## g2-treatment-of-mne-and-intramne-flows

## Source details

**Canonical URL:** [g2-treatment-of-mne-and-intramne-flows](https://www.imf.org/-/media/files/data/statistics/bpm6/gztt/g2-treatment-of-mne-and-intramne-flows.pdf)

## Other formats

- [Markdown version](/-/media/files/data/statistics/bpm6/gztt/g2-treatment-of-mne-and-intramne-flows.pdf.md)
- [Structured JSON version](/-/media/files/data/statistics/bpm6/gztt/g2-treatment-of-mne-and-intramne-flows.pdf.json)

---

### Overview: role of MNEs and SPEs
- Multinational enterprises (MNEs) manage production, trade, direct investment, and international transfer of knowledge and technology to maximize global profits.
- MNEs organize global production among affiliates to leverage lower labor costs, friendlier regulatory environments, more educated workforces, market access, and tax advantages.
- MNEs increasingly establish special purpose entities (SPEs) to channel financial investments globally and increase access to multiple financial markets.

### Measurement challenges for national accounts (NAs) and balance of payments
- Concepts of residence and economic presence central to macroeconomic statistics are less important to MNEs, complicating breakdown of production by economy.
- Current methodological standards posit that foreign affiliates of MNEs are resident in their respective economies of operation to place production where it occurs (fundamental for estimating GDP).
- Transfer pricing on intra-group flows and non-recording of intra-group use of intellectual property products can misallocate GDP between parent and affiliate economies.
- Estimates recorded by MNEs in national statistical surveys may not meet the purposes of NAs and balance of payments statistics.
- Mis-measurement is significant given growing MNE activities and can adversely affect key macroeconomic indicators, including GDP through misallocation between international trade in goods and services and income and depreciation charges associated with movable corporate assets, especially intangible assets.

### Options considered by the Globalization Task Team (GZTT)
- Option 1: Emphasize existing indicators within the SNA framework.
  - Promote use of GNI, GNDI, NNI, and NNDI which are generally less distorted by globalization and less sensitive to MNE impact than GDP.
- Option 2: Increase granularity and scope within the SNA using institutional sector accounts (ISAs).
  - Break down financial and nonfinancial corporations’ sectors to show foreign-controlled corporations and domestic corporations that are part of MNEs.
  - Leverages the G20 DGI-2 sectoral accounts template and the definition of control in BPM6 and the OECD BD4.
- Option 3: Provide more granularity using extended supply and use tables (eSUTs).
  - eSUTs break down industries into enterprises operating domestically, enterprises controlled by domestic MNEs, and foreign-controlled affiliates of foreign MNEs to better analyze trade in value added and global value chains.
- Option 4: Develop indicators beyond the core SNA framework by redefining existing indicators.
  - Example: GNI* (modified GNI) that excludes certain globalization effects such as depreciation related to cross-border additions to intellectual property assets and aircraft used in international aircraft leasing, and retained earnings of corporate inversions or redomiciled PLCs.

### Comparative assessment of options (findings)
- Where GDP is more severely impacted by globalization, indicators beyond GDP (NDP, GNI, GNDI, NNI, NNDI) facilitate in-depth analysis of MNE impacts.
- Transfer pricing tends to distort GDP more than GNI because earnings of wholly foreign-owned MNEs are distributed to or accrued by foreign owners through net primary income.
- Option 1 advantages:
  - International comparability.
  - Limited need for significant adjustments to statistical operations.
  - Preference for gross measures due to challenges compiling net measures (consumption of fixed capital (CFC) estimation difficulties).
- Option 2 advantages and constraints:
  - ISA framework captures full impact of MNE activities across accounts and can identify domestic MNEs and foreign-controlled aggregates.
  - Data intensity and limited support for further financial subsector breakdowns (e.g., Money market funds (S123), Non-MMF investment funds (S124)).
- Option 3 assessment:
  - Conceptually viable but too ambitious and resource intensive for consistent cross-country implementation.
  - A feasible compromise: break down GVA by relevant industries according to domestic MNEs versus foreign-controlled affiliates.

