitted by enterprises in accordance with the Foreign Exchange Act, Korea uses foreign exchange receipts and payments data, and Mexico uses the information available at the National Register of Foreign Direct Investment as the primary sources for their inward and outward data as applicable. (Mexico does not compile outward FDI statistics.)

Other countries

3.28 Enterprise surveys predominate as the primary data source used by these countries for their most timely FDI transactions data and are used by approximately two-thirds of the 31 countries that participated in the 2001 SIMSDI update, often as their sole primary data source. However, one country (Slovenia) uses an enterprise survey as a primary data source for its reinvested earnings data only. Other data sources, such as published company accounts, are used as a primary data source for their inward data by 11 countries, an increase of 8 countries since 1997. All these 11 countries use the other data sources in conjunction with another primary data source, either an ITRS, or exchange control or investment approval records. An ITRS is used by nine countries as their primary data source, although only three countries (Israel, Kuwait, and Thailand) use an ITRS as their sole primary data source for both their inward and outward data. Somewhat fewer countries now use exchange control or investment approval records as a primary data source than in 1997, and only one country (Colombia) uses such records as its sole primary data source.

Table 3.7.Primary Data Sources for the Most Comprehensive FDI Transactions Data
Number of CountriesEnterprise SurveysInternational Transactions Reporting System (ITRS)Exchange Control or Investment Approval AuthoritiesOther (Published Sources, Press Reports, Bilateral Sources, etc.)
InwardOutwardInwardOutwardInwardOutwardInwardOutward
Total 2001 (61)1916322241
Total 1997 (61)2821121011713
Change–9–5–9–8–9–5+3–2
OECD 2001 (30)1211210010
OECD 1997 (29)1311882000
Other 2001 (31)75112231
Other 1997 (32)1510429713

Most comprehensive data

3.29Table 3.7 shows the primary data sources used for the most comprehensive FDI transactions data for those countries that still compile such data and indicates a significant move away from the use of an ITRS as the primary data source toward the use of enterprise surveys. The 2001 update indicates that only 3 of the 23 countries that compile at least one component of most comprehensive inward FDI transactions data and 2 of the 18 countries that compile at least one component of most comprehensive outward FDI transactions data now use an ITRS as a primary data source. Table 8 of Appendix I gives the details by country for the 2001 update.

OECD countries

3.30 The 2001 results indicate that all but 2 of the 13 OECD countries that still compile either inward or outward most comprehensive FDI transactions data for at least one component use an enterprise survey as their sole primary data source. Italy and Spain use an ITRS: in conjunction with an enterprise survey for reinvested earnings in the case of Italy, and in conjunction with information from the Ministry of Finance in the case of Spain. (Spain does not compile outward most comprehensive FDI transactions data.) Half of the OECD countries that compile most comprehensive FDI statistics using enterprise surveys as their primary data source are countries that use an ITRS, records of investment approval authorities, or company accounts and press reports as primary data sources for their most timely FDI transactions data. No OECD countries now use exchange control or investment approval records as a primary data source for their most comprehensive transactions data.

Other countries

3.31 Seven of the 10 countries that still compile most comprehensive inward FDI transactions data for at least one component use enterprise surveys as a primary data source—four as the sole primary data source. Only one country (Kuwait) now uses an ITRS as a primary data source for its most comprehensive FDI statistics. Only one country (Colombia) uses exchange control and investment approval records as its sole primary data source, while Chile uses such records in conjunction with an enterprise survey as its primary data sources. The seven countries that still compile most comprehensive outward FDI transactions data use the same primary data sources as used for their most comprehensive inward FDI transactions data.

Sources for the FDI Position Data

Most timely data

3.32 The results of the 2001 SIMSDI update indicate that enterprise surveys continue to be used as a primary data source for approximately three quarters of the 47 countries that compile either inward or outward most timely FDI position data. However, a significant number of the countries have also begun to use an ITRS as a primary data source, with the result that an ITRS is used as a primary data source by approximately one quarter of the countries compiling most timely FDI position data. Table 3.8 summarizes the results of the 2001 update compared with the 1997 SIMSDI survey. Table 9 of Appendix I gives details of countries’ practices in 2001.

