This chapter describes several special cases where it is not obvious whether a particular type of enterprise or activity qualifies as foreign direct investment. These cases include special purpose entities; investment in and by collective investment institutions; investment in land, buildings and other immovable objects; construction enterprises; mobile equipment, such as aircraft, ships and drilling rigs; insurance companies; and cash-pooling services. Each case is described and recommendations are made for its recording.
OECD Benchmark Definition of Foreign Direct Investment (Fifth Edition)
Abstract
382. There are a number of special cases where it may be unclear whether or why a particular type of enterprise or activity qualifies as a direct investment enterprise (DIE) or direct investor. The following cases are discussed below: special purpose entities (SPEs) and pass-through funds more generally; collective investment institutions (CIIs); land, structures and other immovable objects; construction enterprises; mobile equipment; insurance companies; and cash-pooling services.
383. Multinational enterprises (MNEs) often have complex ownership structures to create and manage their regional and global production networks but also to minimise their tax burdens and overcome regulatory barriers. These complex structures can make it difficult to interpret foreign direct investment (FDI) statistics. This section will first describe SPEs, which are often embedded in the ownership structure. It then discusses the topic of pass-through funds, which result from these complex ownership structures, more generally. Many of the standard and supplementary FDI series presented in subsequent chapters are meant to address some of the issues with the interpretation of FDI statistics raised by SPEs and pass-through funds.
6.2.1. Special purpose entities
384. MNEs often diversify their investment geographically, through organisational structures that may include a SPE. A SPE is defined as: A formally registered and/or incorporated legal entity recognised as an institutional unit and resident in an economy, with no or little employment, up to a maximum of five employees, no or little physical presence, and no or little physical production in the host economy. SPEs are directly or indirectly controlled by non-residents. SPEs are established to obtain specific advantages provided by the host jurisdiction with an objective to (i) grant its owner(s) access to capital markets or sophisticated financial services; and/or (ii) isolate owner(s) from financial risks; and/or (iii) reduce regulatory and tax burdens; and/or (iv) safeguard confidentiality of their transactions and owner(s). SPEs transact almost entirely with non-residents and a large part of their financial balance sheet typically consists of cross-border claims and liabilities (Integrated Balance of Payments and International Investment Position Manual, Seventh Edition (BPM7, (IMF, Forthcoming[1]), Chapter 5, Section 2, paragraph 5.86)). SPEs used by MNEs can include captive and specialised financial entities, such as financing subsidiaries, conduits, holding companies, shell companies, shelf companies, brass-plate companies, factoring and invoicing companies and securitisation vehicles, as well as non-financial entities, such as royalty and licensing companies, and merchanting companies (Box 6.1).
385. As legal devices, SPEs may be relatively cheap to create and to maintain. Incorporation or registration of SPEs is often associated with offshore financial centres1, but they may also be found in other jurisdictions.
386. SPEs are residents of the economies in which they are incorporated or registered and, therefore, they may be direct investors or DIEs, even if they are shell companies or pass-through entities without any other productive economic activity of their own. If a shell company is incorporated abroad but all of its physical assets, operations, or management activities are in a second foreign economy, then it is classified as a direct investment holding company in its country of incorporation, and its activities in the second foreign economy would be evaluated to determine whether a separate institutional unit exists and, thus, be treated as a DIE. Similarly, if a shell company is incorporated abroad and has some physical assets or operations in two or more other foreign economies, the resident unit in the country of incorporation is classified as a holding company, and the activities in each foreign economy are evaluated individually to determine whether they meet the criteria for treatment as DIEs.
In its final report, the International Monetary Fund (IMF) Task Force on Special Purpose Entities provided a typology of SPEs. The typology was intended to help compilers to identify and classify SPEs based on their economic functions, but it should be noted that not all entities of the types included in the typology are SPEs; only entities meeting the definition of a SPE using the decision tree (Box 6.2) should be classified as such.
Captive financial entities
conduits that raise or borrow funds, often from unaffiliated enterprises, and remit those funds to its direct investor or another affiliated enterprise
holding companies that own or control equity in other enterprises on behalf of its owner without actively directing them
shell companies, also known as shelf or brass plate companies, are formally registered, incorporated, or otherwise legally organised in an economy but do not conduct any operations in that economy other than in a pass-through capacity; these tend to be conduits or holding companies
captive factoring and invoicing companies
captive financial leasing companies
intra-group lending companies
Financial entities
captive insurance companies
securitisation vehicles
companies holding financial and non-financial assets (including real estate) for affiliated companies
Corporate groups non-financial entities
ancillary companies that are wholly owned subsidiaries whose productive activities are confined to providing services to the parent corporation or other affiliated enterprises
operational leasing companies
merchanting companies
royalty and licensing companies
companies that legally hold intangible assets for affiliated enterprise(s)
Wealth management entities
Wealth management entities manage family trusts, foundations, and personal holding companies.
Government-owned financial entities
Central, state, or local governments are direct investors if they hold a 10% or more ownership interest in a non-resident enterprise. SPEs owned by governments for fiscal purposes are entities that raise or borrow funds on behalf of a non-resident general government.
Source: IMF BOPCOM (2018[2]), Final Report of the Task Force on Special Purpose Entities, https://www.imf.org/external/pubs/ft/bop/2018/pdf/18-03.pdf.
387. SPEs qualify as part of a direct investment relationship (and are included with the framework for direct investment relationships – FDIR) by virtue of being resident in one economy and being owned by, or owning, an enterprise in a different economy in compliance with the 10% voting power criterion. In Figure 6.1, for instance, the SPE in economy 2 is used by its parent in economy 1 to indirectly hold enterprises in economy 3 and economy 4. Consequently, economy 2 should include in its FDI assets and liabilities presentation all the FDI investment held by or in resident SPEs as well as the FDI investment held by or in resident non-SPEs.
Source: OECD (2009[3]), OECD Benchmark Definition of Foreign Direct Investment, Fourth Edition (BD4), https://doi.org/10.1787/9789264045743-en.
388. Direct investors use SPEs for the purpose of channelling funds to, and for borrowing funds from, third economies, and for the purpose of holding ownership interests in DIEs. As a result, when an economy hosts SPEs and includes them in its FDI statistics, an increasing part of transactions and positions merely reflects the channelling of funds via this economy. This can lead to a significant overstatement of FDI activity. FDI transactions passing through a SPE generally do not have the expected impact of direct investment, such as through technology transfers and access to competitive markets in the SPE host economies, although even their low levels of employment or physical production can be significant in very small economies. As a consequence, users are interested in series excluding transactions and positions of SPEs, which are purely pass-through funds, and which render these data difficult to interpret for policy or other decision-making purposes. Therefore, as explained further in Chapter 7, countries that host SPEs should explicitly separate FDI statistics on resident SPEs and non-SPEs in standard presentations of FDI according to the directional principle, disaggregated by geography and by industry.
389. In its final report, the International Monetary Fund (IMF) Task Force on SPEs provided a decision tree to help compilers identify these entities (Box 6.2). When reporting to the OECD, compilers should follow the IMF decision tree to identify resident SPEs in their FDI universe.
Following the decision tree in the final report of the IMF Task Force on SPEs, an enterprise is usually considered as a SPE if it meets all of the following criteria:
1. The entity is a formally registered or incorporated institutional unit.
2. The entity is directly or indirectly controlled by a non-resident parent.
3. The entity is established with one or more of the four objectives identified in the definition. To answer this, the compilers will often need to use their knowledge of the legal and institutional setting in their economy to know if entities would gain these advantages.
4. The entity has no or up to five employees.
5. The entity has little or no physical production in the host economy and little or no physical presence.
6. The entity transacts almost entirely with non-residents and a large part of their financial balance sheet typically consists of cross-border claims and liabilities.
Source: IMF BOPCOM (2018[2]), Final Report of the Task Force on Special Purpose Entities, https://www.imf.org/external/pubs/ft/bop/2018/pdf/18-03.pdf.
6.2.2. Pass-through funds
390. Pass-through funds are defined as funds that pass through a direct investment enterprise resident in an economy to an affiliate in another economy, so that the funds do not stay in the economy of the first enterprise. While SPEs play an important role in channelling investment from one economy to another, the phenomenon is not limited to SPEs. In some instances, MNEs may use operational companies to perform functions usually associated with SPEs. Beyond that, the definition above captures any funds that pass-through a DIE to another economy; as such, this definition encompasses pass-through funds arising not just from tax or regulatory motivations but that arise from a number of reasons, including mergers and acquisitions (M&A) and the establishment of regional headquarters (see Annex 6.D).