### GZTT recommended conceptual approach (policy recommendations)
- Encourage use of additional key indicators within the NAs that better highlight MNE activities, with preference for gross measures to promote global comparability.
- Reinforce ISA supplementary presentation supported by the existing SNA framework with added granularity:
  - Separately identify domestic MNEs and an aggregate foreign-controlled nonfinancial/financial corporations category.
  - The foreign/domestic split helps policymakers distinguish economic activities driven by domestic firms from those driven by foreign MNEs.
- Propose a statistical definition of MNEs emphasizing “control” as outlined in BPM6 and BD4 to permit coherent and consistent data collection and implementation of the G20 DGI-2 foreign-controlled breakout.
- Recognize need to distinguish between the MNE enterprise and the MNE group for macroeconomic statistics.

### Definition of control and MNE (explicit definitions)
- Control in DI relationships: the ability of one enterprise to exercise the voting power associated with their shareholdings to affect strategic management decisions of another enterprise.
- Control exists through:
  - an immediate DI relationship where the direct investor owns more than 50 percent of the voting power in the direct investment enterprise (DIE);
  - an indirect DI relationship arising from ownership of voting power in one DIE that owns voting power in another enterprise(s).
- Ownership versus control:
  - Ownership expressed by shareholdings; the BPM6 and the BD4 emphasize the 10 percent ownership rule for DI statistics.
  - Control can exist where less than 50 percent of voting power is held (examples: effective minority control via agreements, cross-shareholding loops, family members acting together).
- Statistical definition of an MNE:
  - An MNE is a legal entity that has at least one nonresident affiliate or branch, and exercises control over its affiliate(s) or branch(es) either directly—by owning over 50 percent of the voting power in the entity—or by indirect transmission of control.
  - The MNE is the ultimate controlling parent (UCP)—the direct investor at the top of the control chain.
  - The MNE group consists of the MNE and the set of legal entities—regardless of their economies of residence—that are under the control of the same UCP.

### MNE group structures, SPEs, and measurement implications
- MNEs operate through subsidiaries, branches, SPEs, and other affiliates across different economies and through vertically and horizontally complex ownership structures.
- Common structures:
  - simple UCP-affiliates structure (no further linkages) — inward/outward DI depending on the residence of the UCP;
  - lengthy ownership chains with multiple cross-border links (e.g., cross-border global value chains (GVCs));
  - joint ventures (JVs) where usually each partner owns 50 percent of voting power; if one JV partner exerts de facto dominance, that partner should be taken as the UCP.
- SPEs in offshore financial centers may have relatively small GDP contribution but large income flows and large financial stocks and flows, exaggerating the role of those centers in international financial transactions (including round-tripping).

### Practical aspects and data sources for compilers
- Official statistics on MNEs remain sparse due to complex ownership structures, absence of uniform data, and limited possibility for NSOs to obtain a complete group view.
- National compilers should rely on administrative data sources and business surveys and extend beyond DI statistics; first step is identifying domestic units that are part of MNE groups.
- Common sources to determine direct/indirect control:
  - business registers (BRs) derived from tax registers, compulsory registration systems, social security data;
  - regulatory institutions: stock exchanges, tax authorities, investment promotion agencies;
  - local enterprise group surveys.
- Consolidated and unit-level financial statements and management accounts are relevant; management accounts often better for high-quality statistical information.
- Case study data sources (ADIMA, UN Global Groups Register (GGR)):
  - commercial data sources such as Orbis, company reports, Legal Entity Identifiers (LEI), big data sources (WikiData, Common Crawl, google and Wikipedia search data).
  - Data merged through fuzzy matching and combined databases to list affiliates related to the UCP.
- Key information in global enterprise registers for identifying MNE groups includes control indicators, economies of registration, identity numbers, addresses, and identity of the UCP.

### Institutional arrangements and Large Case Units (LCUs)
- Countries with significant MNE presence are encouraged to establish LCUs to:
  - provide consistent data from the largest MNE groups to relevant statistical domains;
  - facilitate consistent recording of MNE activities in national accounts and balance of payments;
  - collect timely and accurate data early in the statistical production process;
  - resolve discrepancies before processing by national accounts, trade, or external sector compilers;
  - delineate and classify statistical units of MNEs depending on establishment or enterprise basis.
- LCUs should engage in ongoing consistency assessments across statistical data sources and domains related to particular MNE groups.