OECD countries

3.33 In 1997, only 12 percent of the 25 OECD countries that disseminated most timely inward FDI position data used an ITRS as a primary data source. The 2001 results indicate that 33 percent of the 27 OECD countries that now disseminate these data use an ITRS as a primary data source. There has been a slight decrease in the percentage that uses enterprise surveys as a primary data source—down from 76 percent in 1997 to 74 percent in 2001. One OECD country, the Czech Republic, now uses reports on individual transactions submitted by enterprises under the terms of the Foreign Exchange Act as its sole primary data source. A small number of OECD countries also use other data sources, such as company accounts, as primary data sources, usually in conjunction with an ITRS.

3.34Table 9 of Appendix I indicates that 15 OECD countries7 use the perpetual inventory method, a method of deriving position data from transactions data, for compiling all or part of their most timely inward and outward FDI position data. As might be expected, most of the countries that use the perpetual inventory method are those that use an ITRS or exchange control transactions data as their primary data source. Seven of the 15 countries (Belgium, the Czech Republic, France, Hungary, Italy, the Netherlands, and Portugal) use the perpetual inventory method for all their most timely inward FDI position data, while the remaining 8 countries use it only for parts of the data or for their preliminary data. Although none of the seven countries that use the perpetual inventory method for all their inward data makes all three recommended adjustments for non-transaction changes (foreign exchange changes, price changes, and other changes), only one (Portugal) makes no adjustments for nontransaction changes.

Table 3.8.Primary Data Sources for the Most Timely FDI Position Data
Number of CountriesEnterprise SurveysInternational Transactions Reporting System (ITRS)Exchange Control or Investment Approval AuthoritiesOther (Published Sources, Press Reports, Bilateral Sources, etc.)
InwardOutwardInwardOutwardInwardOutwardInwardOutward
Total 2001 (61)343312115376
Total 1997 (61)2725331030
Change+7+8+9+8+4+3+4+6
OECD 2001 (30)2019991142
OECD 1997 (29)1918330020
Other 2001 (31)1414324234
Other 1997 (32)87001010

Other countries

3.35 The 2001 results indicate that 70 percent of the 20 countries that compile most timely inward FDI position data use an enterprise survey as a primary data source. Three countries, including Thailand and Tunisia, use an ITRS as a primary data source. Four countries (including Colombia, Peru, and Tunisia) use exchange control or investment authority records as a primary data source, and four (Argentina, Israel, Latvia, and Russia) use other sources as a primary data source for either their inward or outward data.

3.36Table 9 of Appendix I indicates that nine countries8 use the perpetual inventory method to compile all or part of their most timely inward position data, and seven use the method for their outward position data, including all the countries that use an ITRS or exchange control transactions data as a primary data source. Of the nine countries that use the perpetual inventory method for their inward FDI position data, seven (including Argentina, Bolivia, Colombia, Peru, Thailand, and Tunisia) use it for all their inward position data while the remaining two (Israel and Latvia) use it for parts of their data only. Of the seven that use the method for all their inward position data, only one country (Thailand) makes all three recommended adjustments for nontransaction changes. Three countries (including Colombia and Tunisia) make adjustments for foreign exchange and price changes, one (Bolivia) makes adjustments only for disinvestments in the previous year, and two countries (Argentina and Peru) make no adjustments for nontransaction changes.

Most comprehensive data

3.37Table 3.9 shows the primary data sources used for the most comprehensive FDI position data and compares the 2001 results with those of the 1997 SIMSDI survey. Table 10 of Appendix I gives the details by country for 2001. Only 11 of the 47 countries that compile either inward or outward position data compile most comprehensive position data.