391. Pass-through funds ‘inflate’ FDI financial transactions, income and positions as each flow into and out of each economy is counted even if the funds are just passing through. As mentioned above, this overstatement of FDI activity can make it difficult to interpret FDI statistics in the sense that the pass-through funds do not represent investment that has a significant impact on the host economy. Moreover, pass-through funds can obscure the ultimate sources and destinations of FDI when the statistics are compiled by immediate partner country. This Benchmark Definition makes several recommendations to address pass-through funds to enhance the interpretability of FDI statistics by recommending that countries compile:
Separate FDI statistics for resident SPEs to estimate a lower bound on the amount of pass-through
FDI position statistics by the ultimate partner economy as well as by the immediate partner economy
FDI statistics according to the residency of the ultimate controlling parent to estimate an upper bound on the amount of pass-through funds.
These presentations are discussed further in Chapter 7.2
392. The Benchmark Definition recommends that when a CII owns at least 10% of the voting power in a non-resident entity that is not a CII, this should be considered as direct investment. However, in a change from the fourth edition of the Benchmark Definition, when an investor in one economy acquires at least 10% of the voting power in a CII in another economy, this investment should be regarded as portfolio investment. More precisely, investment by CIIs, such as mutual funds, hedge funds and distressed funds, should be included in FDI data if the standard 10% voting power threshold is met, but investment into CIIs should always be treated as portfolio investment. This treatment reflects that no single non-resident investor in a CII would likely exert influence over the investment strategy and day-to-day operations due to their collective legal nature. On the other hand, a CII, as a statistical unit, should be regarded as the investor that makes investment decisions, including the possibility that it could influence the operations of an enterprise if it meets the direct investment threshold (see Annex 6.C for examples of CIIs).
393. Ownership of land and structures, including natural resources, in the compiling economy by an entity resident in another economy establishes a direct investment relationship. The one-year rule that is usually used to define the centre of economic interest (see Section 2.2.1) does not apply to land and buildings. Land (defined in Box 6.3) is always considered to be the property of the residents of the economy in which it is located; therefore, a notional enterprise is created when land is acquired by a non-resident, and this is the FDI asset recorded in the international accounts. The only exception to this general concept is the ownership of land and buildings by foreign governments in the form of embassies, consulates, military bases, scientific stations, information or immigration offices, aid agencies, etc.3
394. An entity resident in one economy may acquire direct ownership (that is, without the creation of a separate legal entity) of land or buildings in another economy. The relevant financial transactions, positions and associated income should be recorded in the direct investment statistics of both economies. However, given the specific nature of the assets acquired, a notional DIE is created by the relevant compilers as the immediate owner of the assets in question. This notional enterprise is deemed to be resident in the economy where these assets are located. The actual owner of the assets, i.e., the direct investor resident in another economy, is deemed to own the notional enterprise. In other words, land, structures, and other immovable assets in the host economy that are indirectly owned by a non-resident entity should be regarded as owned by a notional DIE, even if the period of ownership is less than one year. Acquisitions of land, mineral rights, and associated assets, and preparatory expenses for future direct investment units should be regarded as direct investment transactions.
395. Within the context of the previous paragraph, land and buildings owned for personal use but not for production or other business purposes are also considered as direct investment. Holiday and second homes owned by non-residents should therefore be treated in the same way as described above. Similarly, where an investor resident in one economy acquires at least 10% of the ownership of property in another economy for commercial purposes (e.g., rental residence or office accommodation) this investment constitutes FDI.
396. FDI income on these assets includes income from notional DIEs created for holding land and buildings. This income would include rent earned from these assets. When the land or building is rented by the non-resident owners of the notional unit, then an imputation for rent (in the case of use of land), or travel services (such as a holiday home used by non-residents), or operational leasing (if non-resident enterprises own the asset for their own use) should be made. This income is measured net of expenses, such as for upkeep and maintenance of the property and the payment of local taxes. Income from notional DIEs is recorded in FDI income under withdrawals of income from quasi-corporations.
397. In accordance with the 2025 System of National Accounts (2025 SNA, (United Nations et al., Forthcoming[4])), in contrast to an operational lease, the existence of a financial lease on land and buildings gives rise to the creation of a notional DIE with the understanding that it should be for a very long period rather than for a few years.
Land consists of the ground, including the soil covering and any associated surface waters, over which ownership rights are enforced and from which economic benefits can be derived by their owners by holding or using them. The value of land excludes any buildings or other structures situated on it or running through it; cultivated crops, trees and animals; mineral and energy resources; non-cultivated biological resources and water resources below the ground. The associated surface water includes any inland waters (reservoirs, lakes, rivers, etc.) over which ownership rights can be exercised and that can, therefore, be the subject of transactions between institutional units. However, water bodies from which water is regularly extracted, against payment, for use in production (including for irrigation) are included not in water associated with land but in water resources.
Source: United Nations et al. (Forthcoming[4]), 2025 System of National Accounts.
398. Where a DIE is established for the purposes of natural resource exploration, its exploration expenditures should be capitalised (even if the exploration does not prove sufficient economic resources to be viable) as part of the equity investment in the DIE and written off (or written down) as appropriate. This promotes consistency with the SNA, which treats these expenditures as fixed capital formation.
399. For example, assume an oil company incorporates a DIE that drills an oil well. Assume also that the direct investor makes an equity investment in its DIE of 100 in period 1 (including expenditure on bonus preliminary fee payments made by the direct investor to a resident of the host country, usually the government for the right to undertake exploration for natural resources), 30 in period 2, and then closes down the operation at the end of period 2 when the well proves to be dry. There is a financial transaction from the direct investor to the DIE of 100 in period 1 and 30 in period 2 in direct investment accounts of both economies. No further entries for cross-border transactions are recorded after the shutdown of the operation in period 2. Instead, a negative stock adjustment of 130 is made in the direct investment asset position of the economy where the oil company parent is located, and an equal reduction is made in the direct investment liability position of the economy where the affiliate was located.4
400. Assume in contrast that the direct investor makes an equity investment in its DIE of 100 in period 1 (including expenditure on bonus preliminary fee payments made by the direct investor to a resident of the host country, usually the government, for the right to undertake exploration for natural resources), 30 in period 2, and receives natural resources (payments in-kind) as 40 of dividend in periods 3 to 5 and 10 of disinvestment in the last period. In this scenario, there are receipts of direct investment income and a withdrawal of direct investment (equity capital transaction) by the direct investor; a decrease in the direct investment asset position is recorded accordingly.
401. A construction enterprise in one economy may undertake the construction of plant, buildings, etc. in another economy through subsidiary or associate companies in that economy, through a foreign branch in that economy, or by directly undertaking the work itself. The construction work abroad is to be regarded as a direct investment activity in the economy in which it is being carried out in the first two circumstances. If the construction enterprise undertakes the work itself (such as through an unincorporated site office or a branch), its activities may be regarded either as a direct investment activity or as an export of construction services by that enterprise, depending on circumstances. If certain criteria are met – such as the project extending over a period of at least one year; the maintenance of a complete and separate set of accounts for the activity (i.e., income statement, balance sheet, transactions with the parent company, etc.); the activity being subject to tax and other legislation in the host country; the existence of a substantial physical presence; the receipt of funds for its work for its own account, etc. – the work undertaken is to be treated as having involved the creation of a separate institutional unit, a branch, resident in the economy where the activity is being carried out and, therefore, a DIE. If these criteria are not met, the activity is to be treated as an export of services by the construction enterprise. Construction involved with major projects (bridges, dams, power stations, etc.) that are carried out through unincorporated site offices, in most cases, meet the criteria that requires treatment as the production of a resident unit, i.e., a DIE, and thus, as part of the production of the host economy, not as an export of services to that economy.
402. Where an enterprise resident in one economy installs machinery and equipment in another economy, the Benchmark Definition recommends that the work be regarded as services provided to another economy if the installation is carried out entirely, or primarily, by employees of the enterprise who travel abroad to do the work and they complete the installation in less than one year (that rule to be applied flexibly).
403. The operator of mobile equipment (such as ships, aircrafts, gas and oil drilling rigs) that operate within a single economy for at least one year and that fulfil the criteria for treatment as an unincorporated DIE should be treated as a direct investment branch (see also Chapter 2).