### Selection criteria, confidentiality, and data exchange
- Selection of MNEs for national accounts should focus on firms with complex ownership structures, large volume of activities, rearrangements and relocations (including corporate inversions), and firms that own intellectual property products (IPPs).
- Global micro data linking and bilateral exchange of business demographic, accounting, or business microdata are needed to fully understand MNE operations and avoid asymmetries while maintaining confidentiality.
- Compiling institutions should exchange aggregated data along with information on adjustments that deviate from company accounts, while maintaining confidentiality, to avoid asymmetries.

### Changes proposed to the 2008 SNA and institutional sectors
- Work supports adding a new chapter to the 2008 SNA to include the definition of MNEs and agreed approaches, replacing fragmented discussion in the 2008 SNA.
- Changes to 2008 SNA paragraphs 4.81 and 21.47 would be required to fully align with BPM6 and BD4 on control.
- Additional institutional sub-sectors need identification, specifically national private corporations that are part of domestic multinationals for both domestically controlled nonfinancial/financial corporations’ sectors.
- The GZTT proposes adoption of Figure 4 in Annex V (including allocation of MNE units and SPE units to institutional sectors) as a replacement for Figure 4.1 of the 2008 SNA.

### AEG outcomes and consensus on measurement approaches
- The AEG agreed with Recommendations I and II to emphasize existing “national and net” indicators within the SNA framework and to recommend additional breakdowns to highlight MNE activities.
- The AEG unanimously agreed that concepts outside the core SNA framework such as GNI* or mainland GDP should not be included in the SNA update; such measures are left to countries to implement based on policy needs.
- The AEG favored net measures (NDP and NDI) over gross measures for analyzing MNE impacts, while recognizing need for improved guidance to produce internationally comparable measures of CFC.
- The AEG agreed that national accounts should define MNEs by emphasizing the control aspect shown in the Foreign Direct Investment Relationships (FDIR) as in BPM6 and BD4.
- The AEG recommended GZTT coordinate with the United Nations Committee of Experts on Business and Trade Statistics to ensure consistent recommendations.

### Direct Investment Task Team (DI TT) consultation outcomes (Annex VIII)
- Consultation 1 (communication, MNE definition, prioritization of breakouts):
  - DITT members all agreed on need for more focused communication by NSOs to emphasize indicators beyond GDP within the SNA framework.
  - Members agreed increased user education regarding GNI, GNDI, NNI, and NNDI would be analytically useful.
  - Members generally agreed with defining MNEs by emphasizing control as per the FDIR, but noted issues delineating chains of relationships and enterprise versus establishment data compilation.
  - On approaches to highlighting MNEs (ISA, GVA, eSUTs):
    - eSUTs least favored due to resource intensity and confidentiality concerns, especially for smaller economies.
    - ISA approach may be less feasible where statistical operations rely on surveys/models rather than financial records.
- Consultation 2 (definition of MNEs, decision tree, classification by UCP):
  - Need to clearly distinguish between the MNE and the MNE group led to restructuring of definitions in the GN:
    - MNE: “a legal entity that has one nonresident affiliate or branch and exercises control over its affiliate(s) or branch(es) either directly—by owning over 50 percent of the voting power in the entity—or by indirect transmission of control.”
    - MNE group: “...the parent (MNE) and the set of legal entities—regardless of their economies of residence—that are under the control (direct or indirect), of the same UCP.”
  - DITT members generally agreed with classifying institutional units according to the residence of the UCP and with the decision tree in Annex V, with some uncertainties about round-tripping treatment.
  - GN revised to show that a domestic affiliate is treated as a national corporation that is part of domestic MNEs even though the immediate parent is nonresident; classification according to UCP dictates this treatment.

*Guidance Note of the Globalization Task Team (GZTT), International Monetary Fund.*

### SECTION I: INTRODUCTION

### g2-treatment-of-mne-and-intramne-flows - SECTION I: INTRODUCTION

### Overview: role of MNEs and SPEs
- Multinational enterprises (MNEs) manage production, trade, direct investment, and international transfer of knowledge and technology to maximize global profits.
- MNEs organize global production among affiliates to leverage lower labor costs, friendlier regulatory environments, more educated workforces, market access, and tax advantages.
- MNEs increasingly establish special purpose entities (SPEs) to channel financial investments globally and increase access to multiple financial markets.

