## d14 — Financial Conduits — SECTION I: THE ISSUE

## Source details

**Canonical URL:** [d14 — Financial Conduits — SECTION I: THE ISSUE](https://www.imf.org/-/media/files/data/statistics/bpm6/approved-guidance-notes/d14-financial-conduits.pdf)

## Other formats

- [Markdown version](/-/media/files/data/statistics/bpm6/approved-guidance-notes/d14-financial-conduits.pdf.md)
- [Structured JSON version](/-/media/files/data/statistics/bpm6/approved-guidance-notes/d14-financial-conduits.pdf.json)

---

### Background and problem statement
- IMF Guidance Note (D.14) describes Special Purpose Entities (SPEs) used by multinational enterprises (MNEs) to raise funds in financial centers and remit them to parents or related enterprises; these SPEs are termed Financial Conduits (FCs).
- Empirical observations cited:
  - BIS found a correlation between intercompany lending components of external debt and International Debt Statistics on a nationality, rather than residency, basis for some large, emerging markets (2006 to 2014).
  - Coppola et al. (2020) estimate that U.S. holdings of Brazilian corporate bonds in 2017 would be USD 50 billion, compared with USD 8 billion in residency-based official statistics.
- Core problem: FCs can obscure financial linkages between countries by inserting a resident conduit between open-market creditors and MNE headquarters, complicating assessment of exposures, partner economies, and exchange-rate risk.

### Functional classification and typical conduit behavior (Issue 1)
- Classification principle:
  - Instrument classification is determined first; functional category follows (BPM6 Table 6.1 links instruments to functional categories).
  - Supplementary identification of debt securities in DI is recommended (GN D.7).
- Typical FC balance-sheet behavior:
  - Liability side:
    - FCs issue debt securities on open markets (classified as Portfolio Investment per BPM6, paragraph 5.44).
    - FCs may take loans from resident or non-resident banks (Other Investment).
    - Presence of unrelated counterparties on the liability side distinguishes FCs from “Intra group lending companies.”
  - Asset side:
    - FCs typically do not transact on open markets; they remit funds raised to parents or related enterprises, primarily through loans and sometimes trade credits or advances (repayable in goods in certain commodity-linked arrangements).
- Typical mapping of instruments to functional categories:
  - Equity — Held by affiliated parties — DI. (Equity should not represent the main part of FC’s balance sheet.)
  - Debt — Loans — Held by unaffiliated — Other Investment (OI).
  - Debt — Loans — Held by affiliated — DI.
  - Debt — Debt securities — Held by unaffiliated — Portfolio Investment (PI).
  - Debt — Debt securities — Held by affiliated — DI.
  - Asset-side instruments:
    - Loans to affiliated parties — DI.
    - Trade credit and advances to affiliated parties — DI.
- Debt instrument complexity:
  - Hybrid instruments (convertible bonds, contingent convertible bonds (CoCo’s), non-participating preferred shares) raise classification questions because some are negotiable securities while others are non-negotiable and directly held by affiliates.

### Analytical implications of current DI recording for FCs
- Under current BPM6 standards:
  - Loans from an FC to the head office and to affiliates are recorded as DI; FCs are typically Captive Financial Institutions and are not excluded from DI (BPM6 paragraph 6.28).
  - This recording shifts capital-market lenders’ exposure onto the country of the FC rather than the MNE headquarters, understating market creditors’ exposure to the headquarters.
  - Interposition of FCs obscures risk characteristics: FCs often lack independent revenue to repay debt, implying parents likely bear repayment responsibility and making such DI riskier than standard intercompany lending (BPM6 paragraph 6.26 notes intercompany lending is identified separately for debt analysis).
  - Exchange-rate risk assessment can be distorted if FCs issue debt in foreign currencies while MNE revenues are in home currency, understating resident-country exposure to foreign-currency debt.