404. Various entities may be involved in the provision of shipping and other transportation services. For shipping, the following entities may be involved:
Owner – holds an asset (the ship) and may be engaged solely in a leasing activity, i.e., the leasing of the ship to an operator
Operator – is involved in shipping activities such as fishing, drilling or transporting freight and/or passengers
Ticket sales or business promotion offices – established either by the owner of the ship (who is seeking operators) or by the operator of the ship (who is seeking passengers, freight, or other business).
405. The flag of a ship determines the authority that is responsible for overseeing the operations of the ship and may help determine the jurisdiction where business disagreements are litigated. The flag is often a “flag of convenience” – that is, neither the ship owner nor its operator may have any business operation in the country whose flag is flown. Therefore, the country of registration of the mobile equipment (e.g., of the ship) is not considered in determining the residency of any of the units involved in the shipping activity (that is, the owner or operator of the ship, and the ticket sales and business promotion offices), and so it is not relevant to the discussion of whether there may be a direct investment relationship.
406. If the owner and the operator are the same entity and the owner operates the ship in its own economy of residence, then there is no direct investment. If the owner and the operator are not the same entity (or are not related), then they are separate institutional units. The owner will typically receive a fee from the operator, reflecting the payment (rental services) for the use of the vessel. In this case, the owner is a lessor and is not a provider of transportation services. The operator of the vessel provides transportation services; it receives revenues for transporting passengers and/or freight, pays wages to crew members (who often are residents of a different economy than the operator), and incurs other transportation-related expenses including port expenditures. The relationship between the owner and the operator of the ship is essentially that of a lessor and its lessee, not that of direct investor and a DIE if the owner and the operator reside in different economies and are not related; although, it is possible that they could be in a direct investment relationship. If the lessor owns at least 10% of the voting power in the lessee, any related transaction is recorded in FDI and not trade.
407. An owner may establish an incorporated or unincorporated enterprise that operates the ship. The country of the (affiliated) operator may differ from that of the owner, in which case a direct investment relationship exists. When the ship operations occur in international waters, the activities should be attributed to the economy in which the operator maintains residence.
408. The classification of management offices on the one hand, and business promotion and ticket sales offices on the other, also depends upon the criteria listed for the determination of, firstly, the existence of an institutional unit and then for a DIE. In some circumstances, they will qualify for classification as direct investment and, in other circumstances, they will not. Under the criteria used to determine the existence of direct investment (see Chapter 2), a distinction should be made between those enterprises that engage in real economic activities and have income and other financial statements, and units that are set up to increase sales of the institutional units that established them but that have no sales of their own, such as ticket sales offices and business promotion offices.
409. Determining the residence of shipping companies is often a difficult issue. Different types of leasing arrangements may make it difficult to determine whether the ship is being used under an operational lease or whether it has been effectively sold to the enterprise that operates it (that is, under a financial lease). Also, a ship may operate entirely in an economy outside the one in which its operator is incorporated. In this circumstance, the activity in the economy where the ship operates is a DIE (if it meets the criteria of a branch) that is owned by a direct investor located in the economy where the operator is incorporated.
410. Issues related to owners and operators of ships and complex leasing arrangements and their solutions are applicable to other types of mobile equipment, such as aircraft.
411. The treatment of insurance raises many complex methodological questions. Most of these questions have applicability beyond direct investment, such as to portfolio investment, other investment, other investment income, and services trade (see Integrated Balance of Payment and International Investment Manual, Seventh Edition (BPM7, (IMF, Forthcoming[1])). The Benchmark Definition is fully consistent with the other major international standards where cross-cutting questions concerning the handling of insurance and insurance enterprises are more thoroughly discussed.
412. In general, the Benchmark Definition recommends that direct investment statistics for insurance companies be defined in the same way as those for industrial and commercial companies. However, compilers should take special steps for their appropriate treatment as positions and transactions involving insurance companies may be quite significant, and the accounting principles they follow may not always be fully consistent with preferred direct investment concepts.
413. One complex area involves the treatment of non-life insurance company technical reserves (prepayments of premiums and reserves against unsettled claims). Normally, these reserves should not be included in the direct investment equity position because these amounts are considered to belong to policyholders. In the special case of a captive insurance affiliate, however, the policyholder may be affiliated with the insurance enterprise. Businesses may establish or acquire a “captive” insurance company in order to obtain insurance coverage for unusual types of risk, or for lower premiums than may be available commercially. In this case, a 10% or more ownership level often is reached, and the captive insurance company then meets the criteria for treatment as a DIE. In this circumstance, the technical reserves should be included in the direct investment debt component of the position.
414. For clarification, non-life insurance companies tend to classify insured losses according to general types. The first type includes losses that arise from events that are expected (from an actuarial perspective) to occur sometime in the future. Insurance companies establish and maintain sizable technical reserves largely for the purpose of satisfying losses from these future events. The second type includes losses that are “incurred but not reported”. These are losses from insured events (such as natural disasters) that have already occurred, but the insurance company does not know the size of the losses because claims have not yet been filed (that is, these losses were incurred but were not reported to the insurer). The third are known losses (incurred and reported to the insurer). To be consistent with the treatment of technical reserves (these belong to policyholders and not to stockholders or owners), liabilities arising from all three types of losses should be regarded as debt obligations of the insurance company to its policyholders and not as equity. As noted, in the case of a captive insurance affiliate, the policyholder is the direct investor, and so the debt of affiliated enterprises should be recorded in direct investment.
415. Mutual insurance companies are owned by their policyholders who tend to be unrelated to one another. Typically, no one resident owns at least 10% of the voting power (or the equivalent, for an unincorporated business enterprise) of a mutual insurance company, and, so, these companies do not usually meet the criteria for treatment as DIEs.
416. Cash-pooling arrangements provided by banks allow corporations to externalise the intra-group cash management, and thus, manage their global liquidity more effectively and with lower costs. Financial innovation has responded to the needs of company groups to manage funds in a centralised way by creating different cash-pooling arrangements. These are agreements between a bank and the entities of a group, which can be located in the same or in different economies, allowing for pooling of cash in real time. Cash-pooling appears to have become more popular after the 2007 global financial crisis and is currently very relevant in certain European countries.
417. The main types of cash-pooling arrangements are as follows:
1. Single legal account: Consists of (a) a set of virtual transactions/operational sub-accounts, which are used directly by the individual companies (part of a group), as well as the parent company, for their day-to-day operations; and (b) a top/master group account (usually held by the parent company), which constitutes an obligation of the pooling bank vis-à-vis the beneficiary, and concentrates the funds of the group. The virtual sub-accounts only track the intra-group positions and are not in a direct relationship with the bank. The information on virtual transactions usually is part of the service provided by the pooling bank to the client but are not necessary for the bank’s accounting system. From the point of view of the bank, only changes in the top/master account should be reported as this will reflect changes in its claims vis-à-vis the parent company, the only direct client in this type of cash-pool.
2. Physical cash-pool: Each company that participates in the cash-pooling arrangement holds an account with the pooling bank. In addition, there is a master account that usually is held by the parent company. The balances of the surplus accounts are transferred to the master account on a regular basis (e.g., on a daily basis at close of business). Conversely, the parent company transfers liquidity from the master account to the accounts in deficit at the end of the period (day). In that sense, they are all counterparties of the bank, and the deficit balances from pooling participants appear only temporarily as assets on the bank’s balance sheet.
There are two types of physical cash-pools. In a zero-balancing cash-pool, the full balance of the surplus accounts is transferred to the master account on a regular basis. The target-based cash- pooling specifies a (positive) threshold whereby liquidity is transferred to the master account from individual accounts when the balances of individual accounts exceed the threshold, and conversely, liquidity is transferred from the master account to individual accounts when their balances are below the threshold.
3. Notional cash-pool: As in the physical cash-pool, all the bank accounts represent a legal relationship between the pooling bank and the participating entities, which are thus direct counterparties of the bank. The pooling is performed by the bank by creating a notional top/master account that virtually consolidates the positions of the pooling participants but does not represent a resource or an obligation of the bank. As a result, no liquidity transfers resulting in inter-company loans take place. The funds will remain in the assets of the bank as a loan to a particular participating entity. However, following the structure of the cash-pooling, the loan is guaranteed by the cash-pooling members, subject to lower charges, restrictions and implicit interest, and typically can only be drawn upon to the extent that the overall pool has a positive net balance.