### Measurement challenges for national accounts (NAs) and balance of payments
- Concepts of residence and economic presence central to macroeconomic statistics are less important to MNEs, complicating breakdown of production by economy.
- Current methodological standards posit that foreign affiliates of MNEs are resident in their respective economies of operation to place production where it occurs (fundamental for estimating GDP).
- Transfer pricing on intra-group flows and non-recording of intra-group use of intellectual property products can misallocate GDP between parent and affiliate economies.
- Estimates recorded by MNEs in national statistical surveys may not meet the purposes of NAs and balance of payments statistics.
- The extent of mismeasurement is not easily quantified, but it is significant given the growing size and importance of MNE activities; mis-measurement can adversely affect key macroeconomic indicators, including GDP through misallocation between international trade in goods and services and income and depreciation charges associated with movable corporate assets, especially intangible assets.

### Issues for discussion and four options considered by the GZTT
- The Globalization Task Team (GZTT) discussed four options to highlight MNE activities in the NAs:
  - Option 1: Emphasize existing indicators within the SNA framework.
    - Promote use of gross national income (GNI), gross national disposable income (GNDI), net national income (NNI), and net national disposable income (NNDI) which are generally less distorted by globalization and less sensitive to MNE impact than GDP.
  - Option 2: Increase granularity and scope within the SNA using institutional sector accounts (ISAs).
    - Break down financial and nonfinancial corporations’ sectors to show foreign-controlled corporations and domestic corporations that are part of MNEs.
    - Leverages the G20 DGI-2 sectoral accounts template and the definition of control in BPM6 and the OECD BD4.
  - Option 3: Provide more granularity using extended supply and use tables (eSUTs).
    - eSUTs break down industries into enterprises operating domestically, enterprises controlled by domestic MNEs, and foreign-controlled affiliates of foreign MNEs to better analyze trade in value added and global value chains.
  - Option 4: Develop indicators beyond the core SNA framework by redefining existing indicators.
    - Example: GNI* (modified GNI) from the Irish Economic Statistics Review Group (ESRG) that excludes certain globalization effects such as depreciation related to cross-border additions to intellectual property assets and aircraft used in international aircraft leasing, and retained earnings of corporate inversions or redomiciled PLCs for Ireland.

### Existing guidance and literature
- The 2008 SNA and BPM6 clarified principles related to institutional units, residence, economic presence, and economic ownership; they discuss MNEs, control, and transfer pricing (2008 SNA, paragraphs 21.47; 4.81; 3.131–3.133; BPM6 paragraphs 6.15; 11.101–11.102).
- The 2008 SNA generally treats a nonresident unit as controlling a resident corporation if it owns 50 percent of the equity, but control may exist with less than half the equity if the unit can exercise powers indicating possible control.
- Additional guidance includes:
  - Chapter 2 of The Impact of Globalization on National Accounts addressing MNE statistical challenges.
  - IMF Committee on Balance of Payments Statistics (BOPCOM) analyses on economic ownership, transfer pricing, and residence.
  - G20 DGI-2 Recommendation 8 encouraging separation of foreign-controlled corporations and corporations that are part of domestic MNEs through the sequence of accounts.
  - Academic and statistical proposals such as Ahmad (2018) proposing eSUTs with ownership breakdowns.

### Considerations from GZTT consultations (options compared)
- Where GDP is more severely impacted by globalization, indicators beyond GDP (NDP, GNI, GNDI, NNI, NNDI) facilitate in-depth analysis of MNE impacts and better reflect economies highly affected by globalization.
- Transfer pricing tends to distort GDP more than GNI because earnings of wholly foreign-owned MNEs are distributed to or accrued by foreign owners through net primary income.
- Option 1 advantages:
  - International comparability.
  - Limited need for significant adjustments to statistical operations.
  - Preference for gross measures due to challenges compiling net measures (consumption of fixed capital (CFC) estimation difficulties).
  - NSOs can highlight analytical indicators but will require extensive education and communication to refocus users.
- Option 2 advantages and constraints:
  - ISA framework captures full impact of MNE activities across accounts and can identify domestic MNEs and foreign-controlled aggregates, assisting policy analysis of domestic versus foreign-driven economic activity.
  - Corporate inversions can be separately identified as an “of which” category in the domestic nonfinancial corporations sector.
  - Data intensity and limited support for further financial subsector breakdowns (e.g., Money market funds (S123), Non-MMF investment funds (S124)).
- Option 3 assessment:
  - Conceptually viable to highlight MNE activities but considered too ambitious and resource intensive for consistent cross-country implementation.
  - Implementation would require significant redesign of surveys, compilation systems, methods, and data.
  - A feasible compromise: break down GVA by relevant industries according to domestic MNEs versus foreign-controlled affiliates.