### Pattern, consequences, and analytical needs (Issue 2)
- Simplified flow described:
  - F3: FC raises funds from open markets (debt securities).
  - F4: FC passes funds to headquarters and affiliates through intercompany lending (loans).
  - F5: Productive assets are funded downstream as equity.
  - Affiliates in other countries may hold parts of the debt issued by the FC, resulting in DI classification for those holdings.
- Consequences:
  - Total amount of MNE debt is recorded correctly, but functional categories and partner-economy attribution do not reflect economically meaningful linkages between open-market funding and ultimate borrowers.
  - Linking DI debt with corresponding open-market funding would better reveal exposures and the true nature of the debt.

### Outcomes considered and recommendations (summary of proposals)
- Option 1 — No change to standards:
  - Continue relying on researcher datasets and methods (e.g., Coppola et al. (2020)), BIS, national ad-hoc statistics (Bank of Russia). The IMF SPE template addresses some data gaps. Conclusion: no necessary change to current reporting framework.
- Option 2 — Add an “of which” item to identify DI lending of FCs:
  - Introduce an “of which” item to make intercompany lending by FCs visible within DI statistics.
  - SPE reporting template will collect some FC information; additional items identifying FC intercompany lending should be added.
  - Feasibility depends on countries’ ability to identify FCs in the Statistical Business Register (SBR) and distinguish on the liability side between debt securities and loans; feasibility studies are recommended, particularly in financial centers.
- Option 3 (rejected) — Identify the economy of the ultimate sponsor (Ultimate Investing Economy, UIE):
  - Collecting UIE in SBR was judged too burdensome at this time, though selective implementation by host countries of FCs could be considered later.
- Option 4 (rejected) — Imputation approach analogous to government-sponsored nonresident issuers:
  - Imputing FC borrowing to resident government is consistent with SNA 2008 and Government Finance Statistics Manual 2014 for government-sponsored FCs but was rejected for corporations as too complex and unnecessary.

### Decision tree, committee views, and consultation outcomes
- DITT outcomes:
  - The DITT unanimously supported the decision tree for deriving the functional category of instruments issued by FCs (Issue 1).
  - On Issue 2, the majority of the DITT supported maintaining the status quo.
  - Concerns raised by members included confidentiality risks from separate reporting and the sufficiency of additional information to capture FC risks.
  - Minority positions:
    - Two members supported Option 2 (separate “of which FCs” identification under DI/debt in SPE annual reporting, if feasible).
    - One member suggested Option 3 be considered as part of BPTT work on the nationality concept.
    - One member asked whether Option 1 should also apply to SPEs owned by public corporations because SPEs of public corporations are indirectly government-owned and could serve quasi-fiscal purposes.
  - Written consultation outcome:
    - Committee members largely supported the decision tree for Issue 1 and maintaining the status quo for Issue 2.
    - A large majority preferred maintaining the status quo for Issue 2 but including the enhanced imputation for FCs owned by public corporations as described in Option A in GN D.5.

### Reporting scope, SPE template, and data-collection challenges
- Reporting scope and timing:
  - Reporting template collects data only for resident SPEs, recognizing difficulties collecting data for non-resident SPEs.
  - Template collects financial account components of the Balance of Payments and IIP related to SPEs for flows during the current year 20XX and positions as at end of current year 20XX.
  - In 2021, the IMF will begin collecting data from countries on resident SPEs for calendar year 2020 and any earlier years for which information is available.
- Template components requested (selected):
  - Direct Investment related to SPEs:
    - 1.1 Equity and investment fund shares /1
      - 1.1. Equity other than reinvestment of earnings (breakdowns by direct investor, direct investment enterprises, between fellow enterprises with ultimate controlling parent resident/nonresident/unknown)
      - 1.1. Reinvestment of earnings
    - 1.2 Debt instruments (Direct investor in direct investment enterprises; Direct investment enterprises in direct investor; Between fellow enterprises with ultimate controlling parent resident/nonresident/unknown)
  - Portfolio investment related to SPEs:
    - 2.1 Equity and investment fund shares
    - 2.2 Debt securities (Short-term; Long-term)
  - Other Investment related to SPEs:
    - Short-term; Long-term
  - Financial Derivatives related to SPEs
- Data-collection challenges for UIE breakdowns:
  - Breakdown of FC liabilities (debt securities issued) by UIE is challenging for countries using the “residual approach” to PI compilation.
  - Residual approach: PI liabilities vis-à-vis RoW equal total liabilities (first source) net of resident holdings (second source), not the sum of contributors’ liabilities vis-à-vis RoW.
  - Identifying UIE in SBR suffices for the first source, but residual approach requires UIE of the issuer at security level, ideally via a reference database (for instance Legal Entity Indicator (LEI) level 2).
  - Conceptual representation provided for securities identified by ISIN (or CUSIP) and resident holdings; liabilities of FC vis-à-vis RoW defined as FCrow = ΣISINix − ΣΣISINih, with decomposition by UIE similarly specified.

