A. Residence

The residence of each institutional unit is within the economic territory with which it has the strongest connection, expressed as its center of predominant economic interest.⁴ Each institutional unit is a resident of one and only one economic territory⁵ determined by its center of predominant economic interest. Most entities have strong links with only one economy, so their residence is clear, but with increasing international economic openness, a growing number of institutional units have connections to more than one economy. The concepts of economic territory and residence are designed to ensure that each institutional unit is a resident of a single economic territory such that cross-border transactions and positions of each institutional unit are recorded in a single economic territory.

For entities such as SPEs, the residence is determined by their place of incorporation (BPM6, paragraph 4.115d). SPEs require particular attention because for an entity with little or no physical presence, its residence is determined according to the economic territory under whose laws it is incorporated or registered (legally constituted or legally domiciled), consistent with paragraph 4.56 of the System of National Accounts 2008 (2008 SNA). According to international statistical guidelines, residence is not based on nationality or legal criteria, contrary to the concepts of residence used in many countries for exchange controls, tax, or other purposes. The concept of residence for SPEs⁶ in these guidelines follows the BPM6 and 2008 SNA standards.⁷ BPM6 paragraph 4.135 clearly states that the incorporation and registration of such entities represent a substantial degree of connection to the economy, associated with jurisdiction over the enterprise's existence and operations. In contrast, other connections such as ownership, location of assets, or location of managers or administration may be less clear-cut.

B. Institutional Unit

An institutional unit is an economic entity that is capable, in its own right, of owning assets, incurring liabilities and engaging in economic activities and in transactions with other entities.⁸ Because they have legal responsibility for their actions, institutional units are considered to have autonomy of decision for all aspects of economic behavior. The concept of an institutional unit is the same in BPM6 and the 2008 SNA.

However, there are some special treatments of SPE units in cross-border situations. The general rule, according to the 2008 SNA paragraph 4.69, is that for being considered as separate institutional units, they have to be responsible and accountable for the decisions and actions they take, that is, they should have autonomy of decision. Consequently, SPEs which 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, paragraph 4.61 of the 2008 SNA states that [emphasis added] “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… .” The residence of an SPE is, therefore, of critical importance to its statistical treatment. Also, because SPEs have legal responsibility for their actions, they can be accountable for all aspects of economic behavior.

Therefore, for the purpose of operationalizing the definition, the residence of the SPE is an important factor for its recognition as a separate institutional unit. Based on the current statistical methodological standards (2008 SNA and BPM6), SPEs controlled by nonresidents meet the definition of an institutional unit and are treated as residents of their territory of incorporation. Although SPEs are controlled by nonresidents and their autonomy of decision could be questioned, to the extent that they are legally independent, they are considered to be separate institutional units from their nonresident parents. The 2008 SNA, paragraph 2.16, does recognize that legal independence to hold assets and liabilities and autonomous behavior do not always coincide such that SPEs are treated implicitly as responsible, and accountable, for their decisions and actions.

SPEs owned and directly controlled by residents in the same economy, even if indirectly controlled by nonresidents, typically would not meet the statistical definition of an institutional unit. In practice, their accounts would be consolidated with those of the resident owners, which may or may not be an SPE. In these instances, the cross-border transactions of the SPE will be consolidated with the resident parent and not be recorded as SPE-related cross-border flows. However, from the perspective of the need to untangle SPE activities from those of other economic agents, the Guidelines encourage compilers, to the extent possible, to separately identify the cross-border transactions of these resident SPEs (indirectly controlled by nonresidents) from the resident non-SPE immediate parent. This guidance to compilers is only meant to identify separately SPEs related cross-border transactions or positions from the rest and should not be viewed as an exception to the definition of an institutional unit as laid out in the 2008 SNA. Figure 2 illustrates an example.