### Recommended conceptual approach by the GZTT
- Encourage use of additional key indicators within the NAs that better highlight MNE activities, with preference for gross measures to promote global comparability.
- Reinforce ISA supplementary presentation supported by the existing SNA framework with added granularity:
  - Separately identify domestic MNEs and an aggregate foreign-controlled nonfinancial/financial corporations category.
  - The foreign/domestic split helps policymakers distinguish economic activities driven by domestic firms from those driven by foreign MNEs, which may have different welfare implications.
- Propose a statistical definition of MNEs emphasizing “control” as outlined in BPM6 and BD4 to permit coherent and consistent data collection and implementation of the G20 DGI-2 foreign-controlled breakout.
- Recognize the need to distinguish between the MNE enterprise and the MNE group for macroeconomic statistics to reflect the scale and scope of MNE group activities.

*Source: Globalization Task Team (GZTT), IMF.*

### 23.      Control, as defined within the context of direct investment (DI) relationships refers to the

### g2-treatment-of-mne-and-intramne-flows - 23.      Control, as defined within the context of direct investment (DI) relationships refers to the

### Definition of control in DI relationships
- Control is the ability of one enterprise to exercise the voting power associated with their shareholdings to affect strategic management decisions of another enterprise.
- Control is determined to exist through:
  - an immediate DI relationship where the direct investor owns more than 50 percent of the voting power in the direct investment enterprise (DIE);
  - an indirect DI relationship arising from the ownership of voting power in one direct investment enterprise that owns voting power in another enterprise(s)—indirectly through a chain of control.
- The definition of control is aligned with the Framework of Direct Investment Relationship (FDIR), a generalized methodology for identifying and determining the types and extent of direct investment relationships, including rules for indirect transmission of control and influence along chains of ownership.
- The GZTT proposes defining control as laid out above to ensure consistency and avoid subjective judgement.

### Distinguishing ownership and control
- Ownership is expressed by the investor’s shareholdings, giving rights to distributed earnings and voting rights; ownership has a lower threshold than control.
- Both the BPM6 and the BD4 emphasize the 10 percent ownership rule for DI statistics.
- Control can exist where less than 50 percent of voting power is held; examples include:
  - effective minority control via agreements among minority shareholders actively working together;
  - effective cross-shareholding loops;
  - family members acting together can be treated as a single unit, combining ownership shares to show control.
- Practical detection of effective minority control is difficult due to limited access to information such as internal voting power; therefore, the Guidance Note prefers the clear threshold definition in paragraph 23.

### Definition of MNE and MNE group
- Statistical definition of an MNE is based on the concept of institutional unit and control (direct or indirect) to identify foreign-controlled units and units that are part of domestic MNE groups.
- The GN proposes defining an MNE as a legal entity that:
  - has at least one nonresident affiliate or branch, and;
  - exercises control over its affiliate(s) or branch(es) either directly—by owning over 50 percent of the voting power in the entity—or by indirect transmission of control.
- The MNE is the ultimate controlling parent (UCP)—the direct investor at the top of the control chain.
- The MNE group consists of the MNE and the set of legal entities—regardless of their economies of residence—that are under the control of the same UCP.

### MNE group structures, SPEs, and measurement implications
- MNEs operate through subsidiaries, branches, SPEs, and other affiliates across different economies and through vertically and horizontally complex ownership structures.
- Common structures:
  - simple UCP-affiliates structure (no further linkages) — inward/outward DI depending on the residence of the UCP;
  - lengthy ownership chains with multiple cross-border links (e.g., cross-border global value chains (GVCs));
  - joint ventures (JVs) where usually each partner owns 50 percent of voting power; if one JV partner exerts de facto dominance, that partner should be taken as the UCP.
- Certain types of SPEs, often in offshore financial centers, are used to leverage tax and fiscal advantages; these SPEs may have relatively small GDP contribution but large income flows and large financial stocks and flows, and can exaggerate the role of those centers in international financial transactions (including round-tripping).
- Practical implementation aids:
  - the GZTT developed a decision tree allocating institutional units, including units that belong to an MNE group, to institutional sectors; the decision tree extends Figure 4.1 of the 2008 SNA and includes classification of MNE units and SPE units.