### Country example (Bank of Russia) and analytical outputs
- Bank of Russia example (Annex III):
  - Twenty percent of the corporate debt of the Russian Federation in the form of loans and deposits was initially raised through placements of Eurobonds and other debt securities by FCs abroad on behalf of Russian residents.
  - At users’ request, the Bank of Russia developed an analytical table presenting funds raised on behalf of Russian banks and other sectors by non-resident FCs.
  - Liabilities are broken down by major currencies for currency risk analysis.
  - The information is published quarterly by the Bank of Russia and highlights the correlation between intercompany lending and issuance of debt securities.

### Rejected alternatives and rationale
- Rejected: Reclassify intracompany debt of FCs from DI to Other Investment (as done for selected affiliated financial corporations, BPM6 paragraph 6.28).
  - Rationale: FC loans differ from operations excluded under that exception; reclassification would blur institutional-sector lines and impede identification of non-bank financing relevant for financial stability and macroprudential analysis.
- Rejected: Consolidate the FC with the entity that controlled it.
  - Rationale: GZTT rejected consolidating SPEs across national boundaries; conceptual and practical constraints (cross-border consolidation and lack of data exchange) make consolidation infeasible, except possibly for selected government-controlled entities.
- Rejected: Develop supplemental nationality-based MNE statistics (re-attribute borrowing by foreign subsidiaries to parent’s home country).
  - Rationale: Would require group-level consolidation to eliminate double-counting of intra-group flows, goes beyond the balance of payments framework, and is more appropriate for BPTT consideration of a nationality concept.

### Implementation and next steps
- GN D.7 implementation may highlight debt securities between affiliated parties; separate identification will help users interpret DI debt securities.
- The SPE reporting template will collect certain FC-related information; additional SPE-template items may be required to identify intercompany lending of FCs.
- Feasibility studies should assess whether distinguishing FCs in national compilation systems and SBRs (liability-side separation between debt securities and loans) is practicable, especially in financial centers.

*Source: IMF Guidance Note (D.14) on Financial Conduits — SECTION I: THE ISSUE*

### SECTION I: THE ISSUE

### d14 — Financial Conduits — SECTION I: THE ISSUE

### Background and problem statement
- Multinational enterprises (MNEs) can finance activities through foreign affiliates, including Special Purpose Entities (SPEs) established to minimize taxes or access financial services.  
- MNEs can issue bonds in financial centers and lend proceeds back to the head office or other parts of the MNE.  
- BIS found a correlation between intercompany lending components of external debt and International Debt Statistics on a nationality, rather than residency, basis for some large, emerging markets (Figure 1, 2006 to 2014).  
- Coppola et al. (2020) developed a methodology linking holders of bonds issued in financial centers to MNE home countries and estimate, using that method, that U.S. holdings of Brazilian corporate bonds in 2017 would be USD 50 billion, compared with USD 8 billion in residency-based official statistics.  
- Financial Conduits (FCs) are SPEs established to “raise funds, often from unrelated enterprises, and remit those funds to their parents or to another related enterprise.” The main distinguishing feature from “Intra group lending companies” is the presence of unrelated counterparties on the liability side.  
- Use of FCs can obscure financial linkages between countries, complicating assessment of exposures and where capital is deployed. Concerns exist that countries may record these transactions differently, with some countries looking through the FC.

### Issues for discussion (scope of the Guidance Note)
- Two issues are explored:
  - Under which circumstances some debt securities issued by FCs should be recorded as Direct Investment (DI) and how to interpret potential DI statistics.
  - Alternatives for recording FC activities to better link their portfolio investment liabilities and DI assets.