In Figure 2, B.2 and B.3 resident in the same economy B as their non-SPE parent entity will be consolidated with B.1 for macroeconomic statistics purposes, if both do not qualify as separate institutional units as defined in the 2008 SNA. Subsequently, the cross-border transactions and positions of SPE B.3 will not be separately identified, as the resulting consolidated entity would not meet the definition of SPEs. For the purpose of ESS, compilers are highly encouraged, to the extent possible, to consider the cross-border flows and positions of B.3 as SPE-related, separate from their resident non-SPE parent, B.1. This approach is considered more useful in understanding SPE cross-border activities, particularly in those jurisdictions where SPE activities get combined with non-SPE operating units. It is recognized, however, that in this approach foreign funds that will flow into B.3, through B.1, will not be captured as SPE flows, while the outflows from B.3 will.

View Full Size

SPEs within Enterprise Groups

Citation: Technical Notes and Manuals 2022, 006; 10.5089/9798400231490.005.A001

• Download Figure
• Download figure as PowerPoint slide

C. Direct or Indirect Control of SPEs by Nonresidents: How to Apply It in Practice?

Control or influence may be achieved directly by owning equity that gives voting power in the enterprise, or indirectly by having voting power in another enterprise that has voting power in the enterprise.⁹ BPM6 and the Benchmark Definition of Foreign Direct Investment, fourth edition (BD4) of the Organisation for Economic Co-operation and Development (OECD) define DI based on control or significant degree of influence on the management of an enterprise.

For SPEs, associated particularly with flexible corporate structure of MNEs, the focus is on the control aspect of the direct investor. According to the definition of SPEs, these entities should be directly or indirectly controlled by nonresidents. Control is determined to exist if the direct investor owns more than 50 percent of the voting power in the DI enterprise either directly or indirectly through a chain of DI relationships.

Direct and Indirect Relationship

Control may be obtained directly (for example, company A is the owner of equity of company B that gives control) or indirectly (company A owns equity of company B and company B owns equity of company C, so company A exercises indirect control on company C though company B).

Figure 3 illustrates an example in which Company A controls company B and, because company B controls company C, company A also controls company C. In this example, company A directly controls company B, company B in its turn directly controls company C, and thus company A indirectly controls company C. Indirect control can continue through a series of companies as long as control exists at each stage in the ownership chain.

View Full Size

Indirect Relationship

Citation: Technical Notes and Manuals 2022, 006; 10.5089/9798400231490.005.A001

• Download Figure
• Download figure as PowerPoint slide

Framework for Direct Investment Relationships

The framework for DI relationships (FDIR) sets the rules for indirect transmission of control and influence through a chain of ownership.¹⁰ Control can be passed down a chain of ownership as long as control exists at each stage or when the investor and its subsidiaries combined own more than 50 percent of the voting power of an enterprise (see cases 4 and 6 in Figure 4).¹¹ Whereas the FDIR applies a criterion of 10 percent or more of voting power for immediate DI, transmission through chains of ownership is not linked to a particular equity share, but a chain of control. For example, a chain of ownership of enterprises with each link involving 60 percent of the voting power involves a chain of control, even though the indirect equity by the top enterprise is 36 percent at the second level (that is, 60 percent of 60 percent), 21.6 percent at the third level (that is, 60 percent of 36 percent), and so on. The application of these principles is explained in BPM6—see Box 6.1.

The flow chart in Figure 4 provides examples of ownership and control, identifying SPEs that meet the definition. The ownership and control of SPEs is liable to the jurisdiction of incorporation of the SPE such that the latter will be required to meet the specific legal, tax, accounting, and regulatory requirements of the relevant jurisdiction. The SPE should be directly or indirectly controlled by a nonresident as emphasized and presented in Figure 4.

SPEs can also exist within a chain of entities, where the immediate parent and the affiliated SPEs are operating in the same economy. Case 5 presented in Figure 4 depicts this chain and warrants special attention. It showcases the presence of intricate chains involving some sort of layering whereby a resident SPE controlled by nonresidents establishes another resident SPE or there may be a mixture of SPEs and non-SPEs in the chain.

