The term “comparability analysis” is used to designate two distinct but related analytical processes:

1. Developing an understanding of the accurately delineated trans-action, which includes:
   1. Identifying the economically significant characteristics and circumstances of the controlled transaction, i.e. the trans-action between associated enterprises; and
   2. Identifying the respective roles and responsibilities of the parties to the controlled transaction. This is generally considered as part of the functional analysis, see further section 3.4.4.
2. Comparing the prices and other conditions of the controlled transaction (as established in step 1 immediately above) with those in uncontrolled transactions (i.e. transactiWons between independent enterprises) taking place in comparable circumstances. The latter are often referred to as “comparable uncontrolled transactions” or “comparables”.

​​​​​​The incremental profit, if any, derived from the exploitation of LSAs is known as “location rent”. Thus, the term “location savings” represents “cost savings” whereas “location rent” represents the incremental profits derived from LSAs. The value of “location rent” is at most equal to, or less than, the value of LSAs.

Figure 3.D.4

Determining Location Rent

​​​​​​​The concept of comparability analysis is used in the selection of the most appropriate transfer pricing method, as well as in applying the selected method to arrive at an arm’s length price or financial indicator (or range of prices or financial indicators). It plays a central role in the overall application of the arm’s length principle.

​​​​​​A practical difficulty in applying the arm’s length principle is that associated enterprises may engage in transactions that independent enterprises would not undertake. Where independent enterprises do not undertake transactions of the type entered into by associated enterprises, the arm’s length principle is difficult to apply because there is little or no direct evidence of what conditions would have been established by independent enterprises. However, the mere fact that a comparable transaction may not be found between independent parties does not of itself mean that the controlled transaction being analyzed is not at arm’s length.

​​​​​It should be kept in mind that the relative lack of comparables for a taxpayer’s controlled transaction does not imply that the arm’s length principle is inapplicable to that transaction. Nor does it imply anything about whether that transaction is or is not, in fact, at arm’s length. In a number of instances, it will be possible to use “imperfect” comparables, e.g. comparables from another country with similar economic conditions or comparables from another industry sector. Such comparables may need to be adjusted to eliminate or reduce the differences between that transaction and the controlled transaction as discussed in 3.1.6 below, provided such adjust-ments can be done reliably.

In other instances, where no comparables are found for a controlled transaction between associated enterprises, it may become necessary to use approaches not depending directly on comparables to find an arm’s length price (see 2.5.3.2). It may also be necessary to examine the economic substance of the controlled transaction to determine whether its conditions are such that it might be expected to have been agreed between independent parties in similar circumstances— in the absence of evidence of what independent parties have actually done in similar circumstances.

​​​​​Practical guidance is needed for cases without sufficient comparables. There are two distinct problems relating to comparables for developing countries’ tax authorities. The first is lack of access to existing sources of data, such as existing non-local company databases. The second is the lack of reliable local country comparables. Each of these causes problems for local tax administrations (e.g. the lack of data impedes the reliable and efficient determination of appropriate arm’s length results) and also causes problems associated with double taxation and dispute avoidance (e.g. the lack of adequate data impedes a country’s ability to reach agreement with taxpayers and with other tax authorities in mutual agreement negotiations).

The first step in undertaking a transfer pricing analysis always involves the accurate delineation of the transaction; this includes an awareness of the industry and market context in which the transaction takes place. From this, the most appropriate transfer pricing method can be selected (bearing in mind the likely existence of necessary data) and where appropriate, a tested party will be chosen. This process should determine the kind of comparables to be sought. Where comparables operating in the same jurisdiction as the tested party are available there is no need to consider whether geographic differences might have a material impact on the prices or profits under review. However, in the absence of such information, foreign comparables should not automatically be rejected as all transfer pricing cases require a solution. A pragmatic approach, making use of the best available comparables, will often be required. Adjustments may need to be considered and made where they improve the reliability of the comparison.

​​​​​​The extent to which LSAs will lead to location rents depends on competitive factors relating to the end product and to the general access to LSAs. It is possible that in a particular case, even though LSAs exist, there are no location rents. For example, in situations in which the market for the end product is highly competitive and potential competitors also have access to the LSAs, much or all of the benefits of LSAs would be passed on to the customers through lower prices of products, resulting in little or no location rent. However, circumstances where extra profits are passed on to customers are varied, and may be permanent or temporary. Where this is temporary, at the end of this period of competition, the MNE may possibly achieve a larger market share in the local market with an increased ability to sell products at a higher price. Alternatively, if an MNE has exclusive access to the LSAs, then the MNE may derive significant location rents associated with the LSAs, as the LSAs reflect a competitive advantage. These location rents may dissipate over time due to competitive pressure, depending on the facts and circumstances of each case.

​​​​​This chapter discusses a possible procedure to identify, screen, select and adjust comparables in a manner that enables the taxpayer or tax administration to make an informed choice of the most appropriate transfer pricing method and apply that method correctly to arrive at the appropriate arm’s length price or profit (or range of prices or profits).

A typical approach that can be followed while performing a comparability analysis is outlined below. The steps listed below are by no means exhaustive but rather suggest an outline based upon which a comparability analysis could be carried out. It should be noted that the process is not necessarily linear, for example, a number of the steps may need to be carried out repeatedly until a satisfactory result is achieved.

​​​​​A summary of the comparability analysis in operation steps are set out below. The subsequent sections of this chapter deal with each of these steps in more detail.

​​​​​​The first step is understanding the economically significant characteristics of the industry, taxpayer’s business and controlled transactions; see below. This involves:

• Gathering of basic information about the taxpayer;
• Identifying and accurately delineating the controlled transaction in question; and

Deciding whether transactions should be evaluated separately or on a combined / aggregated basis.

The next step is an examination of comparability factors of the controlled transaction, see 3.4:

• Characteristics of the property or service transferred;
• Contractual terms of the transaction;
• Functional analysis of the controlled transaction under examination;
• Economic circumstances of the transaction; and Business strategies of the parties.

As with the determination of whether location rents exist, the arm’s length attribution of location rents depends on competitive factors relating to access to the LSAs, and on the realistic alternatives available to the associated enterprises given their respective bargaining power. To the extent that competitors would not have access to the LSAs, the relevant question is why this is so. There are a number of possibilities. For example, the MNE could have production intangibles that allow it to manufacture at a lower cost than competitors. At arm’s length, the owner of the intangible would typically be entitled to the rents associated with this cost saving, as it would have a realistic alternative to undertake its production elsewhere at similarly low costs. As another example, it might be that the low-cost producer is the first to operate in the low-cost jurisdiction and there are no comparable low-cost producers in its jurisdiction or other jurisdictions, implying that, for a time at least, it is well-placed to extract a part of the location rents.

The remaining steps in a comparability analysis (see 3.5) are:

• Selecting the tested party/parties (if applicable under the most appropriate method selected);
• Identifying potentially comparable transactions—internal and external;
• Comparability adjustments where appropriate;
• Selection of the most appropriate transfer pricing method;
• Determination of an arm’s length price or profit (or an arm’s length range of prices or profits); and
• Documentation of the comparability analysis and monitoring.

​​​​​An essential first step to enabling effective transfer pricing analysis is the collection of information about the taxpayer to understand its business operations and activities. This factfinding process should include identification of associated enterprises involved in the controlled transaction, and gathering information about relevant cross-border controlled transactions in the context of the commercial and financial relations between the relevant members of the MNE (including the functions performed, assets used or contributed (including intangibles, see Chapter 6) and risks assumed, by each party, the nature of products/services transferred, the terms and condi-tions of the transaction, the economic circumstances etc.).

An analysis should be performed of the circumstances specific to each taxpayer and industry. This includes, but is not limited to, an analysis of the industry, competition, economy, regulatory factors and other elements that may significantly affect the taxpayer and its environment.

Information about the taxpayer from its annual report, product brochures, news articles, research reports prepared by independent agencies, management letters and internal reports could act as a good starting point for understanding the taxpayer’s circumstances. A study of these documents will provide an idea of the industry to which the enterprise belongs, the nature of its business activities (i.e. manufacturer, wholesaler, distributor etc.), its market segment, market share, market penetration strategies utilized, type of products/services dealt in, etc.

​​​​​​The next question would be the appropriate split consistent with the arm’s length principle. As discussed above, the bargaining power of the associated enterprises which reflects the arm’s length nature of two independent parties negotiating over their respective shares of savings/rents may be well suited as the key metric for this. This can be used to determine the arm’s length surplus (savings/rents) allocations when comparable uncontrolled transactions or benchmarks are not available.

The arm’s length price for a transaction between two or more controlled entities must be established in relation to transactions actually undertaken. Thus, the critical first step in any comparability analysis is to accurately define those transactions by analyzing their economically relevant characteristics, as reflected not only in the contracts between the parties, but also their conduct and any other relevant facts. In this regard, the contractual terms will generally be the starting point for the analysis (as clarified or supplemented by the parties’ conduct); and to the extent that the conduct or other facts are inconsistent with the written contract, the parties’ conduct (rather than the terms of the written contract) should be taken as the best evidence of the transaction(s) actually undertaken.

​​​​​Tax authorities should not substitute other transactions in the place of those that have actually been undertaken. They should also not disregard actually undertaken transactions other than in exceptional circumstances. Such circumstances may exist, for example, where the arrangements viewed in their totality are not commercially rational, thereby preventing the determination of an arm’s length price for each party to the transaction (taking into account their own perspectives and the options realistically available to each of them). This test is substantive and looks at the nature of the arrangements entered into; a lack of independent transactions does not, of itself, indicate that the controlled transaction lacks commercial rationality.

The test for commercial rationality must be considered from each entity’s own perspective, as an arrangement that is commercially rational at group level is not necessarily arm’s length from the perspective of each party.

In addition, an arrangement that is expected to leave the MNE worse off as a whole on a pre-tax basis than it would be if it had not entered into the arrangement will raise questions. The question is whether it is primarily tax driven and it may need further examination as to whether it is commercially irrational thereby preventing the determination of an arm’s length price for each party to the transaction.

​​​​​​Government rules and regulations should be treated as conditions of the market in the particular country if they apply in the same way to controlled and uncontrolled transactions. Such rules would include government interventions in the form of price controls, interest rate controls, exchange controls, subsidies for certain sectors, anti-dumping duties etc., and should be taken into account in arriving at an appropriate transfer price in that market. The question becomes whether, in light of these conditions, the transactions between associated enterprises are consistent with comparable uncontrolled transactions between independent enterprises.

​​​​​Where a transaction that was actually undertaken is not commercially rational, any alternative transactions that are substituted for transfer pricing purposes should correspond as closely as possible to the actual facts of the case while achieving a commercially rational expected result: i.e. one which would have enabled the parties to come to a price acceptable to both at the time the arrangement was entered into.

In general, non-recognition or substitution of transactions should not be undertaken lightly as this would create significant uncertainty for taxpayers and tax administrations. Non-recognition or recharacterization of transactions may also lead to double taxation due to the divergent views taken by countries on how any substitute transactions are structured. The ability of tax authorities to disregard or substitute transactions will depend on their powers under applicable domestic law and should be considered in developing domestic transfer pricing legislation and administrative rules. See further Chapters 6 and 10.

An important aspect of transfer pricing analysis is whether this analysis has to be carried out with respect to a taxpayer’s individual international controlled transactions or whether a group of international controlled transactions that have a close economic nexus and are carried out between the same parties can be analyzed together for transfer pricing purposes.

​​​​The transfer pricing analysis should ideally be made on a transaction-by-transaction basis. However, there are cases where separate transactions are so closely linked that such an approach would not lead to a reliable result. Where transactions are so closely interrelated or continuous that application of the arm’s length principle on a transaction-by-transaction basis would become unreliable or cumbersome, transactions are often aggregated for the purpose of the analysis.

​​​​​​An example of where government rules affect the market is that of certain pharmaceutical formulations, which may be subject to price regulation in a particular country. Another example is Export Oriented Units which may be subject to beneficial provisions under the taxation laws of a country; ideally, companies that enjoy similar privileges should be used as comparables, and if that is not possible, comparability adjustments may need to be made as part of the comparability analysis. Another example is where foreign exchange regulations limit the amounts of the payments that can be made for services or intangibles. However, such regulatory limits may not set arm’s length prices for services or intangibles. For example, assuming that all the transactions are denominated in the same currency, certain countries have restrictions on the payment of interest on external commercial borrowings and the exchange control regulatory requirements authorize the borrower to pay interest at LIBOR plus say 200 basis points. The country of the lender may, however, not agree to use this as a basis for benchmarking the transaction when the lending enterprise itself borrows in its domestic market at a higher rate.

• • • •  
• An example can be the case of transactions involving the licensing of know-how to associated manufacturers together with the supply to the licensed associated manufacturers of components needed to exploit such know-how. In such a case, the transfer pricing analysis may be more reliable if it takes into account both the license and the supply of components together, compared to a consideration of each separate activity without recognizing that they are closely interrelated transactions. Similarly, it may be difficult to separately analyze pricing of sales of products and closely linked long-term service supply contracts.

​​​​​Another important aspect of combined transactions involves the use of composite contracts and “package deals” among the members of an MNE group. A composite contract and/or package deal may contain a number of elements including leases, sales and licenses all packaged into one deal. Generally, it will be appropriate to consider the deal in its totality to under-stand how the various elements relate to each other, but the components of the composite contract and/or package deal may or may not, depending on the facts and circumstances of the case, need to be evaluated separately to arrive at the appropriate transfer price. In certain cases, it may be more reliable to allocate the aggregate compensation among the elements of the composite contract or package deal.

“Aggregation” issues also arise when looking at potential comparables. Since third-party information is not often available at the transaction level, entity level information is frequently used in practice when looking at external comparables (e.g. in the absence of reliable internal comparables; “external comparable” and “internal comparable” are defined in section 3.5.2 below). It must be noted that any application of the arm’s length principle, whether on a transaction-by-transaction basis or on an aggregation basis, needs to be evaluated case by case, applying the most appropriate transfer pricing method to the facts in that particular case.

​​​​​​The first part of a comparability analysis for transfer pricing purposes involves understanding and defining the controlled transaction to be tested. In addition to the contextual information on the industry and the overall business of the taxpayer, this analysis is typically structured around five so-called comparability factors.

The market level of the company is another key factor. For example, the price at the wholesale and retail levels would generally differ so that a retail price should not be compared to a wholesale price. Other market conditions which may influence the transfer price may include: costs of production (including costs of land, labour and capital); availability of substitutes (both goods and services); level of demand/supply: transport costs; the size of the market, and the extent of competition.

The factors are:

• • • •  
• The characteristics of the property or service transferred;
• The contractual terms; the functions performed by each of the parties (taking into account the assets employed an risks assumed);
• A functional analysis of the controlled transaction under examination;
• The economic circumstances; and
• The business strategies followed by each of the parties.

​​​​Property, whether tangible or intangible, as well as services, may have differing characteristics which may lead to value and price differences in their values in the open market. Therefore, these differences must be accounted for and considered in any comparability analysis of controlled and uncontrolled transactions.

Characteristics that may be important to consider are:

• In the case of tangible property: physical features, quality, reliability, availability and the volume of supply;
• In the case of services: the nature, quality and extent of the services provided; and
• In the case of intangible property: form of the transaction (e.g. licensing or sale) and the type and form of property, duration and degree of protection and anticipated benefits from use of the property.

For example, comparability analysis should take into account the differences between trademarks and trade names that aid in commercial exploitation (marketing intangibles) as opposed to patents and know-how (trade intangibles).

​​​​​​The conduct of the contracting parties is generally a result of the terms of the contract between them. The contractual relationship thus requires careful analysis when computing the transfer price. While there may be a written contract, the terms of the transactions may also be found in correspondence and communications between the parties involved. In cases where the terms of the arrangement between the two parties are not explicitly defined in written contracts, the contractual terms have to be deduced from the parties’ economic relationship and their conduct.

​​​​​An important point to note is that associated enterprises may not hold each other fully to the terms of the contract since they have common overarching interests. This contrasts with independent enterprises, which are expected to hold each other to the terms of the contract. Thus, it is important to determine whether the contractual terms between the associated enterprises are a “sham” (something that appears genuine, but when looked at more closely lacks reality, and is not valid under many legal systems) and/or have not been followed in reality.

​​​​​​Also, explicit contractual terms of a transaction involving members of an MNE may provide evidence as to the form in which the responsibilities, risks and benefits have been assigned among those members. For example, the contractual terms might include the form of consideration charged or paid, sales and purchase volumes, the warranties provided, the rights to revisions and modifications, delivery terms, credit and payment terms etc. In addition to an examination of these contractual terms, it will be important to check that the actual conduct of the parties conforms to them.

On a general level, business strategies are one of the important factors in a comparability analysis. However, the examination of the legitimate business strategy of an MNE will depend on the facts and circumstances of each case. The business strategy of an MNE is dependent upon the structural characteristics of its industry. Nonetheless, MNEs with different business strategies may exist within the same industry. The business strate-gies pursued by different MNEs may differ because of their different levels of global integration, different corporate histories, internal efficiencies and competitive advantages. Business strategies would take into account many aspects of an enterprise such as innovation and new product development, degree of diversification, risk aversion, assessment of political changes, the impact of existing and planned labour laws, duration of arrangements and other factors bearing upon the daily conduct of business. Such business strategies may need to be taken into account when determining the comparability of controlled and uncontrolled transactions.

Where there are material differences in economically significant contractual terms between the taxpayer’s controlled transactions and the potential comparables, such differences should be evaluated, in order to judge whether comparability between the controlled and uncontrolled trans-actions is nevertheless satisfied and whether comparability adjustments need to be made to eliminate the effects of such differences.

Functional analysis typically involves identification of functions performed, assets employed and risks assumed (also called FAR analysis) with respect to the international controlled transactions between members of an MNE. Functional analysis seeks to identify and compare the economically significant activities and the responsibilities undertaken by each of the associated enterprises and by any independent entity engaging in a potentially comparable transaction. An economically significant activity is one which materially affects the price charged in a transaction and/or the profits earned from that transaction.

​​​​Functional analysis is the cornerstone of any transfer pricing exercise. Its purpose is to gain an understanding of the operations of an enterprise in connection with its transactions with associated enterprises. The functional analysis examines the respective roles of the parties to the controlled transaction under examination. The roles undertaken by each of the associated enterprises will affect the determination of an arm’s length remuneration for the controlled transaction since compensation in transac-tions between two independent enterprises will usually reflect the functions that each enterprise performs, taking into account assets employed and risks assumed. Generally, the more valuable those functions and assets, and the greater the risks, the greater the expected remuneration.

​​​​​​​

​​​​​The functional analysis also examines the activities of the parties to uncontrolled, potentially comparable transactions. The functional anal-ysis of parties to potentially comparable transactions will assist in evaluating whether such transactions appropriately can be utilized as comparables in the analysis since close comparability of functions performed, assets used or contributed and risks assumed is often a prerequisite for a valid comparison.

Functional analysis is a process of finding and organizing facts about the transaction in terms of the functions, risks and assets in order to identify how these are divided between the parties involved in the transaction. The functions, risks and assets are analyzed to determine the nature of functions performed, degree of risks undertaken and the nature of the assets employed by each party. This analysis helps to select the tested party/parties where needed (as explained below), the most appropriate transfer pricing method, the comparables, and ultimately to determine whether the profits (or losses) earned by the entities are appropriate to the functions performed, assets employed and risks assumed.

The functional analysis is important because the expected return of the entities involved in a transaction depends on the importance of the functions performed, the nature and degree of risks assumed and the nature and value of assets employed. Generally, the more valuable the functions performed, assets employed and the greater the risks assumed by a party to a transaction the greater its expected return (or potential loss). It is therefore extremely important to map the functions performed, assets employed and risks assumed by all the associated enterprises in relation to the controlled transaction under examination.

Functions performed are the activities that are carried out by each of the parties to the transaction. In conducting a functional analysis, economi-cally significant functions are to be considered, as such functions add more value to the transactions and are therefore expected to fetch higher anticipated returns for the entity performing such functions. Thus, the focus should not be on identifying the maximum number of functions but rather on the identifica-tion of critical functions performed by the associated enterprises.

The above market share strategies depend on various factors like market power and the business life cycle of the MNE in a particular market. Market penetration occurs when an MNE is a relative newcomer to a particular market and is seeking to enter and establish its products/services in the new market. An MNE might actively pursue a market expansion strategy to increase its market share in highly competitive markets. Market maintenance occurs when an MNE has already entered a market and is aiming at maintaining its market share.

Some of the relevant functions that are often observed and that may be economically important in determining the price for a transaction are:

• Research and development;
• Product design and engineering;
• Manufacturing, production, assembly, and process engineering;
• Purchasing, materials management and other procurement activities;
• Transportation, warehousing and inventory;
• Marketing, advertising, publicity and distribution;
• Market intelligence on technological developments; and
• Intragroup services, for example managerial, legal, accounting and finance, credit and collection, training and personnel man-agement services.

It should be emphasized that this list is purely indicative. The extent to which each of these functions (or other functions not listed above) is economically significant and contributes to the creation of value depends on the particular facts and circumstances of the case.

A functional analysis can be approached by evaluating all the economically significant activities performed in relation to the controlled transaction under examination (such as the list indicated above) and in potentially comparable uncontrolled transactions. In general, a taxpayer should prepare this list for both parties to the relevant controlled transaction (e.g. for the producing and selling/distributing activities in this example) to ultimately support the selection of the most appropriate transfer pricing method and to assist in identifying useful comparables.

The following common functions are required to manage the busi-ness in the above example. While both parties are likely to participate in performing these functions, one may perform one or more of the functions more intensively than the other party.

• Corporate strategy determination: Generally, all policies within the MNE are determined by the management of the respective entities which continuously monitor the economic environment surrounding the entity, assess their strategic position within the industry and set targets to achieve their corporate objectives;
• Finance, accounting, treasury and legal functions: The manage-ment of the respective entity is responsible for managing the finance, treasury, legal and accounting functions. Each entity is also responsible for all local statutory compliance; and
• Human resource management function: The HR function of each entity is coordinated by its management, which is respon-sible for recruitment, development and training of the personnel including the pay structure.

A market penetration strategy may involve a combination of strategies for attracting existing users of a competitive brand to new products and attracting non-users to the product category to which the new product belongs.

In addition to identifying the specific functions undertaken by each party, the analysis should also evaluate the relative intensity with which each party performs its functions. Table 3.T.2. indicates one way such a qualitative evaluation of relative functional intensity might be summarized in a tick chart.

Table 3.T.2

Qualitative Relative Assessment of Functions Performed

(by P Co. and Q Co. in relation to Q Co.’s Market)

​​​​​​Assets (tangible as well as intangible) that are used by, or transferred between, the associated enterprises in the course of an international controlled transaction need to be identified as part of the functional analysis.

The analysis should involve the identification of the type of capital assets employed (e.g. plant and equipment, intangible assets, financial assets etc.) and their significance to the controlled transaction. For economically significant assets it may be necessary to perform a more detailed analysis of the assets employed, such as their age, location, property right protections available, market value etc.

​​​​​​In the case of capital-intensive industries, the employment of a capital asset such as property, plant and equipment, etc. is costly and has to be financed either internally or externally. However, there can also be cases where the entities are involved in activities for which the assets employed may not require such a large capital investment. Depending on the circum-stances and applicable accounting standards, costs associated with acquiring rights to use an asset may be recorded as operating expenses (e.g. in the form of leasing or rental costs) or as financial expenses below the operating profit line (e.g. as interest expenses). Where an operating profit measure is used to determine if the controlled transaction(s) are arm’s length, it will be impor-tant to ensure such amounts are considered in the comparability analysis and that their accounting treatment is consistent. Adjustments might be required to ensure consistency of accounting standards between the controlled trans-action and the comparable. Differences in the use of assets can sometimes be eliminated or reduced to a significant extent by making comparability adjustments on account of working capital or capacity utilization.

Where the transactions involve the use or transfer of economically significant intangibles, the special considerations set out in Chapter 6 should be borne in mind.

When an MNE pursues a market maintenance/expansion strategy it may focus on combining multiple strategies of:

• Attracting users of competitive brands;
• Pursuing current users to increase usage; and
• Attracting non-users of the product category.

It can be hypothetically assumed for the purpose of the example that B Co. owns the following tangible assets. In any actual case the list of relevant economically important assets would depend on the specific facts of the case.

• Land and buildings;
• Plant and machinery;
• R&D equipment;
• Office equipment;
• Furniture and fixtures;
• Vehicles;
• Computers; and
• Testing equipment.

It can be assumed for purposes of the example that:

• Q Co. has established a research and development department which tries to increase the level of its performance by improv-ing technologies so as to achieve further efficiencies. This would also reduce dependence on outside sources of technology in the future and achieve cost savings. The department also conducts R&D programmes to support Q Co.’s business and to provide technical assistance to its customers. These efforts help to increase production efficiency and product quality;
• Q Co. has established its own marketing intangibles in Country B by incurring significant expenditure on marketing and has penetrated the market for the new product in the territory of Country B. As noted above, these marketing intangibles are separate from the intangibles of P Co. in Country A for which a technology agreement is in place with P Co.;
• Q Co. has entered into a technology license agreement with P Co. for procuring technology for the manufacture of specified products. Thus Q Co. uses the process, know-how, operating/ quality standards etc. developed/owned by P Co. Q Co. leverages value from these intangibles for continued growth in revenue and profits;
• ​​​​​​​P Co. is the market leader in the development and supply of electronic meters, as well as related software, energy monitoring, billing solutions and payment systems. Over the years the com-pany has amassed a wealth of proprietary technical knowledge. This includes product specifications, designs, the latest manufac-turing processes and empirical data on the usage of products by customers in the industry;
• P Co. enjoys a reputation for quality products. In the inter-national utility markets, product supplies from international players from developed countries are preferred by the custom-ers and utilities as compared to direct product supplies from suppliers located in developing countries. Q Co. leverages on P Co.’s established brand name and reputation for high technology products. P Co.’s commitment to quality also pro-vides Q Co. with an edge while selling products in the domes-tic markets.

Risk analysis is an important part of the functional analysis. The detailed guidance provided in this section on the analysis of risks as part of a functional analysis does not mean that risks are more important than functions or assets. The relevance of functions, assets and risks in a specific transaction will need to be determined through a detailed functional analysis.

​​​​​​Risks are an inherent part of commercial activities. Businesses exist and undertake commercial activities in order to pursue opportunities to make profits. Simply put, risk is the effect of uncertainty on the objectives of the business. Greater risks are associated with higher expected returns—profit-seeking enterprises would only take on the risks in commercial opportunities if they anticipate a positive return. But such opportunities are inherently uncertain: costs may be higher than anticipated; revenues may be lower; circumstances may change and therefore actual results may be better or worse than those which were expected.

All these three market share strategies use two fundamental tactics:

1. Lowering the price of their products on a temporary basis by offering discounts on the product to become extremely competitive in the market; and
2. Increasing marketing and selling expenses through increased advertisement; sales promotion activities like offering rebates, free samples, offering extended warranties, and increased marketing activities such as increasing the number of salespersons, commission agents or distributors and increasing the amount of incentive payments made to commission agents or distributors.

It may be desirable to isolate the costs related to the pursuit of the above tactics as precisely as possible so that the allocation of costs at arm’s length can be computed.

​​​​​​Since the assumption of economically significant risks will be relevant to the pricing of a transaction, a transfer pricing analysis must first identify such risks, and then determine which entity assumes them. This analysis of risk should start from the contractual terms that exist between the parties but should also have regard to the conduct of the parties, including the functions they perform and any other relevant facts. Only then can the controlled transaction be properly understood and defined, and from there, appropriately priced. For transfer pricing purposes, the analysis of risk can be broken down into six steps.