OECD countries

3.38 All seven OECD countries that compile most comprehensive data use enterprise surveys as a primary data source, six as their sole primary data source. (France uses enterprise surveys as the sole primary data source for the data on other capital, but uses company accounts as the sole primary data source for the data on equity capital and reinvested earnings.) No OECD country uses the perpetual inventory method to compile its most comprehensive inward or outward position data.

Table 3.9.Primary Data Sources for the Most Comprehensive FDI Position Data
Number of CountriesEnterprise SurveysInternational Transactions Reporting System (ITRS)Exchange Control or Investment Approval AuthoritiesOther (Published Sources, Press Reports, Bilateral Sources, etc.)
InwardOutwardInwardOutwardInwardOutwardInwardOutward
Total 2001 (61)88001023
Total 1997 (61)1715111001
Change–9–7–1–100+2+2
OECD 2001 (30)77000011
OECD 1997 (29)1212110000
Other 2001 (31)11001012
Other 1997 (32)53001001

Other countries

3.39 Of the four countries that still compile either inward or outward most comprehensive FDI position data, enterprise surveys are used by one (Botswana). Two countries (Argentina and Israel) use other data sources, and one country (Peru) uses exchange control transactions. (Israel does not compile most comprehensive inward position data.) Only one country (Peru) uses the perpetual inventory method for compiling its most comprehensive inward position data, with no adjustments for nontransaction changes. (Peru does not compile most comprehensive outward position data.)

Table 3.10.Availability of Geographic Breakdowns of FDI Income, Financial Flows, and Position Data
Number of CountriesCountries That Compile Geographic Breakdowns of:
Direct investment incomeDirect investment financial flowsDirect investment positions
InwardOutwardInwardOutwardInwardOutward
Total 2001 (61)373253434137
Total 1997 (61)302742303025
Change+7+5+11+13+11+12
OECD 2001 (30)252428262523
OECD 1997 (29)232225222219
Other 2001 (31)12825171614
Other 1997 (32)7517886

Geographic and Industrial Classifications

3.40 The SIMSDI results provide information on the geographic and industrial sector breakdowns that are available, as well as information on the principles used to allocate the data by country or by industrial sector.

Availability of Geographic Breakdowns of FDI Statistics

3.41Table 3.10 indicates that there have been significant increases since 1997 in the number of countries that compile geographic breakdowns, particularly for the FDI financial flows data and for the FDI position data—increases that affect both the inward data and the outward data. These increases primarily reflect the larger numbers of countries disseminating FDI statistics. The percentages of countries that compile geographic breakdowns of their disseminated data have not changed so markedly. Eighty-eight percent of the countries that disseminate FDI transactions data now compile geographic breakdowns for the inward financial flows data and 72 percent for the outward data—only slightly higher than the percentages in 1997. Eighty-eight percent of the countries that now disseminate inward FDI position data compile geographic breakdowns of those data, the same percentage as in 1997. Eighty-two percent of the countries that now disseminate outward FDI position data compile geographic breakdowns for those data, a slightly higher percentage than in 1997. The percentage of countries that compile geographic breakdowns for the FDI income data is lower than for the financial flows and position data (66 percent for both the inward and the outward data), but is slightly higher than the 1997 percentages. However, these figures disguise sometimes marked differences between the percentages for the OECD countries and those for the other IMF member countries. Table 11 of Appendix I gives the details by country for 2001.

OECD countries

3.42 Not surprisingly, given the reporting requirements of the OECD for geographic breakdowns, a significant percentage of the OECD countries compile geographic breakdowns for all three types of FDI statistics and for both inward and outward FDI statistics. The 2001 results indicate that 86 percent of the relevant OECD countries compile geographic breakdowns for their inward and outward FDI income data; 97 percent for their inward FDI financial flows data; 93 percent for their outward FDI financial flows data; 89 percent for their inward FDI position data; and 85 percent for their outward FDI position data.