418. Following the debtor/creditor principle, the recording of cash-pooling arrangements in the balance of payments is complex and depends on the identification of the actual debtor and creditor instead of the payment service provider. The parent company and its subsidiaries should be able to report the internal booking entries as inter-company lending (assets and liabilities) depending on the direction of the funds, that is, if they become creditors or debtors. The presence of cash-pooling arrangements between affiliated parties needs to be identified via direct investment surveys when they take the form of single legal account or physical cash-pool.
Note: 1. Except for the debt between affiliated financial intermediaries that should be included under other investment. Debt instruments refer only to deposits and loans following the convention that if a bank is the debtor, the instrument takes the form of deposits, otherwise the instrument is a loan.
419. Table 6.1 above summarises the proposed statistical treatment for the different types of cash-pooling arrangements from the perspective of the compiling economy of the participant subsidiaries. If an overdraft or loan is received by a pooling participating subsidiary and covered by a single legal account or a physical cash-pool, the creditor is the parent company (the holder of the top/master account) and the recording is done as intra-group loans in FDI (except for the debt between affiliated financial intermediaries). However, if the overdraft or loan is based on a notional cash-pool, the creditor is the pooling bank and, thus, the recording is done as ‘other investment’ between the pooling bank and participating entities.
References
[5] IMF (2022), Special Purpose Entities: Guidelines for a Data Template; IMF Technical Notes and Manuals (TNM/2022/006), http://www.imf.org.
[1] IMF (Forthcoming), Integrated Balance of Payments and International Investment Position Manual, Seventh Edition (BPM7), International Monetary Fund, Washington D.C.
[2] IMF BOPCOM (2018), Final Report of the Task Force on Special Purpose Entities, International Monetary Fund, Washington D.C, https://www.imf.org/external/pubs/ft/bop/2018/pdf/18-03.pdf.
[6] Lewellen, K. and L. Robinson (2013), “Internal Ownership Structures of U.S. Multinational Firms”, SSRN Electronic Journal, https://doi.org/10.2139/ssrn.2273553.
[3] OECD (2009), OECD Benchmark Definition of Foreign Direct Investment, Fourth Edition (BD4), OECD Publishing, Paris, https://doi.org/10.1787/9789264045743-en.
[4] United Nations et al. (Forthcoming), 2025 System of National Accounts, United Nations et al., New York.
SPEs and pass-through funds
420. Financial corporations such as special purpose entities (SPEs) or conduits that raise funds in open markets on behalf of their parent corporation or fellow enterprises are usually encompassed in the SNA definition of “Other financial corporations”. Therefore, non-equity transactions/positions between these financial corporations should be included in FDI. A more detailed discussion of SPEs and related issues may be found in Annex 6.B.
421. Fiscal SPEs are entities owned or controlled by general government that are resident in another territory and used to support government revenue or expenditure. Such entities are resident in their economy of incorporation or registration, not in the economy of their owner. For example, a government may use a non-resident special purpose or other entity to issue securities to fund its expenditures. The Benchmark Definition recommends that fiscal SPEs should be treated as DIEs. All financial transactions and positions between them and their parent government should be recorded as direct investment.
422. More generally, cross-border transactions/positions that are pass-through funds via structures put in place to facilitate the financing and transfer of investment within multinational enterprises are included in direct investment as they are integral parts of a direct investment relationship identified according to the framework of direct investment relationships (FDIR). Nevertheless, these transactions often transit through an economy without producing the same impact as FDI in operating affiliates in this economy when the final destination of investment is in a different economy. Since these transactions may distort the analysis of direct investment positions/transactions, this edition of the Benchmark Definition recommends that countries in their standard FDI presentation i) exclude funds going through resident SPEs from the key FDI statistics but to present them as a separate series; (ii) also present inward positions statistics by ultimate investing economy. It also recommends that countries include in their supplementary FDI presentations, (i) outward position statistics by the ultimate host economy; and (ii) aggregate FDI statistics by the residence of the ultimate controlling parent (i.e., if domestic or not). These statistics should be presented according to the directional principle, and the first series should be disaggregated by economy of immediate counterpart and by industry to facilitate the economic analysis of direct investment.
Round-tripping
423. Round-tripping is defined as the channelling abroad by direct investors of local funds and the subsequent return of these funds to the local economy in the form of direct investment. From the perspective of the reporting economy, the simplest example of round-tripping occurs when a domestic investment involves FDI through a subsidiary or associate located abroad. For example, in Annex Figure 6.A.1, company A in the reporting economy provides FDI funds to a non-resident related company (company B) in the routing economy for it to invest them back in another company (company C) in the reporting economy.
Source: OECD (2009[3]), OECD Benchmark Definition of Foreign Direct Investment, Fourth Edition (BD4), https://doi.org/10.1787/9789264045743-en.
424. The simple group structure in Annex Figure 6.A.1 can be easily extended to cover other more complex group structures that are conducive to round-tripping. Two examples of such structures are included in Annex Figure 6.A.2 and Annex Figure 6.A.3 below:
Source: OECD (2009[3]), OECD Benchmark Definition of Foreign Direct Investment, Fourth Edition (BD4), https://doi.org/10.1787/9789264045743-en.
Source: OECD (2009[3]), OECD Benchmark Definition of Foreign Direct Investment, Fourth Edition (BD4), https://doi.org/10.1787/9789264045743-en.
425. There are many reasons for round-tripping, such as:
Tax and fiscal advantages: some economies provide preferential policies to attract FDI, including low taxation, favourable land use rights, convenient administrative support, etc. Since it is not always easy for local enterprises to attract foreign investors, they may channel domestic capital abroad, which is then repatriated as foreign capital for local investment to take advantage of the preferential treatments only available to foreign investors.
Property right protection: infrastructure for property right protection in some economies is not well established. Therefore, the enterprises in these economies may have the motivation to locate their wealth in related enterprises set up in other economies having better legal and institutional settings for property right protection. As well, some investors may prefer to conceal their identities by investing through companies set up in offshore financial centres. Capital will then be brought back to host economies in the form of FDI if there are profitable investment opportunities.
Expectations on exchange control and exchange rate: some economies have control of financial movements and exchange rates. Expectations on changes in exchange control and exchange rates may generate round-tripping for greater flexibility in foreign exchange management.
Accessing better financial services: Financial markets of some economies are not well developed. Enterprises resident in these economies have to access overseas financial markets for better financial services, such as listing of companies in overseas stock markets. The funds raised will be brought back to host economies in the form of FDI. Round-tripping may occur as part of this process.
From mergers and acquisitions: If a domestic company acquires control of a foreign enterprise with pre-existing links back to the domestic economy, this can create a round-tripping relationship and subsequent transactions in the newly-formed domestic multinational enterprise.
Regional headquarters: a domestic company may control other domestic affiliates through a non-resident company to profit from knowledge and skills located abroad and by benefitting from the spillovers that similar companies can generate together.
Reducing transaction costs: companies may provide goods and services through supply chains consisting of their affiliates, including affiliates in the domestic economy. Ownership links directly between these affiliates may help reduce transaction costs and facilitate interactions along these supply chains.
426. Round-tripping of funds flowing between subsidiaries, associates or fellow companies have to be recorded as FDI transactions/positions. For the local economy, they appear as FDI assets for the local funds channelled to routing economies, and as FDI liabilities for the subsequent return of the funds to the local economy. For the routing economy, they appear as FDI liabilities for the funds received from the local economy, and as FDI assets for the return of these funds to the local economy.
427. It may be argued that these round-tripping funds lead to an overstatement of the genuine magnitude of FDI. The Benchmark Definition recommends therefore separate standard and supplementary breakdowns when this phenomenon significantly affects FDI data of an economy. From the point of view of the routing economy, round-tripping may be partly linked with pass-through funds transactions/positions. In this case, the Benchmark Definition recommends that round-tripping that passes through SPEs in the routing economy along with all transactions and positions to and from resident SPEs be separately distinguished. From the point of view of each reporting economy, the geographical breakdown according to the ultimate host economy (UHE) and the ultimate investing economy (UIE) could provide users with very interesting information on the extent of round-tripping into the compiling economy.
Investment funds
428. This Benchmark Definition recommends that investment by collective investment institutions (CIIs) that meets the operational definition of FDI be treated as direct investment (see Section 6.3) despite the fact that the motivations and investment strategies of CIIs may differ from those of other direct investors. For example, CIIs may not be interested in managing the strategic and day-to-day operations of the DIE; they may also have a shorter time horizon for their investment. This treatment is aligned with the strict application of the 10% ownership of voting power threshold and ensures consistency in recording across economies. However, in a change from the fourth edition of the Benchmark Definition, when an investor in one economy acquires at least 10% of the voting power in a CII in another economy, this investment should be regarded as portfolio investment. Annex 6.C provides a detailed discussion of collective investment institutions.