In Case 5, there is a group of SPEs in the same enterprise group controlled by a foreign investor. Units B.52 and B.53, as shown in Figure 4, are directly controlled by residents B.51 (operating unit) and B.52, respectively. The operating SPE unit B.51 is, on its part, controlled by a nonresident entity A.5. Consequently, both units B.52 and B.53 are indirectly controlled by nonresident A.5. B.52, in this case, has a resident parent and has another resident affiliate (B.53). Whereas units B.52 and B.53 have domestic equity capital and B.51 has domestic equity assets, they all may have mostly cross-border assets and liabilities, with very little interaction with the economy of incorporation. In such instances, units B.52 and B.53 would be consolidated with B.51, and the resulting consolidated entity would be classified as an SPE for ESS purposes. This is because all three units altogether fulfil the SPE criteria as described in the definition and have their operations dominantly related to the rest of the world.

View Full Size

Ownership and Control of SPE

Citation: Technical Notes and Manuals 2022, 006; 10.5089/9798400231490.005.A001

Source: IMF Committee on Balance of Payments Statistics (2018); prepared by the ECB.

• Download Figure
• Download figure as PowerPoint slide

In case unit B.51 is a non-SPE in the above example but units B.52 and B.53 meet the definition of SPEs, the guidance provided in paragraphs 18 and 19 should be followed.

National compilers, have recourse to different sources to be able to determine direct/indirect control of entities, of which:

• a. Business registers in the context of ESS compilation (DI enterprises) with information on entities with nonresident participation. The adequacy of the registers should, however, be carefully refined when targeting SPEs.

• b. Regulatory institutions, which license the SPEs. Usually when incorporated, SPEs should provide information about their shareholding structure and may be required to update it regularly.

• c. Local enterprise group surveys, provided the resident SPEs are known. However, this kind of survey-based collection can complicate the identification of SPEs that belong to larger, local enterprise groups (see paragraph 82).

D. Objectives for Establishing SPEs

The SPE definition determines that SPEs are established to obtain specific advantages provided by the host jurisdiction. A key aspect of the SPE, going by its name, is the special purpose with which the entity is incorporated. This special purpose can only be achieved through the registration or incorporation of a legal entity in a given specific jurisdiction by making use of the “special” advantages that this jurisdiction offers. National compilers, in trying to operationalize the objectives included as part of the definition, should therefore direct their efforts toward documenting the legal provisions for SPEs offered by their countries rather than seeking the answers from the entities.

Each national jurisdiction has its own law which governs incorporation of such entities, their operations, and holdings located in that jurisdiction. The presence of host supervisory or regulatory bodies can be an important source of information. Having knowledge of the institutional arrangements between countries can also help—for instance the list of double taxation treaties that exist.

Attractive features in the tax system are also important arrangements. These include relatively generous participation exemption regime and low withholding tax (WHT) rates on dividends, interest, and royalties. In most advanced economies, qualifying dividends and capital gains received from foreign-owned related companies are exempt from corporate income tax to prevent international double taxation. However, the conditions under which such exemptions are granted vary across countries. WHTs on dividends, interest, and royalties are often used by countries to tax passive income at source; or they serve as a safeguard against base-eroding payments by MNEs.

Bilateral tax treaties (BTTs) typically lower the WHT rates specified in domestic law on a reciprocal basis, facilitating bilateral foreign DI. Certain jurisdictions also offer a preferential treatment of income from intellectual property to stimulate innovation.

These advantages may be known to compilers. Thus, intra-group lending SPEs might domicile in a jurisdiction due to the zero WHT on interest and/or because of the low WHT rates agreed in BTTs with other countries. Holding SPEs may be attracted due to a relatively generous participation exemption and low or zero WHT rates on dividends.

The choice to locate SPEs in certain jurisdictions may also have to do with non tax factors as well. A well-developed international financial center (IFC) with a well-integrated ecosystem providing easy access to its capital market or access to sophisticated financial services, a skilled multilingual workforce, and a stable political and social climate can contribute to a jurisdiction being an attractive hub for international business. Compilers should be able to determine that these investments in financial assets that flow into and out of these jurisdictions usually have limited contribution to the domestic economy.