Other countries

3.43 Smaller percentages of the other IMF member countries compile geographic breakdowns of their FDI statistics. Only 40 percent of the relevant countries compile geographic breakdowns of their inward FDI income data and 30 percent for their outward FDI income data. Significantly higher proportions compile geographic breakdowns for their financial flows and position data—81 percent compile geographic breakdowns for their inward FDI financial flows data and 61 percent for the outward data, while 70 percent compile geographic breakdowns for their inward FDI position data, and 67 percent compile these breakdowns for their outward FDI position data. In part, the lower percentages for the outward data reflect confidentiality constraints faced by some countries that have relatively small amounts of outward FDI or a limited number of direct investors abroad.

Principles Used for the Geographic Allocation of FDI Statistics

3.44 The principle used for the geographic allocation of FDI transactions will have, all other things being equal, an important impact on the bilateral comparison of the data. FDI transactions may be allocated to the country to which the funds were paid or from which the funds were received, even if this is not the country of the direct investment enterprise or direct investor—this allocation principle is referred to as the transactor principle. On the other hand, the geographic allocation may be based on the country of the direct investment enterprise or direct investor, even if the amounts are paid to or received from another country—this is known as the debtor/creditor principle. Although there are no definite recommendations regarding the regional allocation of FDI transactions, the BPM5 recommends that FDI position data be allocated using the debtor/creditor principle. Consequently, countries that allocate FDI transactions on the basis of the transactor principle require reconciliation items to bridge the differences between the geographic breakdowns for the transactions data and those for the position data. Equally important, the allocation of FDI position data may vary considerably depending on whether the geographic allocation was based on the immediate host or investing country or on the ultimate host or investing country. The BPM5 and the Benchmark recommend that position data be allocated according to the immediate host or investing country although the Benchmark also recommends that separate accounts be maintained based on the country of ultimate ownership.

Table 3.11.Geographic Breakdowns: Principles Used for Allocating FDI Transactions Data by Country
Number of CountriesDebtor/Creditor PrincipleTransactor Principle
InwardOutwardInwardOutward
Total 2001 (61)4338105
Total 1997 (61)31221311
Change+12+16–3–6
OECD 2001 (30)242343
OECD 1997 (29)181699
Other 2001 (31)191562
Other 1997 (32)13642

Debtor/creditor principle versus transactor principle

3.45Table 3.11 shows the numbers of countries that use the debtor/creditor principle as opposed to the transactor principle for their geographic allocation of transactions data and indicates the changes since the 1997 SIMSDI survey. Table 12 of Appendix I gives the details by country for 2001. Significantly higher numbers of countries now use the debtor/creditor principle for both the inward and outward FDI statistics (43 of the 53 countries that compile geographic breakdowns for their inward FDI transactions data). Only five countries (Australia, Belgium, Botswana, Greece, and the Philippines) now use the transactor principle for both their inward and outward transactions data, and five countries (including Italy and Nigeria) use the transactor principle for their inward transactions data only. (Italy uses the debtor/creditor principle for its outward data, Nigeria does not compile outward transactions data, and three other non-OECD countries do not compile geographic breakdowns for their outward transactions data.) The percentage of OECD countries that use the debtor/ creditor principle has improved significantly since 1977, both for the inward FDI statistics and for the outward FDI statistics—86 percent of the relevant OECD countries now use the debtor/creditor principle for their inward data, compared with 66 percent in 1997; and 88 percent now use it for their outward data, compared with 64 percent in 1997. The percentage for the other countries that participated in the 2001 SIMSDI update has remained unchanged since 1997 for the inward data, with 76 percent of the relevant countries using the debtor/creditor principle. However, the use of the debtor/creditor principle by these countries for the outward data has increased from 75 percent in 1997 to 88 percent in 2001.