Payments associated with the acquisition of a right to undertake a direct investment
429. In many developing or transition economies, the government requires the payment of an agreed amount of money by direct investors for the right to undertake a direct investment in the host economy. Often, but not always, these operating or concession rights are related to the extraction of natural resources. In transition economies, compilers refer to these payments as "bonuses". They are legal transactions and should not be associated with poor governance.
430. The Benchmark Definition recommends that such bonus payments should be recorded as “direct investment: equity” when there is an intention to establish a DIE (such as in the case of a contractual arrangement between the investor and the government).5
Other borderline cases
431. In the process of globalisation of economic activities, cross-border transactions might at first appear to be foreign direct investment when in fact they do not meet the criteria. For example:
1. An enterprise undertakes to build for a foreign client, usually a government, a complete manufacturing plant, to provide technical knowhow, and to manage and operate a plant for a number of years, without an ongoing on-site managerial presence and without other criteria for the existence of a DIE being met. It has complete control over day-to-day operations and receives a management fee, paid either in cash or in goods produced by the plant. However, the enterprise has no equity stake in the plant and is performing a cross-border service.
2. An enterprise has a long-term contract with a foreign company, provides it with technical knowhow, and has considerable influence over the quality and quantity of output. The enterprise may provide a loan to the foreign company and sometimes will have a member on the company's board. However, there is no equity stake. It is once again a cross-border service.
3. Some host countries have made agreements with a number of foreign enterprises where the host country supplies factory accommodation, electricity, staff accommodation, administration and labour. The foreign enterprise supplies all production machinery, fixtures and fittings for the building and production materials, and is responsible for the initial training of the labour force. The foreign enterprise then pays an agreed piecework rate for each item produced. Where the production machinery and fixtures and fittings remain the property of the foreign enterprise, there is technically a direct investment branch, though the branch's profits will be zero. There is no direct investment interest if the machinery becomes the property of the host country.
4. Some professional companies operate much like a multinational enterprise, but do not hold equity in one another. For example, unrelated (in an equity sense) accounting or management consulting firms may operate globally under a single name, refer business to one another and receive fees in return, share costs (or facilities) for such items as training or advertising, and may have a board of directors to plan business strategy for the group. This is not direct investment and would be difficult or impossible to account for as such, but it does have much in common with direct investment.
5. Other cases might include foreign sales and representative offices, as well as foreign stations, ticket offices, and terminal or port facilities of domestic airlines or ship operators. Such offices or activities can be treated as direct investment only if they meet the requirements of residence and the attribution of production in an economy as described in Chapter 2.
Transactions/positions/income between related financial intermediaries
432. The Benchmark Definition recommends that all inter-company flows – with the exception of those pertaining to equity finance – between certain types of related financial intermediaries should be excluded from FDI financial and income transactions and positions. Deposits and other amounts lent by a financial intermediary to its financial intermediary subsidiary or associate located abroad, as well as deposits and other borrowings between such companies, should be classified as "other investment" rather than direct investment. A similar treatment applies to investment between fellow enterprises that are financial intermediaries. Debt securities between related financial intermediaries are classified as “portfolio investment”.
433. The definition of the scope of enterprises included under “financial intermediary” should be equivalent to the 2025 SNA definitions. The SNA classifies financial corporations under three categories, namely financial intermediaries, financial auxiliaries and other financial corporations. Financial intermediaries are institutional units that incur liabilities on their own account for the purpose of acquiring financial assets by engaging in financial transactions on the market. Financial auxiliaries are institutional units principally engaged in serving financial markets, but that do not take ownership of the financial assets and liabilities they handle. Other financial corporations are institutional units providing financial services, where most of their assets or liabilities are not available on open financial markets.
434. The financial corporation sector can be divided into nine sub-sectors according to its market activity and the liquidity of its liabilities (see Annex 2.C). Hence, non-equity transactions/positions (and investment income) between two related financial intermediaries that are part of deposit-taking corporations, money market funds (MMFs), non-MMF investment funds or other financial intermediaries, except insurance corporations and pension funds (ICPFs), would be excluded from FDI. While they are financial intermediaries, insurance corporations and pension funds are not treated in the same manner as other financial intermediaries for the purposes of this exclusion.
435. Annex Table 6.A.1 summarises the transactions/positions included in or excluded from FDI statistics for financial corporations:
Source: OECD (2009[3]), OECD Benchmark Definition of Foreign Direct Investment, Fourth Edition (BD4), https://doi.org/10.1787/9789264045743-en.
Financial derivatives
436. A financial derivative contract is a financial instrument that is linked to another specific financial instrument or indicator or commodity and through which specific financial risks (such as interest rate risk, foreign exchange risk, equity and commodity price risk, etc.) can be traded in their own right in financial markets. There are two broad types of financial derivatives: forwards (including futures and swaps, other than gold swaps) and options. Financial derivatives are excluded from FDI statistics.
437. The purpose of this annex is to assist compilers to identify some common types of special purpose entities (SPEs) relationships. The identification of SPEs allows foreign direct investment (FDI) compilers to segregate transactions and positions that hardly affect domestic economic activity and do not reflect genuine investment activities in or of the reporting economy itself.
Source: OECD (2009[3]), OECD Benchmark Definition of Foreign Direct Investment, Fourth Edition (BD4), https://doi.org/10.1787/9789264045743-en.
438. Referring to Annex Figure 6.B.1, this Benchmark Definition addresses problems with the interpretation of FDI statistics caused by SPEs as follows:
1. for the standard presentations:
a. economy 2 should report separate transactions and positions vis-à-vis economy 1 (inward) and 3 (outward) of its resident SPEs and non-SPEs;
b. economies 3 and 4 should report their inward investment positions by the ultimate investing economy (economy 1);
2. for supplementary presentations:
a. economy 1 is recommended to reallocate its outward investment position in the SPE in economy 2 to the ultimate host economies of 3 and 4;
b. economy 2 is recommended to report its aggregate FDI statistics according to the residency of the ultimate investor to estimate the upper bound of the total pass-through funds.
439. FDI statistics under points (1) and (2) above are based on the directional principle to facilitate the economic analysis of foreign direct investment.
440. The next section of this annex will provide some guidance for the implementation of the definition of SPEs. The last section will provide a few examples of SPEs in ownership structures that may pose difficulties for compilers and explain the recommended recording (IMF, 2022[5]).
441. An entity must meet six criteria to be classified as a SPE (Box 6.2). This section will consider each of these criteria and offer some advice for applying them.
A formally registered or incorporated resident institutional unit
442. There are two aspects to this criterion: i) that it be resident in the economy, and ii) that it be an institutional unit. As SPEs have little or no physical presence, their residence is determined according to the economic territory under whose laws it is incorporated or registered (BPM7, Chapter 4, Section 2, paragraph 5.86). An institutional unit is responsible and accountable for the decisions and actions that it takes. Consequently, SPEs that do not meet the criterion of autonomy of decision should in principle be regarded as part of the parent and be consolidated with it for statistical purposes. However, BPM7 (Chapter 4, Section 2, paragraph 5.93) states that “An entity of this type that cannot act independently of its parent and is simply a passive holder of assets and liabilities… is not treated as a separate institutional unit unless it is resident in an economy different from that of its parent… .” Therefore, the residence of a SPE and its parent is a critical determinant of its statistical treatment. If a SPE is resident in the same economy as its parent, even if indirectly foreign-controlled, then it will likely not meet the definition of a statistical unit and would be consolidated with its parent. This case is discussed further in Section ‘Types of SPEs relationships’ below.
Directly or indirectly controlled by a non-resident
443. While FDI covers both influence and control relationships, for identifying SPEs, the entity must be controlled by the direct investor. Control is determined to exist if the non-resident direct investor owns more than 50% of the voting power in the DIE either directly or indirectly through a chain of FDI relationships following the FDIR.
Objectives of the SPE
444. The definition of a SPE requires that the SPE be established to benefit from advantages offered by the host economy. Specifically, the direct investor is seeking to achieve at least one of four objectives in establishing the SPE to:
gain access to capital markets or sophisticated financial services
isolate owners from financial risk
reduce tax or regulatory burden
safeguard confidentiality of transactions and owner(s).