### First issue — functional category of instruments issued by conduits (key observations)
- Typical FC balance sheet behavior:
  - Asset side: FCs typically do not transact on open markets; they remit funds raised to parents or related enterprises primarily through loans and sometimes trade credits or advances (repayable in goods in certain commodity-linked arrangements).  
  - Liability side: FCs issue debt securities on open markets, which are classified as Portfolio Investment (BPM6, paragraph 5.44). They may also take loans from resident or non-resident banks (Other Investment).  
- Debt instrument complexity:
  - Sophisticated hybrid instruments exist (convertible bonds, contingent convertible bonds (CoCo’s), non-participating preferred shares). Some hybrids are negotiable; others are non-negotiable and directly held by affiliates, raising questions whether they should be classified as “securities” or as loans.  
- Classification principle:
  - Instrument classification must be determined first; the functional category follows from that. BPM6 Table 6.1 provides links between financial instrument classification and eligible functional categories.  
  - Supplementary identification of debt securities in DI is recommended (GN D.7), as debt securities between affiliated parties may differ in nature from those between unaffiliated parties.

### Typical functional-category mapping for FCs (as set out in the GN)
- Typical instruments issued by an FC and their functional categories:
  - Equity — Held by affiliated parties — DI. (Equity should not represent the main part of FC’s balance sheet.)  
  - Debt — Loans — Held by unaffiliated — Other Investment (OI). FCs may take bank loans from unaffiliated foreign banks on a smaller scale.  
  - Debt — Loans — Held by affiliated — DI.  
  - Debt — Debt securities — Held by unaffiliated — Portfolio Investment (PI). (Securities held by market participants are probably the major part.)  
  - Debt — Debt securities — Held by affiliated — DI. (Securities held by affiliates are probably a minor part.)
- Typical instruments held by an FC on the asset side:
  - Loans to affiliated parties — DI. (Typically how FCs remit funds raised.)  
  - Trade credit and advances to affiliated parties — DI.

### Analytical implications of current DI recording for FCs
- Under current standards:
  - Loans from an FC to the head office and to affiliates are recorded as DI. FCs are typically Captive Financial Institutions and are not excluded from DI per BPM6 paragraph 6.28.  
  - This recording understates capital market lenders’ exposure to the MNE headquarters by placing exposure on the country of the FC rather than the headquarters.  
  - The insertion of the FC obscures the nature and riskiness of intercompany lending: FCs lack their own revenue to repay debt, so parents likely bear repayment responsibility, making such DI riskier than standard intercompany lending. BPM6 paragraph 6.26 highlights that intercompany lending is identified separately for debt analysis because of different implications for risk and vulnerability.  
  - Exchange rate risk assessment is affected: if the FC issues debt in a foreign currency (e.g., US dollars) but the MNE earns revenue mainly in its home currency, resident-country exposure to foreign-currency debt may be understated.

### Second issue — contribution of conduits to Direct Investment (pattern and concerns)
- Simplified flow: FC raises funds from open markets (F3: debt securities), passes funds to headquarters and other affiliates through intercompany lending (F4: loans), and productive assets are funded downstream (F5: equity). Affiliates in other countries may hold parts of the debt issued by the FC, leading to DI classification for those holdings (Figure 2).  
- Consequences:
  - Amount of MNE debt is correctly recorded but not characterized by functional category or partner economy in an economically meaningful way.  
  - Statistics that link DI debt with corresponding open market funding would better reveal the true nature of the debt and exposures.