There are many sources and types of risk, the significance of which will vary depending on the nature of the business transaction. The significance of a risk will depend on a combination of its likelihood and its potential impact on the profits (or losses) of the business. For example, the risk associated with the design of new packaging to improve visibility of a product may be relatively small compared to the risk associated with the development of a completely new product line. Changes to a “flagship” product are likely to carry more risk than changes or variations to a less important product or to one product among many sold by the business, and risks associated with developments based on novel technologies or wholly new applications are likely to be higher than those which build on existing, proven products or technologies.

An examination of the key functions and commercial context of a transaction will help to identify significant risks. In many cases, an examina-tion of the functions performed, assets used or contributed and risks assumed by other associated enterprises in the MNE contributing to the group’s crea-tion of value may help in this process since risks also represent opportunities and businesses will generally allocate resources to manage significant risks.

An illustrative list of risks that may be assumed by the parties to the transaction is provided below. However, the relevance and significance of each individual risk factor listed below will depend on the nature of the individual transaction and the particular facts of the individual case.

Table 3.T.4

Illustrative List of Risks Assumed

​​​​​​It should be emphasized that this list is purely illustrative, and that the extent to which each of these risks (or other risks not listed above) is economically significant and contributes to the creation of value depends on the industry and on taxpayer-specific circumstances. Hence, real life knowledge of how a particular MNE is functioning vis-à-vis its associated enterprise is very crucial in determination of the risk. For instance, not all industries involve the same level of product liability risk.

• • • •  
• Which entity is the initiator of the strategy?
• Which entity is the intended beneficiary of the strategy?
• Are unusually intense advertising, marketing and sales promotion efforts taking place, since these would provide a signal of market penetration or market share expansion strategies;
• What is the nature of the relationship between the related parties, i.e. their responsibilities and risk profile?
• Does the strategy involve intangibles? and
• Which party is the legal and economic owner of such intangibles?

​​​​​​Once economically significant risks have been identified, the analysis turns to consideration of which party assumes such risks. In this regard, the starting point for the analysis is usually the contractual terms between the parties (STEP 2). Parties transacting at arm’s length would be expected to agree on the allocation of significant risks between them before the outcome of the risk-taking is known. However, contracts between associated enter-prises may not specify the allocation of all the economically significant risks.

Figure 3.D.1

The Process of Delineating and Pricing Transactions

Most of the commonly assigned risks in the contract are risks which can be mitigated against, for example inventory risk, bad debts, foreign exchange risk etc. Market circumstances, price competition, the supply of raw materials, rises in wages etc. are risks which typically are more difficult to mitigate, and which may not be identified in the contract. Volatility in the global market in the past decade has demonstrated that risks which are difficult to mitigate are often economically more significant than the kinds of contractual risks as mentioned above.

​​​​​Moreover, in some cases, written contracts may be inconsistent or may not be followed in practice. For example, a contract may say that a manufacturer of electronic goods bears warranty risk; however, in practice the reseller habitually pays for the cost of customer repairs made under warranty. In this case, it is the reseller that is bearing the risk. The determination of the risk assumption between the parties must therefore be based on the actual conduct of the parties, rather than merely to the legal form of the agreement.

​​​​​Even where a comprehensive and consistent contract is in place, an analysis of the conduct of the parties and other facts is critical. In particular, it is important to consider which party or parties control the economically significant risks, and whether a party assigned a risk in fact has the financial capacity to assume it. Both control (see sections 3.4.4.33 to 3.4.4.35) and financial capacity (see sections 3.4.4.36 to 3.4.4.38) are necessary for one party to be treated as having assumed a particular risk; but neither of them is by itself sufficient.

​​​​​​The next step in the risk analysis process gathers facts on the actual conduct of the parties through a functional analysis. As has been noted above, information relating to the exercise of control over risk and the financial capacity to assume risk are particularly important. This information will then be analyzed in the remaining steps.

For example, a limited risk company acting solely as a sales agent with little or no responsibility for market development would be less likely to share the costs of a market penetration strategy initiated by its parent company than would a full risk distributor that might benefit more from a successful market penetration or expansion strategy.

When an MNE enters a new market with its product or expands market share of its product in an existing market through its subsidiary, questions of the creation of marketing intangibles and increases in the value of product-related intangibles such as trademarks, trade names etc. follow closely behind. It is important to examine and follow the process of creation of intangibles in a market, as well as the legal ownership of such intangibles and the right to share in the return from such intangibles (the notion which some countries refer to as “economic ownership”). Transfer pricing aspects of these issues are discussed in greater detail in Chapter 6 on intangibles.

Control over risk involves the process of real decision-making. Decision-makers must be able to understand the risk and the impact the decision could have on the business. They must have access to relevant information. If information or analysis is provided by others, the decision-maker must be able to assess whether the right information has been provided, and whether the analysis being relied upon are adequate. Without the foregoing, the mere formalization of a decision, for example in the form of the signing of documents or the minutes of a meeting reflecting a decision effectively already made elsewhere, are insufficient.

​​​​​The setting of the broad policy framework in which to assess risks is also not enough. For example, the setting by Company Y of broad MNE objectives or a general company “image” would not mean that Company Y controls risks relating to specific marketing strategies. Similarly, a requirement to analyze and report on certain risks in a certain way, or in accordance with a particular framework or template, does not necessarily constitute control over risk for the purposes of a transfer pricing analysis.

​​​​​​Where a risk has materialized it will be a question of fact as to which party bore the consequences. However, since any analysis of risks must take into account temporality (i.e. past risks where outcomes are known are no longer risks at all), it will be relevant to consider whether a party has the financial capacity to assume a risk. Financial capacity to assume risk can be defined as access to funding to take on or to lay off the risk, to pay for the risk mitigation functions and to bear the consequences of the risk if it materializes. Access to funding takes into account the available assets of the party, as well as the options realistically available to it to access additional liquidity, if needed, to cover the costs anticipated to arise should the risk materialize. The consideration of whether an entity has the necessary financial capacity should be done on the basis that it is operating as an unrelated party in the same circumstances as the entity. For instance, if an entity has the right to exploit income-generating assets, it is likely to be able to access liquidity against its income stream.

It should be noted that the financial capacity to assume the risk is not necessarily the financial capacity to bear the full consequences of the risk materializing (e.g. the full loss): the risk-bearer may have the capacity to protect itself from the consequences of the risk materializing (e.g. by hedging the risk or insuring against the impact of the risk). However, because financial capacity to assume a risk is not by itself sufficient to assume a risk, a high level of capitalization does not necessarily mean that the highly capitalized party assumes the risk.

​​​​​​It is relevant to mention here that in a multinational enterprise associated entities may work together to exert control over the risks of the entire MNE group. Precise distribution of risk among the associated enterprises may be extremely difficult to achieve. The transfer pricing analysis regarding which entity assumes certain risks should therefore be done considering all the facts and circumstances of each case.

​​​​​Steps 4 to 6 of the risk analysis framework analyze the information gathered in the earlier steps to determine the assumption of risk for the purposes of the transfer pricing analysis.

In cases where the contractual assumption of risk is fully supported by the parties’ conduct, including an alignment with the exercise of control and financial capacity to assume the risk, the analysis will be straightforward. That is, where a party, which is assigned a risk under a consistent contract (i.e. one that is followed in practice) (STEP 4(i)) also controls that risk and has the relevant financial capacity (STEP 4(ii)), it will be regarded as assuming the risk for the purposes of understanding and defining the transaction and pricing it under a TP analysis (STEP 6). The fact that another party also performs control functions or has financial capacity will not affect the determination of the assumption of risk under the transfer pricing analysis. In some cases, risks may be contractually shared by more than one party.

​​​​​​When applying the Cost Plus Method (CPM), Resale Price Method (RPM) or Transactional Net Margin Method (TNMM) (see further Chapter 4) it is necessary to choose the party to the transaction for which a financial indicator (mark-up on costs, gross margin or net profit indicator) is tested. The choice of the tested party should be consistent with the functional analysis of the controlled transaction. Attributes of controlled transaction(s) will influence the selection of the tested party (where needed).

​​​​​​In other cases, where the contractual assumption of risk is not aligned with the exercise of control or the financial capacity to assume the risk, the analysis will require an additional step (Step 5). That is, where a party is contractually assigned a risk (or is made to actually bear the costs of the risk when it materializes) (STEP 4(i)) but does not control it, or does not have the relevant financial capacity (STEP 4(ii)), it cannot be regarded as truly assuming the risk. Instead, the party which does exercise control over the risk and has the relevant financial capacity should be allocated the risk (STEP 5). If multiple associated enterprises are identified that both exercise control and have the financial capacity to assume the risk, then the risk should be allocated to the associated enterprise or group of enterprises exercising the most control. This allocation of risk is what should be used to define the transaction and determine the arm’s length price for transfer pricing purposes (STEP 6). The other parties performing control activities should be remunerated appropriately, taking into account the importance of the control activities performed.

​​​​​In exceptional cases, it may be the case that there is no party that both exercises control and has the financial capacity to assume the risk. Such a scenario would rarely occur between independent enterprises and therefore a thorough analysis should attempt to identify the reasons for this. An assessment of the commercial rationality of such a transaction may be necessary (see sections 3.3.2.3 to 3.3.2.6 above).

The assumption of risk based on the analysis above should be compensated with an appropriate anticipated return. Normally, this means that the party or parties assuming the risk will enjoy the potential upside consequences resulting from the playing out of the risk, for example, the profits that result from a successful venture risk; but would also bear the potential downside consequences if the risk materializes resulting in greater costs or lower than expected profits. In a proper transfer pricing analysis, associated enterprises should always be appropriately remunerated for their contributions— the functional analysis considers functions and assets and not only risks. For example, parties performing risk mitigation functions on behalf of an entity assuming risk should be adequately compensated at arm’s length for those functions. Similarly, where a party is performing control functions, this should be taken into account even if it does not assume the risk relating to those control functions. The form of this compensation will depend on the arrangements between the enterprises and the nature of the contribution: it may be appropriate for such a party to share in the potential upside and downside consequences resulting from the playing out of the underlying risk. Alternatively, the contribution might be compensated in a manner that is not contingent on the underlying risk.

Continuing the example from section 3.4.4.6, it is assumed for the purpose of the example that the following are the risks assumed by the respective parties.

Table 3.T.5

Risk Assumption—Exposure and Control Decisions

​​​​​​Based on the foregoing, the risks assumed by P Co. and Q Co., combined with a qualitative evaluation of the intensity of those risks, can be summarized in a table like the one below. It should be emphasized that one type of risk may be more important or have a larger economic impact on prices than another type of risk. Thus, P Co.’s technology risk may affect pricing far more than Q Co.’s inventory or credit risk, even though all three types of risk are assigned three ticks in the chart. Thus, this type of summary cannot be used as a mechanical tool for allocating profit.

Table 3.T.6

Summary of Risks Borne by Each Party 

Economic analysis deals with industry analysis and the circum-stances that may be relevant for determining market comparability. The relevant information on the industry can be broadly classified as follows:

• Global economic trends and developments relating to the indus-try to which the enterprise belongs;
• Economic trends in each taxpayer’s country for the same industry; and
• Market position of the enterprise and surrounding economic conditions.

Care must be exercised while considering global economic trends, as the market trends in the taxpayer’s country and in the country of its associated enterprise and/or of the potential comparables (in the case where foreign comparables are used) could be significantly different. For example, in the 2008 global financial crisis some of the banks and automobile companies reported huge losses globally, but significant profits in emerging economies. Where there are such significant differences between the economic circumstances prevailing in different markets such that it is not possible to eliminate them by making reliable comparability adjustments, then companies from such different markets might not be retained as reliable comparables.

Comparable uncontrolled transactions (“comparables”) are of two types:

• Internal comparables, i.e. transactions between one of the par-ties to the controlled transaction (taxpayer or foreign associated enterprise) and an independent party; or
• Third-party or external comparables, i.e. comparable uncontrolled transactions between two independent parties, neither of which is a party to the controlled transaction.

Undertaking a more detailed classification of the above broad headings may suggest that the following specific factors should be evaluated in performing an industry analysis if they are economically significant for the examined controlled transaction:

• Geographic location of the market;
• Market size;
• Level of the market (e.g. retail or wholesale);
• Competition in the market and the relative competitive positions of the buyers and sellers;
• Availability of substitutes;
• Government regulations of the market;
• Levels of supply and demand;
• Consumer purchasing power;
• Location-specific costs of production including the costs of land, labour, capital, transportation costs etc.;
• Economic conditions of the overall industry, the key value driv-ers in the industry and the date and time of transactions;
• The existence of a cycle (economic, business or product cycle); and 
• Other relevant factors.

Market prices for the transfer of the same or similar property may vary across different markets owing to cost differentials and/or differences in purchasing power and habits prevalent in the respective markets which may affect the market price. Markets can be different for numerous reasons; it is not possible to itemize exhaustively all the market conditions which may influence transfer pricing analysis but some of the key market conditions which influence such an analysis are discussed below.

In general, uncontrolled comparables would first be sought from the geographic market in which the controlled taxpayer operates, because there may be significant relevant differences in economic conditions between different markets. If reliable comparables from the same market are not available, an uncontrolled comparable derived from a different geographical market may be considered if it can be determined that (i) there are no differences between the two markets that would materially affect the price or profit of the transaction or (ii) reasonably reliable adjustments can be made to account for such material differences between the two markets.

An example of a potential issue relating to geographic location is that of “location savings”, which may come into play during a transfer pricing analysis. Location savings are the net cost savings that an MNE realizes as a result of relocation of operations from a high-cost jurisdiction to a low-cost jurisdic-tion. Typically, the possibility to derive location savings may vary from one jurisdiction to another, depending for example on the following:

• Labour costs;
• Raw material costs;
• Transportation costs;
• Rent;
• Training costs;
• Subsidies;
• Incentives including tax exemptions; and
• Infrastructure costs.

It is quite possible that part of the cost savings may be offset at times by “dis-savings” on account of the poor quality and reliability of the power supply, higher costs for transportation, quality control etc. Accordingly, only the net location savings (i.e. savings minus dis-savings) may give rise to an extra profit arising to an MNE due to the relocation of its business from a high-cost to a low-cost jurisdiction.

This Chapter describes several transfer pricing methods that can be used to determine an arm’s length price and describes how to apply these methods in practice. Transfer pricing methods (or “methodologies”) are used to calcu-late or test the arm’s length nature of prices or profits. Transfer pricing methods are ways of establishing arm’s length prices or profits for transactions between associated enterprises. As described in Chapter 3 on Comparability Analysis, the transaction between related enterprises for which an arm’s length price is to be established is referred to as the “controlled transaction”. The selected transfer pricing method is the mechanism by which the prices or results of the controlled transaction are compared to the prices or results of comparable uncontrolled transactions. In this way, the application of transfer pricing meth-ods helps assure that transactions conform to the arm’s length standard. It is important to note that although the term “profit margin” is used, companies may also have legitimate reasons to report losses at arm’s length. Furthermore, transfer pricing methods are not determinative in and of themselves. If an associated enterprise reports an arm’s length amount of income, without the explicit use of one of the recognized transfer pricing methods, this does not mean that its pricing should automatically be regarded as not being at arm’s length and that adjustments should be imposed.

The computation of location savings typically involves the quantification of the net cost savings derived from relocating in a low-cost country, as compared to the relevant high-cost country. In theory, the cost savings computation includes selection of a pre-transfer manufacturing or servicing base in the relevant high-cost country compared to the comparable manufacturing or services cost in the low-cost country, taking into account such things as total labour cost per unit of output (adjustment on account of difference in labour productivity), cost of raw material, costs of land and rent costs; tax benefits etc. The cost savings can be partially offset by higher cost of infra-structure such as less reliable power supplies etc. in certain cases.

Figure 3.D.2

Determining Net Location Savings

Location-specific advantages and location savings are defined as a type of benefit related to geographical location. The relocation of a business may in addition to location savings give some other location-specific advantages (LSAs). These LSAs could be, depending on the circumstances of the case:

• Highly specialized skilled manpower and knowledge;
• Proximity to growing local/regional market;
• Large customer base with increased spending capacity;
• Advanced infrastructure (e.g. information/communication net-works, distribution system); or
• Market premium

Figure 3.D.3

Determining Location-specific Advantages

Even though internal comparables may display a higher degree of comparability, there is a need to subject internal comparables to as rigorous a scrutiny as external ones regarding comparability factors, and to make comparability adjustments when necessary. Use of internal comparables may have advantages but also requires caution as mentioned below; accordingly, this will require careful consideration of the facts and circumstances of each case.

In each case, the objective is to identify the most appropriate method for the particular case. Considerations involved in selecting a method can include: the respective strengths and weaknesses of each method; the nature of the controlled transaction; the availability of reliable information (in particular on uncontrolled comparables) needed to apply the selected method; and the degree of comparability between the controlled and uncon-trolled transactions.

The advantages of internal comparables are:

• Internal comparables may have a more direct and closer relationship to the transaction under review than external ones due to one party to the transaction being the same. This may make functional comparisons easier and may make it easier to conclude that identical accounting standards are being applied;
• Transaction specific financial and other information is more likely to be available; and
• Comparability analysis involving internal comparables may be less expensive for the taxpayer as no public database search is required.

Method selection is an important part of the comparability analysis described in Chapter 3. The starting point in selecting a method is to develop an understanding of the controlled transaction (inbound or outbound), by means of the functional analysis. The functional analysis establishes the foundation necessary to select the most appropriate transfer pricing method as it helps:
►    To identify and accurately delineate the intragroup transactions (the wording “accurate delineation” connotes the act of estab-lishing the facts of the transaction based on available evidence, including the relevant acts of the parties);
►    To identify the characteristics that would make a particular uncontrolled transaction or function suitable for use as a comparable;
►    To determine any necessary adjustments to the comparables;
►    To check the relative reliability of the transfer pricing method selected; and
►    Over time, to determine if modification of the transfer pric-ing method is appropriate because the controlled transac-tion, function, allocation of risks or allocation of assets have been modified.
 

The major components of a functional analysis are analyses of the functions undertaken by each of the parties to the transaction, together with the assets they use or contribute and the risks they assume. The functional analysis is described and discussed in detail in Chapter 3, at 3.4.4. A summary is provided below for context in the case of selection of appropriate methods.

The functions performed: The functional analysis identifies and describes the activities performed by each of the relevant parties such as product design, purchasing, inbound logistics, manufacturing, research and development (R&D), assembling, inventory management, outbound logistics, marketing and sales activities, after sale services, supporting activities, services, advertising, financing and management etc. The functional analysis must spec-ify which party performs each activity and the intensity with which each party performs that function. Where both parties are involved in performing an activity the functional analysis should identify the relevant differences; e.g. if both have inventories but Company A holds inventories for a period of up to two years whereas Company B holds inventories for a period of one month the functional analysis should identify that fact. The activities that add most value must be identified and should be discussed in more detail.

The risks assumed: The functional analysis should identify risks assumed and undertaken by each of the parties. Examples are: financial risk (currency, interest rate, funding risks etc.) credit and collection risk (trading credit risk, commercial credit risk), operational risk (systems failure risk), commodity price risk, inventory risk and carrying costs, R&D risk, environmental and other regulatory risks, market risk (country political risk, reli-ability of customers, fluctuation in demand and prices) and product risk (product liability risk, warranty risk and costs, contract enforceability and the risk of product obsolescence). The party which assumes a risk would expect to have higher earnings than one that does not; and will incur the expenses and perhaps related loss if and when risk materializes.

Internal comparables, where available and reliable, may allow the taxpayer to consider the use of the CUP Method which is the most direct and reliable method if it can be applied. Internal comparables may also be used with the other recognized transfer pricing methods.

The assets used or contributed: The functional analysis must identify the assets used or contributed by each of the parties and should distinguish between tangible and intangible assets. Tangible assets such as property, plant and equipment have to be financed and an investment in such capital assets would usually be expected to earn a long-term return based on the use and risk level of the investment. Intangible assets are very important as substantial competitive advantage is often achieved by the creation and use of intangible assets. Some intangibles have legal protection (e.g. patents, trademarks, trade names) but other intangibles with less legal protection may be equally important and valuable (e.g. know-how, trade secrets, marketing intangibles etc.).²⁸

However, reliable internal comparables may not exist. If that is the case, it will be necessary to examine external sources of potential comparable transactions among third parties.

Interplay of the above factors: Today, in an MNE, operations tend to be more integrated across jurisdictional boundaries and the functions, risks and assets are often shared between entities in different jurisdictions. This makes functional analyses both more difficult and more necessary. The functional analysis can help identify which functions, risks and assets are attributable to the various related parties. For example, the functional analysis may reveal that one company performs one particular function but the cost of this is borne by the other party to the transaction. The functional analysis could highlight that situation and consider the legal allocation of risk and the economic substance of the transaction. Another example would be where a company performs one particular function and bears the cost thereof but the benefit also accrues to the other party to the transaction. The functional analysis could emphasize that situation and consider which party bears the risk in legal terms and which party assumes the risk according to the economic substance of the transaction. The functional analysis typically includes a discussion of the industry in which the tested party operates, the contractual terms of the transaction at issue, the economic circumstances of the parties and the business strategies they employ. The functional analysis helps to identify the operations that would be expected to benefit a related party and require an arm’s length return.

Selecting a method after the functional analysis: Once the functional analysis is performed the application of a transfer pricing method, with the associated evaluation of comparable transactions, may be considered. Transfer pricing methods typically use information on comparables; the lack of such comparables can make a particular method— even one that might seem initially preferred— inapplicable, and a different method more reliable. These comparable transactions are also referred to as “uncontrolled transactions” because the parties involved in the transactions are independent of each other. Although uncontrolled transactions of independent unrelated companies are usually used as comparables for transfer pricing purposes, in practice it is sometimes not possible to identify reliable, closely comparable data in the same markets. In such cases practical solutions should be sought in good faith by taxpayers and the tax administration. Comparability issues are discussed in more detail at Chapter 3.

There are two types of third-party or external comparables. The first type relates to transactions between two independent parties, neither of which is a party to the controlled transaction. For example, it might be possible to apply the CUP Method based on the price of a comparable product sold under comparable circumstances by uncontrolled parties.

Solutions for cases where comparables are difficult to find may include the following:
►    Searching for comparables in other industries where such comparable companies have similar functions, assets and risks;
►    Searching for comparables in other geographical regions that share certain key similarities with the country in which a company conducts its business; and
►    Using industry analyses (publicly available or conducted internally by the company) to identify profit levels that can reasonably be expected for various routine functions (e.g. production, services, distribution).
The suggestions above are not intended to be exhaustive, neither is any preference implied by the ordering of the alternatives. Rather, the approaches above are presented as examples of what might be done and are included for information purposes only. Due to the difficulty in obtaining access to (publicly available) data, in certain instances methods other than the ones presented above may need to be used. Further guidance on this issue can be found in the PCT Transfer Pricing Toolkit (see 2.5.3.2).

​​​​​​The second type of external comparable relates to the use of the results of comparable uncontrolled companies (engaged in comparable trans-actions) when applying profit-based transfer pricing methods. Typically, such results are identified through the use of commercial databases and the application of “screening” criteria. The determination of appropriate screening criteria is a critical step and should be based on the most economically relevant characteristics of the accurately delineated controlled transaction (see section 3.5.2.28 et seq.). The objective of finding the closest comparables must, however, also be balanced with the need to be pragmatic and to find an answer.

Intangibles: Among the factors to be considered in selecting the most appropriate method in the circumstances of the case. It is important to determine which party has developed or acquired any intangibles used and in what capacity, which party has the legal ownership and which party receives the benefit of the intangibles. The party that developed the intangibles and bore the related intangible development risks should be able to obtain benefits from those intangibles for example through:
►    A sale or licensing of the intangibles to another party who exploits it; or
►    Exploiting the intangible itself, for example by way of an increase in the price of products or services that make use of such intangibles.²⁹

There are various sources of data and information which are available to assist a taxpayer or tax administration in identifying potential external comparables. Possible sources range from commercial or electronic databases to regulatory and other government filings and various analytical reports issued by trade and industry associations. The search objective is to identify the most reliable comparables for the controlled transaction under examination according to the specific set of criteria.

There are two general categories of methods. “Traditional Transaction Methods” consist of the Comparable Uncontrolled Price (CUP), the Cost Plus Method (CPM) and the Resale Price Method (RPM). The “Transactional Profit Methods” consist of the Transactional Net Margin Method (TNMM) and the Transactional Profit Split Method (PSM). A number of jurisdictions also apply “other methods” which are considered to provide arm’s length results; however, it needs to be ensured that such methods are properly applied to be consistent with the arm’s length principle.

No preference for particular methods is advocated in this Manual. The most suitable method should be chosen taking into consideration the facts and circumstances. The taxpayer should e.g. take into account the type of transaction, the functional analysis, comparability factors, availability of comparable transactions and the possibility of making adjustments to the data to improve comparability. For further discussion on this issue, see Chapter 3.

Once a method is chosen and applied, taxpayers are generally expected to apply the method in a consistent fashion. Assuming that an appropriate transfer pricing method is being applied, a change in the method is typically required only if there are any relevant changes in the facts, functionalities or availability of data.

The strengths of the RPM include:
►    The method is based on the resale price, a market price, and thus represents a demanddriven method; in situations where there
is a weak relationship between the costs incurred and the sales price of a product or services (e.g. when demand is inelastic), the resale price may be more reliable; and
►    The method can be used without forcing distributors to inappropriately “make profits”. The distributor earns an arm’s length gross profit margin. However, in some circumstances the distributor could legitimately have operating losses which are not due to the transfer price but rather are the result of commercial factors, for example, high selling expenses caused by business strategies such as a market penetration strategy. By comparison, the application of the TNMM, which analyzes a financial ratio based on operating profits, will generally result in an arm’s length range of positive operating profits. The tested party in the analysis would then generally be expected also to earn a positive operating profit within the range. However, the RPM does not necessarily result in positive operating profits to be earned by the tested party.

The weaknesses of the RPM include:
►    It may be difficult to find comparable data on gross margins due to accounting inconsistencies; and
►    The method involves a onesided analysis, as its focus is on the related sales company as the tested party in the transfer pricing analysis. It is possible that the arm’s length gross profit margin and hence transfer price, which is based on a benchmarking analysis, can lead to an extreme result for the related supplier of the sales company (e.g. the supplier might experience a loss even though its distributor is profitable).

In a typical intragroup transaction involving a “fullyfledged” manufacturer (i.e. as compared, for example, with a limited risk or contract manufacturer) owning valuable patents or other intangibles and affiliated sales companies which purchase and resell the products to unrelated customers, the RPM is an appropriate method to use if:
►    The CUP Method is not applicable;
►    The sales companies do not own or contribute to valuable intangibles; and
►    Reliable comparisons with uncontrolled transactions can be made on COGS.

The data sources provide a vast array of information. Some provide simple leads or contacts, or a starting point to learn more about a particular industry so that appropriate comparables are ultimately selected. Others provide business profiles and detailed financial information about potential comparables. Each source can be important in establishing and documenting the quantitative basis for an arm’s length transfer pricing policy.

It is useful to again consider the example of Figure 4.D.2. It is assumed here that Associated Enterprise 1 owns valuable patents to manufacture the bicycles and has a valuable trade name. Associated Enterprise 2 purchases the bicycles from Associated Enterprise 1 and resells the bicycles to unrelated dealers in the local country. In such a case, the RPM will be selected to determine an arm’s length transfer price between Associated Enterprise 1 and Associated Enterprise 2 if the CUP Method cannot be applied. The CPM (discussed below) will not be selected in this case, because:
►    The fullyfledged manufacturer (i.e. Associated Enterprise 1) owns unique and valuable intangibles, performs R&D activities and generally has operations that are more complex than those of the sales company (i.e. Associated Enterprise 2);
►    The results obtained from applying the CPM will not be as reliable as the results obtained from applying the RPM using the sales company as the tested party; and
►    It will be very difficult, if not impossible, to identify manufacturers comparable to Associated Enterprise 1 (i.e. that own comparable intangibles) when applying the CPM.