Numerous institutions, including international organizations, nongovernmental institutions, and private sector firms, publish indicators of the attractiveness of jurisdictions as hosts of SPEs. These third-party indicators (TPIs)¹² may complement compilers own knowledge and efforts, if deemed useful. In using such indicators/lists compilers should, nonetheless, exercise judgment and must be aware of the different challenges that these may present. Some TPIs are based on hard data, while others are based on qualitative assessment by experts, experience, perception, or composites of various underlying data sources. Issues may arise from the opacity of some indicators, their sources and methodologies, and their quality and reliability that are based on perception or value judgment.

Existing literature has referred to different lists of IFCs or offshore centers. The lists change regularly, and a number of organizations publish their own lists, which have been drawn up according to various criteria. The World Bank, the IMF, other regional organizations, and various other nongovernmental organizations are also constantly compiling lists of these economies according to their own criteria.

In practice, lists of offshore financial centers (OFCs) were published more than a decade ago by the IMF¹³ and recently by the OECD¹⁴; they are based on a qualitative assessment of the jurisdictional regulations and taxation frameworks. In contrast to these qualitative approaches, Zoromé (2007) has defined OFC jurisdictions as those that “provide financial services to nonresidents on a scale that is incommensurate with the size and the financing of their domestic economies.” Cobham, Jansky, and Meinzer (2015) have used flow data on the exports of financial services to calculate ratios that indicate how significantly a jurisdiction acts as an OFC. Fichtner (2015) further expanded this approach by using stock data on international banking assets, portfolio investment, and foreign DI in relation to the gross domestic product (GDP) of a jurisdiction to calculate an offshore-intensity ratio.

The IMF, in practice, has made a distinction in the past between two types of financial centers— OFCs and IFCs. OFCs are usually small jurisdictions, which have very stylized business activity with a view to save on taxes (through low or zero corporate taxes), protect assets, safeguard confidentiality, and provide sophisticated financial services. The bulk of financial sector activity in an OFC is cross-border on both sides of the balance sheet, where the transactions are initiated else-where, and where most of the entities involved are controlled by nonresidents. IFCs are countries, with well-developed financial markets and infrastructure, that are widely perceived as attractive intermediate destinations in the routing of investments, and there may be little distinction between on- and offshore business. Typically, these centers have low or zero taxes imposed on the transfer of capital to other countries, either via interest payments, royalties, dividends, or profit repatriation. In this way, profit from one country can be re-invested in another part of the world paying no or little taxes. Countries such as Ireland, Japan, The Netherlands, United Kingdom, and can be considered IFCs. Nowadays this distinction is not made anymore, with the terms OFCs and IFCs somewhat used interchangeably.

E. The Five-Employee Threshold Applied in the Definition

The definition requires that SPEs should have zero or up to a maximum of five employees¹⁵—a straightforward criterion for countries to implement as long as information about the number of employees is available. In principle, SPEs are considered to have no employment. However, for practical applicability of the definition, a cut off rule of five employees, which also reflects reduced physical presence, was established.

The proposal of establishing an upper limit of up to five employees has been informed by country experiences. Based on a consultation exercise conducted by the IMF in 2018, it came up that following specific legal requirements for minimum employment, in some jurisdictions SPEs may “artificially” employ a reduced number of staff to comply with national law. Applying this criterion, all entities above five employees are not considered as SPEs.

The five-employee threshold allows compilers to distinguish those hybrid companies, which display SPE-like activities, usually referred to as “near-SPEs” (or SPE type or SPE-like) but are not SPEs as they have more than five employees. They may display SPE characteristics, as evidenced by their increasingly disproportionate revenue/balance sheet size per employee. These entities emerged with the legal need to employ more staff as a result of the OECD Base Erosion Profit Shifting (BEPS)¹⁶—an initiative to address tax avoidance strategies that exploit tax gaps and mismatches to artificially shift profits to low- or no-tax locations with little or no economic activity. With the emergence of these near-SPEs and the OECD BEPS initiative, the definition had to cover this attribute more precisely by establishing the cut off rule concerning employment.