445. To apply this criterion, the compiler will likely need to rely on their expert knowledge of the legal and institutional setting in their economy; it may also be useful to discuss with regulatory and supervisory authorities to determine if their economy offers these advantages to foreign investors. In addition, non-governmental organisations, international organisations, and private sector companies may publish information identifying the attractiveness of various economies as hosts for SPEs.
Employment up to five employees
446. A maximum of five employees is used in the definition to establish that the SPE has little or no employment; this limit applies both to those directly employed and those employed through employment service agencies. This criterion is easy for compilers to implement if they have access to information on employment. However, some care should be taken to the stability of the SPE population in applying this limit. For example, if an entity that has five employees and meets the other criteria to be classified as a SPE temporarily increases its employment to six employees, it can still be classified as a SPE as long as it has no more than five employees most of the time.
Little or no physical presence or production
447. Physical production is any activity that uses inputs, such as employees, physical capital, and goods and services to produce outputs that are delivered or supplied to other institutional units. Little physical presence implies the absence of physical assets, such as property, plant, equipment, and sub-soil assets, in the host economy. The physical presence of a SPE is usually expected to be limited to having a physical address, electronic communication addresses, and, if legally required or otherwise necessary, small-scale premises.
Transacts almost entirely with non-residents
448. The purpose of this criterion is to identify entities that have limited links to the domestic economy. For SPEs that engage in pass-through, this criterion implies that the majority of their financial assets and liabilities will be with non-residents (the case of a chain of SPEs is considered in Section ‘Types of SPEs relationships’ below). For other types of SPEs, such as royalty and license fee companies, captive insurers, those engaged in merchanting, and those engaged in operational leasing, this criterion may be more difficult to apply as it requires an understanding of the business model of the SPE.
SPEs in direct relationships with non-resident entities
449. Many SPEs are directly foreign-owned and only have assets and liabilities vis-à-vis non-residents, such as enterprise B in Annex Figure 6.B.2. In this case, the classification as a SPE is straightforward.
Resident chains of SPEs and mixed groups of SPEs and operating entities
450. The part of a chain of FDI passing through a country may also consist of two or more entities as illustrated in Annex Figure 6.B.3. While the entity B1 meets the criteria to be a SPE, it holds both a SPE (B2) and an operating entity (B3) in the domestic economy (economy 2). Therefore, its classification as a SPE will depend on a number of factors. If the chain consisted solely of the two SPEs B1 and B2, then the classification of B1 as a SPE would be straightforward because the assets it holds in B2 are not relevant to the domestic economy. When the chain also includes an operating entity like B3, then one of the key factors in determining if B1 should be classified as a SPE is the composition of its balance sheet; the relative importance of the resident and non-resident assets in its total assets are one consideration for the compiler. Another important consideration is the extent to which the operating entity has links with the domestic economy. If the compiler finds few links between B3 and the domestic economy and/or that the resident assets are only a small proportion of the balance sheet, then B1 should be classified as a SPE.
451. In the case where B1 is an operating entity rather than a SPE, then it is likely that B2 would be consolidated with it, as B2 would not meet the criteria to be considered a separate institutional unit.6 Nevertheless, compilers are encouraged to identify the cross-border positions and transactions of B2 separately from its parent B1, if it is possible to do so, and to classify them as SPE-related to provide a more complete picture of the cross-border activities of SPEs.
452. From the above examples, it may be concluded that for an appropriate classification as SPEs of enterprises owned by non-residents and related with each other in chains or groups, compilers may want to use additional criteria. In the case of resident chains of entities, compilers should take into account whether assets in resident entities, such as those of B1 and B2, are really relevant to the domestic economy. If not, these enterprises would be considered as potential SPEs.
453. The term collective investment institution (CII) generally refers to incorporated investment companies and investment trusts, as well as unincorporated undertakings (such as mutual funds or unit trusts), that invest in financial assets (mainly marketable securities and bank deposits) and/or non-financial assets using the funds collected from investors by means of issuing shares/units (other than equity). Other terms referring to CIIs may also be used, e.g., collective investment scheme, collective investment vehicle, collective investment undertaking, and in certain cases, investment fund.
454. The CII can be open-ended or closed-ended. If open-ended, there is no limit to the number of shares/units on issue and the shares/units can be, at the request of the holders, repurchased or redeemed directly or indirectly out of the undertaking’s assets. If closed-ended, the number of shares/units on issue is fixed and investors entering or leaving the fund must buy or sell existing shares. The shares/units can be quoted or unquoted. The CII may pay periodic dividends, capitalise the income or a combination of those approaches, depending on the terms set out in its prospectus.
455. CIIs may be constituted: (i) under the law of contract (as common funds managed by management companies); (ii) under trust law (as unit trusts); (iii) under a statute (as investment companies); or (iv) otherwise with similar effect. Some CIIs invest in other similar vehicles (e.g., “funds of funds”). It should be noted that occupational pension funds are not CIIs; they are part of the insurance companies and pension funds subsector.
456. The prospectus also sets out the investment policy or strategy of the CII in terms of the types of financial instrument or other assets to be acquired (such as bonds, equities, real estate) and their geographical (e.g., Asian, American, European, emerging markets) or currency make-up, or any other investment strategy. CIIs may have different objectives. For example, some CIIs may target capital growth and be generally characterised as “growth” funds, whereas others may be structured to produce a regular stream of income, through interest and/or dividends and be characterised as “income” funds. The prospectus indicates whether the CII can undertake any leveraging activity and how it might do so (e.g., through straight borrowing, through repurchase agreements, or by using financial derivatives). The terms and conditions for redemption (for open-ended schemes) or the means of sale (for closed-ended schemes) are also set out in the prospectus.
457. CIIs are generally “brass plate” enterprises and are managed by professional investors who may offer a variety of funds with their own market orientation and who make investment decisions on behalf of investors. Administration, management, custodial and trustee services may be provided to the CIIs by separate service providers, some of whom may be located in different jurisdictions. For example, a CII may be domiciled or registered to operate in one economy and may be managed by a service provider in another economy with custodial services being located in a third economy. The location or domicile of a CII may be decided by its promoter in the context of optimising its tax liabilities or of utilising preferential securities regulation arrangements in the process of asset management. Cross-border transactions may, and frequently do, occur.
458. At the outset, CIIs may be classified broadly into two main categories: (a) ordinary or retail type entities and (b) other types of CII. The ordinary/retail type CII may in turn be a contract type or a corporate type. A contract-type retail CII is generally a mutual fund or unit trust or similar vehicle, while a corporate-type CII may be an investment trust or a corporation acting as an investment vehicle. In both cases, the assets acquired tend to be highly liquid and investment subscriptions (and, later, redemptions) are made by a large number and variety of investors. The investment is made generally to gain a short-term interest, via contract-type or corporate-type investment trusts, in portfolio securities. There may be significant diversification of the asset classes, for example, mainly by economy/economic zone, by type of financial instrument (equity, bonds, money market instruments, derivatives, etc.).
459. There is a variety of other types of CII that cannot be considered as retail in nature. The investors into these CIIs can vary from private households to corporate entities and the asset classes acquired can also vary considerably from liquid to illiquid types. Their legal structures, trading practices, investment policies and operations generally need to be considered by compilers to decide the classification for balance of payments statistical purposes of the entities involved and their investment and other economic transactions.
460. A number of different examples of financial institutions fall within the scope of the description of collective investment institution (CII).7 Included are:
investment funds
mutual funds
unit trusts
variable capital companies
investment limited partnerships
feeder/master funds, umbrella funds/sub-funds, funds of funds
hedge funds
professional investor funds
private equity funds
distressed funds
property and real estate funds
money market funds.
461. It must be stressed that the examples of CII listed above can exist under the same or different labels in different economies. The name or label used may give some indication of the type of scheme or its type of investment orientation, or of the main objective of the investment vehicle used (e.g., venture fund, futures and options capital protected fund, bond fund, balanced fund, growth fund, index fund, sector fund, international or regional funds, junk bond fund). However, caution is needed in interpreting from its title or label what a CII is and how it operates.
462. An investment fund is a CII that issues shares (if a corporate structure is used) or units (if a trust structure is used) to the public, and that invests the capital raised in financial and/or non-financial assets.
463. A mutual fund can be defined as an entity that issues shares or units, which are purchased by investors. The basic scheme of a mutual fund is quite common in many other CIIs. The subscriptions collected may be invested in different types of assets (non-financial as well as financial) and the investors may receive either regular income or, at redemption, holding gains (or losses) or a combination of both. The mutual fund can be open-ended or closed-ended, and the shares/units can be quoted or unquoted.