The RPM will establish the transfer price by reference to the resale or gross margins (gross profit/net sales) earned by thirdparty resellers (assuming that no internal comparables are available) and compare them to the gross margin earned by Associated Enterprise 2 on the bicycles purchased from related parties.

​​​​​A key resource among the general sources of information is that of commercial databases including databases in electronic form. These data-bases have been developed by various organizations which compile accounts filed by companies with the relevant administrative bodies and present them in an electronic format suitable for searches and statistical analysis. Some of these databases compile financial data from one country only, while others compile regional or even global data. These products typically provide detailed financial information as well as some textual information such as short business descriptions. The level of detail available in the database largely depends on the country concerned.

In a controlled transaction involving a sale of tangible property, the CPM generally focuses on the related manufacturing company as the tested party in the transfer pricing analysis. The CPM may also be used in the case of services rendered, in which case the service provider is generally designated as the tested party.

The CPM begins with the costs incurred by the supplier of property or services in a controlled sale of property or services to a related purchaser. An appropriate cost plus markup is then added to this cost, to calculate an appropriate gross profit in light of the functions performed, risks assumed, assets used or contributed and market conditions.

The advantage of commercial databases is that they can provide the ability to sort quickly and retrieve selectively only the potential comparables that meet certain qualitative and quantitative screening criteria. Criteria commonly used for initial screening include industry codes, scale or sales volume, ownership and related/associated enterprises, availability of financial data or certain financial ratios.

The CPM is most often used to analyze transfer pricing issues involving tangible property or services. It is typically applied to manufacturing or assembling activities and relatively simple service providers. The method evaluates the arm’s length nature of an intragroup charge by reference to the gross profit markup on costs earned by independent suppliers of tangible property or services in comparable uncontrolled transactions. That is, it compares the gross profit markup earned by the tested party for manufacturing the product or for providing the service to the gross profit markups earned by comparable companies engaged in comparable transactions.
 

​​​​​​​

Figure 4.D.3. explains this further. Associated Enterprise 1, an electrical goods manufacturer in Country 1, manufactures under contract for Associated Enterprise 2. Associated Enterprise 2 instructs Associated Enterprise 1 on the quantity and quality of the goods to be produced. Associated Enterprise 1 will be guaranteed sales to Associated Enterprise 2 and will face little risk. As Associated Enterprise 1 is less complex in terms of functions, assets and risks than Associated Enterprise 2, the analysis under the CPM would focus on Associated Enterprise 1 as the tested party. Since Associated Enterprise 1 is a simple manufacturer, the CPM may be the best method of analysis in this case. The CPM analyzes whether the gross profit markup earned by Associated Enterprise 1 is at arm’s length by reference to the gross profit margins earned by companies manufacturing comparable goods for (or providing comparable services to) unrelated parties. The CPM thus does not directly test whether the transfer price is at arm’s length by comparing prices. As such, it is a less direct (transactional) method as compared to the CUP Method.

Under the CPM (when applied to sales of tangible property) an arm’s length price equals the controlled party’s cost of producing the tangible property plus an appropriate gross profit markup, defined as the ratio of gross profit to cost of goods sold (excluding operating expenses) for a comparable uncontrolled transaction.

The formula for the transfer price in intercompany transactions of products is as follows: TP = COGS x (1 + cost plus markup), where:
►    TP = the Transfer Price of a product sold between a manufacturing company and a related company;
►    COGS = the Cost of Goods Sold to the manufacturing company; and
►    Cost plus markup = gross profit markup defined as the ratio of gross profit to cost of goods sold. Gross profit is defined as net sales minus cost of goods sold.
 

The financial ratio considered under the CPM is the gross profit markup, which is defined as the ratio of gross profit to cost of goods sold of the supplier of products or services. As discussed above, gross profit equals net sales minus cost of goods sold. For a manufacturing company, cost of goods sold equals the cost of producing the goods sold. It includes direct labour costs, direct material costs and factory overheads associated with production.

It is important to note that commercial databases rely on publicly available information. These databases may not be available in all countries, since not all countries have the same amount of publicly available informa-tion about their companies. Further, due to the different disclosure and filing requirements depending on the legal form of the enterprise, the information may not be in a similar format, making it difficult to compare. Most of these databases are used to compare the results of companies rather than of transactions because third-party information at a transactional level is generally not readily available.

As with the RPM, accounting consistency is extremely important in applying the CPM. Application of different accounting principles to the controlled and the uncontrolled transaction may result in inconsistent calculation of the gross profit. Appropriate adjustments of accounting principles may be necessary to ensure that gross profit markups are calculated uniformly for the tested party and the comparable companies. For example, the comparable manufacturers may differ from the related party manufacturer in reporting certain costs (e.g. costs of R&D) as operating expenses or as cost of goods sold. Differences in inventory valuation methods will also affect the computation of the gross profit markup.

The costs and expenses of a company normally fall into the following three groups: (1) direct cost of producing a product or service (e.g. cost of raw materials); (2) indirect costs of production (e.g. costs of a repair department that services equipment used to manufacture different products); and (3) operating expenses (e.g. selling, general & administrative expenses or SG&A expenses). The gross profit margin used in the CPM is a profit margin that is calculated by subtracting only the direct and indirect costs of production from the sales price. In contrast, a net margin analysis would also consider operating expenses. Due to differences in accounting standards between countries, the boundaries between the three groups of costs and expenses are not the same in each and every case. Suitable adjustments may need to be made. In a situation in which it is necessary to consider certain operating expenses to obtain consistency and comparability, a net margin method may be more reliable than the CPM, as discussed below.

The arm’s length price or margin can result from looking at comparable functionality (manufacturers of broadly similar types of product) or from making a transactional comparison by looking at each transaction the tested party engages in involving comparable products (e.g. manufacturing of different types of bicycle).

The arm’s length (range of) gross profit markup(s) can be established in the following two ways:
►    Transactional comparison: the gross profit markup earned by the related party manufacturer when selling goods to an independent enterprise in a comparable uncontrolled transaction, which previously has been rejected as an internal comparable for purposes of applying the CUP Method because for example it involves different models of bicycle. If for example the controlled transaction involves the manufacturing of recreational bicycles, but the unrelated transactions involve bicycle rickshaws etc., these may involve broadly similar products, with comparable accounting measures of COGS making gross margin comparisons sufficiently reliable; and
►    Functional comparison: the gross profit markups earned by independent companies performing functions and incurring risks comparable to the functions performed and risks incurred by the related party manufacturer. Functional comparison involves a search for comparable manufacturing companies.
 

The Comparable Uncontrolled Price (CUP) Method compares the price charged for property or services transferred in a controlled transaction to the price charged for property or services transferred in a comparable uncontrolled transaction in comparable circumstances. The CUP Method may also sometimes be used to determine the arm’s length royalty rate for the use of an intangible, or to determine an arm’s length rate of interest on a loan. CUPs may be based on either “internal” comparable transactions or on “external” comparable transactions. Figure 4.D.1 below explains this distinction in the context of a particular case study.

 

​​​​​​​

In practice, transactional comparisons are more likely to achieve the broad product and accounting consistency required for the CPM than functional comparisons. In a transactional comparison, much more information about the controlled and uncontrolled transactions is available (e.g. contractual terms). In a functional comparison that is based on information provided in publicly available databases and in the annual reports of comparable companies and the tested party, much less specific information is available with respect to the functions performed and risks incurred by the companies. Consequently, it would be more likely in these circumstances that a net margin method would be used, see paragraph 4.5.2.

Based on benchmarking and financial analyses an arm’s length range of gross profit markups earned by comparable independent manufacturers will be determined. If the gross profit markup earned by the related party manufacturer falls within this range, then its transfer price will be considered arm’s length.

An uncontrolled transaction is considered comparable to a controlled transaction in applying the CPM if:
►    There are no differences between the transactions being compared that materially affect the gross profit markup; or
►    Reasonably accurate adjustments can be performed to adjust for the effect of such differences.
 

As with the RPM, and for the same reasons, close similarity of products in the controlled and uncontrolled transactions is less important under the CPM than under the CUP Method, while functional comparability (including comparability of risks assumed and assets used or contributed) is more important. However, because significant differences in products may necessarily result in significant differences in functions, assets and risks, the controlled and uncontrolled transactions should ideally involve the provision of similar services or the manufacturing of products within the same product family.

As the gross profit markup remunerates a manufacturing or service company for performing a manufacturing or service function, the CPM necessarily requires functional comparability. If there are material differences in functions performed, assets used or contributed and risks assumed that affect the gross profit markups achieved on the controlled and the uncontrolled transactions, adjustments should be made to account for such differences. In general, comparability adjustments should be made on the gross profit markups of the uncontrolled transactions. Sometimes the operating expenses in connection with the functions performed will need to be taken into account because differences in functions performed may be reflected in the operating expenses, as service providers may very well have no COGS.

Application of the CPM entails a number of potential difficulties associated with the determination of the costs (in addition to those associated with inconsistent accounting treatments):
►    The link between costs incurred and the market price can be very weak so that gross profit margins can vary greatly each year;
►    It is important to apply a comparable markup to a comparable cost base;
►    Differences between the tested party and potential comparables should be identified. In this respect, it is crucial to consider differences in the level and types of expenses in connection with the functions performed, assets used or contributed and risks assumed between the controlled and uncontrolled transactions. If differences merely represent the differing efficiencies of the parties being compared, no adjustment to the gross profit markup should be made. If, however, additional functions are being performed by the tested party, then it may be necessary to determine an appropriate additional return to such function and permit a separate return for these additional functions.
Similarly, if the comparables perform functions not performed by the tested party, then the return for such functions should be subtracted from the gross profit margin applied to the controlled transactions of the tested party;
►    Careful consideration should be given to what costs should be excluded from the cost base. An example of costs that should
be excluded are particular costs that are passedthrough (that is, costs explicitly not subject to a markup) in both the tested party and comparable transactions;
►    As with the Resale Price Method, accounting consistency is extremely important. Gross profit markups should be calculated uniformly for the associated enterprise and the independent enterprises;
►    Historical costs should in principle be ascribed to individual units of production. If costs differ over a period, average costs over the period may be used;
►    One can use either budgeted cost or actual cost in applying the CPM. On the one hand using actual costs will better reflect the risks faced by a contract manufacturer.³² On the other hand, third parties will usually use budgeted costs in determining prices of products being sold into the market. That is, they will not charge the customer an additional amount at the end of the year if actual costs are higher than budgeted costs; and
►    As the costs considered in using the CPM are only those of the manufacturer of the goods or the service provider, a problem may arise with respect to the allocation of some costs between the manufacturer or service provider and the purchaser of goods or services.

The strength of the CPM is that the method is based on internal costs, the information on which is usually readily available to the multinational enterprise.

​​​​​​Commercial databases can be a practical and sometimes cost-effective way of identifying external comparables and may provide the most reliable source of information, depending on the facts and circum-stances of the case. However, a number of limitations to commercial databases are frequently identified and commercial databases are not avail-able in all countries. Further, they may be costly to use and many developing countries may not have access to them. The use of commercial databases is not compulsory, and it may be possible to identify reliable comparables from other sources of information, including internal comparables as described above, or a manual identification of third parties (such as competitors) that are regarded as potential sources of comparables for the taxpayer’s controlled transaction.

The weaknesses of the CPM include the following:
►    There may be a weak link between the level of costs and the market price;
►    The data on gross profit markups may not be comparable due to accounting inconsistencies and other factors;
►    Accounting consistency is required between the controlled and uncontrolled transactions;
►    The analysis is onesided, i.e. it focuses only on the related party manufacturer or service provider; and
►    Since the method is based on actual costs, there may be no incentive for the controlled manufacturer to control costs.

The CPM is typically applied in cases involving the intragroup sale of tangible property where the related party manufacturer performs limited manufacturing functions or in the case of intragroup provision of services. The method usually assumes the manufacturer or service provider has low risks, because the level of the costs will then better reflect the value being added and hence the market price.

The CPM is also generally used in transactions involving a contract manufacturer, a toll manufacturer or a low risk assembler which does not own product intangibles and assumes little risk. The related customer involved in the controlled transaction will generally be much more complex than the manufacturer (or service provider) in terms of functions performed (e.g. conducting marketing and selling functions, coordination of production and sales, giving instructions to the contract manufacturer about the quantity and quality of production, and purchasing raw materials in some cases), risks assumed (e.g. market risk, credit risk and inventory risk) and assets used or contributed (e.g. product or other intangibles). The contract manufacturer is thus the less complex and as such should be the tested party in the transfer pricing analysis.

The CPM is usually not a suitable method to use in transactions involving a fullyfledged manufacturer which owns or develops unique and valuable product intangibles as it will be very difficult to locate independent manufacturers with comparable product intangibles. That is, it will be hard to establish a profit markup that is required to remunerate the fullyfledged manufacturer for its unique and valuable intangibles. In a typical transaction structure involving a fullyfledged manufacturer and related sales companies (e.g. commissionaires), the sales companies will normally be the less complex entities involved in the controlled transactions and will therefore be the tested party in the analysis. The RPM is typically more easily applied in such cases.

Facts of the Case Study: The controlled transaction in this figure involves the sale of bicycles between Associated Enterprise 1, a bicycle manufacturer in Country 1, and Associated Enterprise 2, a bicycle importer in Country 2, which purchases, imports and resells the bicycles to unrelated bicycle dealers in Country 2. Associated Enterprise 1 is the parent company of Associated Enterprise 2.

Transactional profit methods analyze the profits arising from particular controlled transactions in order to determine whether the transfer price is at arm’s length. Transactional profit methods can be divided into two categories; the TNMM and the PSM. This section covers the TNMM and the next section covers the PSM.
 

These methods differ from traditional methods in that the analysis is not necessarily based on particular comparable uncontrolled transactions involving identical or perhaps even broadly comparable products. Often, and depending on the facts and circumstances, the analysis is based on the net return (generally, the earnings before interest, tax and extraordinary items, i.e. EBIT) realized by various companies engaged in a particular line of business (that is, a series of transactions that are appropriate to be aggregated). Among other situations, one of the transactional profit methods, the PSM, being a twosided method, may be applied when both of the associated enterprises make unique and valuable contributions (e.g. in the form of intangibles). Detailed guidance on when the PSM may be appropriate can be found at section 4.6.3 below.

4.2.1.3    In applying the CUP Method to determine whether the price charged for bicycles sold in this controlled transaction is at arm’s length, the following information is assumed to be available for consideration:
►    The price charged for bicycles sold in a comparable uncontrolled transaction between Associated Enterprise 1 and Unrelated Party C (i.e. transaction #1);
►    The price charged for bicycles sold in a comparable uncontrolled transaction between Associated Enterprise 2 and Unrelated Party A (i.e. transaction #2); and
►    The price paid for bicycles sold in a comparable uncontrolled transaction between Unrelated Party A and Unrelated Party B (i.e. transaction #3).
 

It is rare that enterprises use transactional profit methods to actually determine their prices. However, the profit resulting from a controlled transaction might be quite a good signal to establish whether the transaction was affected by conditions that differ from those that would have been made between independent enterprises in otherwise comparable circumstances. Where complexities make the application of the traditional transaction methods addressed in the previous chapter unreliable, transactional profit methods may prove to be a good solution.

Transactional profit methods and particularly the TNMM are also commonly used by taxpayers for practical reasons. The TNMM often provides a useful check on the accuracy and reasonableness of the traditional transaction methods or is used to supplement these methods. In some situations, it may also be easier to find comparables that can be used to apply the TNMM than it is to find transactional comparables that can be used to apply the CUP, Resale Price or CPM.

The TNMM examines the net profit margin relative to an appropriate base (e.g. costs, sales, assets) that a taxpayer realizes from a controlled transaction (or transactions that are appropriate to be aggregated). The profit margin indicators are discussed below. The TNMM looks at the profits of one of the related parties involved in a transaction, as do the CPM and RPM. The party examined is referred to as the tested party.

The TNMM compares the net profit margin³³ (relative to an appropriate base) that the tested party earns in the controlled transactions to the same net profit margins earned by the tested party in comparable uncontrolled transactions or alternatively by independent comparable companies. As it uses net margins to determine arm’s length prices the TNMM is a less direct method than the CPM and RPM that compares gross margins. It is also an even more indirect method than the CUP Method that directly compares prices. Many factors may affect net profit margins but may have nothing to do with transfer pricing.

The TNMM is used to analyze transfer pricing issues involving tangible property, intangible property or services. It may be applied when one of the associated enterprises employs intangible assets, the appropriate return to which cannot be determined directly. In such a case the arm’s length compensation of the associated enterprise(s) not employing the intangible asset is determined by determining the margin realized by enterprises engaged in a similar function with unrelated parties. The remaining return is consequently left to the associated enterprise controlling the intangible asset. The return to the intangible asset is, in practice, a “residual category” being the return left over after other functions have been appropriately compensated at arm’s length. This implies that the TNMM is applied to the least complex of the related parties involved in the controlled transaction. This approach has the added benefit that generally more comparable data are available and fewer adjustments are required to account for differences in functions and risks between the controlled and uncontrolled transactions. In addition, the tested party typically does not own valuable intangible property. Where the TNMM is applied in cases that involve intangibles the guidance of chapter 6. of this Manual should be noted. See 6.1.4.

Comparable uncontrolled transactions, such as transaction #1 or #2, which involve a transaction between the tested party and an uncontrolled party, are referred to as internal comparables. Comparable uncontrolled transactions such as transaction #3, which involves a transaction between two parties neither of which is an associated enterprise, are called external comparables. The application of the CUP Method involves a detailed transactional comparison whereby the controlled and uncontrolled transactions are compared based on the five comparability factors mentioned in Chapter 3.

The TNMM is more tolerant of, and flexible in, accommodating certain minor differences in functionality (where such functions are reflected in operating expenses), and in accounting practices than the RPM or the CPM. Table 4.T.1. below and the paragraphs following it illustrate this distinction.³⁴

Table 4.T1
Transactional Net Margin Method (TNMM)

Associated Enterprise 1, a bicycle manufacturer in Country 1, sells bicycles to Associated Enterprise 2 which resells the bicycles to an independent enterprise, an unrelated bicycle dealer in Country 2. Assume that Associated Enterprise 1 is the more complex party, controlling a variety of technology and operating intangibles. Associated Enterprise 2, on the other hand, is a routine distributor/ reseller which does not make unique and valuable contributions. The CUP Method would compare the price charged in the controlled transaction between Associated Enterprise 1 and Associated Enterprise 2 with the price charged in comparable uncontrolled transactions. If the CUP Method cannot be applied, the CPM and RPM may be considered. 

The CPM is likely to be relatively unreliable in this case because it would treat the more complex entity, Associated Enterprise 1, as the tested party. Given that Associated Enterprise 2 performs relatively simple activities, the RPM could be considered. Under the RPM, the sales company, the less complex of the two entities involved in the controlled transaction, will be the tested party. The analysis would entail a search for distributors which sell broadly similar products, which perform functions, use or contribute assets and assume risks comparable to those of Associated Enterprise 2, and for which appropriate data relating to gross profits can be obtained.

​​​​​Examining multiple year data may be useful in a comparability analysis but it is not a systematic requirement. Multiple year data may be used where they add value and make the transfer pricing analysis more reliable. Circumstances that may warrant consideration of data from multiple years include the effect of business cycles in the taxpayer’s industry or the effects of life cycles for a particular product or intangible. However, the exist-ence of any such cycle needs to be aptly demonstrated by the taxpayer.

Sometimes it may be more reliable to choose the TNMM and compare net profits. If, for example, there is different reporting of the cost of goods sold and operating expenses for the tested party and the comparable distributors, so that the gross profit margins reported are not comparable and reliable adjustments cannot be made, the RPM may be relatively unreliable. However, this type of accounting inconsistency will not affect the reliability of the TNMM, as this method examines net profit margins instead of gross profit margins. Also, as further discussed below, the fact that the TNMM requires less product comparability than the traditional transaction methods (and as such has a greater tolerance to product differences and cost accounting differences compared to traditional transaction methods) can be a significant practical benefit of using TNMM.

The application of the TNMM would entail an analysis of the least complex party— in this case the distributor. Such an analysis would entail a search for comparable distributors taking into account the comparability standard of this method. An application of the TNMM focusing on a related party manufacturer as the tested party could be appropriate in a situation in which Associated Enterprise 1 is instead, a contract manufacturer and has no unique and valuable intangibles. In such a case, the contract manufacturer may be the least complex entity, as MNEs often separate the ownership of valuable technology intangibles from the manufacturing function. The CPM would normally be considered if the CUP Method cannot be applied. However, due to the accounting inconsistency mentioned above, it may be appropriate to apply the TNMM using a financial ratio based on net profit margin that is appropriate for a manufacturer (e.g. return on total costs).

​​​​​In addition to information from commercial databases of company results, a number of other sources of information may be useful, including in some cases, price databases, publications and exchange quoted prices for commodities. Such publications may provide useful information on market conditions and prices of standard commodities. They can also be useful in understanding relevant market dynamics for the products concerned. In some cases, it may be appropriate to use quoted prices from commodities or futures exchanges in order to benchmark transfer prices for commodities (see 4.7. on the application of a separate transfer pricing method relating to transactions in commodities). However, as with any such source of potential benchmarking data, its reliability in pricing the tested transaction must be carefully considered, particularly in the case of information in relation to less transparent markets, i.e. those in which information on individual transactions is not generally available to those who are not a party to the transaction. In such cases, the published information will typically be based on the publisher’s observations and contacts with key market participants. While this kind of information can be useful, it should be borne in mind that the publisher may have made adjustments to the raw data in ways that may not be apparent. Such data should therefore be used with care.

The next question is how to determine the transfer price based on the application of the TNMM. The mechanism of the TNMM is similar to the mechanisms of the RPM and CPM, as can be seen in the following examples.

Related party distributor: In applying the RPM to establish an arm’s length transfer price, the market price of products resold by the related party distributor to unrelated customers (i.e. sales price) is known, while the arm’s length gross profit margin is determined based on the results of comparable uncontrolled transactions, i.e. a benchmarking analysis. The transfer price or cost of goods sold of the related party distributor is the unknown variable. Assuming a resale price of $10,000 and a gross profit margin of 25 per cent, the transfer price amounts to $7,500:
Table 4.T.2
Mechanism of the Resale Price Method (RPM)

The quantitative screening step involves further screening of the financial information relating to the potential comparables for the relevant period to determine whether their activities are comparable to that of the tested party, and whether they report sufficient data at the level needed to apply the selected transfer pricing method. Under this step, the search process focuses on a rigorous review of each transaction or company in the potential independent comparable pool against the full range of specific screening criteria. The objectives at this stage are verification, final screening and selection. This process is based on trial and error and requires multiple data sources, cross-checks and selected follow-up and confirmation of factual data. It will often be difficult to find ‘perfect’ comparables for a controlled transaction. Therefore, in undertaking the screening process, judgement is required. If the primary screening is applied too rigorously and inflexibly, it may be the case that no apparent comparables remain. In such cases it is particularly important to focus on the most economically significant characteristics of the controlled transaction while dispensing with other, less critical screening criteria for the transaction at hand. Where this is the case, secondary screening can be particularly useful to refine the set of potential comparables.

​​​​​For example, such screening may be done using diagnostic ratios. Diagnostic ratios are financial ratios applied to reject comparables that do not fulfil certain criteria. Particularly in cases where broad primary screening criteria have been used, diagnostic ratios can be used to improve the reliability of a potential set of comparables by helping to distinguish between results from transactions with differing degrees of comparability, and seeking to eliminate those with a lower degree of comparability from the potential comparable set. One or a combination of diagnostic ratios may be used as a kind of additional screen to narrow a range in cases where comparability defects remain in the potential comparables set that are otherwise difficult to eliminate, resulting in a range that would otherwise be overly wide.

The determination of an arm’s length transfer price based on the
 

TNMM is similar. The main difference from a gross margin analysis is that operating expenses are considered in calculating the transfer price. In applying the TNMM to the tested party distributor, the resale price and the operating expenses of the related party distributor are known, while the arm’s length net profit margin (i.e. net profit to sales ratio)³⁵ is found on the basis of a benchmarking analysis. The cost of goods sold and the gross profit are the unknown variables. Assuming a resale price of $10,000, operating expenses of $2,000 and an arm’s length net profit margin of 5 per cent, using the TNMM, the transfer price of $7,500 is determined by working backwards using the available information. That is, a transfer price of $7,500 is required to ensure that the distributor earns a net profit margin of 5 per cent.

Table 4.T.3
Mechanism of the Transactional Net Margin Method (TNMM)

When applying the CUP Method, an uncontrolled transaction is considered comparable to a controlled transaction if:
►    There are no differences in the transactions being compared that would materially affect the price; or
►    Reasonably accurate adjustments can be performed to account for material differences between the controlled and the uncontrolled transaction.

​​​​​​For example, a ratio of marketing and advertising expenses to sales could be an indicator of the intensity of the marketing and advertising function undertaken. This ratio could then be used to refine the arm’s length range based on comparables with similar levels of marketing / advertising intensity in cases where the tested party makes sales to independent customers. Note that it would generally not be reliable to use a diagnostic ratio which comprises elements that are themselves the subject of related party transactions.

The application of diagnostic ratios is based on the assumption that a diagnostic ratio reflects a value driver of a particular line of business and is a reflection of the comparable functional and risk profile. In practice, it also depends on data availability. Most countries with transfer pricing rules acknowledge that the application of a net margin method is less sensitive to product and functional similarity than a traditional transaction method. However, functional comparability is still required in practice so a proper functional analysis and a good understanding of the tested business are essential in determining what diagnostic ratios may be useful, and to help avoid “cherry picking” or subjective use. Diagnostic ratios enable some of the features of a potential comparable that are economically relevant for the comparable search process to be taken into account when performing the comparable search.

Related party manufacturer: In applying the CPM to establish an arm’s length transfer price, the cost of goods sold by the related party manufacturer is known. The arm’s length gross profit markup is based on a benchmarking analysis. The transfer price or sales revenue of the related party manufacturer is the unknown variable. Assuming cost of goods sold of $5,000 and a gross profit markup of 50 per cent, the transfer price amounts to $7,500.

Table 4.T.4
Mechanism of the Cost Plus Method (CPM)
 

​​​​​In order to identify potential comparables with a similar functional and risk profile a diagnostic ratio measuring, for example, the level of wage costs compared to an appropriate base (e.g. total operating costs or total turnover) can be used as a yardstick to measure the level of technical manpower employed by comparable companies engaged in software development. The identification of a diagnostic ratio will depend upon several factors such as geographical location; the nature of the business, product and services; the product and service market etc. Using diagnostic ratios may help to identify comparables which are in line with the functional and risk profile of the tested party.

In applying the TNMM to the tested party manufacturer instead of the CPM, the cost of goods sold and the operating expenses of the related party manufacturer are known. A benchmarking analysis will determine the arm’s length net profit of the related party manufacturer using a profit level indicator such as the ratio of net profit to total cost. The sales price and the gross profit are the unknown variables. Assuming cost of goods sold of
$5,000, operating expenses of $1,000 and an arm’s length net profit to total cost ratio of 25 per cent, the transfer price amounts to $7,500. Table 4.T.4. illustrates that working backwards using the available information leads to the determination that the sales price (i.e. transfer price in this case) is $7,500.