F. No or Little Physical Presence: How to Establish It in Practice?

The definition requires that entities, which are SPEs, should have no or little physical presence and no or little physical production with no specific numerical threshold recommended. Physical production is understood as an 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 buildings, land, machinery and equipment, and sub-soil assets, in the economic territory of incorporation. The physical presence of an 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. Typically, any nonfinancial assets owned by an SPE are restricted to either intangible or mobile tangible assets.

The determination of the existence of little or no physical presence and physical production can be made by using different indicators, such as:

• a. computation of several ratios, of which material fixed assets, trade receivables, and turnover to total assets, with a view to measuring links with the domestic economy

• b. zero or a minimum turnover with domestic parties in the host economy

• c. a minimum cap on the gross book value of property, plant, and equipment

Should national compilers adopt additional specific approaches for added precision for measuring little of no physical presence, these should come in addition to (that is not instead of) the employment criterion of up to five employees.

G. Nonresident Share of Total Balance Sheet Criterion

The SPE definition suggests that a large part of an SPE's financial balance sheet typically consists of cross-border assets and liabilities—viewed from the perspective of the compilers of a given economy. It also recognizes that most of the SPEs transactions are carried out with nonresidents. Nonetheless, this does not exclude that SPEs can transact with residents and hold domestic assets or liabilities.

The criterion that SPEs' assets and liabilities should be mostly vis-à-vis nonresidents is mainly applicable in the case of pass-through DI, which therefore makes the identification of SPEs straight-forward. However, in addition to financial SPEs—which almost exclusively held financial assets and liabilities—nonfinancial SPEs are also set up to engage in production of nonfinancial services, such as those holding intellectual property rights and collecting fees/royalties, engaged in merchanting, and engaged in captive operational leasing. However, if SPEs that are set up to hold intellectual property rights, for example, mostly collect royalties from residents, they should not be treated as SPEs because of their close links to the domestic economy.

Annex 1 provides a typology of SPEs, which aims to delineate the different types of SPEs based on their economic functions and relates them to the appropriate institutional sector following the 2008 SNA. The possibility that portfolio and other investment assets and liabilities can be vis à vis many market participants that may be resident in the host economy exists for many types of SPEs. Furthermore, for SPEs that are owned by residents but indirectly controlled by nonresidents, such as the examples in Figure 4 (Case 5), equity on the liabilities side is held by residents. However, these entities may still have large cross-border assets and liabilities.

View Full Size

Resident Chains of Entities

Citation: Technical Notes and Manuals 2022, 006; 10.5089/9798400231490.005.A001

• Download Figure
• Download figure as PowerPoint slide

Figure 5 elaborates on a possible case of resident chains of entities, where the compilers should take into account whether assets in resident entities, such as those of Y in X, are relevant to the domestic economy. If not, these entities, if they meet the definition of SPEs, including having large cross-border assets and liabilities, would be considered as SPEs.

A resident entity, directly owned by a nonresident entity, may also hold in the same (domestic) economy both SPEs and operating subsidiaries (Figure 5). Whether the resident enterprise (Y), should be considered as an SPE depends on, among other criteria, the composition of its balance sheet. Although it is directly controlled by a nonresident entity, its assets are partially nonresident and partially resident. The assets include investment in a resident enterprise (both SPE (X) and non-SPE (W)) as well as cross-border assets, which it holds directly in nonresident enterprise (V) and indirectly in nonresident enterprise (U). The bearing of the foreign and domestic assets (not necessarily the relative weight) of Y, in this case, is important in classifying Y as an SPE. This depicts a situation where the compiler is expected to look at the relevance of the balance sheet, where though an entity could be resident, it may have no economic links with the domestic economy.

H. Specific Types of SPEs

Securitization Vehicles

Securitization vehicles are entities set up to provide structured finance. These entities, if they exhibit all characteristics that meet the SPE definition, should be classified as SPEs, even though, at times, they may display balance sheets composed of foreign assets (typically loans to its parent company) and a mix of foreign and domestic liabilities (typically bonds held by portfolio investors).