### Recommended approaches and practical aspects for compilers
- The recommended options have varying practical implications; Option I was deemed the most practical by consultations.
- For many economies, implementing Option I involves:
  - increased communication with users;
  - strengthening existing indicators within the SNA rather than developing new statistical products.
- Only a subset of GZTT members construct net income measures on both a quarterly and annual basis; the GZTT highlights gross income measures to promote cross-country comparability while recognizing the importance of producing net measures and the potential role of technical assistance.
- Options 2 and 3 require considerable resources for data collection and linking, especially where ISAs and eSUTs are not produced; a cost-benefit analysis is recommended given varying statistical sophistication across countries.
- Development of recommendations to differentiate material versus immaterial MNE activity for measurement purposes would be useful in the next standards update.
- Practical challenges for sectorizing MNEs and addressing their national-account implications include:
  - identification of units involved given lack of granular information;
  - resource intensity;
  - confidentiality issues;
  - need for techniques to better capture data and harmonize estimates across statistics.
- Compiling institutions should exchange aggregated data along with information on adjustments that deviate from company accounts, while maintaining confidentiality, to avoid asymmetries.

### Changes proposed to the 2008 SNA and institutional sectors
- Work supports adding a new chapter to the 2008 SNA to include the definition of MNEs and agreed approaches, replacing fragmented discussion in the 2008 SNA.
- Changes to 2008 SNA paragraphs 4.81 and 21.47 would be required to fully align with BPM6 and BD4 on control.
- Additional institutional sub-sectors need identification, specifically national private corporations that are part of domestic multinationals for both domestically controlled nonfinancial/financial corporations’ sectors.
- The GZTT proposes adoption of Figure 4 in Annex V (including allocation of MNE units and SPE units to institutional sectors) as a replacement for Figure 4.1 of the 2008 SNA.

### Rejected alternative and consultation questions
- Option 4 was rejected by consultation exercises; while useful for examining globalization impacts, such indicators should be left to individual compiling institutions to define and compile based on specific circumstances and policy needs.
- Emphasis should remain on internationally comparable indicators already within the current SNA framework.
- Questions for discussion include agreement on:
  1. the proposed definition of MNEs in the GN;
  2. aligning the SNA with BPM6 and BD4 on control for defining foreign-controlled corporations;
  3. the proposed decision tree for allocating MNE units to institutional sectors;
  4. adding a new chapter on MNEs in the next update of the SNA and BPM;
  5. (a) including guidance on MNE activity in the next update of the SNA and BPM for economies where MNEs are material and providing supplementary statistics; (b) agreement with adding a new chapter on MNEs in the next update of the SNA and BPM.

*Guidance Note of the Globalization Task Team (GZTT).*

### 2.      However, as noted this level of detail may be administratively burdensome for most

### g2-treatment-of-mne-and-intramne-flows

### Practical aspects and challenges identifying MNE units
- Official statistics on MNEs remain sparse due to complex ownership structures, wide range of activities, absence of uniform data, and limited possibility for NSOs in any one economic territory to obtain a ‘complete’ view of group operations.
- National compilers see only parts of MNEs’ global activities; a complete and symmetric view requires viewing each entity within the group in relation to the others.
- Most national business registers (BRs) identify membership of foreign-controlled MNE groups and the country of the UCP, but few capture economic data on activities outside the domestic economy.
- National statistical authorities need skilled statisticians and business analysts because exploiting internal and external business accounts requires labor-intensive classification, analysis, and integration.

### Data sources used to identify MNE groups
- National compilers should rely on administrative data sources and business surveys and extend beyond DI statistics; the first step is to identify domestic units that are part of MNE groups.
- Common sources to determine direct/indirect control include:
  - business registers: completeness must be assessed; BRs are developed from tax registers, compulsory registration systems, social security data, and other public or private data holdings; linking DI information (usually maintained by central banks) to BRs may be legally or administratively restricted.
  - regulatory institutions: local and international stock exchanges, local tax authorities, and investment promotion agencies may have shareholder-structure information.
  - local enterprise group surveys: provided they contain information on shareholders’ structure.
- Consolidated and unit-level financial statements and management accounts are relevant; management accounts are usually better for high-quality statistical information but are often enterprise-specific and country-specific.