Table 4.T.5
Mechanism of the Transactional Net Margin Method (TNMM)
 

In performing the comparability analysis, the controlled transactions and uncontrolled transactions should be compared based on the comparability factors mentioned earlier and stated in detail in Chapter 3. In determining the degree of comparability between the controlled transactions and uncontrolled transaction #1 in Figure 4.D.1, for example, the following factors should be taken into account: (i) characteristics of goods being transferred or services provided, (ii) contractual terms, (iii) economic circumstances and (iv) business strategies. For the functional analysis it is necessary to analyze the functions performed, the risks assumed and the assets used or contributed.

​​​​​The diagnostic ratio is applied by using cut-off criteria. With this method, financials of the tested party are used to calculate the diagnostic ratios and these ratios are then used to create minimum or maximum values to reject companies. Once a cut-off is determined, generally all the values above or below a particular range of the cut-off will be eliminated, depending upon the facts and circumstances of each case. Subsequently, based on the functional and risk profile of the tested party, all companies with a diagnostic ratio above and below the cut-off range will be excluded.

There is data from a number of independent operators of wholesale distribution businesses. These potential comparables are further narrowed to select companies in the same industry segment that perform similar functions and bear similar risks to BCO. An analysis of the information available on these taxpayers shows that the ratio of operating profit to sales is the most appropriate profit level indicator, and this ratio is relatively stable where at least three years are included in the average. For the taxable years 2007 to 2009, BCO shows the above results.

After adjustments have been made to account for identified material differences between BCO and the uncontrolled distributors, the average ratio of operating profit to sales is calculated for each of the uncontrolled distributors. Applying each ratio to BCO would lead to the comparable operating profit (COP) for BCO shown in the table below.
 

The data are not sufficiently complete to conclude that it is likely that all material differences between BCO and the uncontrolled distributors have been identified. The Country B taxing authority measures the arm’s length range by the interquartile range of results, which consists of the results ranging from $19,760 to $34,840. Although BCO’s operating income for 2009 shows a loss of $4,600, the tax authority determines that no allocation should be made, because BCO’s average reported operating profit of $20,000 ([$27,000 + $37,600 + $(4,600)]/3) is within the interquartile range, which extends from COP observation 19,760 to 34,840.

​​​​​The final stage in the comparables search involves manual consideration of each potential comparable (particularly in the case where the results concerned are the gross or net profits of potentially comparable companies, rather than individual pricing data). For instance, this may involve a review of websites and other publicly available information on the shortlist of potentially comparable companies to ensure they are as reliable as possible.

The interquartile range of comparable operating profits, based on average results from the uncontrolled comparables and average sales for BCO for the years 2008 to 2010, ranges from $15,500 to $30,000. In determining whether an adjustment for the 2010 taxable year may be made, the taxing authority compares BCO’s average reported operating profit for the years 2008 through 2010 to the interquartile range of average comparable operating profits over this period. BCO’s average reported operating profit is determined without regard to the adjustment made with respect to the 2009 taxable year. Therefore, BCO’s average reported operating profit for the years 2008 to 2010 is ($10,000). Because this amount of income falls outside the interquartile range, the tax authority determines that an adjustment may be appropriate. To determine the amount, if any, of the adjustment for the 2010 taxable year, the taxing authority 

compares BCO’s reported operating profit for 2010 to the median of the comparable operating profits derived from the uncontrolled distributors’ results for 2010. The median of the comparable operating profits derived from the uncontrolled comparables results for the 2010 taxable year is $12,000. Based on this comparison, the taxing authority increases BCO’s 2010 taxable income by $22,000, the difference between the median of the comparable operating profits for the 2010 taxable year and BCO’s reported operating profit of ($10,000) for the 2010 taxable year.

Other sources of comparable data may include the following:

1. Government sources—many governments and regulatory agencies maintain databases on several industries. Such sources can be located on the agency’s Internet websites;
2. Trade institutions and organizations—often these institutions or organizations will maintain databases and research reports, and/ or hold files with data on potential comparables. Generally, these institutions or organizations would be:
   • Chambers of commerce;
   • Trade and professional organizations;
   • Embassies, consulates or trade missions; or
   • International organizations (e.g. the UN, the OECD, the World Bank Group, and the IMF).
3. Other sources of knowledge on competitors or other entities which may make suitable comparables.

Certain adjustments may be needed in order to satisfy the requirements for accuracy and reliability of the comparables so that the financial results of the comparables are stated on the same basis as those of the tested party. However, the following important issues should be considered before such a comparability adjustment is made:

• Quality of data being adjusted: a comparability adjustment may be made only where it improves the reliability of comparables. If the search process for comparables has major shortcomings, the addition of unreliable or multiple adjustments will not produce reliable arm’s length outcomes;
• Purpose of adjustment performed: adjustments should not be made for differences that have no material effect on prices or margins being tested;
• Not every transaction being compared is capable of being adjusted: there are transactions where reasonably accurate adjustments can be quantified and be made, but in other situations, the accuracy of an adjustment is uncertain and speculative and the adjustment therefore should not be made; and
• Reliability and accuracy of the adjustment: the adjustment should be calculated based on objective and verifiable data.

​​​​​​​

​​​​​​Comparability adjustments are part of the comparability analysis and should be appropriately documented in order to ensure their reliability.

​​​​​​Comparability adjustments can be divided into three broad categories:

1. Accounting adjustments;
2. Balance sheet/working capital adjustments; and
3. Other adjustments.

​​​​​​Accounting adjustments: There are various types of difference in accounting standards and practices between the tested party and third parties used as comparables which may lead to measurement errors if adjust-ments are not made. The accounting differences can be grouped under the following categories of classification differences and differences under relevant law or standards.

Product comparability should be closely examined in applying the CUP Method. A price may be materially influenced by differences between the goods sold or services rendered in the controlled and uncontrolled transactions. The CUP Method is appropriate especially in cases where an independent enterprise buys or sells products that are identical or very similar to those sold in the controlled transaction or in situations where services are rendered that are identical or very similar to those rendered in the controlled transaction.

Accounting differences may relate to classification where certain operations are recorded in different accounting lines. For example:

• A sales rebate granted to a customer may result in an adjustment to sales or be recorded as negative sales or marketing expenses depending upon accounting practice, and this may affect gross margins (RPM);
• R&D expenditure may be reflected either in operating expenses or in the cost of sales, thus gross margins are not comparable and this requires appropriate adjustment (CPM); or
  • Similarly, the lack of a clear distinction between direct costs and indirect costs affects gross margins. Many of these classification differences are eliminated by applying the TNMM. However, even when using TNMM on a net margin level some accounting differences may exist which can affect net margin in the same way as gross margins resulting in differences between the tested party and comparable. Examples include different depreciation periods, different accounting treatment of employee’s stock options etc.

Other accounting differences that may warrant adjustments relate to situations where a comparable or tested party may have a choice under relevant law or standards to capitalize or expense certain costs like R&D expenses. Thus, a company may have developed significant intangibles but have no intangible property in its assets on the balance sheet. Similarly, different accounting law or standards may be applicable to goodwill recognition and amortization which may create significant discrepancies between the comparables and the tested party. In many cases it is difficult to identify differences in accounting standards due to the following reasons:

• Limited amount of detail available with regard to comparables in the public domain;
• Potential inconsistencies in the reporting of company financial data by private reporting services;
• Inconsistencies among methods of reporting among com-panies; and
• Different accounting standards followed in different countries.

​​​​​​Balance sheet adjustments are intended to account for different levels of inventories, receivables, payables, interest rates etc. The most common balance sheet adjustments, made to reflect differing levels of accounts receivable, accounts payable and inventory, are known as working capital adjustments. The fact that balance sheet adjustments are found most commonly in practice does not mean that they should be performed on a routine or mandatory basis. A significantly different level of asset intensity may require further investigation of the comparability characteristics of the potential comparable and merely making a working capital adjustment would not alleviate the problem.

It is very common for the tested party and each of the potential comparables to differ materially in the amount of working capital (inventory, accounts receivable and payable). Such differences are generally caused by differences in the financing terms of purchases and sales that the company receives from its suppliers and extends to its customers, and by differences in the levels of inventory held by the company. Such differences may generate substantial differences in the working capital structure and may have an impact on the operating profits of the companies due to the financing costs. In order to reduce the effect of differences in terms of purchases and sales and levels of inventory on profitability, adjustments can be made to reflect the time value of the receivables, payables and inventory of the comparables. This, however, should be done only if such adjustments can be reasonably made and they improve comparability.

Although product comparability is important in applying the CUP Method, the other comparability factors should not be disregarded. Contractual terms and economic conditions are also important comparability factors. Where there are differences between controlled and uncontrolled transactions, adjustments should be made to enhance reliability.

​​​​​Adjustments for inventory, accounts receivable and accounts payable follow the same basic mechanics. First a value is calculated as the difference between the ratio of the balance sheet item in question to net sales for the comparables and the same ratio for the tested party. The denominator of these fractions will be an arm’s length amount for the tested party, for example the denominator of a Profit Level Indicator (PLI) can be used. An alternative approach would be to calculate these ratios with respect to operating expenses such as where gross profit/operating expenses are the PLI used. The resulting difference in the ratios is then multiplied by an interest rate and by the net sales of the comparables to generate an amount to adjust the income statements of the comparables. Then the PLI of that comparable is recomputed.

Several profit level indicators (PLIs) are allowed under the TNMM, typically based on operating profit. A PLI is a measure of a company’s profitability that is used to compare comparables with the tested party. A PLI may express profitability in relation to (i) sales, (ii) costs or expenses or (iii) assets. More specifically, the PLI can be the operating profit relative to an appropriate base (e.g. costs, sales or assets). With the help of an appropriate profit level indicator the net profitability of the controlled transaction is compared to the net profitability of the uncontrolled transactions.

Table 4.T.6
Overview of Profit Level Indicators³⁸

Key Definitions:
►    Gross profit is arrived at by deducting from the total sales the cost of sales, including all the expenses directly incurred in relation to those sales;
►    Operating profit or operating income is the income of a company net of direct and indirect expenses but before deduction for nonoperating expenses such as interest and taxes. It is defined as sales minus COGS minus operating expenses (alternatively expressed as gross profit minus operating expenses). “Operating profit” is a better term than “net profit” in this context because net profit is also used to represent the profit of a company after interest and taxes have been subtracted. Further, the term operating profit indicates more clearly that only profits resulting from operating activities are relevant for transfer pricing purposes.
 

Although all of the above PLIs are possible, the three PLIs: (i) return on capital employed (ROCE), (ii) operating margin (OM) and (iii) return on total cost (ROTC) are most used in practice. The Berry Ratio may also be used, but subject to certain concerns about its inappropriate use.³⁹ An OM is typically used for marketing, sales and distribution activities; a Berry Ratio may sometimes be used for service or distribution activities; and full cost plus, ROCE or ROA are typically used for manufacturing activities. The ROA and ROCE divide operating profit by a balance sheet figure. These PLIs are based on assets actively employed in the business. Such tangible assets consist of all assets minus investments (e.g. in subsidiaries), minus cash and cash equivalents beyond the amount needed for working capital. In the case of the ROA a deduction is also made for intangible assets such as goodwill. These two PLIs may, for example, be used for leasing companies. This type of PLI may be the most reliable if the tangible operating assets have a high correlation to profitability. For example, a manufacturer’s operating assets such as property, plant and equipment could have more impact on profitability than a distributor’s operating assets, since often the primary valueadded by a distributor is based on services it provides and these are often less dependent on operating assets. The difference between the ROA and the ROCE is that the ROA focuses on the assets used or contributed while the ROCE focuses on the amount of debt and equity capital that is invested in the company.

Reasonably accurate adjustments may be possible for differences in:
►    The type and quality of the products. E.g. unbranded Kenyan coffee beans as compared with unbranded Brazilian coffee beans;
►    Delivery terms. E.g. Associated Enterprise 1 in Figure 4.D.1 sells similar bicycles to Associated Enterprise 2 and Unrelated Party C. All relevant information on the controlled and uncontrolled transactions is available to Associated Enterprise 1, and hence it is probable that all material differences between the transactions can be recognized.³⁰ The uncontrolled price can be adjusted for the difference in delivery terms to eliminate the effect of this difference on the price;
►    Volume of sales and related discounts. E.g. Associated Enterprise 1 sells 5,000 bicycles to Associated Enterprise 2 for US$90 per bicycle, while it sells 1,000 similar bicycles to Unrelated Party C. The effect of the differences in volume on price should be analyzed, and if the effect is material, adjustments should be made for example based on volume discounts in similar markets;
►    Product characteristics. E.g. the uncontrolled transactions to an unrelated party in Figure 4.D.1 involve bicycles on which modifications have been made. However, the bicycles sold in the controlled transactions do not include these modifications. If the product modifications have a material effect on price, then the uncontrolled price should be adjusted to take into account this difference in price;
►    Contractual terms. E.g. Associated Enterprise 1 sells the bicycles to Associated Enterprise 2 offering a 90day credit term but the contract terms dictate that all sales to Unrelated Party C are cash on delivery;
►    Allocation of risks. E.g. Associated Enterprise 1 is exposed to inventory risk related to sales by Associated Enterprise 2 and the risk that customers of Associated Enterprise 2 will default on their bicycle purchase loans; whereas in the transaction between Associated Enterprise 1 and Unrelated Party C, the latter is exposed to the inventory risk and the risk of its customers’ default. This difference in risk allocation must be analyzed and its effect on price quantified before Associated Party 2’s prices and Unrelated Party C’s prices can be considered comparable; and
►    Geographical factors. E.g. Associated Enterprise 1 sells bicycles to Associated Enterprise 2 located in South Africa, while Unrelated Party C, to which it also sells the same bicycles, is located in Egypt. The only material difference that could be identified between the controlled and uncontrolled transactions concerns the locale. To perform adjustments to account for this
difference one might have to consider, for example, differences in inflation rates between South Africa and Egypt, the competitiveness of the bicycle market in the two countries and differences in government regulations if relevant.

Other PLIs listed above are ratios between income statement items. PLIs based on income statement items are often used when fixed assets do not play a central role in generating operating profits. This is often the case for wholesale distributors and service providers. Operating margin has often been used when functions of the tested party are only broadly similar but not close to those of the comparables, since differences in function have less effect on operating profit than on gross profit.

The Berry Ratio represents a return on a company’s valueadded functions on the assumption that these valueadded functions are captured in its operating expenses. It has been observed in practice that the Berry Ratio is used as a PLI for distributors and service providers. The Berry Ratio assumes that there is a relationship between the level of operating expenses and the level of gross profits earned by distributors and service providers in situations where their valueadded functions can be considered to be reflected in the operating expenses. Consequently, it may be appropriate to use the Berry Ratio if the selling or marketing entity is a service provider entitled to a return on the costs of the provision of its services. It should be noted, however, that the Berry Ratio is very sensitive to functions and cost classification. The Berry Ratio has other limitations, including that it tends to only work if the comparables are functionally similar; and the comparables do not own significant intangible assets.

In general, the gross margin has not been favoured as a PLI because the categorization of expenses as operating expenses or cost of goods sold may be somewhat arbitrary or even subject to manipulation, making comparisons between the tested party and comparables difficult or impossible.

The choice of PLI depends on the facts and circumstances of a particular case. Thus, it may be useful to consider multiple PLIs. If the results tend to converge, that may provide additional assurance that the result is reliable. If there is, on the other hand, a broad divergence between the different PLIs it may be useful to examine important functional or structural differences between the tested party and the comparables.

The arm’s length (range of) net profit margins can be determined by way of:
►    Transactional comparison: the net profit margin that the tested party enjoys in a comparable uncontrolled transaction which initially has been rejected as an internal comparable under the CUP Method due to its strict requirements; and
►    Functional comparison: the net profit margins enjoyed by independent companies performing functions and assuming risks comparable to those of the tested party.
 

In identifying potentially comparable uncontrolled transactions or enterprises two approaches are possible: the “additive” and the “deductive”.

Much more detailed information will be available with respect to the controlled and uncontrolled transactions if a transactional comparison is possible, because the related parties involved have participated in these transactions. The degree of comparability can then be analyzed more thoroughly
 
than in a functional comparison in which only public information is available (e.g. business descriptions in a database, annual reports and Internet data). This may imply that the reliability of transactional comparisons will be higher than that of functional comparisons in practice. In fact, if sufficient data exist to reliably apply a TNMM based on a transactional comparison it may be possible to apply a traditional transaction method.
 

Figure 3.D.6

Working Capital Adjustments

However, functional comparison will be more often used in practice as the data necessary for functional comparison may be available whereas the data needed for transactional comparison is not. Let us assume that a related party distributor is the tested party in the example presented in Table 4.T.7. The TNMM is applied and the profit level indicator is the operating margin. A benchmarking analysis is performed, identifying four comparable independent distributors considering the comparability standard of the TNMM. The arm’s length range of operating margin earned by these comparable distributors falls between 2 per cent and 6 per cent. Because the operating profit margin earned by the related party distributor falls within this range (e.g. 4 per cent), its transfer price is considered to be at arm’s length.

Table 4.T.7
Functional Comparison Example
 

Product comparability is most important in applying the CUP Method, as differences in products will result in different prices. The CPM and the RPM are less dependent on product comparability and focus on func-tional comparability because differences in functions that are reflected in differences in operating expenses may lead to a broad range of gross margins. However, the TNMM is even less dependent on product comparability and functional comparability than the traditional transaction methods, because net margins are less influenced by differences in products and functions. The TNMM focuses on broad product and functional comparability.

However, the comparability standard to be applied to the TNMM requires a high degree of similarity in several factors between the tested party and the independent enterprises that may adversely affect net margins. Net margins may be affected by factors that have no effect, or a less significant effect, on gross margins or prices due to the variation of operating expenses between companies. These factors may be unrelated to transfer pricing.

4.2.2.6    Reasonably accurate adjustments may not be possible for:
►    Unique and valuable trademarks. E.g. assuming Associated Enterprise 1 in Figure 4.D.1 is engaged in manufacturing high value branded goods, and attaches its valuable trademark to the goods transferred in the controlled transaction, while uncontrolled transaction #1 concerns the transfer of goods that are not branded. The effect of the trademark on the price of a watch may be material. However, it will be difficult, if not impossible, to adjust for the effect of the trademark on price since the trademark is an intangible asset that is unique. If reasonably accurate adjustments cannot be made to account for a material product difference the CUP Method may not be the appropriate method for the transaction; and
►    Fundamental differences in the products. E.g. if the products being sold are significantly different from the products sold in the proposed comparable transaction it may not be possible to adjust for the product differences.

Specific factors that may affect net margins include, but are not limited to:

• Barriers to entry in the industry;
• Competitive position;
• Management efficiency;
• Individual business strategies;
• Threat of substitute products;
• Varying cost structures (e.g. the age of plant and equipment); and
• The degree of business experience (e.g. start-up phase or mature business).

If material differences between the tested party and the independent enter-prises are affecting the net margins, reasonably accurate adjustments should be made to account for such differences.

The TNMM is less reliable when applied to the aggregate activities of a complex enterprise engaged in various different transactions or functions. The method should be used to analyze only the profits of the associated enter-prise that are attributable to simpler controlled transactions or functions. The TNMM should thus generally not be applied on a company-wide basis if the company is involved in a number of different controlled transactions or functions which are not properly evaluated on an aggregate basis. However, it may be possible to apply the TNMM when the aggregate activities/transac-tions are sufficiently interlinked, as for example when similar sales functions are conducted for products in similar product lines.

The TNMM should be applied using transactions or functions of independent enterprises that are comparable to the controlled transactions or functions being examined. Furthermore, results attributable to transactions between the tested party and independent enterprises should, if possible, be excluded when evaluating controlled transactions. The latter point is illus-trated in Table 4.T.8. below. In this example, the Related Party Distributor purchases products from both the Related Party Manufacturer and an Unrelated Manufacturer and resells these products to customers. The tax authorities in the country of the Related Party Distributor apply the TNMM to determine whether the transfer prices of the Related Party Distributor are at arm’s length. A benchmarking study performed by the tax authorities shows that comparable distributors earn an operating profit margin between 2 and 6 per cent.

The tax authorities initially apply the TNMM to the profit and loss statement (P&L) of the Related Party Distributor as a whole. The operating profit margin earned by Related Party Distributor is 2 per cent based on aggregate transactions and therefore falls within the arm’s length range. The aggregated transactions therefore appear to be at arm’s length leading to a preliminary conclusion that no transfer pricing adjustment is required. However, if the TNMM was applied only to the results of the distributor from the resale of products acquired in controlled transactions, the conclusions would be very different. The operating profit margin earned by Related Party Distributor on its resale of the products acquired in the controlled transactions is minus 3 per cent, which falls outside the arm’s length range of comparable operating profit margins and therefore appears to merit an adjustment. It appears from the P&L that in this example the resale of products acquired by Related Party Distributor in controlled transactions generated operating losses, which reduced the aggregate income of the distributor as a whole. Consistency is important in quantifying these amounts. Net margins should be calculated uniformly between the tested party and the independent enterprises.

Table 4.T.8
Specific Transactions Versus Company as a Whole
 

An analysis considering multiple year data is sometimes better able to take into account the effects on profits of product life cycles and shortterm economic conditions. Different countries may take different views about when multiple year data should be analyzed and indeed whether that is allowed under a country’s domestic law; see multiple year data analyses at the examples at 4.5.5.3 and 4.5.5.4. Use of an arm’s length range should also be considered, to reduce the effects of differences between the controlled and uncontrolled entities. However, the use of a range may not sufficiently take into account circumstances where the profits of a taxpayer are affected by a factor unique to that taxpayer.

In the additive approach a list is prepared of potentially comparable uncontrolled transactions or of third parties which are believed to be carrying out potentially comparable transactions. As much information as possible on these transactions is then collected to confirm whether they are in effect acceptable comparables, based on the economically relevant characteristics for the controlled transaction. When adopting the additive approach special care should be taken in order to provide a reliable comparable; it is not sufficient that a third-party company be well-known in the relevant industrial sector. Also, one needs to avoid potential third-party companies who themselves have transfer pricing issues.

The strengths of the TNMM include the following:
►    Net margins are less affected by transactional differences than price and less affected by functional differences than gross margins. Product and functional comparability are thus less critical in applying the TNMM;
►    Less complex functional analysis is needed, as the TNMM is applied to only one of the related parties involved;
►    Because the TNMM is applied to the less complex party, it can be used even though one of the related parties holds intangibles for which comparable returns cannot be directly determined;
►    The TNMM is applicable to either side of the controlled transaction (i.e. to either the related party manufacturer or the distributor); and
►    The results resemble the results of what could be referenced as a modified RPM or CPM of analysis.
 

The weaknesses of the TNMM include the following:
►    Net margins are affected by factors (e.g. variability of operating expenses) that do not have an effect, or have a less significant effect, on price or gross margins. These factors affect net profits and hence the results of the TNMM but may have nothing to do with the company’s transfer pricing. It is important to consider these (nonpricing) factors in the comparability analysis;
►    Information challenges, including the unavailability of information on profits attributable to uncontrolled transactions;
►    Measurement challenges, as these may make it difficult to determine sales revenue, operating expenses and assets relating only to the relevant controlled transactions or functions in order to calculate the selected profit level indicator. For example, if a related party distributor purchases products from both a related party and an unrelated enterprise for resale it may be impossible to determine sales revenue, operating expenses and assets attributable to only the controlled transactions to reliably perform a net margin method of analysis. Furthermore, if the companies are engaged in different activities it may also be very difficult to allocate sales revenue, operating expenses or assets between the relevant business activity and other activities of the tested party or the comparables. This measurement problem is an important consideration in practice;
►    The TNMM is applied to only one of the related parties involved. The arm’s length net margin found may thus result in an extreme result for the other related parties involved in the controlled transaction (e.g. operating losses to one of the parties while
the other party is guaranteed a net profit). This weakness also applies to the CPM and RPM but may be more important under the TNMM because net margins are affected by factors that may have nothing to do with transfer pricing. A check of the results of all related parties involved may therefore be appropriate;
►    It may be difficult to “work back” to a transfer price from a determination of the arm’s length net margins; and
►    Some countries do not recognize the use of TNMM. Consequently, the application of TNMM to one of the parties to the transaction may result in unrelieved double taxation when the results of the TNMM analysis are not accepted for the other party.

TNMM is usually applied with respect to broad comparable functions rather than particular controlled transactions. Returns to these functions are typically measured by a PLI in the form of a net margin that may be affected by factors unrelated to arm’s length pricing. Consequently, one might expect the TNMM to be a relatively disfavoured method. Nevertheless, TNMM is often applied when two related parties engage in a continuing series of transactions and one of the parties makes a less complex and more routine contribution to profits than does the other. In fact, it has become the most commonly applied transfer pricing method in crossborder disputes, largely because of the unavailability of good transactional comparables in many circumstances.
 

TNMM may also be appropriate for use in certain situations in which data limitations on uncontrolled transactions make it more reliable than traditional methods. TNMM may be more attractive if the data on gross margins are less reliable due to accounting differences (i.e. differences in the treatment of certain amounts as cost of goods sold or operating expenses) between the tested party and the comparable companies for which no adjustments can be made as it is impossible to identify the specific costs for which adjustments are needed. In such a case, it may be more appropriate to use TNMM to analyze net margins, a more consistent measured profit level indicator than gross margins in case of accounting differences.

Consider the example in Table 4.T.9. below, where the related party distributor earns a gross profit margin of 20 per cent while the comparable distributor earns a gross profit margin of 30 per cent. Based on the RPM one could conclude that the transfer price of the related party distributor is not at arm’s length. However, this conclusion may be incorrect if, due to accounting inconsistency, the related party differs from the comparable distributor in allocating costs between cost of goods sold and operating expenses.

Notwithstanding the difficulties often associated with adjustments to address the sources of noncomparability described above, the need to make adjustments should not automatically prevent the use of the CUP Method. It is often possible to perform reasonably accurate adjustments. Where there are material differences between the transactions, if reasonable adjustments cannot be performed the reliability of the CUP Method is decreased. In these circumstances another transfer pricing method may be more appropriate.

For example, it may be the case that the related party distributor treats warranty costs as cost of goods sold while the comparable distributor treats such costs as operating expenses. If the warranty costs of the comparable distributor can be identified precisely, then appropriate adjustments on the gross profit level can be made. In practice, however, such detailed information about independent enterprises cannot be obtained from publicly available information. It may then be more appropriate to perform a net margin method of analysis where such accounting inconsistency has been removed. The result of applying the TNMM is that the net profit margin of 10 per cent for the related party distributor is similar to that of the comparable distributor. The transfer price is therefore considered to be at arm’s length based on the TNMM.

Also, if the available comparables differ with respect to products and functions, making it difficult to reliably apply the CPM or RPM at gross profit level, it may be more appropriate to apply the TNMM (at operating profit level) because net margins are generally less affected by such differences. For example, in performing a benchmarking analysis for the purposes of the CPM or RPM it may appear that exact product and functional comparables cannot be found. In fact, the comparables differ substantially regarding product and functional comparability. In such a case the TNMM might be more reliably applied using such comparables.

Finally, TNMM may be attractive if the data are simply not available to perform a gross margin method of analysis. For example, this may be the case if the gross profits of comparable companies are not published and only their operating profits are known. The cost of goods sold by companies may also not be available, therefore only a net margin method of analysis can be applied.

In addition to the three situations mentioned above, the TNMM is also used in practice by tax authorities to identify companies for an audit by analyzing their net profit margins. Furthermore, the TNMM is often applied to check and to confirm the results of traditional transactional methods. For example, the TNMM may be used in combination with the RPM to determine an arm’s length compensation for a distribution company.