464. A unit trust operates as a CII established under a trust deed made between the fund’s management company and its trustee. The legal structure of a unit trust may vary among jurisdictions but, in general, it would appear that the trustee acts as the legal owner of the fund’s assets on behalf of the group of investors who are each entitled to an undivided beneficial interest in the fund. Similar to shareholders in an investment company, the unit holders are entitled to attend and vote at meetings on matters affecting the fund. The trust deed is the primary legal document that constitutes the trust, and it sets out the various rights and obligations of the trustee, the management company and the unit holders.
465. A variable capital company is normally set up to invest its funds and property with the aim of spreading investment risk. It is managed by a management company for the benefit of its shareholders who enjoy limited liability status. The characteristics of variable capital companies are that they can repurchase their own shares and that the issued share capital must at all times be equal to the net asset value of the underlying assets.
466. An investment limited partnership is a partnership of two or more persons having as its principal business the investment of its funds in financial and non-financial assets of all kinds and consisting of at least one general partner and at least one limited partner. The limited partner is equivalent to the shareholder in a company whilst the general partner is generally the equivalent of the management company in a unit trust. The investment limited partnership generally does not have an independent legal existence in the way that a company does. All of the assets and liabilities belong jointly to the individual partners in the proportions agreed in the partnership deed. Similarly, the profits are owned by the partners.
467. Feeder/master funds, umbrella funds and funds-of-funds are set up with one of the main objectives being the facilitation of access by investors to greater asset diversification than would be available through the more conventional CIIs. All types of structures exhibit the characteristic that one fund invests in one or more other funds. The arrangement must meet the statutory regulatory requirements of the authorities where the funds are domiciled (i.e., legally registered). In this context, the different funds in a particular investment arrangement may be domiciled in different jurisdictions and may also have different legal structures. In general, the country of registration of the entity concerned is taken to be the country of domicile of the fund.
468. Hedge fund is a term that covers a heterogeneous range of CIIs, typically involving high minimum investment projects, light regulation, and a wide range of investment strategies that tend to involve assets that are highly illiquid. The range of strategies include: hedging against market downturns; investing in asset classes such as currencies or distressed securities; and utilising return-enhancing tools, such as leverage, derivatives, managed futures and arbitrage (e.g., bonds, stocks and risk arbitrage). Many hedge funds target consistency rather than magnitude of return as their primary goal.
469. A professional investor fund is a fund authorised to require a relatively high level of minimum subscription from eligible investors. Borrowing restrictions applied to other more common CIIs may also be relaxed.
470. A private equity fund is established to enable partnerships of qualifying individual investors or groups consisting of up to 100 qualifying individual investors to participate. Such funds can include venture capital funds, as well as buying-out funds, whose investors tend to be mainly institutional investors acquiring and selling shares to gain a long-term interest through participation in the control or management of an enterprise for a specified period to enhance its value.
471. Distressed funds are established to invest at deep discounts in equity, debt, or trade claims, of companies undergoing or facing bankruptcy or reorganisation.
472. In a property or real estate fund, the term property is generally defined as a freehold or leasehold interest in any land or building, with a specified minimum unexpired lease period. Partly paid shares may be issued.
473. A money market fund (MMF) is a very special type of CII. It issues shares or units to the public that are, in terms of liquidity, close substitutes for deposits, and it invests the proceeds primarily in low-risk short-term money market instruments (treasury bills, certificates of deposit, and commercial paper), MMF shares/units, as well as in bank deposits and instruments that pursue a rate of return that approaches the interest rates of money market instruments. In some jurisdictions and depending on their set-up constraints, some MMFs are classified as monetary financial institutions.8
474. It should be noted that occupational pension funds (i.e., those schemes sponsored by employers on behalf of their employees) are generally not regarded as CIIs. The main reasons for this are: firstly, the liabilities of occupational pension funds tend to be long-term and are in the form of insurance technical reserves; the liabilities of CIIs are in the form of equities, shares or units. Secondly, there are very rigid conditions concerning the encashment by investors (employees principally) of their pension funds, e.g., the assets of these funds generally cannot be realised by the individual investor until he/she reaches a certain retirement age. An investor in a CII can cash in his/her investment whenever desired.
475. In general, on the liability side, investment into CIIs involves the pooling of the investment contributions of a number of investors so that the units/shares acquired by each individual account for a very small proportion (generally much less than 10%) of the total number of units/shares issued by the scheme. There are some CIIs, however, that may have a small number of investors (e.g., property/real estate funds, private equity funds, distressed funds) and the investors in individual cases may contribute sizeable proportions (more than 10%) of the overall capital invested. A CII may have resident investors, non-resident investors or a mixture of both. Nevertheless, the collective legal nature of the CII, that is, more than one investor, suggests that no single foreign resident has direct control of the investment strategy and day-to-day operations of the fund. Therefore, investment in a CII should always be classified as portfolio investment even if the voting power of a non-resident investor exceeds the 10% threshold.
476. On the assets side, a CII can invest in both resident and non-resident financial securities or other assets. There tends to be very little or no concentration of the pooled investment into significant proportions (10% or more) of the shares issued by individual enterprises that are acquired by the CII. In the case of FDI by a CII, the fund as a statistical unit should be viewed as the investor that takes investment decisions, suggesting the possibility of control over the strategic and day-to-day operations of the individual businesses it owns. Therefore, if the voting power of a non-resident enterprise by a CII exceeds the 10% threshold, it should be classified as direct investment.9
477. Pass-through funds in foreign direct investment (FDI) are defined as financing that passes through a direct investment enterprise resident in an economy to an affiliate in another economy, so that the funds do not stay in the economy of the first enterprise (BPM7, Chapter 6, Section C, paragraph 6.33). Much pass-through funding is an outcome of the complex ownership structures of multinational enterprises (MNEs). While much attention has been placed on the role special purpose entities (SPEs) play in pass-through funds, the definition would capture funds flowing through any DIE, including an operating affiliate, in one economy to a DIE in another economy. When MNEs channel investment through several economies, FDI financial transactions, income and positions are inflated as each flow into and out of each economy is counted even if the capital, or income, is just passing through.10 Moreover, this behaviour can further obscure the ultimate sources and destinations of FDI when the statistics are compiled by immediate partner economy.
478. The next section of this annex gives some examples of the ways MNEs pass funds along their ownership structures, including the motivations for pass-through funds. It identifies those types of pass-through funding associated with SPEs and those that are more likely to happen through operating affiliates. It concludes with a section on how to differentiate among the various types of pass-through funds. Annex Figure 6.D.1 illustrates an ownership structure that would lead to pass-through; this figure will be used to illustrate some of the examples of pass-through discussed in this annex.
479. Pass-through transactions and positions in FDI can emerge in several different ways and take various forms. This section begins to develop a taxonomy of pass-through in FDI by focusing on the motivations and features of pass-through funding, the characteristics of the economies that attract that type of pass-through, and its likely impacts on the economy hosting the entities that funds are passing through. Much of this discussion draws on (Lewellen and Robinson, 2013[6]). Pass-through funds can take different forms and serve different purposes that are discussed below.
1. Tax minimisation. This is probably the most cited reason for pass-through funds. MNEs can channel funds through affiliates in different economies to both shift income as well as take advantage of opportunities to defer income. This activity can be done through both SPEs and operating affiliates. The importance of tax minimisation strategies through operating affiliates might be growing as a result of initiatives to encourage MNEs to better align where they report income with where they have economic activities, such as the G20/OECD framework to address base erosion and profit shifting.
Pass-through funds due to tax minimisation would be associated with economies that offered tax advantages, including special tax treatments for intellectual property. These economies could offer low tax rates and/or access to a network of taxation treaties. This type of pass-through can also impact the financial mix used by the MNE in cases where debt-financing offers tax advantages not available to equity investment. This type of pass-through funding would often be observed in entities that are characterised by high asset to employment ratios and/or significant royalty and license fee income, but it could also be associated with certain types of entities, such as holding companies. Pass-through funds associated with tax minimisation are likely to result in significant FDI positions for the economies involved but have less direct impact on the real economy (that is, lower levels of employment, value added, and tangible capital in the foreign-owned entities). However, it can have considerable indirect impacts on the host economy by supporting industries offering services to foreign investors, such as financial services and tax planning industries; these indirect impacts can be particularly important in small economies. In Annex Figure 6.D.1, enterprise C could be an example of an entity established to help minimise taxes, especially if economy 3 offered tax advantages.
