The interrelation of transfer pricing and customs

I came across an interesting topic the other day and thought I would write my next blog post on it. This topic has been a discussion point between many authoritative bodies, professional service firms and tax jurisdictions. I thought I would try and shed some light on the topic. As you are well aware there is more to a cross border transaction than just transfer pricing. Especially when structuring certain transactions, there are many different aspects that should be taken into account. One of the factors that must be taken into account when looking at cross border transactions (be it related or independent) is customs.

Customs are the duties levied by a government on imported goods which are payable by the importer of a product. Should the imported product be in relation to a cross border related party transaction, both customs and transfer pricing are applicable. Important to note here is that customs only looks at goods and not at services. Further, customs duties are usually a big part of the total cost in selling a product to a foreign market and as such it is important to comply with the different customs regulation to ensure certainty for total taxes payable on the transaction.

Both disciplines try to create an arms length price/transaction. Therefore the topic that is often discussed in relation to both disciplines is whether both can be complied with by deriving the one from the other. For example, would a MNE be able to determine its customs duties from a detail transfer pricing analysis?

In order to analyse the topic it may be beneficial to look at the similarities and differences of both disciplines first.

What transfer pricing is and how it is dealt with has been discussed previously, but for this blog post it is important to understand that in practice, from a transfer pricing perspective an arms length price is rarely determined for an individual product or service between related parties but it is rather determined on an overall return such as an operating margin. On the other hand, customs is concerned with the valuation of each individual product that is imported (sometimes exported).

In the table below are other important similarities and differences in relation to transfer pricing and customs:



Transfer Pricing


Arms  length

Both disciplines try to ascertain an arm’s length  price/remuneration for the transaction under review (arm’s length connotes  fair value or market value)


The transaction value method from a customs perspective is  similar to the comparable uncontrolled price method from a transfer pricing  perspective


Type of Discipline

Indirect Tax

Direct Tax


Limited window of opportunity (usually at the moment of  invoice or sale)

Usually audited after the tax return is filed

When applicable

Customs duties are applicable at the moment of importation

Forms part of corporate tax which is applied based on taxable  income reported for the year

From the above it seems that transfer pricing and customs are indeed very different but do have a similar goal of trying to ascertain a market value/arms length return for the transaction under review. The big difference between transfer pricing and customs lies in how an arms length price/remuneration (or market value of a unit) is determined. Except for the comparable uncontrolled price (“CUP”) method, transfer pricing methods use a margin to ensure a cross border related party transaction is at arms length (i.e. the party receives a fair remuneration for functions performed, assets used and risks assumed) whereas customs is only looking at a unit price. Even though many courts are arguing in favour of using a CUP method when performing a transfer pricing study, in practice the CUP method is rarely used.

The problem that a taxpayer may face within a normal tax jurisdiction is that both disciplines try to pull the cost per unit in different directions. In other words, a low cost price of an imported good will result in low customs duty but a low cost price will inversely increase the taxpayers profitability resulting in a higher corporate tax payable and vice versa. In practice the customs department of a tax authority will try to increase the price per unit, whereas, the transfer pricing department of the same tax authority will try to decrease overall cost by decreasing the cost per unit. Both departments try to achieve the highest return for the fiscus possible by increasing or decreasing the same cost price respectively.

Depending on how a tax jurisdiction determines an arms length price/remuneration for both disciplines, there may be some room for tax planning due to the different rates applicable. I.e. if customs duties are higher than corporate taxes a taxpayer may be able to lower its cost price through certain structuring exercises as long as he is still within an arms length price/remuneration for both disciplines. On the other hand, because both disciplines do not necessarily use similar methods to determine an arms length price/remuneration it may happen that for each discipline the arms length price/remuneration is different. This may result in double tax for a taxpayer (or if lucky – savings) and therefore it is important to compare the outcome of both disciplines.