Table 4.T.9
Accounting Differences: The Resale Price Method (RPM) as Compared with the Transactional Net Margin Method (TNMM)

 

Transactional profit methods analyze the profits arising from particular controlled transactions in order to determine whether the transfer price is at arm’s length. Transactional profit methods can be divided into two categories; the Transactional Net Margin Method (TNMM) and the Profit Split Method (PSM). This section covers the PSM.

​​​​​​The deductive approach usually commences with a search on a database for comparable companies or transactions. These can be commercial databases developed by editors who compile accounts filed by companies with the relevant governmental authorities, or proprietary databases developed by advisory firms. The approach typically starts with the identification of a wide set of companies that operate in the same sector of activity, perform broadly similar functions, and do not present economic characteristics that are obviously different.

The PSM is a useful, but often complex method of determining transfer prices based on an allocation of the relevant, combined profits made by the related parties in relation to the transaction(s).

The PSM seeks to eliminate the effect on profits of special conditions made or imposed in a controlled transaction (or in controlled transactions that it is appropriate to aggregate) by determining the division of profits that independent enterprises would have expected to realize from engaging in the transaction or transactions.

The PSM may be appropriate where:
►    each related party to the transaction makes unique and valuable contributions; and/or
►    the business operations of the related parties are so highly integrated that they cannot be reliably evaluated in isolation from each other; and/or
►    the parties share the assumption of economically significant risk or separately assume closely related risks.
See section 4.6.3 et seq.

The PSM starts by identifying the relevant profits, or indeed losses in relation to the controlled transactions. It then seeks to split those profits or losses between the associated enterprises involved on an economically valid basis in order to achieve an arm’s length outcome for each party. Typically, the split should reflect the relative value of each enterprise’s contribution, including its functions performed, risks assumed and assets used or contributed.

The PSM is also referred to as the transactional PSM. It can be distinguished from global formulary apportionment approaches in the following ways. The PSM typically does not start with the global or total combined profits of the entire MNE group. Rather, it begins from the relevant profits in relation to particular transactions between two or more associated enterprises. Moreover, in order to comply with the arm’s length principle, the way in which the method is applied should not be arbitrary, but rather should approximate the results achieved had the parties been independent of each other. In particular, the factors by which the relevant profits are split between the associated enterprises to the transaction are typically based on measures of their relative contributions to value creation rather than an arbitrary formula.

It should be emphasized that the exclusive use of either of the two approaches may not yield valuable results. Depending on the facts of each case, one of the above two approaches can be used or both in combination.

The strengths of the PSM include:
►    It can provide a solution in cases where onesided methods are not appropriate because each party to the transaction makes a unique and valuable contribution which cannot be benchmarked;
►    It can be used where the level of integration, or the sharing of risks between the related parties means that the contribution of each party cannot be evaluated in isolation from those of other parties;
►    As it is a twosided method, all relevant parties to the transaction are directly evaluated, helping to ensure an arm’s length result for each entity based on the relative value of its specific contributions, even in cases where there may be specific or unique facts and circumstances which may not be present in transactions between independent enterprises; and
►    It is able to deal with returns to synergies between contributions or profits arising from economies of scale.

The weaknesses of the PSM include:
The PSM is often complex to apply. It may be difficult to measure the relevant revenues and costs to be split between the related parties. In addition to measurement difficulties, the method is typically highly reliant on detailed data from the MNE (see also Chapter 12 of this Manual, which deal with transfer pricing documentation). Determining an appropriate way to split the profits can also be challenging. Care must be taken to ensure the application of the PSM is as objective as possible. Since reliable, direct information on the allocation of profits in comparable independent transactions is relatively rare, the PSM often relies on less direct information or proxies (e.g. relative value of the contributions of each party) in its application of the arm’s length principle.
 

As with any transfer pricing method, the profit split should be used where it is found to be the most appropriate method to the circumstances of the case. Primarily, this determination is based on the nature of the accurately delineated transaction in the context of its circumstances. The analysis to determine the accurately delineated transaction should consider the commercial and financial relations between the related parties, their functions performed, assets used or contributed, and risks assumed, and how the activities of the parties impact the transaction given the market context in which the transaction occurs.

While it is not possible to be prescriptive, as noted above, indicators that a profit split may be the most appropriate method include:
►    Where each related party to the controlled transaction makes a unique and valuable contribution; and/or
►    Where the business operations of the related parties are so highly integrated that the contributions of the parties cannot be reliably evaluated in isolation from each other; and/or
►    Where the related parties either share the assumption of the key economically significant risks associated with the transaction(s), or separately assume closely related economically significant risks associated with the transactions.
 

While as noted above, the PSM can be a complex method to apply reliably, the determination of when it is the most appropriate method should be done as objectively as possible. That is, the PSM should not simply be regarded as a method of last resort. Moreover, while the method may require relatively more, or more detailed information from the taxpayer and its associated enterprise(s) than other methods, where it is indeed found to be the most appropriate method, reasonable efforts should be made to gather such necessary information which, after all, will typically be in the hands of the MNE.

The strengths of the CUP Method include that it:
►    Is a twosided analysis as the price used reflects the agreed price between two unrelated parties to the transaction;
►    Avoids the issue of which of the related parties involved in the controlled transaction should be treated as the tested party for transfer pricing purposes;³¹
►    Involves a direct transactional comparison of a similar transaction between unrelated parties. That is, it is a more direct measure of the arm’s length price than the other methods, all of which
indirectly determine arm’s length prices through analysis of the arm’s length profits. As it is a more direct measure, the CUP Method is less susceptible to differences in nontransfer pricing factors (such as differences in the accounting treatment of costs between controlled and uncontrolled parties); and
►    May be more readily used in instances such as, for example, transactions involving commodity products.

The presence of any one or more of these indicators suggests that the profit split may be the most appropriate method.

Where one or more of the above indicators is present, it is highly unlikely that reliable comparable transactions will be available. However, a lack of comparables per se is insufficient evidence to conclude that a profit split will be the most appropriate method. That is, the PSM should not become a convenient method to be applied in every case where close comparables cannot be identified.

In contrast, where none of the indicators are present and the accurate delineation of the transaction shows that one of the related parties to the transaction performs functions, uses or contributes assets and assumes risks that can be reliably benchmarked by reference to uncontrolled comparables, a profit split is unlikely to be the most appropriate method. In such cases, it is likely to be more reliable to apply a transfer pricing method making use of the uncontrolled comparables, even in cases where ‘perfect’ or closely comparable uncontrolled transactions are lacking (see 4.1.2.9). As with any other method, pricing practices used between independent parties engaged in similar transactions in the same industry or market can provide information relevant to the analysis of the most appropriate transfer pricing method.

It is sometimes argued that since the PSM is seldom used among independent enterprises, its application in controlled transactions should be similarly rare. Whether or not the premise of this argument is correct, where the method is found to be the most appropriate to the circumstances, this should not be a factor. Transfer pricing methods, including the PSM, are not necessarily intended to replicate the way in which independent parties establish prices among themselves; rather, they are a way in which the arm’s length principle can be applied in order to determine appropriate transfer prices in controlled transactions. That said, if there is evidence (e.g. from a joint venture or similar arrangement) that independent parties in comparable circumstances use a PSM among themselves, this may suggest that a profit split will also be the most appropriate method in relation to the controlled transactions.

Unique and Valuable Contributions by Each Party
Perhaps the clearest indicator that the PSM may be the most appropriate method involves situations in which each party to the controlled transaction makes unique and valuable contributions. Such contributions (e.g. functions performed, assets used or contributed, including intangibles) will be “unique and valuable” where:
(i)    they are not comparable to contributions made by uncontrolled parties in comparable circumstances; and
(ii)    they represent a key source of actual or potential economic benefits in the business operations.
Together, these factors mean that the application of other transfer pricing methods may not be capable of reliably determining an arm’s length outcome because neither related party can be reliably benchmarked by reference to comparables.

It is possible that companies identified using the additive approach may not have been identified when using the deductive approach. This may in some cases suggest that the search strategy applied under the deductive approach is not sufficiently robust and should be reassessed, or simply that certain information is not contained in the database selected. Therefore, the additive approach could be useful for assessing whether the deductive search strategy is reliable, comprehensive and appropriate given the economic characteristics being considered.

When evaluating whether certain contributions are unique and valuable such that a PSM may be the most appropriate, a consideration of the context of the transaction, including the industry and market in which it occurs and the factors which affect business performance in that context are particularly relevant. See section 6.2.2 et seq. in relation to unique and valuable intangibles. See also section 6.3.4.2 in relation to transfers of fully developed intangibles and section 6.5.4 in relation to transfers of partially developed intangibles or intangibles whose future profit potential is very uncertain.

Highly Integrated Operations
All MNEs have business operations which are integrated to some degree. However the PSM is likely to be the most appropriate method only in those cases where the integration is so significant that the way in which each party performs functions, uses assets, and assumes risks is interlinked with and cannot be reliably evaluated in isolation from the way in which another related party to the transaction performs functions, uses assets and assumes risks.

One example of highly integrated operations which may warrant the determination that the profit split is the most appropriate method could be where the related parties perform functions jointly, use common assets jointly and/or share the assumption of economically significant risks, and do so to such an extent that their respective contributions cannot be evaluated in isolation.

It is very important that the taxpayer or tax administration using the “additive” and/or “deductive” approaches justifies and documents the criteria used to include or exclude particular third-party data from the pool of potential comparables, in order to ensure a reasonable degree of objectivity and transparency in the process. In particular, the process should be reproducible by the taxpayer and by the tax administration that wishes to assess it. It is also very important that third-party data be refined using qualitative criteria. It would be improper to use financial information relating to the transactions of a large sample of companies that have been selected solely because they are classified in a database under a given industry code.

Another example may be where the integration between the related parties takes the form of a high degree of interdependency. For instance, a profit split may be found to be the most appropriate method where, under a longterm arrangement, each party has made a significant contribution (e.g. of an asset) whose value depends in large degree on the other party. In such cases, a profit split approach could allow for pricing which appropriately takes into account and varies with the outcome of the risks assumed by each party.

Shared Risks
A further indicator that the profit split may be the most appropriate method is where the parties to the controlled transaction share the assumption of the economically significant risks in relation to the transaction (see 14.4.4.). It may also be the most appropriate method in cases where the parties separately assume risks that are so closely related or interlinked that the playing out of the risks of each party cannot be reliably isolated from the risks assumed by the counterparties.

The relevance of risksharing to the determination of the most appropriate transfer pricing method will depend greatly on the extent to which the risks concerned are economically significant such that each party should be entitled to a share of the relevant profits associated with the controlled transaction(s) had the transaction occurred at arm’s length.

​​​​​After having developed a set of comparability criteria that are tailored to the specifics of the controlled transaction at issue, the next step is to conduct an initial identification and screening of potential independent comparables. The objective in this initial screening, where performed using a commercial database, is to identify substantially all companies that have a reasonable probability of demonstrating the threshold comparability requirements and of providing verifiable, objective documentary evidence of market pricing or profits. In other words, the desired initial result is to obtain the largest possible pool of potential independent comparables for subsequent screening, verification, and analysis. Where comparables are selected from information sources other than databases this part of the process may be different.

Availability of Information
It will often be the case that where the profit split is found to be the most appropriate method, direct comparable transactions that may otherwise be used to price the transaction will not be found. However, information from uncontrolled transactions may still be relevant to the application of the PSM, for example in terms of the how the relevant profits should be split amongst the parties, or in the first part of a residual profit split. See 4.6.5.10 et seq. and 4.6.4.7, respectively; see also section 4.6.3.6 on the relevance of market information.

As was noted at the beginning of this section, in general, a PSM first determines the relevant profits, being the total profits in relation to the controlled transactions under examination, and then splits those profits on an economically valid basis. There are a number of different approaches as to how those relevant profits are allocated between the associated enterprises, including the contribution and residual analysis approaches. These are discussed in more detail below.

As with all transfer pricing methods, care should be taken to avoid the use of hindsight in the application of the PSM (see section 4.6.5.10). In general, where it is found to be the most appropriate method, the PSM should be applied consistently to transactions over time, irrespective of the amount of the relevant profits (or indeed if there are losses). Applications of the method which vary depending on the amount of the relevant profits may be found to be arm’s length in some cases, but would be less common. If there are significant unforeseen developments which would have resulted in a renegotiation of the agreement between the parties had they been at arm’s length, a different application (going forward) may be warranted. For example, a different way of determining the relevant profits or how to split them might be agreed. In such cases, documenting the reasons for the different application would be essential.

When applying or evaluating the use of the PSM it is important to ensure that the complexity of the process does not result in losing sight of the intended result: an arm’s length outcome for each related party involved. In some cases therefore, particularly where the process relies on multiple assumptions or complex calculations, it may be useful to perform a ‘reality check’ of the outcomes using alternative methods or means.

 In cases where comparable uncontrolled transactions can be found, the CUP Method is typically a very reliable method to use in determining whether the terms of commercial and financial transactions between associated enterprises are at arm’s length. This implies that an examiner should always consider the feasibility of applying the CUP Method. That is, an examiner should consider whether it is possible to locate acceptable internal comparables or external comparables. With regard to internal comparables therefore, a question that should be asked in any analysis is whether one of the associated enterprises involved is engaged in similar transactions with independent enterprises.

There are several ways in which the PSM can be applied.

Contribution Analysis
Under a contribution analysis, the relevant profits are allocated between the associated enterprises engaged in the controlled transactions in a way that aims to reflect a reasonable approximation of the divisions that would have been agreed by independent enterprises in similar circumstances. Relevant external market data, i.e. from comparable independent transactions between unrelated enterprises or between the taxpayer and an unrelated enterprise, should be used to support this allocation where available. However more commonly, such external data will not be obtainable. In such cases, the arm’s length principle can be applied by using data internal to the taxpayers themselves to determine the relative value of the contributions of each party to the controlled transaction(s). For example, this might be done by comparing the nature and degree of each party’s contributions to the controlled transactions and assigning a percentage based on that relative comparison (and any external market data that may be available).

The way in which the value of such contributions is measured will depend on the facts of each case. The determination of appropriate profit splitting factors is discussed in more detail below (see section 4.6.5.10 et seq). Note that if the relative value of the contributions can be determined, then calculating the actual value of the contribution of each enterprise may not be required.

Residual Analysis

While a contribution analysis takes the relevant profits in relation to the transaction and splits them between the parties in a single step, the PSM can be applied using a staged approach under a residual analysis. Such an approach is likely to be appropriate where one or more parties to the controlled transaction(s) makes a contribution(s) which is routine and could be benchmarked based on comparables.

Step 1: allocation of an arm’s length profit to each enterprise to compensate it for its routine or benchmarkable contributions. Typically this is done by the application of onesided transfer pricing methods such as the TNMM and consideration of the returns earned by independent enterprises engaged in activities which are comparable to those routine or benchmarkable contributions. In this first step, other contributions, such as those which are unique and valuable, are not taken into account. Each related party is allocated an appropriate ‘routine’ return from the pool of relevant profits.

In the example represented in Figure 4.D.1 above, this would involve two distinct questions: (i) whether Associated Enterprise 1 sells comparable bicycles to an unrelated party and (ii) whether Associated Enterprise 2 purchases comparable bicycles from one or more unrelated bicycle manufacturers. If the answer to either one of these questions is in the affirmative, then the next step in the analysis is to determine the degree of comparability between the controlled and uncontrolled transactions based on the comparability factors.

Step 2: allocation of residual profit (i.e. remaining relevant profits after the Step 1 allocation) on an economically valid basis. In the second step, other contributions not already accounted for, including those which are unique and valuable, are considered. As was described above in relation to a contribution analysis, this allocation must be done on an economically valid basis, and aim to achieve a reasonable approximation of the divisions that would have been agreed by independent enterprises in similar circumstances. The second step allocation will thus typically consider the relative value of the contributions of each party to the residual profits, supplemented where possible by external market information on how independent parties would have divided such profits in similar circumstances.

As has been noted above, since reliable, direct information on how profits would have been allocated in comparable uncontrolled transactions might not be available, care is required in applying the PSM. The residual approach to the application of the method aims to reduce possible subjectivity by confining, to the extent possible, the more difficult step 2 allocation (which is typically not based directly on comparables data).

 

The following approaches have been specified in some jurisdictions to determine the relative value of each company’s contributions of intangibles:
►    External market benchmarks reflecting the fair market value of the intangible property;
►    The capitalized cost of developing the intangibles and all related improvements and updates, less an appropriate amount of amortization based on the useful life of each intangible;⁴⁰ and
►    The amount of actual intangible development expenditures in recent years if these expenditures have been constant over
time and the useful life of the intangible property of all parties involved is broadly similar.

The residual approach is used more in practice than the contribution approach for two reasons. Firstly, the residual approach breaks up a complicated transfer pricing problem into two manageable steps. The first step determines a basic return for routine or benchmarkable functions based on comparables and the application of a onesided method or methods. The second step analyzes returns to unique and valuable contributions or other elements which are unbenchmarkable. Rather than trying to determine absolute values for these contributions based on comparables, the method focuses on their relative value which may often be determined more reliably. Secondly, potential conflict with the tax authorities is reduced by using the twostep residual approach since it reduces the amount of profit that is to be split in the potentially more controversial second step.

External comparables may be difficult to find in practice unless the transactions involve a fairly common and homogeneous product or service. However, the advantages of the CUP Method are great enough to warrant a significant effort to apply the method.

Comparable Profit Split Method (PSM)
In some countries, reference is made to the Comparable PSM. This application of the PSM relies on a comparison of the allocation of profits between independent enterprises engaged in comparable activities under comparable circumstances to those of the controlled transaction(s). That is, it relies heavily on external market data to determine how the relevant profits should be split between the related parties. As has been noted above, such information may be very useful, but is rarely available in practice.

The relevant profits to be split under the PSM are those which arise to the associated enterprises as a result of the controlled transaction(s) under examination. It will be important to consider the level of aggregation of transactions in this regard (see 3.3.3.1 to 3.3.3.5) and then to examine the relevant income and expense amounts of each party in relation to those transactions.

In most cases, since the relevant profits will be comprised of income and expense amounts from more than one related party in more than one jurisdiction, the relevant financial data of the entities will need to first be put on a common basis, including with regard to the accounting practice and currency used. As this can materially affect the application of the method, consistency over time is important in this regard.

Other than in cases where the profit split covers all the activities of each of the related parties, the financial data will need to be segmented in accordance with the accurately delineated transaction(s) covered by the profit split approach. In cases where reliable productline or divisional accounts are available, these may be useful to the determination of the relevant profits to be split.

Measures of Profit
The PSM is most commonly used to split net or operating profits. Applying the method in this way means that all the related parties are exposed to both the income and expenses associated with the relevant transactions in a consistent manner. However, depending on the accurate delineation of the transaction, other measures of profits may be appropriate. For own operating expenses. Such an application may be appropriate where the parties do not share the risks associated with the operating expenses relating to the controlled transaction, but do share the risks associated with the volume of sales and prices charged, as well as those associated with the production or acquisition of the goods or services.
 

Experience indicates that the CUP Method will be most useful where:
►    One of the associated enterprises involved in the transaction is engaged in comparable uncontrolled transactions with an independent enterprise (i.e. an internal comparable is available). In such a case all relevant information on the uncontrolled transactions is available and it is therefore probable that all material differences between controlled and uncontrolled transactions can be identified; and
►    The transactions involve commodity type products, and the differences between the products are minor or can be easily adjusted for.

Actual or Anticipated Profits
The PSM is most commonly applied to split the actual relevant profits of the related parties in relation to controlled transactions. Since actual profits will reflect the playing out of the risks which affect the transactions, such a split will typically result in each related party being subject to those risks. It would thus be appropriate where the accurate delineation of the transaction shows that each related party shares such risks. For example, where the parties to the controlled transaction share the assumption of the economically significant risks, or separately assume closelyrelated economically significant risks in relation to the controlled transactions, it would be expected that a split of actual profits would apply.

On the other hand, where the profit split is found to be the most appropriate method but the accurate delineation of the transaction shows that one or more of the related parties does not share in the assumption of the economically significant risks, a split of anticipated profits is likely to be more appropriate.

A common application of an anticipated profit split is in the use of a discounted cash flow valuation technique, which might be used, for example, to determine the present value of a transferred intangible or other asset. For example, Company A transfers all the rights in a fully developed unique and valuable intangible, intangible X, to Company B, its associated enterprise. Company B has its own unique and valuable intangibles which are expected to complement intangible X. Company A expects to have no ongoing involvement in the exploitation of intangible X, as these activities will be wholly undertaken and controlled by Company B. In this case, assume it is determined that the profit split is the most appropriate method since both Company A and Company B make unique and valuable contributions. However, since Company A will not be involved in the ongoing exploitation of the intangible after the transfer, and it does not assume any risks relating to those exploitation activities, at arm’s length, its return should not be subject to those risks. Instead, it should receive a share of the anticipated profits from Company B’s exploitation of the combined intangibles of Companies A and B, discounted to reflect its present value at the time of the transfer. This amount might be calculated using a discounted cash flow valuation technique which analyzes the present value of the likely income from the exploitation of intangible X. The ongoing risks relating to the exploitation of the intangibles are solely assumed by Company B and no adjustment to the remuneration due to Company A needs to be made should the intangible actually be more or less successful than anticipated.

It should be noted that measures of profits which vary to some degree with the playing out of risks, without being fully exposed to such risk, can also be used. In all cases, the measure of relevant profits to be split should be aligned with the accurate delineation of the transaction in order to produce an arm’s length outcome.

Even where a profit split of actual profits is used, the method should be applied without hindsight. That is, unless there are significant unforeseen developments which would have resulted in a renegotiation of the agreement had it occurred between independent parties, the basis for determining how the relevant profits should be calculated and how they should be split amongst the associated enterprises should ordinarily be determined based on information known or reasonably foreseeable at the time of, or prior to the transaction(s). This is the case even though it may only be possible to apply the actual calculations some time thereafter. For example, Company E and its associated enterprise, Company F are so highly integrated that the PSM is found to be the most appropriate method to evaluate the controlled transactions between them. The way in which the relevant profits from their transactions should be determined is established ex ante, that is, at or prior to the time they engage in the transactions. At that time, they also determine that the residual PSM of actual net profits should be applied, and that the residual profits should be split between them on the basis of the value of current year marketing expenses of each party, after having allocated basic or ‘routine’ returns on the routine sales and distribution activities conducted by both Companies E and F. In this example, the way in which the PSM is to be applied is determined at the start of the period. However, the agreed method can only be applied at year end, once the amount of sales, marketing expenses, and the amount of the relevant actual net profits has been determined. If, in a subsequent period, these intragroup transactions are subject to a transfer pricing audit, the tax administration would not be precluded from examining the selection of the transfer pricing method or the way in which it was applied in order to confirm compliance with the arm’s length principle. In doing this, the tax administration may also examine what information was actually known or reasonably foreseeable at the time of the transaction.

​​​​​The process of screening, verification and selection of comparables will largely depend upon the availability of databases in the public domain in the country. Public databases may be available in some countries whereas other countries may not have these databases. In such cases, one of the options could be to rely on a database from a comparable economy with reasonable and reliable adjustments.

Profit Splitting Factors
The PSM aims to determine transfer prices by reference to the manner in which independent parties would have divided profits amongst themselves had they engaged in comparable transactions. However, information on comparable profit splits or similar arrangements is often not available, so the method is more often applied by reference to some other measure of the relative contributions to those profits of each associated enterprise, as a way of approximating the outcome that would have been achieved between independent parties.

It would not be appropriate to provide prescriptive guidance as to the measure or measures to be used to split the relevant profits, as this will depend on the facts of each case. However, whatever factor(s) are selected, they should be capable of objective measurement and not themselves subject to nonarm’s length pricing or valuation. The measures should also be verifiable and supported by data. While these considerations need to be borne in mind, amounts based on the taxpayer’s own internal information (e.g. from their financial accounts) are commonly used.

In some cases, a multifactor approach to splitting profits may be adopted. However, it may also be the case that a single measure of the key contributions to value of each enterprise to the transaction will be sufficient as a proxy for the relative value contributed.

In this regard, information from the functional analysis is likely to be particularly important. Other information in the taxpayer’s local file may also be useful. In addition, where the master file is available, the information therein on key valuedrivers, considered in the context of the business and industry environment, may also be helpful to the extent that the value drivers for the transactions under examination are similar to those for the MNE or business line that is the subject of the master file.

Depending on the circumstances, profit splitting factors might be based on the value of (certain types of) assets or capital, where there is a strong correlation between tangible assets or intangibles, or capital employed, and the creation of value in the controlled transaction. In such cases, care should be taken to ensure reliable and consistent measures of the value of the asset(s) concerned.

In other cases, costbased factors may be found to be appropriate,
e.g. costs related to the unique and valuable contributions such as R&D, engineering, design, marketing, etc., or the development of unique and valuable intangibles. Note that although cost is often a poor measure of the absolute value of unique and valuable intangibles, the relative costs incurred by each party may provide a reasonable approximation of the relative value of their respective contributions. In some instances, it may be appropriate to adjust the cost amounts, e.g. where they are incurred in different periods, to ensure they represent reliable measures of the respective contributions of each party.
 

Other examples of profit splitting factors could include incremental sales, employee remuneration or bonus payments, time spent, headcount, etc. Such factors may be found to be appropriate where they provide a strong and sufficiently consistent correlation to the creation of value represented by the relevant (residual) profits.

Transfer pricing rules require associated enterprises to price their intragroup transactions in accordance with the arm’s length principle. The five methods described at sections 4.2 to 4.6 inclusive of this Practical Manual are used to calculate or test the arm’s length nature of intragroup prices or profits earned from intragroup transactions. As described at 4.1.2.1, the starting point in selecting a transfer pricing method is an understanding of the controlled transaction (inbound or outbound) based on the comparability (including the functional) analysis. This is necessary regardless of which transfer pricing method is selected. The CUP Method is a suitable method when prices from comparable transactions of the same or similar products are available. For instance, the quoted prices on a commodities market or exchange may be comparable uncontrolled prices for transactions involving commodities performed by related parties under comparable circumstances.

Tax authorities may find themselves unable to successfully determine whether a controlled transaction between associated enterprises is comparable with uncontrolled transactions observed in the market if the taxpayer does not provide sufficient supporting documentation to its controlled transactions. For developing countries this concern may be even more pressing than for developed countries, as the former countries may not have data on companies doing business in their countries to perform a comparability analysis in addition to potentially having limited knowhow and resources to conduct transfer pricing studies.

In view of this difficulty, several countries have a rule in place that is generally referred to as the socalled “sixth method” or “commodity rule” although the name of the rule or the method applied may differ from country to country. Historically this approach has been used for commodities by several countries. The common feature of these rules is that they rely on the quoted prices on the commodities market to price commodity transactions between associated enterprises. For practical purposes, in this section the approach will be referred to as the “sixth method”.⁴¹

Because versions of the “sixth method” have been more widely adopted and applied in recent years, this section discusses the rationale for and country experiences with the “sixth method”. However, this section does not seek to address the relationship between the sixth method as implemented in any particular domestic legislation and the arm’s length principle as defined under Article 9 of UN Model Convention. The workings of the sixth method may resemble the Comparable Uncontrolled Price (CUP) Method. The sixth method generally considers certain characteristics and substance requirements, as defined in relevant domestic practices, for the method to be applicable to a particular transaction. Absent reliable and timely confirmation from the taxpayer about the relevant characteristics of the controlled transaction, some countries apply a default rule which includes the application of a transfer price based on publicly available pricing information and typically an assumption of the time the relevant tested transaction took place (which corresponds to the commodity price listing used). Despite its more general title, considering its operation, the sixth method—depending on how it is applied—could be considered as an antiabuse rule, an abusedeeming rule or even presumed to be a form of safe harbour. Some countries consider the sixth method an (imperfect) application of the CUP Method.⁴² This section will refer to the practice as “the sixth method” without addressing whether the term method, antiabuse rule or abusedeeming rule or safe harbour is more appropriate or legally more correct.