### Box case studies: ADIMA and GGR data sources
- ADIMA (example sources):
  - Commercial data sources such as Orbis.
  - Company reports and regulatory submissions (audited quarterly and annual reports; regulator filings).
  - LEI Relationship Records (Legal Entity Identifier, 20-character reference code).
  - Big data sources (WikiData, Common Crawl) to find business names and harmonize IDs.
  - Affiliates from all sources combined into one database with extracted metadata and connections.
- UN Global Groups Register (GGR) (example sources):
  - Company reports and regulatory submissions.
  - Legal Entity Identifiers.
  - Big data sources (google and Wikipedia search data).
  - Data merged through fuzzy matching to list companies related to the UCP.

### Data exchange and shared business registers
- Global micro data linking and bilateral exchange of business demographic, accounting, or business microdata are needed to fully understand MNE operations and avoid asymmetries while maintaining confidentiality.
- Direct access and international data sharing are limited due to strict confidentiality and legal constraints.
- Ongoing efforts include the UNSD Global Groups Register (GGR) and the EuroGroups Register (EGR) in Europe; the EGR is the main source used by NSOs in Europe for MNE statistics.
- Key information in global enterprise registers for identifying MNE groups:
  - Control of the unit: (a) resident legal units controlled by the entity; (b) resident legal unit that controls the entity.
  - Economies of registration, identity number(s) or name(s), and address(es) of nonresident legal units controlled by the entity.
  - Economies of registration, identity number/name/address of the nonresident legal unit that controls the entity.
  - Identity number of the UCP (legal entity) if resident; if nonresident, identify its country of registration (optionally identity number, name and address if available).
  - These data can be found from MNE websites, annual reports, corporate directories, investor relations information, or company profiles.

### Institutional arrangements and Large Case Units (LCUs)
- More structured institutional arrangements are important for compiling MNE estimates and aligning access to data with national accounts guidelines.
- Countries with significant MNE presence are encouraged to establish Large Case Units (LCUs) to:
  - provide consistent data from the largest MNE groups to relevant statistical domains;
  - facilitate consistent recording of MNE activities in national accounts and balance of payments;
  - collect timely and accurate data early in the statistical production process;
  - resolve discrepancies before processing by national accounts, trade, or external sector compilers;
  - delineate and classify statistical units of MNEs depending on whether compilation is on an establishment or enterprise basis.
- Several countries have established LCUs; LCUs should engage in ongoing consistency assessments across statistical data sources and domains related to particular MNE groups.

### Selection criteria and additional data needs
- Selection of MNEs for national accounts should focus on firms with complex ownership structures, large volume of activities, rearrangements and relocations (including corporate inversions), and firms that own IPPs.
- Once the population of MNE groups is identified, the data items to be shared should be identified.
- Key variables to share include key globalization indicators, monetary flows between countries, restructuring and relocations, and accounting standards information to enable proper reconciliation.

### Outcome of AEG discussion: recommendations and consensus
- The AEG agreed with Recommendations I and II to emphasize existing “national and net” indicators within the SNA framework and to recommend additional breakdowns (e.g., new sub-sectors) to highlight MNE activities.
- The AEG unanimously agreed that concepts outside the core SNA framework such as GNI* or mainland GDP should not be included in the SNA update; such measures are left to countries to implement based on policy needs.
- The AEG favored net measures (NDP and NDI) over gross measures for analyzing the impact of MNE activities, while recognizing the need for improved guidance to produce internationally comparable measures of CFC.
- The AEG agreed that national accounts should define MNEs by emphasizing the control aspect shown in the Foreign Direct Investment Relationships (FDIR) as in BPM6 and BD4, while noting conceptual tensions among control, ownership, and residency; members highlighted the need for a clear control framework to facilitate data collection.
- Regarding approaches to highlighting MNE activities, the AEG acknowledged the analytical usefulness of:
  - institutional sector accounts disaggregated into foreign-controlled affiliates and domestically controlled affiliates of MNEs (as in recommendation 8 of G20 DGI.II);
  - gross value added by industry disaggregated into foreign-controlled and domestically controlled units;
  - extended/satellite supply and use tables.
- The AEG expressed that NSOs should decide which method to adopt based on statistical production processes and users’ analytical and policy needs.
- The AEG recommended that the GZTT coordinate with the United Nations Committee of Experts on Business and Trade Statistics, which is working on global groups registers and statistical business registers, to ensure consistent recommendations.