Even though transfer pricing and customs can rarely leverage an arms length price from each other it is still important to analyse both disciplines when looking at cross border related party transactions. Tax authorities should (and do) use information collected from customs for transfer pricing audits and vice versa to ensure compliance with both disciplines.

Lastly, from the above one can see that a joint process for the two disciplines could benefit both tax authorities and taxpayers through simplification and cost/time savings.

Attribution of profits – notional transactions

Tax_1The OECD has provided a new business profits article in its 2010 Model Tax Convention (“MTC”) and discusses the concept of notional transactions within the Commentary to the MTC. A notional transaction in short can be defined as “An estimate of a real transaction, not based on direct measurement”. In other words, for transfer pricing purposes a notional transaction is the hypothetical creation of a transaction between a head office and its permanent establishment (“PE”) to simulate a transaction that would generally exist between third parties.

The new “functionally separate entity” approach implemented by the OECD changed the previous business profit article as follows:

  • It requires a much further/deeper separation of the PE from its head office; and
  • The PE can now claim certain expenditure relating to notional interest, notional royalties/licences and mark ups on notional management fees, which it previously couldn’t.

So how come the new “functionally separate entity” approach is not followed by all OECD member countries or UN countries? I am not certain as to why some countries have elected to not follow the new approach but the following is a plausible answer.

When creating/simulating a notional transaction in order to attribute profits to a PE this will create more expenditures than previously available. The creation of extra expenditure will result in less taxes payable by that PE within its tax jurisdiction. This in itself means that the countries who ‘house’ the PEs will be worse off than previously. Now one might argue – but what if I create a subsidiary, then all the expenses would also be deductible. That is correct, but with real transactions other taxes are applicable. For example a ‘real’ transaction involving interest payments usually triggers withholding taxes, a notional transaction will not trigger such a tax.

From the above it seems that tax jurisdictions which generally have more PEs, such as the UN countries would be worse off implementing the new “functionally separate entity” approach.

What are your thoughts on the above? Have you got another plausible answer?

Intellectual Property – The Beginning

rubik cubeIntellectual property and intangibles are one of the hot topics and as such I wanted to touch on the subject. Please note that intellectual property and intangibles sometimes are used interchangeably, however, there are instances where an intangible is not an intellectual property (for example most websites are not considered to be intellectual property). Because intangibles encompass intellectual property, in transfer pricing we usually talk of intangibles when referring to IP.

A lot of work around IP has been done by the OECD Guidelines and as such I do not want to copy exactly what has already been done but rather touch on some issues or discuss some problems that one may encounter. I do not have all the answers to the problems and hope to get some discussion going rather than just giving my view points.

Please refer to the OECD’s discussion draft on IPs for an in depth analysis on:

  • Identifying Intangibles
  • Identification of Parties Entitled to Intangible Related Returns
  • Transactions Involving the Use or Transfer of Intangibles
  • Determining Arm’s Length Conditions in Cases Involving Intangibles

The above is still a discussion draft and the OECD published comments on the draft which can be found here.

When doing a transfer pricing analysis we must not forget two important things, firstly transfer pricing is not an exact science and secondly the arm’s length principle. From a transfer pricing perspective the steps performed to find an arm’s length price/arm’s length range for a cross-border related party transaction involving IP is similar to that of other goods or services. The biggest problem usually lies within the valuation of the IP. Putting the statement differently, how do you value something like the coca cola brand? Yes there are many different methods on how to value brands but do they give us a fair result? The problem is that we are trying to compare a cross-border related party transaction involving IP to that of independent companies involved in a comparable transaction, however, independent parties may not be involved in such transaction.

Before going further into the above let’s start from the beginning. As mentioned by the OECD’s discussion draft on IP, a transaction involving IP would fall under Article 9 of the Model Tax Convention, if a double tax treaty is in place. Further, the OECD provides a definition of IP and what it entails. This may result in the following problem. The definition provided by the OECD may be much wider than local legislation in  different countries and as such the OECD definition may encompass transactions that would otherwise not fall within the ambit of IP. As we all know double tax treaties merely give a taxing right but cannot create one. So what happens if one country that has a double tax agreement with another country would define a transaction as dealing with IP and the other country does not?