The sixth method approach has benefits and disadvantages. A benefit of the sixth method is its relative certainty and relative ease of application for the tax authorities and its tax collection efficiency. Some disadvantages of the sixth method include that it is not one of the traditional transaction methods or traditional transactional profit methods described in this chapter and thus may not be recognized by the country of the associated enterprise at the other end of the transaction. Disadvantages also include its inaccuracy considering the general standards of application of the arm’s length principle, that the method does not consider the economic circumstances of the actual transaction that is being tested, and that the method may result in overcompensation of one associated enterprise to the detriment of another associated enterprise and therefore may give rise to potential double taxation. Benefits and disadvantages are discussed further in section 4.7.5.

This chapter describes the socalled sixth method as it is observed in practice in several countries. It also discusses OECD guidance and an InterAmerican Center of Tax Administrations (CIAT) study relevant to this method or approach that generally is used to tackle abusive transfer pricing practices. The chapter aims to provide a basic description of the application of the socalled sixth method in case of commodity transactions and a description of steps that can be taken to mitigate some identified disadvantages of using the sixth method.

At the date of drafting this section, the sixth method has been adopted in countries such as Argentina, Bolivia, Brazil, Costa Rica, Dominican Republic, Guatemala, Honduras, Peru, Uruguay and some Caribbean countries. Practical experience with the rule exists mainly in Argentina, Brazil, Ecuador and Uruguay.⁴³ However, it should be noted that the method is not applied unequivocally in the same manner in all these countries. In Argentina, the method was first introduced in 2004. Other countries have followed in implementing (a version of) the sixth method since its first implementation in Argentina.

The fact pattern traditionally targeted by the sixth method is one where associated enterprises engaged in the business of exporting commodities such as grains, oil and oilseeds, oil and gas, mining and fishing products, invoice an associated enterprise related to the sale of the commodities yet ship the commodities to a jurisdiction (and party) different from the associated enterprise that is being invoiced by the seller. The actual shipment date usually is at a later point in time than the date of the original sale between the associated enterprises and the intercompany invoice date. In general, the associated enterprise being invoiced is a trading entity that obtains title to the shipped goods only for a limited period of time and the later shipment is to a destination determined by an unrelated party that has bought the commodities from the associated trader (not to the residence of the associated trading entity).

If based on clear criteria that can be readily ascertained it is established that the associated trading entity has insufficient substance to perform relevant functions related to the acquisition and sale of commodities and the trading entity is located in a low tax jurisdiction, there is a risk that the associated trading entity transaction will erode the tax base of the jurisdiction of the seller of the commodities. This may result from paying a price for the commodities that is lower than that received by the associated trading entity from a thirdparty in a later sale, despite the fact that the trading entity does not have the substance to perform the trading functions. To curtail that risk, the sixth method determines the intragroup price by reference to the quoted price of the commodities on the shipment date, i.e. it looks through the intermediary trading entity as if the transaction had taken place between the seller and the thirdparty customer. The sixth method in some countries specifies that the intragroup sales price will be deemed to be the higher of either the exchange price of the commodity on the intragroup sale date or on the shipment date of the commodities to the unrelated buyer. Applied under these circumstances and in this way the sixth method can be considered to function as an antiabuse rule.

As stated earlier, several countries that have implemented the sixth method have implemented it differently. CIAT describes the respective permutations of the sixth method observed in practice related to the different aspects that make up the rule as follows:⁴⁴

Table 4.T.10
Implementation of the Sixth Method

Considering that the arm’s length principle requires prices between associated enterprises to be comparable with those used between unrelated parties, it would generally be required that market prices be used that would apply in the same or similar circumstances as those that apply to the transaction between associated enterprises. Where the sixth method offers taxpayers the opportunity to provide evidence that the intragroup transactions are not abusive and the prices are at arm’s length, and the method allows comparability adjustments as need be, it may be more generally consistent with the arm’s length principle and provide for legal certainty. It can also help by providing the tax authorities with more transparency as regards the transaction between the associated enterprises.

The following analytical needs and constraints should, however, be kept in mind:

• The search process should avoid any systematic biases;
• The screening process must be executed and documented in a manner consistent with the general requirement for due diligence; and
• It should be recognized that some of the initial comparables will be eliminated in subsequent stages of screening and analysis.

The Group of Twenty (G20) and OECD countries have examined the transfer pricing aspects of crossborder commodity transactions between associated enterprises as part of Action 10 of the BEPS Action Plan. As a result, Chapter II of the OECD Transfer Pricing Guidelines has been amended to include new guidance especially applicable to commodity transactions.⁴⁵ The new guidance includes a clarification of the existing guidance on the application of the CUP Method to commodity transactions. The new guidance states that:
1)    The CUP Method would generally be an appropriate transfer pricing method for commodity transactions between associated enterprises;
2)    Quoted prices can be used under the CUP Method, subject to a number of considerations, as a reference to determine the arm’s length price for the controlled commodity transaction; and
3)    Reasonably accurate adjustments should be made, when needed, to ensure that the economically relevant characteristics of
the controlled and uncontrolled transactions are sufficiently comparable.

The guidance also includes a provision on the determination of the pricing date for commodity transactions that, under certain conditions, provides for tax administrations to impute a pricing date. This provision acknowledges the difficulties that can arise for tax administrations in verifying the correct date for pricing the goods sold and serves to prevent taxpayers from using pricing dates in contracts that enable the adoption of the most advantageous quoted price. Although the pricesetting date used by the parties is the prima facie date, in certain circumstances, for example if there is no evidence available as to what that date really is, the tax authorities can impute a pricing date, (e.g. the shipment date) in a manner similar to the sixth method.

Tax authorities in developing countries may find it easier to test the arm’s length nature of intragroup commodity transactions if taxpayers are required to provide more information on such transactions. The OECD guidance referenced above emphasizes that taxpayers can assist tax administrations in conducting an informed examination of their transfer pricing practices related to commodity transactions by providing:⁴⁶
►    Documentation on the pricesetting policy for commodity transactions;
►    The information needed to justify price adjustments based on the comparable uncontrolled transactions or comparable uncontrolled arrangements represented by the quoted price; and
►    Any other relevant information, such as pricing formulas used, thirdparty endcustomer agreements, premiums or discounts applied, pricing date, supply chain information and information prepared for nontax purposes.

The date of pricing commodity transactions is particularly relevant in determining whether such transactions between associated enterprises are priced appropriately. Tax administrations should determine the price for commodity transactions by reference to the pricing date agreed by associated enterprises, where the taxpayer can provide reliable evidence of the pricing date agreed by such associated enterprises in the controlled commodity transaction. Such evidence should include:
►    The time the transaction was entered into (e.g. proposals and acceptances, contracts or registered contracts, or other documents setting out the terms of the arrangements may constitute reliable evidence), and
►    That this is consistent with the actual conduct of the parties or with other facts of the case.
 

If the pricing date specified in any written agreement between the associated enterprises is inconsistent with the actual conduct of the parties or with other facts of the case, tax administrations may determine a different pricing date consistent with those other facts of the case and what independent enterprises would have agreed in comparable circumstances (taking into consideration industry practices). When the taxpayer does not provide reliable evidence of the pricing date agreed by associated enterprises in the controlled transaction and the tax administration cannot otherwise determine a different pricing date under the guidance on accurately delineating the transaction, tax administrations may deem the pricing date for the commodity transaction on the basis of evidence available to the tax administration. Such evidence may include the date of shipment as evidenced by the bill of lading or equivalent document depending on the means of transport. This would mean that the price for the commodities being transacted would be determined by reference to the average quoted price on the shipment date, subject to any appropriate comparability adjustments based on the information available to the tax administration. It would be important to permit resolution of cases of double taxation arising from application of the deemed pricing date through access to the mutual agreement procedure under the applicable treaty.⁴⁷

CIAT conducted a study to analyze the operation of the socalled sixth method 10 years after its introduction in Argentina.⁴⁸ In particular, it conducted a comparative analysis of the implementation of the sixth method in other Latin American countries. The study notes that transfer prices are calculated using the methods developed under the OECD Transfer Pricing Guidelines and that the methods are based on comparability of the transactions under analysis and that of other similar transactions carried out between or with independent parties. It also notes that a main problem encountered is the difficulty in assessing useful and timely information for carrying out such a comparison.

The CIAT study notes that authorities may react differently to distortions caused by companies being under common control. Some authorities may wish to disregard a legal entity making use of a substanceoverform analysis in case an enterprise is under the control of another enterprise in economic, financial and corporate terms. Other authorities may wish to make use of transfer pricing rules and determine the intercompany price that should apply by eliminating any pricing distortions that result from the influence of being associated enterprises. Although the traditional approach to the arm’s length principle may be preferred, the CIAT study also notes that the complexity and practical difficulty in applying the transfer pricing methods and with finding comparables should allow for consideration of alternatives for developing countries, which may have limited resources to address the complexity that these exercises present or overcome practical difficulties in this respect.

When considering the legal nature of the sixth method, the CIAT study points out that the characterization of the method under domestic law may have consequences as regards the application of a treaty for the avoidance of double taxation. That is, to the extent the sixth method qualifies as a transfer pricing method, the workings of the Associated Enterprises article, usually Article 9 of a treaty for the avoidance of double taxation, may govern the international consequences of its application and to the extent it is seen as being a domestic antiavoidance rule it may fall outside the scope of such a treaty and taxpayers may be unable to successfully claim avoidance of double taxation. If the method is characterized as an antiabuse rule or a safe harbour, it will be relevant to get a treaty partner to acknowledge that, before a treaty can successfully be availed of to reduce any potential double tax.

In many cases, the sixth method may have been incorporated into domestic law to help prevent tax base erosion and facilitate the collection of corporate taxes. The method is most often applied to intragroup exports and/ or imports of commodities and applies the publicly available market price of the commodity on the actual date of shipment as the transfer price. The method generally disregards whether the buyer or seller pays for the shipping costs and at what date delivery of the goods finally takes place. The method may also apply a cruder standard of comparability, in that the grade or quality of the commodity and sometimes even the volume shipped may not need to be the same as applies in a standard contract traded on the exchange. However, some countries and authorities allow for comparability adjustments to correct for this. The lack of substance of the intermediary is generally a given for the method to apply, although in practice there also are countries where the method is applied to all exports and imports regardless of (i) foreign intermediaries and/or (ii) the involvement of associated intermediaries.

The CIAT study goes further than the original sixth method in that it considers the position of the associated intermediary that makes the thirdparty sale. It recommends that tax authorities try to obtain an understanding of the value chain related to the commodity and its subsequent processing or transformation to help with determining the arm’s length compensation for the functions performed, and risks assumed by the entities involved in that value chain.

Use of the sixth method offers some benefits but also carries with it some disadvantages as discussed above. The sixth method, in its most rudimentary form, operates as a pricesetting mechanism that may roughly resemble the CUP Method. However, depending on how it is applied, it may not meet the rather strict requirements that the CUP Method traditionally requires for its application. Some countries consider it a benefit that the method can be used when no exact transfer price or comparable transaction is available for the commodities or products involved. This approach may serve as a practical means to raise revenue and requires relatively limited audit activity related to taxpayers engaged in intragroup commodity trading activities. A disadvantage of the approach is that, depending on how the sixth method is applied, there is a potential risk of divergence from the arm’s length principle and resulting double taxation. Another disadvantage of the approach is that it does not apply a traditional transaction method or transactional profit method and hence may not be recognized by the country at the other end of the transaction. As a country’s tax system and tax authorities develop and benefit from more transfer pricing– related knowhow and resources, the sixth method may not or may no longer be necessary, or it may be adjusted or updated to achieve greater consistency with the arm’s length principle.

It should be unequivocally clear in what situations the sixth method applies; this will avoid disadvantaging taxpayers and will foster legal certainty. If a (lack of) substance test applies, clear and objective criteria should be provided. Where a substance test is applicable, only when the taxpayer clearly fails to meet the test should the sixth method approach be applied. Potential double taxation resulting from application of the sixth method approach can also be a disadvantage. This disadvantage could be mitigated to the extent tax authorities would allow any double taxation resulting from the application of the sixth method to be eligible for potential relief under an applicable treaty.

​​​​​The person performing the search for comparables may have to use a variety of information sources for third-party or external comparables. These can include company-specific information sources including annual reports, regulatory and other government filings, product literature and securities analyst reports, as well as various trade and industry association materials. Once intermediate screening has been completed a complete set of company financial statement data should be generated and reviewed for adequacy, period coverage and general consistency. Sometimes details may even be obtained through telephone or personal interviews with company management and it is also possible to use the knowledge of internal operating personnel to identify comparables. For example, sales and marketing personnel can be asked to assist in identifying independent third-party resellers whose financial statements may be used as a basis for establishing comparable profit margins.

Publicly available prices may be based on standardized agreements setting forth basic aspects of the transaction, such as quality, quantity and terms of delivery of the commodities marketed. A reliable application of the CUP Method would require that the raw materials being compared are sufficiently similar. Therefore, a closer application of the arm’s length principle can be achieved to the extent that the sixth method allows for comparability adjustments to be made to the quoted prices for features such as physical characteristics and the quality of commodities as well as the volumes traded, the terms and conditions of the contracts, other relevant variables, the delivery date and conditions (CIF, FOB etc.) and whether the transaction between the associated enterprises is carried out at the same level of the supply chain as the one that served to set the publicly available price.

As stated earlier, the application of the sixth method in some countries is generally conditioned on the lack of (evidence on) substance of the intermediary trading entity. To the extent taxpayers may provide evidence as to the substance of (their associated) intermediary entities to opt out of the application of the sixth method, it could be expected that taxpayers will make an effort to provide the requisite additional information. As a result, the more accurate and arm’s length income allocation may become that which applies between the associated enterprises.

Allowing for evidence or proof of substance is not uncommon with this approach. Doing so adds an administrative burden on taxpayers and tax authorities, however. As an example, one country’s law and regulations provide in relevant part:
“the [sixth] method does not apply if and when: (a) the (associated) enterprise that constitutes the international intermediary, demonstrates effective presence in the jurisdiction of residence, having a commercial office or premises where the business is managed, complying with the legal inscription and filing of balance sheets; the assets, risks and functions undertaken by the international intermediary are adequate in view of its commercial operations; (b) the international intermediary’s main business does not consist of receiving passive income or the trading of goods to or from the country of origin of the commodities or with other members of the economic group⁴⁹;
(c) the international intermediary’s crossborder trade with other members of the same multinational group does not exceed 30% of the total annual transactions conducted by the intermediary.”

Considering the observations resulting from the CIAT study mentioned in the previous paragraphs and the guidance issued by the OECD with respect to commodity transactions, to encourage access to treaty provisions providing potential relief from double taxation two observations can be made: First of all, in those situations where the rule considers the status of (foreign) intermediaries, it would be preferable if taxpayers are allowed to present evidence that their intermediaries have the requisite substance in the other jurisdiction(s). Second, taxpayers could be allowed to provide evidence that their intragroup pricing is at arm’s length and appropriate and reasonable adjustments to publicly available commodity prices could be allowed in order to reflect possible transaction and product differences. This is important as commodity prices are known to fluctuate significantly and be highly dependent on grade, quality and specifics of the respective commodity. In addition, import/export conditions are likely to influence prices.

The less arbitrary the sixth method and its criteria for application can be, the more legal certainty may result and the more efficient and successful audits of transactions and taxpayers subject to the method can be. Clarifying the definition of intermediaries subject to the sixth method (if and when it refers to intermediaries), specifying the transactions subject to the sixth method and specifying the criteria for its application will assist in reducing uncertainty and potential tax disputes. To the extent taxpayers can provide evidence as regards to the necessary substance of intermediaries and what would constitute an arm’s length price, and the legislation includes provisions for adjustments to the publicly available commodity prices to assure improved comparability, the sixth method becomes more sophisticated and in line with the arm’s length principle. Taxpayers may benefit from (improved) access to avoidance of double taxation in that event as well.

​​​​​​Other adjustments are those proposed by the taxpayer or tax administrator to adjust for specific economic circumstances that affect the transactions being compared. There can be significant differences in the mix of functions performed by the potential comparables vis-à-vis the tested party, or in the assets used or contributed, risks assumed or capital employed. When such differences exist and are not adjusted, they may affect the reliability of the comparables in establishing an appropriate arm’s length profit range.

​​​​​​The financial results of the comparables may need to be adjusted to eliminate the effect of such differences. Such adjustment is possible only when reliable and accurate segmented detailed information is available. An adjustment is made to the revenue and costs relevant to the functions performed by the comparables but not by the tested party. If an arm’s length return is established for additional functions performed by the tested party, it is not necessary to adjust the comparables. That arm’s length return based on another set of comparables may be applied to the tested party for those functions. Care should be exercised while making a functional adjustment which involves a subjective assessment.

There can be significant differences in the mix of functions performed by the potential comparable vis-à-vis the tested party. For example, a controlled distribution company may differ from a set of independent distribution companies in that it performs import and regulatory functions not performed by the independent distributors (notwithstanding that the independent distributors have been determined to be the best available comparables); performs only first-tier distribution functions and performs limited manufacturing and assembly functions. To adjust for such differences, the financial results of the comparable may be adjusted to account for the revenue, costs, and associated profits associated with the functions performed by the comparable but not by the tested party, or vice versa.

​​​​​Adjustments performed to compensate for material differences in the mix of functions performed by a controlled storage device distributor and a set of independent storage device distribution comparables is considered here to illustrate this point. It is assumed that the independent device distributors (determined to be the best available comparables) also perform manufacturing/assembly operations and downstream distribution functions that are not performed by the controlled storage device distributor. In this case, the financial results of the comparables may need to be adjusted to eliminate the profits associated with manufacturing/assembly operations and with downstream distribution functions based upon the profitability earned in uncontrolled comparable storage manufacturing and downstream distribution transactions. In other words, for comparability purposes, only the functions comparable to the functions carried out by the controlled storage device distributor should be taken into consideration.

To contrast the treatment above with a different set of circumstances, it could be assumed that the controlled storage device distributor above performs some import functions which are not performed by the independent distributors. The margins of those comparables that did not perform import functions could perhaps be adjusted to reflect an arm’s length profit associated with these functions if sufficient data exist to make an accurate adjustment.

Examination of the economically relevant characteristics of the accurately delineated controlled transaction will help in the selection of the most appropriate transfer pricing method and in developing search criteria to identify reliable comparables with which to apply the selected method. This examination is the primary focus of the functional analysis described above in section 3.4.4.

The RPM is one of the traditional transaction methods that can be used to determine whether a transaction reflects the arm’s length principle. The RPM focuses on the related sales company which performs marketing and selling functions as the tested party in the transfer pricing analysis. This is depicted in Figure 4.D.2 below.

The RPM analyzes the price of a product that a related sales company (i.e. Associated Enterprise 2 in Figure 4.D.2) charges to an unrelated customer (i.e. the resale price) to determine an arm’s length gross margin, which the sales company retains to cover its sales, general and administrative (SG&A) expenses, and still make an appropriate profit. The appropriate profit level is based on the functions it performs, the assets it uses and the risks it assumes. The remainder of the product’s price is regarded as the arm’s length price for the intragroup transactions between the sales company (i.e. Associated Enterprise 2) and a related company (i.e. Associated Enterprise 1). As the method is based on arm’s length gross profits rather than directly determining arm’s length prices (as with the CUP Method) the RPM requires less direct transactional (product) comparability than the CUP Method.
 

​​​​​​​

Consequently, under the RPM the starting point of the analysis for using the method is the sales company. Under this method the transfer price for the sale of products between the sales company (i.e. Associated Enterprise 2) and a related company (i.e. Associated Enterprise 1) can be described in

Resale price    =    US$100
Resale price margin (25%)    =    US$ 25
Arm’s length transfer price    =    US$ 75
the following formula:

TP = RSP x (1GPM), where:
►    TP = the Transfer Price of a product sold between a sales company and a related company;
►    RSP = the Resale Price at which a product is sold by a sales company to unrelated customers; and
►    GPM = the Gross Profit Margin that a specific sales company should earn, defined as the ratio of gross profit to net sales. Gross profit is defined as Net Sales minus Cost of Goods Sold.

Comparable search or “screening” criteria are developed based upon the results of the above-mentioned examination of the economically relevant characteristics in relation to the controlled transaction. These criteria must be defined so as to identify those external uncontrolled trans-actions that satisfy comparability vis-à-vis the controlled transaction and the tested party. The search criteria should be set so as to select the most reliable comparables. At the same time, the initial search criteria should not be overly restrictive, in order not to set unrealistic expectations in terms of comparability. Once potential comparables have been selected comparability adjustments should be considered, and in cases where they improve the reliability of the comparison, they should be made. The selection of the most appropriate transfer pricing method will primarily be driven by the nature of the accurately delineated transaction, but of course, the availability of reliable comparables will influence the choice.

Other approaches are possible. For example, if the associated enterprise acts as a sales agent that does not take title to the goods, it is possible to use the commission earned by the sales agent (represented as a percentage of the uncontrolled sales price of the goods concerned) as the comparable gross profit margin. The resale price margin for a reseller should always be determined by taking into account the functions performed, assets used or contributed and risks assumed by the reseller.

Database Screening (Primary Screening)

The financial ratio analyzed under the Resale Price Method (RPM) is the gross profit margin. Gross profit is defined as net sales minus cost of goods sold. It is easiest to determine where the reseller does not add substantially to the value of the product. The net sales figure of a sales company is the sales revenue obtained from selling products to unrelated customers, while the cost of goods sold equals the cost of purchasing the goods sold plus certain additional nonoperating costs. Thus, if we are determining the gross margin for products purchased from a related company, the cost of goods sold will include the transfer price paid to the related party (often a manufacturer).

The determination of appropriate screening criteria will depend on the most economically relevant characteristics of the accurately delineated transaction. Typically, they will begin with the industry code and include screens to ensure the transactions engaged in by the potential comparables are indeed comparable and uncontrolled, and that sufficient financial information is available and can be relied upon. While screens based on industry codes and geographic market are commonly applied, it is always important to consider what characteristics are most economically relevant to the accurately delineated tested transaction(s). For instance, functional comparability may be more important than similarity of industry or market. In such cases, indiscriminate application of the less relevant criteria may be unhelpful, resulting in no comparables being left with which to apply the transfer pricing method.

Accounting consistency is extremely important in applying the RPM. Gross profit margins will not be comparable if accounting principles and/or practices differ between the controlled transaction and the uncontrolled transaction. For example, the comparable distributors may differ from the related sales company in reporting certain costs (e.g. discounts, transportation costs, insurance and costs of performing the warranty function) as operating expenses or as cost of goods sold. Differences in inventory valuation methods will also affect the gross margins. It is thus important that the analysis does not compare “apples with oranges” but rather “apples with apples”. Appropriate adjustments may need to be applied to the data used in computing the gross margin to make sure that “similar” gross margins are compared.

The arm’s length price or margin can result from looking at comparable functionality (distributors of broadly similar types of product) or from making a transactional comparison by looking at each transaction the tested party engages in involving comparable products (i.e. sales of different types of bicycles).

The arm’s length (range of) gross profit margin(s) to be earned by the sales company in the controlled transaction can therefore be determined in the following two ways:
►    By transactional comparison: for example, one could determine the gross profit margin that Associated Enterprise 2 earns when reselling bicycles purchased from an independent manufacturer in a comparable uncontrolled transaction. This uncontrolled transaction may initially have been rejected as an internal comparable for purposes of applying the CUP Method because, for example, the transaction involves a different type of bicycle. If the sale of recreational bicycles is at issue, but the unrelated transactions involve bicycle rickshaws (pedicabs) or the like this may involve broadly similar products with comparable accounting measures of Costs of Goods Sold (COGS) making gross margin comparisons sufficiently reliable; and
►    By functional comparison: the gross profit margins earned by independent companies in comparable uncontrolled transactions performing functions, using assets and assuming risks comparable to the functions performed, assets used or contributed and risks assumed by Associated Enterprise 2. Functional comparison thus involves a search for comparable distribution companies rather than comparable transactions. This could, for example, include comparable distributors of wheelbarrows and carts.
 

In practice, transactional comparisons are more likely to achieve broad product and accounting consistency than functional comparisons. However, it is sometimes not necessary to conduct a resale price analysis for each individual product line distributed by a sales company under this method. Instead, the RPM is used in those situations to determine the gross margin a sales company should earn over its full range of (aggregated) products.

Information derived from external comparables should reflect the economic environment at the time the controlled transaction was undertaken. 

In principle, information from external comparables contemporaneous with the controlled transaction might be expected to reflect the same economic environment, but there can be practical difficulties in obtaining contemporaneous information given the time required for such information to be prepared, reported, and uploaded on to databases. For a discussion on timing issues, see section 3.6.2.

An uncontrolled transaction is considered comparable to a controlled transaction if:
►    There are no differences between the transactions being compared that materially affect the gross margin (for example, contractual terms, freight terms etc.); or
►    Reasonably accurate adjustments can be performed to eliminate the effect of such differences.

As noted above, the RPM is more typically applied on a functional than on a transactional basis so that functional comparability is typically more important than product comparability. Product differences will probably be less critical for the RPM applied on a functional basis than for the CUP Method, because it is less probable that product differences will have a material effect on profit margins than on price. One would expect a similar level of compensation for performing similar functions across different activities.

While product differences may be more acceptable in applying the RPM as compared to the CUP Method, the property transferred should still be broadly similar in the controlled and uncontrolled transactions. Significant differences between the nature of the products sold in the controlled and uncontrolled transactions may reflect differences in functions performed, assets used or contributed or risks assumed. Such differences might suggest differences in arm’s length gross margins.

The compensation for a distribution company should generally be the same whether it sells washing machines or dryers, because the functions performed (including risks assumed and assets used or contributed) are similar for the two activities. It should also be noted, however, that distributors engaged in the sale of markedly different products cannot be compared. The price of a washing machine will, of course, differ from the price of a dryer, as the two products are not substitutes for each other. Although product comparability is less important under the RPM, greater product similarity is likely to provide more reliable transfer pricing results. It is not always necessary to conduct a resale price analysis for each individual product line distributed by the sales company. Instead, the RPM can be applied more broadly, for example based on the gross margin a sales company should earn over its full range of broadly similar products.

As the gross profit margin remunerates a sales company for performing marketing and selling functions; the RPM especially depends on comparability regarding functions performed, risks assumed and assets used or contributed. The RPM thus focuses on functional comparability. A similar level of compensation is expected for performing similar functions (using similar assets and assuming similar risks) across different activities. If there are material differences that affect the gross margins earned in the controlled and the uncontrolled transactions, adjustments should be made to account for such differences. In general, comparability adjustments should be performed on the gross profit margins of the uncontrolled transactions. The operating expenses in connection with the functions performed, assets used or contributed and risks assumed should be taken into account in this respect, as these differences are frequently reflected in different operating expenses.