### Outcomes considered and recommendations (summary of proposals)
- Option 1 — No change to standards:
  - Rely on researcher datasets and methods (e.g., Coppola et al. (2020)) and existing BIS or national ad-hoc statistics (Bank of Russia). The IMF SPE template significantly addresses data gaps on FCs. Conclusion: no necessary change to current reporting framework.
- Option 2 — Add an “of which” item to identify DI lending of FCs:
  - Include an “of which” item to identify intercompany lending by FCs so DI lending of FCs is visible. The SPE reporting template will collect some needed information, but additional information identifying FC intercompany lending should be added.  
  - Feasibility depends on the ability of countries to spot FCs in their Statistical Business Register (SBR) and to distinguish on the liability side between debt securities (“FC” type) and loans (“Intra group lending” type). Feasibility studies are necessary, particularly in financial centers.
- Option 3 (rejected) — Identify the economy of the ultimate sponsor (Ultimate Investing Economy, UIE):
  - Collecting UIE in SBR for FCs would yield a more complete residency-based picture but was rejected as too burdensome at this time, although it could be implemented selectively by countries hosting FCs.
- Option 4 (rejected) — Imputation approach analogous to government-sponsored nonresident issuers:
  - Impute transactions and positions between resident government and the economy of the FC so borrowing by the FC is reflected as resident government borrowing. While valid for government-sponsored FCs and consistent with SNA 2008 and Government Finance Statistics Manual 2014, this approach was rejected as too complex and unnecessary for corporations.

### Implementation and next steps
- GN D.7 implementation may highlight debt securities between affiliated parties; separate identification will help users interpret DI debt securities.  
- The SPE reporting template will collect certain FC-related information; additional SPE-template items may be required to identify intercompany lending of FCs.  
- Feasibility studies should assess whether distinguishing FCs in national compilation systems and SBRs (liability-side separation between debt securities and loans) is practicable, especially in financial centers.

*Guidance Note prepared by IMF staff (Maria Borga, Paul Feuvrier, Fedor Kharlashin, Mirco Lattwein, Francesca Spinelli) — SECTION I: THE ISSUE*

### 25.      On the first issue, the DITT unanimously supported the decision tree. On the second issue,

### Financial Conduits (D.14)

### Decision tree, Committee views, and consultation outcomes
- The DITT unanimously supported the decision tree for deriving the functional category of instruments issued by FCs (Issue 1).
- On Issue 2, the majority of the DITT supported maintaining the status quo.
- Concerns raised by members:
  - Separate reporting of FCs may raise confidentiality issues.
  - Additional information from separate reporting may not be enough to capture the risks of FCs and would require additional research and analysis.
- Minority positions:
  - Two members supported Option 2—separately identifying FCs through an “of which FCs” under DI/debt in the annual reporting of SPEs if it is feasible for countries.
  - One member suggested Option 3 be considered as part of the Balance of Payments Task Team (BPTT) work on the nationality concept.
  - One member raised whether Option 1 should apply to SPEs owned by public corporations as well as private corporations, noting SPEs of public corporations are indirectly owned by the government and could serve quasi-fiscal purposes.
- Outcome of the written consultation (paragraph 26):
  - Committee members largely supported the proposed decision tree for Issue 1 and maintaining the status quo for Issue 2.
  - It was agreed that the decision tree, with a few minor modifications, will be helpful for compilers.
  - A large majority preferred maintaining the status quo for Issue 2 but including the enhanced imputation for FCs owned by public corporations as described in Option A in GN D.5 (i.e., treatment of FCs owned by public corporations should align with the final decision on GN D.5).