*Source: IMF GZTT chapter on treatment of MNE and intramne flows (annexes and AEG outcome).*

### Annex VIII. Outcome of the Direct Investment Task Team (DI TT) Consul tati on

### Annex VIII. Outcome of the Direct Investment Task Team (DI TT) Consultation

### Consultation 1 — Communication, MNE definition, and prioritization of breakouts
- Questions posed to the DITT included:
  - Do members agree that more focused communication efforts will be needed to emphasize the use of existing indicators?
  - Do members agree with the proposal to define MNEs by emphasizing the control aspect shown in the FDIR as in the BPM6 and BD4?
  - Please rank (1–3) which additional breakouts should be given the highest priority:
    - Institutional sector accounts (following the breakdown into foreign-controlled and domestically controlled corporations as designed in recommendation 8 of G20 DGI 2)
    - Gross value added by industry broken down into foreign-controlled and domestically controlled units
    - Extended supply and use tables

- Outcome — key findings from the DITT:
  - The DITT members all agreed that there should be more focused communication by NSOs to emphasize the indicators—beyond GDP—that currently exist within the SNA framework.
  - Members agreed that regardless of whether NSOs choose to adopt either the ISA, GVA, or ESUTs approach to provide additional granularity, increased user education regarding indicators such as GNI, GNDI, NNI, and NNDI would be analytically useful.
  - The DITT members generally agreed with the proposal to define MNEs by emphasizing control as defined in the FDIR.
    - However, issues were noted concerning how to delineate a chain of relationships and the issue of compiling data for the enterprise relative to the establishment.
    - The DITT noted that a precise definition of control and clearer definition MNEs are required.
  - On the three approaches to highlighting MNEs (ISA, GVA, ESUTs) there was no clear preference from the DITT, although:
    - The extended supply and use table approach was least favored because it may be resource intensive and may raise issues of confidentiality, especially for smaller economies.
    - The ISA approach may be less feasible where statistical operations are based on surveys or models instead of financial records; the DITT pointed to practical issues that may impact implementation of these approaches.

### Consultation 2 — Definition of MNEs, decision tree, and classification by UCP
- Questions posed to the DITT included:
  - Do members agree with the proposed definition of MNEs as provided in this GN?
  - Do members agree with classifying institutional units within the institutional sector accounts according to the residence of the ultimate controlling parent (UCP) (e.g., if the UCP is nonresident then the institutional unit is foreign controlled and if the UCP is resident the unit is part of the domestic MNE)?
  - Do members agree with the proposed decision tree for allocating MNE units to institutional sectors?
  - Taking into consideration the extensive discussion on defining more comprehensive statistics on MNEs, do members support that the next update of the SNA and BPM6 include guidance on MNE activity, for economies where MNEs are material and provide the supplementary statistics?

- Outcome — key findings and revisions:
  - Most respondents highlighted a need to more clearly distinguish between the MNE and the MNE group.
    - The consultation noted the need to more explicitly cover inward direct investments, global value chains, the treatment of SPEs including round-tripping.
  - In response, the GZTT restructured the definition as reflected in paragraph 26 and paragraph 27 to clearly state the difference between the MNE and the MNE group:
    - The MNE is defined as “a legal entity that has one nonresident affiliate or branch and exercises control over its affiliate(s) or branch(es) either directly—by owning over 50 percent of the voting power in the entity—or by indirect transmission of control.”
    - The MNE group consists of “...the parent (MNE) and the set of legal entities—regardless of their economies of residence—that are under the control (direct or indirect), of the same UCP.”
    - Additional information was provided in the GN (paragraph 28) to reflect the various ownership or organizations of MNE groups.
  - Members of the DITT generally agreed with the geographical classification of MNE groups according to the residence of the UCP—such that institutional units that belong to MNE groups can be classified as purported by the decision tree (Annex V of the GN).
  - The DITT members generally agreed with the decision tree as shown in Annex V.
    - Some uncertainties were identified regarding the treatment of round-tripping—pass-through funds invested in an affiliate in a second economy, before being re-invested into a third affiliate in the ultimate investing economy (UIC).
    - The GN was revised to show that the domestic affiliate is treated as a national corporation that is part of domestic MNEs—even though the immediate parent is nonresident. The classification of MNE units according to UCP dictates this treatment.

---


_Source: https://www.imf.org/-/media/files/data/statistics/bpm6/gztt/g2-treatment-of-mne-and-intramne-flows.pdf_