The following issues should be considered in determining whether the functions performed, assets used or contributed and risks assumed by an uncontrolled entity are comparable to those of a controlled entity for purposes of applying the RPM:
►    In contrast to the CUP Method, the reliability of the RPM can be influenced by factors that have less effect on the price of a product than on the costs of performing functions. Such differences could affect gross margins even if they do not affect the arm’s length prices of products (e.g. the composition of COGS). These factors could include cost structures (e.g. accounting practices), business experience (e.g. startup phase or mature business) or management efficiency;
►    A resale price margin requires particular attention where the reseller adds substantially to the value of the product, for example by assisting considerably in the creation or maintenance
of intangibles related to the product (e.g. trademarks or trade names) or where goods are further processed into a more complicated product by the reseller before resale;
►    The amount of the resale price margin will be affected by the level of activities performed by the reseller. For example, the distribution services provided by a reseller acting as a sales agent will be less extensive than those provided by a reseller acting
as a buysell distributor. The buysell distributor will obviously obtain a higher compensation than the sales agent;
►    If the reseller performs a significant commercial activity in relation to the resale activity itself, or if it employs valuable and unique assets in its activities (e.g. valuable marketing intangibles of the reseller), it may earn a higher gross profit margin;
►    The comparability analysis should try to take into account whether the reseller has the exclusive right to resell the goods, because exclusive rights may affect the resale price margin;
►    The analysis should consider differences in accounting practices that apply to the reseller and to comparable companies in order to make appropriate adjustments to enhance comparability; and
►    The reliability of the analysis will be affected by differences in the value of the products distributed, for example, as a result of a valuable trademark.

It should be recognized that returns to similar functions (bearing in mind assets used or contributed and risks assumed) may not be the same in different markets. Generally, reliability is enhanced when the reseller and the comparable companies are operating in the same market.

​​​​Where a significant part of the potential comparables profits is attributable to significant, unique intangibles, such as unique product design or unique engineering, that are not present in the tested party, it may not be possible to eliminate the effects of such intangibles on operating profits by performing reliable comparability adjustments. If accurate adjustments cannot be made, the potential comparable may need to be rejected.

​​​​​​The degree of comparability between a tested party and an uncontrolled taxpayer is impaired when the entities assume different economically significant risks which may require making a risk adjustment. For example, a contract manufacturer in certain circumstances may not assume the market risks that fully-fledged manufacturers customarily do and therefore might be expected to earn lower profit margins.

​​​​​​There is no universally accepted method for risk adjustment. However, in practice MNEs carry out risk adjustment through the application of certain methods that attempt to quantify on an ex ante basis (i.e. before the event) the effect of risk on anticipated profitability based on, for example, the weighted average cost of capital/capital asset pricing model. However, it is worth mentioning that both models are based upon risk models used mainly in relation to the risk of securities. Most statistical methods have their inherent limitations. Therefore, risk adjustment must be made carefully, only where needed and only if a reasonable and accurate adjustment is possible.

It has to be recognized that problems can arise due to significant differences in the transactional structure between associated party sales in a controlled company and similar transactions involving independent companies.

​​​​​In some cases material differences may exist in the way transactions are structured by potential comparables and by the tested party, due to the fact that the latter operates with associated enterprises in an MNE. In such cases it may not be possible to find comparable transactions that have the same transactional structure as the controlled transaction. In these circumstances, adjustments may be needed to eliminate the effects of these differences. For example, the margins of independent distributors operating on short-term contracts may not be comparable to those of associated enterprises on long-term contracts, unless an adjustment is made to account for the short duration of the former.

​​​​​It has to be stressed that comparability adjustments should be considered if and only if they are expected to increase the reliability of the results. Relevant considerations in this regard include the materiality of the differences for which an adjustment is being considered, the quality of the data used in the adjustment, the purpose of the adjustment and the reliability of the approach used to make the adjustment.

​​​​​Comparability adjustments are only appropriate for differences that have a material effect on the comparison. A comparison may be appropriate despite an unadjusted difference, provided that the difference does not have a material effect on the reliability of the comparison.

​​​​​No specific rules or guidelines can be given that may be applicable to every transaction. In each case, the critical factors that have a material impact on the price of the product (if the CUP Method is used) or on profit (if the RPM, CPM, TNMM or PSM is used) should be identified. Ultimately, decisions regarding whether adjustments are required depend entirely on the facts and circumstances surrounding the transactions, on the availability of information needed for the analysis and on the accuracy and reliability of any adjustments that may be made.

Available information is often not complete enough to enable a review to be made of each possible comparability factor. The analysis almost always takes place with imperfect information. That realization can be helpful in deciding whether a particular difference is material enough to make adjustments, or whether the comparability difficulties should affect the selection of the most appropriate method.

​​​​​​​See Chapter 4 for an explanation of transfer pricing methods and selection of a suitable method.

​​​​​A comparability analysis may result in an “arm’s length range” of financial indicators (prices or margins), all of which may considered to be equally reliable. It should be noted that the domestic law in some countries will specify how such a range is to be derived from the final results of the comparables, for instance by the use of particular statistical techniques. Where the transfer price is within this range, it is normally accepted as arm’s length.

​​​​​However, it may be difficult to determine whether the search process has indeed resulted in a range of results, all of which are equally reliable. Uncertainty may also arise in cases where the range of results from a comparables search is very wide. Where such concerns exist therefore, it can be helpful to consider whether it is possible objectively to determine whether some potential comparables are more reliable than others. The (further) use of diagnostic ratios and qualitative screening can sometimes be helpful in this regard.

​​​​​The search for reliable comparables is at the heart of most transfer pricing analyses. In many cases, it may not be straightforward, but rather, require the application of judgement. Care should thus be taken to consider potential screening criteria as objectively as possible, and avoid “cherry-picking” data. Similarly, absent factual changes, it would be expected that such criteria would be used consistently over time.

The degree of comparability between the controlled and the uncontrolled transactions, including the reliability of comparability adjustments needed and the availability of reliable information (especially on uncontrolled comparables) are key factors in selection of the most appropriate transfer pricing method. Other factors include the strengths and weaknesses of the method, the appropriateness of the method in the light of the nature of the controlled trans-action (based upon a functional analysis) etc. See further Chapter 4, Methods.

​​​​​​Once the transfer pricing method is selected, the next step is to apply the selected method to arrive at the correct arm’s length price or profit (or range of prices or profits) and to document the results of the analysis. The topics mentioned in this section are discussed in greater detail in Chapter 4 and 12.

​​​​​​The comparability analysis should be as reliable as possible. However, on many occasions a careful comparability analysis will not yield perfect matches in terms of comparable enterprises or comparable transactions to those carried out by the associated enterprises. The nature, type, quality etc. and number of comparables along with the adjustments made during a comparability analysis may be the subject of debate, interpretation and contention between the taxpayer and tax authorities. Some of the key concerns surrounding comparability analysis are described below.

​​​​​​Timing of origin of the transactions needs to be considered. In principle, information relating to the conditions of comparable uncontrolled transactions undertaken or carried out during the same period of time as the controlled transaction (“contemporaneous uncontrolled transactions”) is expected to be the most reliable information to use in a comparability analysis, because it reflects how independent parties have behaved in an economic environment that is the same as the economic environment of the taxpayer’s controlled transaction.

Timing of collection of the relevant comparable data is also a potential issue. In some cases, taxpayers implement transfer pricing documentation to demonstrate that they have made reasonable efforts to comply with the arm’s length principle at the time their intragroup transactions were undertaken, i.e. on an ex ante basis (hereinafter “the arm’s length price-setting” approach), based on information that was reasonably available to them at that point. Such information includes not only information on comparable transactions from previous years, but also information on economic and market changes that may have occurred between those previous years and the year of the controlled transaction. In effect, independent parties in comparable circumstances would not base their pricing decision on historical data alone. This ex ante analysis of the arm’s length price is, however, not the most common approach.

​​​​​In other instances, taxpayers might test the outcome of their controlled transactions to demonstrate that the conditions of these transactions were consistent with the arm’s length principle, i.e. on an ex post basis (hereinafter “the arm’s length outcome testing” approach). This test typically takes place as part of the process for establishing the tax return at the year-end. An ex post (after the event) analysis is the most commonly used method to test the arm’s length price of international transactions.

​​​​​​The arm’s length price setting and the arm’s length outcome testing approaches, as well as combinations of these two approaches, are found among countries that have implemented transfer pricing rules. Country views differ as to whether data on contemporaneous transactions which only become available to the taxpayer and tax administration at the time of filing of the tax return, or conducting ex post analysis of transfer pricing, is permitted or represents improper use of hindsight.

​​​​​Another potential question is whether and how to take into account future events in the transfer pricing analysis. Such events were not predict-able at the time of the testing of a controlled transaction, in particular where valuation at that time was highly uncertain. The question should be resolved, both by taxpayers and tax administrations, by reference to what independent enterprises would have done in comparable circumstances to take account of the valuation uncertainty in the pricing of the transaction.

The main issue is to:

• Determine whether the valuation was sufficiently uncertain at the outset that the parties at arm’s length would have required a price adjustment mechanism; or
• Whether because the change in value was so fundamental, or other developments arose, this would have led to a re-negotiation of the transaction.

Where this is the case, the tax administration would be justified in deter-mining the arm’s length price for the transaction on the basis of the adjustment clause or re-negotiation that would be provided at arm’s length in a comparable uncontrolled transaction. In other circumstances, where there is no reason to consider that the valuation was sufficiently uncertain at the outset that the parties would have required a price adjustment clause or would have renegotiated the terms of the agreement, there is no reason for tax administrations to make such an adjustment as it would represent an inappropriate use of hindsight. The mere existence of uncertainty should not require an ex post adjustment without consideration of what independent enterprises would have done or agreed between them.

​​​​​Data from years following the year of the transaction may also sometimes be relevant to the analysis of transfer prices, but care must be taken to avoid the use of hindsight, perceiving the significance of facts and events with the benefit of knowledge accruing after they have occurred.

​​​​​One of the most frequent problems taxpayers and administrations face with comparability analysis is the lack of reliable comparables with respect to the transaction(s). This issue can be particularly acute in devel-oping countries. Issues relating to lack of reliable comparables are discussed above and in the PCT Transfer Pricing Toolkit (see 2.5.3.2).

​​​​​​When products, property or services are offered by first-movers in specific segments there may be a dearth of comparables. These transactions typically involve use of new technology, cutting-edge research, bundled intangibles etc. which may not have satisfactory comparables. An example is intellectual property content relating to high-tech computer software. Such situations may be dealt with either by using a one-sided method (CPM, RPM or TNMM) for which the tested party is the one that does not contribute such intangibles; or, in those cases where unique intangibles are contributed by both parties to the transaction, by using a profit split method.

​​​​​​Owing to business consolidation and vertical integration, it may be extremely difficult in some industries to find reliable internal or external comparables. An example is the pharmaceutical industry where there exists a high level of vertical integration and consolidation in order to drive up efficiencies. In such scenarios the controlled transactions are part of a larger global supply-chain and it can be difficult to find comparable transactions between independent enterprises. In such cases also, it may be possible under certain circumstances to use comparables from other industries, possibly adjusted, in order to address this issue.

It is frequently not possible to obtain information on perfect comparables in practice, and it is therefore often necessary to use broad search criteria when identifying third-party comparables. It must be ensured that potentially relevant external comparables are not excluded because of “cherry picking” of favourable third-party information by either the taxpayers or the tax authorities, ignoring other information that does not support the position argued for.

​​​​​For example, extreme results may be rejected as comparables after careful consideration by the tax authorities as they tend to skew the data. While this could be a correct application of the arm’s length principle in certain circumstances, in other cases the reasons for the extreme result may be genuine and may not always justify rejecting the company from the pool of comparables.

The need for careful consideration can be particularly important for a comparable company that loses money in a given year. This may occur because of a recession year which hit the controlled and uncontrolled trans-actions in the same way, but the loss may also be attributable to unique factors with the potentially comparable company which may warrant a conclusion that strict comparability standards are not satisfied.

To come to a correct conclusion, an unbiased analysis of the facts and circumstances surrounding the transactions should be carried out. Where one or more of the potential comparables is loss-making, further examination would be needed to understand the reasons for such losses and confirm whether the loss-making transaction or company is a reliable comparable.

​​​​​Well-documented search procedures and comparability criteria make the application of the comparability standard transparent, in that the comparability standard that was applied is clearly stated and its scope can be evaluated. This will ensure that results are less susceptible to “cherry picking” since the reasons for rejection of each potential comparable are provided.

A deliberate or intentional set-off occurs when an associated enterprise has provided a benefit to another associated enterprise within the MNE group and is compensated in return by that other enterprise with some other benefits. These enterprises may claim that the benefit each has received should be set-off against the benefit each provided and that only the net gain or loss if any on the transactions needs to be considered for tax assessment.

Set-offs can be quite complex. They might involve a series of trans-actions and not just a single transaction or they may involve several members of the MNE group. Ideally the parties identify and disclose all set-offs accurately and have enough documentation to substantiate their set-off claims so that after taking account of these the conditions governing the transactions are consistent with the arm’s length principle.

​​​​​​The tax authorities may evaluate the transactions separately to deter-mine whether the transactions satisfy the arm’s length principle. However, the tax authorities may also choose to evaluate the set-off transactions together, in which case comparables have to be carefully selected. Set-offs in international transactions and in domestic transactions may not be easily comparable due, for example, to the asymmetries in the tax treatment of the set-offs under the taxation systems of different countries.

​​​​​​The price paid or payable for the goods (which under certain limited circumstances may also include the costs of services and royalties) in import transactions is the starting point for determination of any applicable customs duties. A higher price on import reduces the profit of the importer (all other things being equal) and thus the direct tax that might be due in the importing country. However, where customs duties apply, this would also result in higher duties being payable. Accordingly, there may be an inherent conflict between the revenue implications and the motivation of the customs and direct tax authorities. While the direct tax authority would focus on overvalued import prices, the customs authority will seek to ensure that the declared customs value has not been undervalued to reduce duty liability.

​​​​​The WTO Agreement on Implementation of Article VII of the General Agreement on Tariffs and Trade 1994 (“the WTO Valuation Agreement”) sets out the methodology for determining the customs value of imported goods. Customs valuation is the procedure applied to determine the customs value of imported goods for the purpose of calculating ad valorem customs duties. Article 22 of the WTO Valuation Agreement requires that each Member of the WTO shall ensure the conformity of its laws, regulations and administrative procedures with the provisions of the Agreement. In contrast, for direct tax purposes, the tax authorities in most countries use the “arm’s length principle” as a standard for valuing cross-border related party transactions.

​​​​​​Both customs valuation and transfer pricing approaches aim at determining an appropriate price for cross-border transactions. However, the approaches differ and in some cases may be incompatible due to different motivations, theoretical frameworks, documentation requirements or other factors, causing practical difficulties for importers. Therefore, better coordination and exchange of information between direct tax and customs authorities is encouraged. However, the extent to which this is possible may depend on how the customs services and tax administrations are organized in each country. For some countries the two organizations are more integrated, and for others they are completely separate.

In appropriate circumstances the verified customs value may be useful to tax administrations in evaluating the arm’s length character of the transfer prices of imported goods in international transactions between associated enterprises. In particular, customs may have contemporaneous information regarding the transaction that could be relevant for transfer pricing purposes, while tax authorities may have transfer pricing documentation which provides detailed information on the circumstances of the transaction.

​​​​​​Some customs administrations are now also making use of transfer pricing data, where relevant, to ensure that the price of an associated party transaction has not been affected by the special relationship between the parties.

​​​​​There are similarities between customs valuation and transfer pricing methods, for instance, one method permitted for the purposes of verifying customs values uses a comparison between the value of the goods imported by a related party with the value of identical or similar goods imported by independent parties, which may be considered as analogous to the application of a CUP Method for transfer pricing. Examining customs values may thus provide relevant information and a useful starting point for transfer pricing purposes in some cases and may also help in reducing the compliance burden for taxpayers. However, it should be borne in mind that customs valuation methods are highly prescriptive and may not be fully aligned with the arm’s length principle as it applies for direct tax purposes.

​​​​There has been a great deal of focus internationally on the inter-play between transfer pricing and customs valuation methods. Following two joint World Customs Organization (WCO)– OECD conferences in 2006 and 2007, it became clear that harmonization of the two systems was not a realistic proposition; particularly given the fact that the WTO Valuation Agreement is not expected to be updated in the short to medium term. Discussions have therefore focused on the extent to which Customs may use transfer pricing information when carrying out examination of related party transactions. The principle of the customs valuation in cases involving related party exporters and importers and where there are doubts as to the reliability of the price paid or payable for the goods, is to judge whether the special relationship between the parties influenced the price by exam-ining “the circumstances surrounding the sale” (WTO Valuation Agreement Article 1, paragraph 2(a)).

​​​​​​The WCO’s Technical Committee on Customs Valuation, which has the mandate for ensuring, at the technical level, uniformity in interpretation and application of the WTO Valuation Agreement, has issued several instruments on this topic.²⁶ These are briefly summarized below.

​​​​​Commentary 23.1 recognizes the principle that a transfer pricing study may, in some cases, be used by customs as a basis for examining the circumstances of the sale. Following this general principle, Case Study 14.1, sets out a scenario where customs use transfer pricing documentation, based on the TNMM, to confirm that the prices applicable in a related party trans-action have not been influenced by the relationship between those parties.

​​​​​​Case Study 14.2 also provides an example of customs authorities making use of transfer pricing information (based on the RPM) but in contrast concludes that the declared import price was not settled in a manner consistent with normal pricing practices of the industry but rather had been influenced by the relationship between the buyer and seller.

The WCO has produced the WCO Guide to Customs Valuation and Transfer Pricing (2018) mentioned in 3.6.6.8 above, which includes all relevant technical information on the two methodologies and explores the interaction between them. It includes good practices for customs and tax administrations, and businesses. In particular, customs and tax administrations are encouraged to work more closely together and the guide emphasizes that businesses should consider customs’ needs when developing transfer pricing strategies. To this end, the WCO has produced Guidelines for Strengthening Cooperation and the Exchanging of Information between Customs and Tax Authorities at the National Level (October 2016).²⁷ These Guidelines endeavour to provide guidance and ideas to customs and tax authorities for formalizing the contacts and strengthening the existing cooperation at the national level, on a range of issues of mutual interest.

Concern is often expressed by taxpayers, especially MNEs, over aspects of data collection by tax authorities and its confidentiality. Tax authorities have access to, as they need to, very sensitive and highly confidential information about taxpayers, such as data relating to margins, profitability and business contracts. Confidence in the tax system means that this information needs to be treated carefully, especially as it may reveal sensitive business information about that taxpayer’s profitability, business strategies and so forth.

​​​​​A secret comparable generally refers to the use of information or data about a taxpayer by the tax authorities to form the basis of transfer pricing scrutiny of another taxpayer. The taxpayer under scrutiny is not given access to that information — it may, for example, reveal confidential information about a competitor (i.e. the first taxpayer— to which the data relates).

​​​​​There is a need to exercise caution against the use of secret comparables unless the tax administration is able, within the limits of its domestic confidentiality requirements, to disclose the data to the taxpayer whose transactions are being reviewed. This would enable an adequate opportunity for the taxpayer to defend its own position and to safeguard effective judicial control by the courts. Taxpayers contend that the use of such secret information is against the basic principles of equity, as the taxpayer is required to benchmark its controlled transactions with comparables not available to it, without the opportunity to question comparability or argue that adjustments are needed. Taxpayers contend that it would be unfair if they face the consequences of adjustments made on this basis, such as additions to income, typically coupled with interest, penalties etc. Furthermore, double taxation may not be relieved if secret comparables cannot be disclosed to the compe-tent authority of another country.

​​​​​Comparability analysis looks simple in theory but in practice it can be a laborious, difficult, time-consuming and, more often than not, expen-sive exercise. Seeking information, analyzing all the data from various sources, documenting the analysis and substantiating adjustments are all steps that require time and money. It is therefore important to put the need for comparability analyses in perspective. The aim should be to ensure that the compliance burden and costs borne by a taxpayer to identify possible comparables and obtain detailed information thereon are reasonable and proportionate to the complexity of the transaction. It is recognized that the cost of obtaining information can be a real concern, especially for small to medium-sized operations, but also for those MNEs that deal with a very large number of controlled transactions in many countries. However, it should be observed that the burden of cost cannot be a reason for the dilution of comparability standards.

​​​​​​These resource considerations apply at least as much to many devel-oping countries, and efforts must be made to ensure that their position is not prejudiced by a lack of such resources in ensuring the arm’s length pricing of transactions in their jurisdictions.

​​​​​When undertaking comparability analysis there is no requirement for an exhaustive search of all possible relevant sources of information. Taxpayers and tax administrations should exercise judgment to determine whether particular comparables are reliable.

​​​​Transfer pricing theory meets practice in comparability analysis— the translation of the arm’s length principle into the selection of reliable comparables and of the appropriate transfer pricing method, eventually yielding the transfer price. This is all facilitated by comparability analysis.

​​​​​​A good comparability analysis is an essential step in any transfer pricing analysis in order to gain a correct understanding of the economically significant characteristics of the controlled transaction, and of the respective roles of the parties to the controlled transaction. This will assist in the selection of the most appropriate transfer pricing method in the circumstances of the case. This part of the process is fact-based and requires the taxpayer or tax administration to demonstrate an understanding of how business operates.

​​​​​In most cases, the application of the selected transfer pricing method will then rely on the identification of uncontrolled comparable transactions. This part of the process may be particularly complicated, especially in countries that have limited access to information on potential comparables. It is worth emphasizing that solutions exist to deal with this problem, including the collection of information on internal comparables (i.e. transactions between the taxpayer or its associated enterprise and a third party) where they exist; the collection of public information on third parties (e.g. competitors) that are likely to be involved in uncontrolled transactions comparable to the taxpayer’s controlled transaction, or the possible use of databases from other countries.

​​​​​​It is clear that the comparability analysis should be as reliable as possible so as to arrive at the correct arm’s length price or profit (or range of prices or profits). In performing this comparability analysis, it may be necessary for the taxpayer or the tax authorities to undertake a detailed functional analysis taking into consideration a wide variety of data sources, other factors and, if necessary, a series of comparability adjustments while arriving at a suitable set of benchmarks (or comparables). The choices made in the course of this analysis have to be substantiated and the overall process has to be thor-oughly documented.

​​​​​​Further, as noted in the introduction, the lack of comparables for a given controlled transaction does not mean that it is or is not at arm’s length or that the arm’s length principle cannot be applied. This is especially important given the growing importance of integrated business models and of transactions involving unique intangibles for which comparables may not be available. The need for a reliable analysis must therefore be balanced with a pragmatic approach and one should not set unrealistic expectations for comparability analyses.

Financial transactions between independent enterprises are based on various commercial considerations. Members of an MNE, however, have the flexibility and discretion to decide the conditions that will apply to financial transactions within the group. As a result, in an intragroup situation, the tax consequences may be a consideration which has an influence on the nature or structure of those financial transactions.

​​​​​Financial transactions are an important part of the operations of MNEs. Typically, financial transactions play a role in supporting the value creation process of MNEs. The responsibilities of corporate treasurers include managing cash to help MNEs meet their financial and busi-ness obligations and objectives. The aims of the corporate treasury function will include ensuring the availability of necessary cash flows, and will often extend to evaluating investment strategies to achieve an appropriate balance of risk and reward. Debt management is an integral part of their responsibility, as it is common for MNEs to finance part of their operations through loans. Similarly, the corporate treasury function may also manage the costs of external funding through the use of intragroup guarantees or through cash pooling arrangements. MNEs may operate on a decentralized basis with respect to their treasury management whereby each entity or group of entities within the group is responsible for its own financing arrangements; or it may instead centralize the treasury function at a regional or global level.

​​​​​Intragroup financial transactions are subject to the arm’s length principle just as intragroup services, intragroup sales of products and components, and other intragroup transactions are. As is the case in any other intragroup arrangement, the application of the arm’s length principle to financial transactions first requires the accurate delineation of the actual transaction (see 3.3.2), including consideration of the purpose of the financial transaction in the context of the business of the MNE. Guidance on these matters is provided in section 9.6.

Regulated financial institutions like banks and insurance companies are governed by supervisory authorities, central banks and multinational banking institutions and require a license to operate. Such institutions are subject to a strict regulatory regime (Basel III rules) which may influence the intragroup financial transactions they can enter into. This chapter does not address transfer pricing of financial transactions conducted within a regulated financial institution. Instead, the discussion and guidance in this chapter are tailored to non-financial services MNEs that engage in intragroup financial transactions. The chapter does not attempt to cover the full range of financial transactions that may occur. However, as a general principle, the transfer pricing analysis of any intragroup financial transaction follows the same analytical framework as would apply to any other intragroup transaction. These are the principles laid out in Chapter 3 (Comparability Analysis), which describes the process by which the actual transaction can be accurately delineated and priced by reference to the arm’s length principle.

​​​​​​Several factors combine to make intragroup financial arrangements important for both taxpayers and tax administrations:

• The significance (in terms of amounts involved and frequency) of these transactions for MNEs;
• The fact that money is mobile and fungible, which makes it relatively simple for an MNE to shift or locate debt in a particular MNE entity; and along with that debt, the ability to deduct associated interest. This can have the effect of reducing taxable profits of the borrowing entity, and can, depending on the situation of other MNE members, reduce the MNE overall tax liability;
• The difficulty that tax administrations may face in determining the true character and the specific features of certain intragroup financial instruments; and
• The concern that excessive interest deductions provide an opportunity for tax base erosion.

For the above reasons, many countries have introduced tax measures aimed at reducing the tax advantages of debt financing.

This chapter will introduce the transfer pricing considerations for intragroup financial transactions. First, it will describe the commercial considerations relating to corporate financing decisions. It will then present some of the more common types of intragroup financial transactions (section 9.3) as well as describing the operations of group financing departments/entities (section 9.4). The chapter goes on to describe various relevant corporate income tax approaches taken by tax administrations (section 9.5). Next, the chapter discusses the application of the arm’s length principle to financial transactions in general (section 9.6 to 9.11), and finally, there are sections specifically covering intragroup loans (section 9.12) and intragroup financial guarantees (section 9.13).⁶³

Corporate financing decisions are of fundamental relevance for an MNE. When an MNE identifies a business opportunity for which it requires funding, it will need to consider whether it should make use of internal funding (to the extent it is available) and /or external funding. Where external funding is used, consideration of the balance between equity financing and debt financing may also be required. Each has its own advantages and disadvantages that extend beyond tax considerations. Interest payments deriving from debt financing are generally deductible for the borrower (payer) and taxed as ordinary income in the hands of the lender (payee), whereas dividend payments, or other equity returns are generally not tax deductible to the payer and are often subject to some form of tax relief (exemption, credit, etc.) in the hands of the payee. It is beyond the scope of this Chapter to address the economic benefits or disadvantages of corporate financing decisions.

​​​​​Although there are many theories that have attempted to hypothesize the relevant factors defining an optimal corporate capital structure, it should be noted that numerous factors influence the decision of a company’s Management Board when defining the capital structure of their firm. Transfer pricing considerations do not serve to determine what capital structure is optimal for a company.

​​​​​However, the capital structure of an entity that is a part of an MNE may impact the transfer pricing analysis of intragroup financial transactions. That is, to assess the impact of an entity’s capital structure on intragroup financial transactions, it may be important to analyze the entity’s debt capacity. This specific aspect is not discussed further in this Chapter. The commentary to Article 9 is relevant in this respect.

Many ordinary activities of an MNE require consideration of funding. These include: assuring cash flow for day-to-day operations, funding mergers or acquisitions, or making available credit facilities for operating companies seeking to purchase plant or equipment. Different financial instruments may be utilized, depending on the amount of funding needed and length of time for which the funding is required. A wide range of instruments can be regarded as financial transactions. Typical examples include equity instruments (e.g. common stocks), debt instruments (e.g. ordinary and special bank loans, ordinary and special bonds, commercial paper and money market instruments, debentures, government securities), and financial derivatives (e.g., foreign exchange transactions, stock options, futures, forward contracts, notional principal contracts, investment derivatives and other hybrids).

​​​​​In an intragroup context, more common financial transactions include intragroup loans, financial guarantees by a parent or group financing company for third-party loans to subsidiaries, cash pooling arrangements, hybrid financing, derivatives, and treasury services (e.g. foreign exchange risk management, factoring and forfeiting, netting arrangements, payment factories, commodity risk management, captive insurance, asset management, carbon trading, etc.). Intragroup loans and financial guarantees are discussed in more detail in sections 9.12 and 9.13, respectively.