### Reporting scope, template, and data-collection challenges
- Reporting scope:
  - In recognition of collection difficulties for non-resident SPEs, the reporting template only collects data for resident SPEs.
  - The reporting template collects financial account components of the Balance of Payments and IIP related to SPEs for flows during the current year 20XX and positions as at end of current year 20XX.
  - In 2021, the IMF will begin collecting data from countries on resident SPEs for calendar year 2020 and any earlier years for which information is available.
- Template components requested (selected items as presented):
  - Direct Investment related to SPEs:
    - 1.1 Equity and investment fund shares /1
      - 1.1. Equity other than reinvestment of earnings
        - Direct investor in direct investment
        - Direct investment enterprises in direct
        - Between fellow enterprises
          - if ultimate controlling parent is resident
          - if ultimate controlling parent is nonresident
          - if ultimate controlling parent is unknown
      - 1.1. Reinvestment of earnings
    - 1.2 Debt instruments
      - 1.2. Direct investor in direct investment enterprises
      - 1.2. Direct investment enterprises in direct investor
      - 1.2. Between fellow enterprises
        - if ultimate controlling parent is resident
        - if ultimate controlling parent is nonresident
        - if ultimate controlling parent is unknown
  - Portfolio investment related to SPEs:
    - 2.1 Equity and investment fund shares
    - 2.2 Debt securities (Short-term; Long-term)
  - Other Investment related to SPEs:
    - Short-term; Long-term
  - Financial Derivatives related to SPEs
- Data-collection challenges for breakdowns by UIE:
  - The breakdown of FCs’ liabilities (debt securities issued) by UIE (country of MNE’s headquarter) is challenging for countries collecting portfolio investment (PI) information at security level when implementing the “residual approach.”
  - Residual approach definition: PI liabilities vis-à-vis rest of the world (RoW) equal total liabilities (first source) net of resident holdings (second source), not the sum of contributors’ liabilities vis-à-vis RoW.
  - Identifying UIE in the SBR is sufficient for the first source, but the residual approach requires the UIE of the issuer at security level, ideally through a reference database (for instance Legal Entity Indicator (LEI) level 2).
  - Conceptual representation (as presented):
    - Assume FCs in a given country issue i = 1 to I securities identified by ISIN (or CUSIP). Each security i is “ultimately issued” by UIE u. h = 1 to H institutional units (including FCs) in the same country potentially hold those securities (resident holdings).
    - ISINix denotes outstanding amount associated to ISINi and ISINih the amount of ISINi held by holder h.
    - Liabilities of FC vis-à-vis RoW are defined by: FCrow = ΣISINix − ΣΣISINih (as specified in source).
    - Liabilities of FC (vis-à-vis RoW) ultimately held by country u restrict securities to i(u): FCrow_u = ΣISINi(u)_x − ΣΣISINi(u)_h (as specified in source).
  - Footnote references retained: SBR link option; LEI Level 2 reporting both direct and ultimate parent’s LEI.

### Country example and analytical outputs
- Bank of Russia example (Annex III):
  - Twenty percent of the corporate debt of the Russian Federation in the form of loans and deposits was initially raised through placements of Eurobonds and other debt securities by FCs abroad on behalf of Russian residents (standard usage of FCs).
  - By request of statistical users, the Bank of Russia developed an analytical table presenting funds raised on behalf of Russian banks and other sectors by non-resident FCs.
  - For currency risk analysis, liabilities are broken down by major currencies.
  - The information is published quarterly by the Bank of Russia and the table highlights the correlation between intercompany lending and issuance of debt securities.
  - Table referenced: Table 2. Corporate Sector External Debt of the Russian Federation (source indicated).

### Rejected alternatives and rationale
- Rejected option: Reclassify intracompany debt of FCs from DI to Other Investment (as is done for selected affiliated financial corporations, BPM6, paragraph 6.28).
  - Rationale for rejection:
    - FC loans have little to do with operations currently excluded and would blur lines between institutional sectors “Other financial intermediaries except insurance corporations and pension funds” and “Captive financial institutions and money lenders”.
    - Mixing FC loans with securitized loans would impede identification of non-bank financing important for financial stability and macroprudential analysis.
- Rejected option: Consolidate the FC with the entity that controlled it.
  - Rationale for rejection:
    - The Globalization Task Team (GZTT) already rejected consolidating SPEs across national boundaries.
    - Conceptual core disallows consolidating entities across national boundaries.
    - Practical constraints include lack of data exchange between countries that would make consistent, comparable consolidation difficult.
    - Possible exceptions may exist for selected government-controlled entities.
- Rejected option: Develop supplemental statistics on MNEs based on nationality rather than residency (re-attribute borrowing by foreign subsidiaries to the home country of the parent).
  - Rationale for rejection in this GN:
    - Would require consolidation of financial measures for the group to eliminate double-counting of funds in transit or round-tripping.
    - Consolidation would involve netting investments between affiliates from the group’s total assets to remove intra-group flows.
    - Goes beyond the balance of payments framework; could be considered by the BPTT as part of work on introducing a nationality concept into the balance of payments/IIP framework.

---


_Source: https://www.imf.org/-/media/files/data/statistics/bpm6/approved-guidance-notes/d14-financial-conduits.pdf_