The problem would be solved if we all used the definition provided by the OECD but this does not seem to be the case. In some countries double tax agreements become law by being written into the income tax act or are given the power through constitutional sections. But then we still haven’t solved the issue that a double tax treaty cannot create additional tax.

What are your thoughts on this? The OECD is trying to deal with some of these issues with the well known BEPS (Base Erosion and Profit Shifting) action plan which will be discussed in another post.

The arm’s length principle

In this post I will discuss the arm’s length principle and different viewpoints thereof. Please note the arm’s length principle is not easily implemented, even if it may sound easy from a theoretical perspective. Further, I will touch on why the arm’s length principle may not be the best approach or as some say it is inherently flawed. Please note there are many discussions around the arm’s length principle and I am merely trying to summaries it here. I would love to discuss your viewpoints around the arm’s length principle and its application in the comment section below. 


From the transfer pricing definition discussed previously, it is clear that it is of the utmost importance to understand the arm’s length principle, as the whole transfer pricing methodology is based upon the arm’s length principle.

The arm’s length principle is the basis of transfer pricing, which is not an exact science but rather a methodology which can be interpreted differently by different people. The arm’s length principle, simply stated, requires that each inter-company transaction is remitted to the same level that would have applied had the transaction taken place between independent parties, all other factors remaining constant. Although this may sound simple, Raby (2009) believes the actual application of the arm’s length principle in practice is notoriously difficult.

Why do we have an arm’s length principle?

The reason for establishing an arm’s length principle was to ensure that each tax jurisdiction will get its fair share from an affected transaction between related parties. The OECD further states that independent parties normally deal with each other within financial and commercial relations which are determined by market conditions. These market forces that determine the price between independent parties are comparable to the arm’s length principle. This means that two independent parties do not have much choice in determining a price between them. If the price is too high the one party will find someone else to contract with and if the price is too low the other party may make losses and will not be interested in doing business under those terms.

When related parties transact with each other, the financial and/or commercial relations may not affect a transaction in the same way as between independent parties. For example an MNE may try to increase profits on a global basis rather than on a company to company basis, regardless of where the companies are incorporated. This means that the MNE tries to be as profitable as possible on a global level even if a single entity of the group has to make losses. This is one of the reasons why it may be so difficult to ascertain an arm’s length principle for some transactions. It is very difficult if not impossible to compare such an MNE transaction to an independent party transaction because an independent party would not enter into a contract which will only guarantee losses or fail to provide a proper return on investment. An MNE as a whole can achieve savings overall due to the loss on one transaction and therefore the transaction makes commercial sense from an MNE perspective. Usually tax jurisdictions would seek to adjust such loss making cross-border related party transactions to profit making as the tax jurisdiction is only concerned with the transaction and not the MNE as a whole (For example a loss making product within a product range – if this still is not clear please ask me for further clarification in the comment section).

There may be justifiable reasons for losses from a tax jurisdiction’s perspective. For example, during the start-up stage of an MNE there may have been high capital costs for the new manufacturing plant and/or manufacturing equipment or if the MNE plans to penetrate a new market, and therefore offers its products at a lower price to gain a niche in the market. The OECD (2001) further acknowledges that this is one of the reasons why tax administrations should not automatically assume that related parties of a global MNE have sought to manipulate their profits, but should recognise that the MNE does have certain loss-leaders for valid economic and commercial reasons.  

There are other reasons for MNEs to deviate from an arm’s length consideration but this are left for later blogs/discussions.

Is the arm’s length principle flawed?

Some argue that the arm’s length principle is inherently flawed. This is mainly due to the fact that the arm’s length principle does not account for the economies of scale related to integrated systems when compared to independent parties. MNEs are known to have great cost savings through centralised management structures and cost centres (as discussed previously in ‘The reason for the existence of MNEs’). These savings are, however, not considered in the determination of the arm’s length range.