Securitization involves raising funds by selling a security backed by specific assets or income streams.¹⁷ It is the process where a financial institution or other business converts existing assets or future cash flows into marketable securities, by placing those assets in a structure where the risks and rewards of owning the assets are transferred to a third party. SPEs are an integral part of many structured finance transactions—securitizations. A securitization vehicle is created for a specific objective, which is to isolate financial risk.

View Full Size

Securitization and the SPE

Citation: Technical Notes and Manuals 2022, 006; 10.5089/9798400231490.005.A001

• Download Figure
• Download figure as PowerPoint slide

In a cross-border context, a securitization vehicle is, therefore, the SPE that is directly or indirectly controlled by nonresidents, usually acts as a subsidiary company to enable an asset and liability structure, and holds up a legal status and secures obligations, on the off chance that the parent company goes bankrupt. Generally, these are the elements to securitization (Figure 6): (1) assets generating the cash flow that are to be securitized (underlying asset), (2) the SPE functioning as the conduit linking the underlying assets and investors, and (3) investors investing in the cash flows (securities) generated by the underlying asset, usually grouped in tranches and sold to meet the credit risk preferences of a wide range of investors. These securities are termed asset-backed securities. In the ESS context, most of the involved flows are cross-border. However, given that those SPEs transact assets and/or liabilities on the open market, the balance sheet can be a mix of cross-border and domestic investors, where the conduits issue debt securities and have no control over who holds the securities.

If the securitization vehicle set up, be it a registered trust or an incorporated company, meets the definition of an SPE, with no employees, is directly or indirectly controlled by a nonresident, and has a large share of its balance sheet with nonresidents, the compilers should include them into the SPEs population. Securitization vehicles are, however, also examples where the functional classification of the cross-border assets and liabilities can vary between DI, other investment, and port-folio investment.

Compilers need to distinguish cases (BPM6 paragraph 4.78) in which the originator issues asset-backed securities on its own books, then securitization may take place without the creation of a separate entity. When the portfolio is not transformed, or the vehicle does not bear market or credit risks, then it can be combined with its parent (if resident in the same economy) or treated as a captive intermediary (if in a different economy to that of its parent—see typology of SPEs, Annex 1).

Additionally, there are securitization vehicles that are created independent from their parents such that they have no relationship with the originator of the assets. In those instances, the assets are sold to entities, which then become entirely independent of the originator, with the latter holding no share capital. The aspect of control, as defined for DI purposes, is clearly absent in these cases. These entities are usually called “orphan” SPEs with their shares settled on a charitable trust and with professional directors provided by an administration company to ensure that there is no connection with the originator. These securitization vehicles have no employees, but have a manager, who assumes the responsibility of the overall securitization framework, including the liaison between the originator, investors, lenders, and other related parties.

These securitization vehicles have, in these instances, autonomy of decision and are assessed to be institutional units “on their own right.” Consequently, based on the current definition of SPEs, these orphan entities would not be SPEs in the host economy as they do not have direct or indirect nonresident control.

Compilers are, therefore, advised to be particularly attentive to understanding the framework of the securitization vehicle and exercising caution when defining the sample of SPEs for securitization. This is important because different terms are used interchangeably in different countries dependent on the forms of SPEs: sponsor, originator, administrator. These concepts/terms can be difficult to understand, when deciding on direct or indirect ownership.

A. Collaboration Between the Respective Regulatory Body and the ESS Compiling Agency

Each host economy/jurisdiction has its regime to license/regulate SPEs, with specific agencies, procedures, legal basis, means and protocols such that different types of data can be made available and collected. Some countries may have a political or administrative structure that complicates their data collection process (for example, complex jurisdictions or federations); other countries may simply not have the capacity to collect data in the first place.

As indicated in the metadata produced and through consultation with compilers, in countries already compiling cross-border data on SPE activities, in general, the latter are collected as part of the general balance of payments data collection framework. Collecting data from SPEs, in these reporting countries, is associated with exposing confidential data. In this respect, legal arrangements along with the adoption of memorandums of understanding (MOUs) between institutions help preserve the confidentiality of information and facilitate data sharing (Annex 3).