2. Managing expropriation or other risks to the value of their investment. MNEs that are worried about “political” or expropriation risk or that wish to benefit from an advantageous protection regime might decide to hold that investment indirectly through an economy that offers treaty-based protections to investors. For example, if the investor’s home economy does not have a bilateral investment treaty or equivalent arrangement with the host economy, they may choose to hold the investment indirectly through one of their DIEs in an economy that does have such a treaty with the host economy of the investment. This type of pass-through could be carried out through a SPE or an operating affiliate. As a result, this type of pass-through funds would be more likely to occur in host economies with a large network of bilateral investment treaties or in economies that have treaties with unconventional partners. It would also likely have limited direct impacts on the host economy but could possibly have indirect impacts by supporting an industry offering services to foreign investors.
3. Reduction in transaction and/or administrative costs through subgroup structures. Affiliates in the MNE’s production network that have a significant number of interactions may forge closer ownership links to reduce transaction costs and further strengthen such interactions. For example, a MNE investing in the United States may invest in Mexico and Canada through their U.S. operation, forming an integrated, regional operation. While not necessarily thought of as in the same class as the financial activities discussed above, this still involves the passing of funds through intermediate links.
This type of arrangement would be more likely to happen with more complex production processes and products, where the transaction costs would be expected to be higher. This type of pass-through capital, in which the MNE could establish regional headquarters, would be most likely to happen between DIEs with strong commercial links; thus, it is more likely to happen between affiliates in economies that share strong commercial ties as evidenced by significant trade flows and that share preferential trade agreements. It would also be more common in economies that are physically or culturally close. The entity serving as host for the regional headquarters may be expected to have a highly-skilled labour force, be better able to organise production across economies, and to conduct research and development (R&D) on their own behalf. Headquarter services, such as financial planning, may be a more important part of the activity of the regional headquarters. This form of pass through could also be based on specific corporate functions (such as R&D, marketing, or logistics). Enterprise D in Annex Figure 6.D.1 could be an example of this type of pass-through if it helps coordinate activities between itself and G and H.
With this type of pass-through funds, which are most probable to happen through operating affiliates, the financial structure overlaps with the reporting and operational structure within the MNE. As such, there is likely to be more direct impact in the host economy as the regional headquarters will have significant employment, value added, R&D, and trade in both goods and services. The regional headquarters are likely to have indirect impacts on the host economy in a number of areas, including integrating domestic suppliers into the production networks they coordinate.
4. Inherited pass-through structures. When a MNE acquires an existing MNE through a merger or acquisition (M&A), it also acquires the ownership structure and any existing pass-through positions within the acquired MNE. While the acquiring MNE can choose to change such structures either immediately or over time, it may not do so. As a result, the acquisition will create an ownership chain involving existing links through which funds are passing. This form of pass-through would be more likely to occur in economies where more of the inward FDI transactions were the result of M&A rather than of greenfield investment and, more specifically, where inward investment was in an economy hosting the MNEs that have since been acquired. This type of pass-through is likely to happen through an entity that has become an operating affiliate following the M&A, but it is hard to assess the impact of this type of pass-through capital on the host economy as it depends on what happens to those affiliates after the M&A has been completed. In Annex Figure 6.D.1, this type of pass-through could result if enterprise B had been a MNE that owned foreign affiliates D, G, and H that was then acquired by enterprise A.
5. Protection of the parent from claims against the affiliate. If the parent is concerned that a DIE is exposing them to financial claims, they may be more likely to own it indirectly to limit those claims. This might be the case, for example, with joint ventures or other cases of shared ownership. This might be more likely to happen in host economies that provide greater protections to investors as discussed above. This type of pass-through can happen through a SPE and is likely to be less associated with direct impacts on the host economy, in such terms as employment, but could have other impacts, such as generating tax revenue.
6. Intra-group lending centres. The MNE can use a DIE in one economy to borrow and lend funds to other parts of the MNE. These entities can draw on equity to lend funds to other parts of the MNE (either equity capital or reinvested earnings). They can also intermediate in loans between affiliated parties. Finally, they can raise financing from outside of the enterprise; for example, MNEs can use their foreign affiliates to raise capital by issuing debt securities or taking loan(s) and then channel the funds raised to other parts of the MNE, including back to the parent; these entities are often a type of SPE called a financial conduit. The first part of this transaction is either domestic, or either portfolio or other investment, but only the second part is a FDI transaction. In this case, pass-through capital goes beyond FDI. This form of pass-through funds would be more likely to happen in economies with deep capital markets or that offer offshore or sophisticated financial services. It would not have significant direct impacts in the host economy (e.g., in terms of employment) but could have significant indirect impacts in the financial services sector. In Annex Figure 6.D.1, enterprise C could be an example of this type of pass-through as it borrows from and lends to other affiliates.
7. Private financing vehicles. A private investor may establish an affiliate in a foreign economy for the purpose of engaging in portfolio investment from the host economy. It is the first leg of this case that brings the transaction within the scope of FDI, while the other part would be in the other functional categories of international investment. This kind of pass-through capital would be more likely to happen in economies that offer offshore or sophisticated financial services or secrecy. The home economy would likely be one with higher tax rates and that had stronger controls on outflows of portfolio capital than on direct investment. It would not have significant direct impacts in the host economy but could have significant indirect impacts in the financial services sector.
8. Corporate re-domiciliations. Corporate re-domiciliations often involve substantial FDI financial transactions that are almost completely offset by portfolio investment flows. It is possible that these kinds of transactions, when they result in offsetting flows, could be treated as a form of pass-through funding. In this case, the characteristics that would make an economy likely to host a re-domiciliation would be those offering relative tax, regulatory, and other benefits to foreign investors. The extent of the impact on the host economy would depend on the extent to which the redomiciled company actually shifted operations to its new headquarters economy.
Differentiating types of pass-through funds
480. In existing data collection systems, it is not possible for the statistician to ask the motivation behind the pass-through funds to determine to what extent it aligns with one of these categories listed above. It could be that pass-through funds serve several different purposes within the MNE structure at the same time. As such, it is difficult to completely differentiate between all of the different types of pass-through funding. In this Benchmark Definition, the separate identification of financial transactions, income, and positions of resident SPEs is considered a lower bound on pass-through funds and is likely associated with pass-through that has fewer direct impacts on the host economy. As discussed in Chapter 8, the identification of financial transactions, income, and positions based on the residency of the ultimate controlling parent can be considered an upper bound estimate of pass-through funds and would encompass all forms of pass-through.
← 1. The definition and the geographical coverage of offshore centres may differ depending on the purpose of the analysis. This Benchmark Definition does not apply any specific description of offshore centres. The focus is rather the entity that qualifies as direct investor or direct investment enterprise (DIE) and not the geographical location.
← 2. The fourth edition of the Benchmark Definition of Foreign Direct Investment (OECD, 2009[3]) included a recommendation for supplementary series of FDI financial transactions, income, and positions looking through non-resident SPEs. This recommendation has been dropped in this edition because no OECD members disseminated these series, but this edition does include a recommendation for a supplementary series of FDI positions by ultimate host economy, which is discussed further in Chapter 8.
← 3. The creation or relinquishment of territorial enclaves in the rest of the world should not be included in direct investment.
← 4. Both adjustments should be recorded under “volume changes” in the international investment position. See Chapter 5 and Annex 5.A for more details.
← 5. See the BPM7 (IMF, Forthcoming[1]), for the treatment of such bonus payments when no DIE is or will be established.
← 6. Most SPEs owned and directly controlled by residents in the same economy would not meet the statistical definition of an institutional unit and, thus, for statistical purposes would be consolidated with those of the domestic parent.
← 7. The types of collective investment institutions described in this annex are for readers' information and mainly based on the experiences of the editor of the fourth edition of the Benchmark Definition of Foreign Direct Investment (OECD, 2009[3]); thus. details or names of each type as well as legal structures and so on could differ among countries.
← 8. Monetary financial institutions cover two types of entities: (a) credit institutions whose business is to receive deposits or other repayable funds from the public and to grant credit for their own account, and (b) other financial institutions whose business is to receive deposits or close substitutes for deposits from entities other than monetary financial institutions and to grant credit for their own account or to make investment in securities.
← 9. Unless it is an investment in a CII, in which case it is recorded in portfolio investment per the guidance given above.
← 10. However, the balance, that is, the net FDI financial transactions, net FDI income, and net FDI positions would not be affected as the funds passing through the economy would be netted out.