The OECD Guidelines (2001) acknowledge that the arm’s length principle may not always be simple to use in practice but it is sound in theory and gives the closest approximation to a fair price between related parties. The arm’s length principle usually allocates appropriate levels of income between off-shore related parties and is therefore accepted by tax administrations. There may be instances when the arm’s length principle is flawed but it is the closest method of establishing a fair principle for each tax administration. There are no other acceptable principles or methods to determine values for cross-border related party transactions that are fair and sound in theory. The arm’s length principle has been accepted internationally by the major corporations and tax administrations and the experience with the arm’s length principle has become “sufficiently broad and sophisticated to establish a substantial body of common understanding among [them]” (OECD, 2001:I-6). This understanding ensures that each tax administration receives its fair share of taxes and in addition the corporation does not suffer double taxation.

 Global formulary apportionment

The global formulary apportionment has been suggested as an alternative to the arm’s length principle to compare a cross-border related transaction and its related profits for each participating tax administration. The OECD (2001) expresses the opinion that some tax administrations have tried to use the method without success. The global formulary apportionment method is not seen as a suitable method to determine the arm’s length price and the OECD Guidelines do not recommend the use of the global formulary apportionment (Please note some still believe that the global formulary apportionment is the way forward and will take over the arm’s length principle in the future).

The global formulary apportionment method is not accepted by OECD member or observer countries and therefore is not a realistic alternative for the arm’s length principle. The reason given by the OECD Guidelines (2001) is that the global formulary apportionment does not achieve the protection against double taxation or ensure taxation of the profit by a single fiscal authority. In order to achieve this, it would require extensive international coordination and consensus on the global formulary apportionment method. The difficulty in this is that every single tax administration must agree to the method and its predetermined formula. A common accounting system would have to be chosen and adopted within all the tax jurisdictions, even the non-member countries. To achieve this may be very time consuming, extremely difficult and there is no guarantee of no double taxation, because if one tax administration does not apply the method in its jurisdiction there would be a problem.

Closing words

I hope the above made sense. My opinion is that the arm’s length principle isn’t 100% correct but it is the best we have and as such it is the closest we will get to a fair tax share between all the tax jurisdictions involved in a cross-border related party transaction. If the issues of the global formulary apportionment can be resolved it may be a better option but only time will tell.

Please feel free to share your experiences in the comments below.

Don’t forget to follow me on twitter for the latest updates on the blog

The reason for the existence of MNEs

GlobeIn order to be able to apply a correct transfer pricing method to satisfy the arm’s length principle, it is important to understand the reason for the existence of MNEs when compared to corporate entities going into the market alone. Once the reason is understood, a further analysis is needed to establish how this affects the arm’s length principle. 

Vannoni (n.d.) states that the relevant theoretical approaches which discuss the nature and existence of MNEs are the transaction costs theory, the monopolistic advantage theory and the resources theory, which are discussed in detail below.

Transaction Cost Theory

Vannoni (n.d.) defines the ‘transaction cost’ as the cost involved in making an economic exchange between two parties and is the result of imperfect information in markets which behave irrationally and therefore lead to bounded rationality and opportunism. This implies that the cost of gathering the information, bargaining prices and the enforcement costs of a transaction fall within the transaction cost. MNEs exist because they manage to organise these costs more efficiently and therefore have transaction cost savings.