​​​​​Company treasurers are generally concerned with how to ensure MNEs have access to cash to meet their anticipated needs, to secure cost-effective financing, and to provide financial risk management appropriate to the level of risk the MNE wishes to assume. For example, if an MNE operates internationally, it is likely to receive payments in different currencies. For planning and budgeting purposes, different currencies present variability of future cash flows (usually at a cost). Entering into a forward contract can hedge (and effectively fix) future cash flows. Not hedging would leave the company exposed to currency fluctuations and to uncertainty as to the actual cash flows in the future. Group treasury functions may monitor these risks, evaluate any natural hedges that may exist within the MNE, and price hedging contracts where appropriate. Similarly, the need to buy inputs for production such as commodities that are subject to price fluctuations can cause substantial profit and loss volatility for a company. It is not always possible to enter into fixed price contracts for commodities; when it is possible, fixed price contracts may exclude the possibility to obtain further cost savings. The company’s procurement and treasury departments may therefore work together to evaluate hedging arrangements. This chapter on financial transactions does not discuss hedging transactions.

​​​​​​In addition to management of cash flows, entering into intragroup loans or using revolving lines of credit, MNEs may also issue bonds or securities in the market to fund activities or to refinance existing loans. To make securities issued by the company more attractive to external investors, a parent company guarantee may be provided in support of the subsidiary that is the issuer of record. Similarly, a parent company may issue a guarantee to an independent bank that provides funds to a subsidiary company which has a low credit rating, in order to improve the terms and conditions of the subsidiary’s bank loan (e.g. to reduce interest expenses). Intragroup financial guarantees come in many forms and are discussed in more detail in section 9.13.

Where an MNE has subsidiaries in different countries, the different parts of the business may be independently responsible for the cash management functions of each entity or group of entities. If these areas each act prudently, ensuring they have adequate cash, the group may, as a whole, end up holding more cash than they would have needed for operating purposes had they pooled their resources. Where these individual cash requirements are managed through a centralized treasury department which has a revolving credit facility with a third-party bank, the total amount drawn down will be greater than it would otherwise need to be. The MNE’s treasurer may decide to implement a centralized cash pooling arrangement to net off the facility (i.e. a target-balancing or zero-balancing cash pool). This would reduce the cost of the credit facility (or avoid having to take out a loan for other needs) and make optimal use of the average cash balances sitting idle in the accounts of each area.

​​​​​There are also cash pooling arrangements where a bank combines the debit and credit balances of different entities or departments of the MNE to derive net balances on a real or notional basis. As a result, interest is credited on a positive overall balance and debited on a negative overall balance (i.e. notional or interest compensation cash pool).

​​​​​​An intragroup cash pooling arrangement can generate numerous advantages for participants in the arrangement. These include minimizing the liquidity requirements, minimizing external interest costs, ensuring flexible day-to-day financing, reducing transaction costs related to local bank accounts, increasing the participants’ bargaining power relative to external banks and thus helping them to obtain more advantageous conditions (e.g. interest rates) on the common bank account, and helping to centralize financing decisions. This chapter does not however discuss cash pooling transactions in further detail.

​​​​​​Another common type of intragroup financial transaction is captive insurance. An MNE may create an insurance company to provide coverage for participating MNE entities. Typically, the main purpose for doing so is to avoid using third-party insurance companies, which may have volatile pricing, or may not meet the specific needs of the company. By creating its own insurance company, the MNE may aim to stabilize premiums, reduce costs, cover difficult-to-insure risks, have direct access to reinsurance markets, and increase cash flow. When a company creates a captive insurer, it is indirectly able to evaluate the risks of subsidiaries, write policies, set premiums and ultimately either return unused funds in the form of profits, or invest them for future claim pay-outs. Captive insurance companies are also sometimes set up to insure the risks of the MNE’s customers. This is an alternative form of risk management. This chapter on financial transactions does not discuss captive insurance transactions in any detail.

​​​​​The scope of this chapter will be limited to the analysis of intra-group loans and intragroup financial guarantees, since they are the most commonly seen financial transactions in practice. However, the general guidance on these transactions may also be relevant to other types of financial transactions.

​​​​​Financial transactions can be performed and organized in different ways within an MNE. The organization of the treasury function will depend on the structure of a given MNE and the complexity of its operations. Different treasury structures involve different degrees of centralization. In its most decentralized form, each entity within the MNE has full autonomy over its own financial resources. At the other end of the spectrum, a centralized treasury has full control over the financial resources of the entire group. That is, it centralizes some or all activities relating to cash and liquidity management, management of foreign exchange risk and interest rate risk, etc. In these situations, individual group members remain responsible for operational matters, while responsibility for financing matters is held by the central treasury. Centralization of financing and treasury functions can offer significant scale benefits and financing cost savings for an MNE.

• Treasury departments/entities operating as profit centres. Such departments/entities seek profits by actively creating market positions, as well as actively managing operational exposures. To be able to manage operational exposures they tend to be centralized in terms of structure and control of financial activities/ requirements. They may operate as in-house banks, maximize the profits of their own operations, and assume the risk of capital.

In practice, a combination of the profiles above is often seen.

The category of treasury department/entity that renders the specific financial transactions that are in place may be relevant and provide an initial indication of the most appropriate method to be used to assess the arm’s length nature of intragroup financial transactions. To determine an arm’s length remuneration for services rendered, an accurate delineation of the actual transaction (including a functional analysis) is required. See further Chapter 5 on Intragroup Services.

​​​​​​Treasury departments/entities operating as service centres are typically remunerated by applying the CUP Method, the CPM, or the TNMM based on cost. In contrast, treasury departments/entities operating as profit centres are typically remunerated based on pricing the transactions conducted and allocating the credit risk of those transactions to the treasury department. Consequently, it may be appropriate for the ‘spread’ between costs of funding and return on cash invested to be largely allocated to that treasury department/entity. To determine the arm’s length remuneration for financial transactions such as loans and guarantees, reference is made to sections 9.12 and 9.13. It should also be remembered that the ‘substance’ of centralized activities generally requires careful review and is an important element of the accurate delineation process.

​​​​​Raising corporate tax revenue can be especially important for devel-oping countries. To the extent that a country’s tax systems provide for income tax deductions for interest and not for dividends or other returns to equity capital, there is an economic incentive for companies doing business in those countries to use debt financing.

​​​​​To reduce the base erosion effect of debt financing and the relevance of the tax factor in choosing between equity and debt financing, some countries have made the tax policy choice to introduce in their domestic tax laws measures aimed at either reducing the advantage of debt financing or increasing the advantages of equity financing. Measures which seek to limit base erosion associated with excessive debt financing can be broadly grouped into General Anti-Avoidance Rules (GAARs) and Specific Anti-Avoidance Rules (SAARs). For a more in-depth discussion on the specific available measures to counter excessive interest deductions claimed by residents, reference is made to the UN Practical Portfolio on Protecting the Tax Base of Developing Countries against Base-eroding Payments: Interest and Other Financing Expenses.⁶⁴ The Practical Portfolio includes a discussion of the pros and cons of various measures⁶⁵ to counter excessive interest deductions.

One common approach is to implement a rule that would limit net interest expense deductions based on their ratio to earnings before interest, taxes, depreciation and amortization (EBITDA).⁶⁶ Banks, insurance companies and other financial businesses (leasing companies, asset management companies, companies subject to special tax regimes) might require special consideration under such regimes.⁶⁷ Another common approach is to implement thin capitalization rules which limit interest deductions by reference to a maximum allowable debt to equity ratio.

​​​​​​The interaction between the corporate income tax measures addressing financing decisions such as rules on thin capitalization and the transfer pricing rules might need careful consideration under domestic law, since both sets of rules might address similar issues and result in completely or partially denying deductions for interest and similar expenses.

​​​​​The assessment of the arm’s length nature of an intragroup financial transaction essentially follows the same approach that applies for other intra-group transactions discussed in section 3.2. It requires the identification of the commercial or financial relations (including an understanding of the economically significant characteristics of the controlled transactions) leading to the accurate delineation and recognition of the actual transaction, and, after that, the selection and application of the most appropriate transfer pricing method. In this chapter, for practical purposes, references are often made to loan transactions since they are a common type of intragroup financial transaction. However, similar considerations apply to other types of intra-group financial transactions.

​​​​​​From a policy perspective the question regularly arises as to whether base erosion through excessive debt may be tackled through application of the arm’s length principle. Article 9 of the UN Model Convention embodies the arm’s length principle. The commentary to this Article references the OECD Commentary on Article 9, which in turn clarifies that the Article is relevant not only in determining whether the rate of interest provided for in a loan contract is an arm’s length rate but also whether a prima facie loan can be regarded as a loan or should be regarded as some other kind of payment, in particular a contribution to equity capital. Based on the analysis in the UN Article 9 Commentary, (developing) countries have expressed the desire to use the concepts of Article 9 as embodied in their domestic transfer pricing rules for purposes of analyzing the arm’s length nature of intragroup financial transactions and determining not only whether interest charges are excessive, but also whether the financial transaction is accurately delineated as debt. In this respect, reference is also made to paragraphs 3.1.4 to 3.1.9.

​​​​​The decision to characterize a transaction that is presented as a loan (in its entirety or partly) as something other than a loan requires careful analysis and should be based on adequate information, as such a conclusion may lead to double taxation (see section 3.3.2.2). The approach taken to analyzing financial transactions is up to the tax authorities of the relevant jurisdiction, although it is recommended that they clarify which approach is routinely followed under their domestic transfer pricing rules and guidance.

​​​​​​Considering the above, the analysis of the arm’s length nature of financial transactions can arguably be conducted from several perspectives. First, it could be undertaken by initially accepting the taxpayer’s characterization of the transaction as a loan at face value, until the facts and circum-stances of the transaction (including any additional available evidence or conduct of the parties) leads to a conclusion that the transaction is commercially irrational. If the latter conclusion is arrived at, the transaction may be disregarded as a loan for transfer pricing purposes. That conclusion does not necessarily affect the civil law or common law denomination of the transaction; it only affects the transfer pricing analysis. In this first scenario, the transaction essentially is treated as it is presented, until and unless it can be considered commercially irrational.

​​​​​Alternatively, a second possibility is that the analysis of the financial transaction could be conducted from the perspective of determining whether the economically significant characteristics of the transaction lead to the conclusion that it sufficiently resembles and has the features or hallmarks of a loan, or alternatively, it more resembles something other than a loan. At a certain point, the review of the characteristics (together with the conduct of the parties or any additional evidence) may lead to the conclusion that the financial transaction is not a loan. If so, it may be that for transfer pricing purposes, the transaction ought to be treated as something other than a loan. This conclusion does not necessarily affect the civil law or common law denomination of the transaction, or its classification for accounting purposes; it only affects the transfer pricing analysis. Furthermore, if the facts and circumstances of the transaction lead to the conclusion that the transaction is commercially irrational, it may be disregarded as a loan for transfer pricing purposes.

​​​​​A third possibility involves the same process as the second in terms of determining the characteristics of the financial transaction, however it also considers whether only part of the transaction should be treated as a loan. Relevant evidence might, for example, include a debt capacity analysis of the borrower. In such cases, it may be that for transfer pricing purposes the transaction is treated partly as a loan and partly as something other than a loan, such as a contribution to equity (see also the guidance at 3.1.8). Once again, this conclusion does not necessarily affect the civil law or common law characterization of the transaction or its classification for accounting purposes; it only affects the transfer pricing analysis. Furthermore, if the facts and circumstances of the transaction lead to the conclusion that the transaction as a whole is commercially irrational, the transaction may be disregarded as a loan for transfer pricing purposes.

​​​​​​Before deciding to question the nature of a financial transaction and split it into its component parts, tax authorities would be expected to have conducted a detailed analysis of the transactions and relevant associated enterprises. This may include consideration of the purpose to which they intend to put the borrowed funds, their economic circumstances, relevant business strategies, creditworthiness, debt capacity and security offered, etc. as outlined at 9.7.2.2 below. In all three scenarios mentioned above, the treatment of a transaction as something other than a loan would, for tax purposes, result in a limitation in the deductibility of interest expenses (entirely or partially) and not necessarily imply characterizing the transaction as some-thing else (e.g. an equity instrument).

The following section provides an overview of economically significant characteristics of a financial transaction that may be considered when assessing such transactions for transfer pricing purposes.

​​​​​​Some of the economically significant characteristics of a financial transaction include the following:

​​​​​​Determining the arm’s length nature of a financial transaction requires that the perspectives of both parties to the transaction are considered. With respect to an intragroup loan, for example, this means that the economically relevant characteristics of the transaction should be analyzed from the perspective of both the lender and borrower. At arm’s length, a lender will conduct a credit assessment of the borrower to make the decision on whether to provide a loan, as well as on the amount and the terms of the loan. A borrower will generally assess whether the term of the loan will meet its commercial needs and fall within its debt capacity. It will also need to have the capability to make decisions relating to the risks it is purported to assume.

​​​​​​The arm’s length nature of a transaction should initially be considered by reference to the transaction actually undertaken by the associated enterprises as it has been structured by them. Tax administrations should examine the conduct of the parties and base the analysis of the financial transaction under review on the actual conduct of the parties. Based on domestic law or tax treaty considerations, it may be that the “label” applied to an intragroup financial transaction is not correct. Where this is the case, as discussed in 9.7.1 above, the arm’s length principle may be applied to characterize an intragroup financial transaction as being different from that which was initially presented by the taxpayer.

​​​​​​Separately, it should be noted that in many jurisdictions there may be domestic jurisprudence on the above aspects as well, and this may impact on the nature of transactions involving funding. Domestic jurisprudence may be relevant or even determinative for the characterization of a financial transaction. However, in instances where the character of a financial transaction as debt or equity is not clear and where jurisprudence does not provide persuasive guidance, consideration of the relevant aspects mentioned in this chapter may be useful to the analysis of the transaction.

​​​​Once the financial transaction is accurately delineated, the most appropriate transfer pricing method can be selected and applied. Within this process, potentially comparable financial transactions can be identified, and comparability adjustments might be applicable, to determine the arm’s length price or profit (or range of prices or profits) for the financial transaction(s) under review.

​​​​​​To accurately delineate the actual transaction and to be able to seek reliable comparables to test the arm’s length nature of the intragroup financial transaction, the creditworthiness of the associated enterprises involved in the transactions may need to be considered. The creditworthiness of an enterprise has regard to the potential that the enterprise will not be able to meet its payment obligations in accordance with the terms of the transaction (in this respect mention is also made of “debtor” or “issuer” credit ratings, where the term “issuer” indicates the debtor). In the case of intragroup loans, this essentially involves, inter alia, consideration of the security of the lending (that is, what collateral the borrower can offer) and consideration of the borrower’s likely future cash flows to pay interest and repay the principal amount of the debt. One way to assess debt capacity is to look at the credit rating of the debtor, which reflects the credit risk for a creditor extending debt to the debtor.

​​​​​​Credit risk may be measured by assigning a rating (i.e. credit rating) to the debtor or to a specific instrument issued by a debtor. In some cases, these ratings may be derived from independent commercial credit rating agencies. The rating expresses the probability of default and thus the risk to the potential lender. In addition to specialist credit rating agencies, some companies have also developed in-house commercial tools that are intended to be used for credit rating purposes. Credit ratings provided by independent credit rating agencies generally consider both qualitative and quantitative factors. Whereas credit rating methodologies used in in-house commercial tools generally prioritize quantitative factors and may not include qualitative factors such as industry forecasts, MNE strategies or the risk profile resulting from the MNE’s management style.

​​​​​​Determining a credit rating is not an exact science and can be particularly difficult for certain types of issuers such as start-ups, special purpose vehicles, or indeed for members of an MNE. In the case of a credit rating determination for a member of an MNE, the financial metrics used in the process may be influenced by related party transactions.

A robust process for determining a credit rating will typically include consideration of both quantitative and qualitative factors. It is important to remember even amongst issuers with the same credit rating there is likely to be a range of creditworthiness represented. It should be considered that, while they can be useful, credit ratings are only an indication of an entity’s probability of default and they may not be perfect. For example, in the 2009 financial crisis, some entities with high credit ratings defaulted on their debts and ended up bankrupt. Furthermore, in some developing countries the government may have official prescribed interest rates in place and no use is made of international commercial credit rating approaches.⁶⁹

A summary of the credit rating categories used by the credit rating agen-cies Moody’s, Standard and Poor’s and Fitch is provided in the following table.

Table 9.T.1

Credit Ratings

Source: Bhatia, A. V. (2002). Sovereign Credit Ratings Methodology: An Evaluation. IMF Working Paper 02/170. Washington: International Monetary Fund. Table 2, p.8

​​​​​​When applying the arm’s length principle, the starting point is that the related parties involved in the financial transaction should be treated as if they were entities independent of each other, but otherwise in the same circumstances. However, “the same circumstances” must include any incidental benefits and group synergies that derive from the fact that the related entities belong to an MNE. This would include the impact of any implicit support (sometimes also referred to as ‘passive association’, ‘parental support’, or ‘group support’). To the extent that a borrower that is a member of an MNE benefits from an improved credit rating solely on the basis of implicit support, no payment is required for this benefit.

Credit ratings from independent professional rating agencies such as Standard & Poor’s, Moody’s or Fitch, are typically only available for the parent company of a group. Where no such independent credit rating is available for the borrower of the funds, consideration will therefore need to be given as to how to evaluate the credit risk of that borrower. The following approaches may be considered:⁷⁰

• Beginning with the parent’s credit risk, adjust this credit risk (if required) to approximate the credit risk of the borrower; and
• Derive the borrower’s credit risk by using various credit scoring tools.

The effect of any implicit support available to the borrower would need to be factored into the analysis irrespective of the approach taken.

​​​​​​When assessing the credit rating of the associated enterprise and the extent of any implicit support that needs to be taken into account, consideration should be given to the specific circumstances, including (i) that the associated enterprise belongs to an MNE (which may have a higher credit rating than the enterprise) and (ii) that the parent or other group companies may be reasonably expected to support their affiliates (especially core affiliates) in their financial needs (referred to as ‘stewardship by the parent company’). These factors may significantly influence the analysis of the arm’s length conditions of the overall transaction. An improved credit rating for an associated enterprise based merely on implicit support or passive association does not require a return or payment at arm’s length.

​​​​​​Expanding on the considerations of implicit support presented above, it might be relevant to consider the following questions:

• To what extent (if any) would implicit support be taken into account by independent institutions (e.g., independent lenders or credit agencies) when assessing the credit risk of the borrower?
• How would the implicit support be quantified?

​​​​​​In practice, the answers to the above questions will depend in large part on the level of strategic importance that the borrower has in the MNE (including the potential consequences of a default by the borrower on the rest of the MNE). In this regard, the following aspects could be considered:

• If the consequences of not supporting the borrower would create negative impacts on other parts of the group (for example due to legal obligations, operational impacts, effect on group reputation, etc.);
• If there are explicit statements of policy/intent by the parent/ group to support the borrower;
• If there is a history of support to MNE entity borrowers in cases where they get into financial difficulty.

The following table is an example of the possible effects on the credit rating of an MNE entity borrower based on its level of strategic importance to the MNE:⁷¹

Table 9.T.2

Group Rating Approaches

It should be noted that implicit support is not equivalent to an explicit guarantee and is generally unenforceable by a creditor of the borrower. See 9.13 which deals with financial guarantees.

​​​​​​It is also important to note that although implicit support is typically associated with improving the credit rating of the borrower, it might also be the case that the borrower’s credit rating is negatively influenced by the MNE’s credit risk (i.e. as a result of negative synergies).

In addition to the credit rating of the borrower entity, for accurate delineation purposes the credit rating of the specific debt instrument is also relevant. See 9.12.⁷²

Where there are significant difficulties in determining the extent and effect of any implicit support, and in cases where there is substantial information asymmetry, challenges may be created in the transfer pricing analysis which, if not resolved, may result in outcomes that are not reliable. In such cases, the credit rating of the MNE may also be used for pricing the accurately delineated loan where the facts so indicate, particularly in situations such as where the borrower entity is important to the group, and where the borrower’s indicators of creditworthiness do not differ significantly from those of the MNE.

​​​​​A further question arises as to whether the credit rating of the borrower entity should be established based on its creditworthiness before the financial transaction under review is put in place or afterwards. In most cases, the situation after the new financing transaction takes place should be considered.

​​​​​In addition to the considerations above in determining the credit rating of a borrower entity it may also be relevant to consider the risk of an entity operating in a particularly risky country (i.e. the risk deriving from a country’s business environment including legal environment, levels of corruption, and socioeconomic variables such as income disparity), to the extent that this is not already reflected in the credit rating of that entity. The country risk for developing countries tends to be higher than for developed countries due to greater perceived or real risk of currency fluctuations; political instability; economic instability such as recessions or higher inflation; the risk of default by the government on sovereign debt and the effect of foreign exchange and other controls. A high level of country risk will impact the business risk of a borrower located in that country and therefore also (likely decrease) its credit rating.

​​​​​In addition to considering the creditworthiness of the borrower, in order to accurately delineate the actual transaction and to be able to seek reliable comparables, the specific features of the financial instrument must also be taken into account. For example, if ACo makes available a loan to associated enterprise BCo, but BCo already has three other loans (regardless of their sources), and the loan from ACo is subordinated to the earlier loans, then the “status” of the loan from ACo is lower than that of the other three loans. That is, if, in the event of BCo’s bankruptcy, ACo is only entitled to claim repayment from BCo after the repayment of the other three loans, ACo holds a subordinated loan instrument with a higher risk. It may therefore be relevant to consider the credit rating for a specific financial instrument (e.g. a specific loan) and not just the credit rating of a borrower more generally.

​​​​​Note that in the case of bonds (which are often used as comparables for intragroup loans) this risk “status” is generally expressed as the “issuance credit rating”, that is the credit rating that applies to a particular bond issue rather than to an entity.

​​​​​In practice, the credit rating of a particular financial instrument is generally notched down from the credit rating of the borrower entity, (usually) based on methodologies provided by credit rating agencies. That is, when comparables are sought for the financial instrument, first the credit rating of the borrower is considered, and subsequently the credit rating of the particular financial instrument is estimated by adjusting the credit rating of the borrower, taking into account the features of the instrument.

​​​​​​For example, assume that the credit rating of BCo is BBB, and the financial instrument provided by ACo to BCo is subordinated. Assume also that in line with the methodology provided by credit rating agencies, it is considered appropriate to apply a one-notch credit rating downgrade to reflect the subordinated nature of this financial instrument. Now, the credit rating of this financial instrument is BBB-, which is a one-notch credit rating downgrade based on the investment grade ratings (in this example of S&P and Fitch) presented in Table 9.T.1. Note however that different rating agencies have different approaches. While there is no universal approach, the effect of subordination merits consideration. See also 9.7.2.2.

​​​​​Any of the transfer pricing methods described in this Manual can be used to price financial transactions. With respect to intragroup loans, the most commonly used transfer pricing method is the CUP method. The CUP method may be employed when comparable transactions exist between one party to the intragroup loan transaction and an independent party (“internal comparable”) or between two independent parties, neither of which is a party to the intragroup loan transaction (“external comparable”). This is discussed further in section 9.12.

Separate and apart from the pricing of individual financial transactions or instruments, treasury services provided within an MNE are likely to require an arm’s length remuneration (see section 9.3.3). For these services, the CPM or a cost-based TNMM may be appropriate (or, if the treasury services entity is simply a conduit and adds no value, remuneration at cost). It is common that one entity of the group (e.g., the financing department/ entity) acts as a general treasury service provider or intermediary for other entities in the group. See chapter 5 on intragroup services. However, if a financing department/entity provides financing to group members and refi-nances these with deposits from other group members or external sources and therefore is exposed to a mismatch in timing and/or currencies as well as credit risk exposure, the CPM might not be the appropriate transfer pricing method to reward the financing function.

​​​​​​Another method that could be appropriate in some cases is the transactional PSM. However, in practice, the use of this method for financial transactions is quite limited (some exceptions being for global trading of financial instruments or for certain cash pooling transactions).

To simplify the determination of the arm’s length price for intra-group financial transactions, some countries have introduced safe harbours, most of which concern interest rates. More specifically, some countries issue annual official interest rates or margins that, if applied to intragroup loans, extinguish the obligation for the taxpayer to prove that the compensation related to those transactions is arm’s length, while providing some assurance that the rate applied will not pose a significant risk of base erosion.⁷³ See also section 9.12.2.8 on the application of “sixth method” approaches to intra-group loans.

​​​​​Access to the credit rating of individual associated enterprises and the determination of the impact/effect of implicit support on intragroup financial transactions may be difficult and is often based on judgements or determinations that can be very hard to verify for tax administrations. Therefore, another possible simplification could be to use the MNE (or parent) credit rating as a basis if taxpayers do not provide evidence to substantiate the credit rating used. This approach has the added benefits of providing certainty and reducing the administrative burden to both tax administrations and taxpay-ers. See section 9.8.14.

​​​​​When determining the arm’s length compensation for an intragroup financial transaction, the use of simplification measures or safe harbour rules should be carefully considered. Furthermore, it should be considered how the simplification or safe harbour interplays with the definition and application of the arm’s length principle both on a domestic and on an international level. In some countries, taxpayers maintain the right to rebut a safe harbour or simplification rule by demonstrating that an alternative amount is at arm’s length. In other countries, no such option exists. As regards to the use of safe harbours, reference can be made to sections 5.5 and 10.2.2.

​​​​​This section sets out a number of relevant characteristics of intra-group loans. An intragroup loan is the provision of financial resources from one related party (the lender) to another (the borrower) to be repaid at a later date. With an intragroup loan, the borrower will obtain the financial resources; the lender will generally assume the credit risk related to the loan and needs to be compensated for the liquidity provided and the risk taken on by an arm’s length interest payment. Ideally, the relevant terms and conditions of the loan will be specified in the loan agreement and supported by the conduct of the parties. If, and to the extent that an MNE has specific (explicit) group polices in place with respect to the (target) cost of financing, the likely impact thereof (if any) on the characteristic of a particular loan might also be considered relevant.

In practice, many different types of loans exist. Two examples are provided below:

• Term loan: a loan with a specified schedule for the payment of interest and the principal amount,⁷⁴ and a maturity ranging from one to ten plus years. Such loans are often used to fund medium- and long-term assets such as plant and equipment. A term loan may be secured or unsecured, carry a fixed interest rate or a floating interest rate, and contain general or specific performance covenants; and
• Revolving loan or revolving credit facility: a secured or unse-cured credit line with a maturity ranging from six months to five plus years that a borrower can draw down and repay multiple times. A typical facility requires the borrower to pay the bank an annual commitment fee on the entire line in order to keep it available for future use; those without a fee are typically not committed and may be withdrawn by the bank at will. In some instances, banks require borrowers to repay the facility in full before allowing further draw-downs or renewals (a process known as a clean-up call).

Apart from the credit risk, the most common risks relevant to an intragroup loan will be interest rate risk, reinvestment risk, call/prepayment risk, inflation (or purchasing power) risk, liquidity risk, exchange rate (or currency) risk, volatility risk, political or legal risk, event risk, sector risk and country risk. During the accurate delineation process, the allocation of these risks will generally be considered. See Table 3.T.1 in section 3.4.4.6.

​​​​​​When analyzing an intragroup loan, relevant characteristics that may be considered include the following: conversion rights, currency, any applicable guarantees, timing and calculation of interest payments, options, repayment clauses, security provided, seniority and other terms of the loan. Loan characteristics that benefit the borrower generally have the effect of increasing the interest rate and those that have the impact of benefitting the lender tend to decrease the interest rate.