Table 3.12.Geographic Breakdowns: Basis Used for Allocating FDI Position Data by Country
Number of CountriesImmediate Country BasisUltimate Country Basis
InwardOutwardInwardOutward
Total 2001 (61)413753
Total 1997 (61)322775
Change+9+10–2–2
OECD 2001 (30)252342
OECD 1997 (29)242175
Other 2001 (31)161411
Other 1997 (32)8600

Immediate country versus ultimate country basis

3.46Table 3.12 indicates the number of countries that use the immediate country basis or the ultimate country basis for the geographic allocation of their FDI position data, both for the 1997 SIMSDI survey and for the 2001 update. Table 12 of Appendix I gives the details by country for 2001. The 2001 results show that all countries that compile geographic breakdowns for their FDI position data use the immediate country basis. However, a small number of countries also compile FDI position data on an ultimate country basis. Five countries compile these data for their inward position data (Denmark, Estonia, Luxembourg, Portugal, and the United States). Only Denmark and the United States disseminate these data; Luxembourg’s data are disseminated by the OECD but not in the national publications. Three countries (Denmark, Estonia, and Luxembourg) compile data on an immediate country basis for their outward FDI position data, but only Denmark disseminates these data. (Estonia’s and Luxembourg’s data are not disseminated.)

Availability of Industrial Breakdowns of FDI Statistics

3.47Table 3.13 shows the results of the 2001 update regarding the compilation of industrial breakdowns for the various components of the inward and outward FDI statistics and indicates the changes since 1997. (Table 13 of Appendix I gives the details by country for 2001.) Table 3.13 indicates that there have been significant increases in the number of countries that compile industrial sector breakdowns across all components of FDI statistics and for both the inward and the outward data. There have also been increases in the percentages of relevant countries that compile industrial sector breakdowns for FDI income data and the FDI financial flows data, both inward and outward.

OECD countries

3.48 As with the geographic breakdowns, a significantly higher percentage of the relevant OECD countries compile industrial sector breakdowns for the FDI financial flows data than for the FDI income data (86 percent for the inward financial flows and 79 percent for the outward financial flows in 2001, compared with 66 percent for the inward FDI income and 61 percent for the outward FDI income data). For all categories, there have been sizable increases since 1997 in the percentage of relevant OECD countries that compile these breakdowns. However, while the percentage of relevant OECD countries that compile the industrial sector breakdowns for the FDI position data is relatively high, the percentages of 82 percent for the inward position data and 74 percent for the outward position data have remained largely unchanged since 1997.

Table 3.13.Availability of Industrial Breakdowns of FDI Income, Financial Flows, and Position Data
Number of CountriesCountries That Compile Industrial Breakdowns of:
Direct investment incomeDirect investment financial flowsDirect investment positions
InwardOutwardInwardOutwardInwardOutward
Total 2001 (61)332749364032
Total 1997 (61)221737242924
Change+11+10+12+12+11+8
OECD 2001 (30)191725222320
OECD 1997 (29)151421182118
Other 2001 (31)141024141712
Other 1997 (32)7316686

Other countries

3.49 The other IMF member countries that participated in the 2001 SIMSDI update show a similar pattern. The percentage of relevant countries that compile industrial sector breakdowns is significantly lower for the FDI income data (47 percent for the inward data and 37 percent for the outward data) and the outward FDI financial flows data (50 percent) than for the inward FDI financial flows data (77 percent). The lower percentages for the outward data reflect, in part, the confidentiality constraints faced by some countries with smaller amounts of outward FDI transactions or number of direct investors abroad. As with the OECD countries, the percentage of other IMF member countries that compile industrial breakdowns is relatively high for the inward FDI position data (74 percent). However, the percentage for the outward FDI position data is lower (57 percent) in part because of confidentiality constraints if the total amount of FDI assets is relatively small, or if there are only a few investors involved in FDI abroad.

Basis for Industrial Allocation of FDI Statistics

3.50 In the case of the inward FDI statistics, the data can be allocated on the basis of the industry of the resident direct investment enterprise or on the basis of the nonresident direct investor. A country can also compile two separate datasets, each using a different allocation basis. In the case of the outward FDI statistics, the data can be allocated on the basis of the resident direct investor or on the basis of the nonresident direct investment enterprise. Two separate sets of data can also be compiled, each using a different allocation basis.

3.51 The compilation of industrial sector breakdowns based solely on the resident direct investment enterprise in the case of the inward data, or on the resident direct investor in the case of the outward data, can create difficulties for bilateral comparisons. Such difficulties can arise if (1) the industry of the resident direct investment enterprise differs from the industry of the nonresident direct investor in the case of the inward data, or (2) the industry of the resident direct investor differs from the industry of the nonresident direct investment enterprise in the case of the outward data.

3.52Table 14 of Appendix I shows the results of the 2001 update by country regarding the basis used for the industrial allocation of the FDI statistics. The table shows that all of the 50 countries that compile industrial sector breakdowns for their inward FDI transactions data allocate the data on the basis of the industry of the resident direct investment enterprise. However, three countries (the Netherlands, Portugal, and the Slovak Republic) also compile, if not disseminate, data on the basis of the industry of the nonresident direct investor. For the inward FDI position data, all 40 of the countries that compile industrial sector breakdowns do so on the basis of the industry of the resident direct investment enterprise, while four countries (Iceland, the Netherlands, Portugal, and the Slovak Republic) also compile, if not disseminate, data on the basis of the industry of the nonresident direct investor.

3.53 The results of the 2001 update indicate that all but 3 of the 36 countries that compile industrial sector breakdowns for their outward FDI transactions data compile these data on the basis of the industry of the resident direct investor. Three countries, Colombia, Iceland, and Poland, compile their breakdowns based solely on the industry of the nonresident direct investment enterprise. Seven countries (Austria, Italy, the Netherlands, Portugal, the Slovak Republic, the United Kingdom, and the United States) compile, if not disseminate, data on both bases. However, in the case of the United States the data that are based on the industry of the resident direct investor cover outward FDI income only. For the outward FDI position data, 27 of the 32 countries that compile industrial sector breakdowns do so on the basis of the industry of the resident direct investor; five do so solely on the basis of the industry of the nonresident direct investment enterprise (including the Czech Republic, Germany, Iceland, and Poland); and seven countries compile, if not disseminate, data using both bases (Austria, Italy, the Netherlands, Portugal, the Slovak Republic, the United Kingdom, and the United States [reinvested earnings only]).

1The Belgian authorities have compiled joint Belgo-Luxembourg Economic Union (BLEU) FDI transactions data for Belgium and Luxembourg through 2001. Luxembourg will compile its own FDI transactions data from 2002 onward.
2“Most timely” data are the direct investment data that are first disseminated; that is, the data with the shortest lapse of time between the end of the reference period (or the reference date) and dissemination of the data. Although disseminated, such data may be preliminary and subject to revision.
3The European Union includes Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden, and the United Kingdom.
4“Most comprehensive” data are the direct investment data disseminated that are based on the most comprehensive, regularly available data sources. These data may be preliminary and subject to revision. (See also Appendix II.)
5An ITRS measures individual balance of payments cash transactions passing through the domestic banks, and may also measure foreign bank accounts of enterprises and noncash transactions and stock positions. Data are compiled from forms submitted by domestic banks to the compilers and may also be compiled from forms submitted by enterprises to the compiler.
6See paragraphs 696–698 of the Balance of Payments Compilation Guide. An electronic version of the publication can be viewed on the balance of payments section of the IMF website at http://www.imf.org/bop, under publications.
7Austria, Belgium, the Czech Republic, Denmark, Finland, France, Greece, Hungary, Italy, the Netherlands, Portugal, Spain, Sweden, the United Kingdom, and the United States.
8Including Argentina, Bolivia, Colombia, Israel, Latvia, Peru, Thailand, and Tunisia.

    Other Resources Citing This Publication