Monopolistic Advantage

The ‘monopolistic advantage’ which is established through a monopoly is defined by Philip Mohr, Louis Fourie and Associates (2008:246) as “a market structure in which there is only one seller of a good or service that has no close substitutes. A further requirement is that entry to the market should be completely blocked”. It may be argued that there is no such thing as a real monopoly but MNEs may come very close. Examples of MNEs in ‘near-monopolies’ are: Microsoft, SABMiller or Coca-Cola. From such a ‘near-monopoly’ an MNE has certain advantages over smaller companies. One of the obvious reasons is that an MNE within a near-monopoly can charge higher prices for goods or services to its clients or customers when compared to a competitive market. If there is no substitute for a commodity or service, prices are higher because the customer cannot obtain the same good or service more cheaply anywhere else. MNEs gain most advantages by effectively managing inefficiencies that arise in markets, by communicating and assisting each other with information or resources, which is not the case
between two competing independent parties.

Resource Theory

The final theory, the resource theory, is defined by Wade and Hulland (2004) as follows:

“firms possess resources, a subset of which enable them to achieve competitive advantage, and a subset of those that lead to superior long-term performance. Resources that are valuable and rare can lead to the creation of competitive advantage. That advantage can be sustained over longer time periods to the extent that the firm is able to protect against resource imitation, transfer, or substitution. In general, empirical studies using the theory have strongly supported the resource-based view.”

From the theories discussed above one can conclude that the reason for establishing an MNE is to save costs, acquire better information in a market, reduce risks and to be more competitive when compared to an independent corporate entity. 

When discussing the arm’s length principle in our next post the above will become more evident and important.


The start of a journey through transfer pricing

It is probably not news anymore that around the world, tax jurisdictions are facing budgetary crises and are doing everything in their powers to generate additional tax revenues. Transfer pricing is seen, by most tax jurisdictions, as an effective way to increase said tax revenues. As such, more and more tax jurisdictions are tightening or implementing transfer pricing regulations and strengthening enforcement thereof.

JourneyOn the other hand multinational enterprises (“MNEs”) are under pressure by shareholders to become more profitable which in current economic times is a tough task. Many MNEs look at transfer pricing in order to structure their tax affairs in the most efficient, yet legal, way. A correctly applied transfer pricing policy can save MNEs millions of dollars in taxes whereas a wrong strategy or implementation may result in taxes and penalties which are far greater than any tax estimates which may render a profitable transaction loss making.

The purpose of this post is to give a brought overview on transfer pricing and therefore will not go further into detail on transfer pricing policies and/or strategies. We will leave that for a little later.

So what is transfer pricing? A colleague of mine once described transfer pricing as the art of splitting a pie which has been created by friends (or otherwise related people) residing in different jurisdictions into fair pieces for each tax jurisdiction involved. This sounds much easier than it really is. To give you some food for thought, how do you determine who gets how much? I mean, who is to say that the base of the pie is more important than its filling and therefore the person who made the base should get a greater share of the pie. What if the different jurisdictions provide different ingredients to make the pie, which is more important, the eggs or the flour? Again, I cannot stress enough that this is merely a very simplified example but I am just trying to get the point across that transfer pricing is tricky, to say the least.

Let’s get into the theoretical stuff. Transfer pricing is defined by the Tax Foundation (2008) as follows:

“The price that is assumed to have been charged by one part of a company for products and services it provides to another part of the same company, in order to calculate each division’s profit and loss separately. Generally, transfer pricing rules indicate that one affiliate must charge another affiliate the same price as would be demanded in an ’arm’s length transaction’ [,a] transaction between two related or affiliated parties that is conducted as if they were unrelated, so that there is no question of a conflict of interest.” 

One of the reasons why a company would sell products more cheaply or more expensively than the prevailing market price to a related part of its enterprise may be for the purpose of “earnings stripping” Please note there are other legitimate reasons but earnings stripping or also known as profit stripping is a tax authority’s main concern.

The Tax Foundation (2008) defines earnings stripping as follows:

“Earnings stripping is a process by which a firm reduces its overall tax liability by moving earnings from one taxing jurisdiction, typically a relatively high-tax jurisdiction, to another jurisdiction, typically a low-tax jurisdiction. Often, earnings stripping arrangements involve the extension of debt from one affiliate to another. Debt is accumulated in a high-tax jurisdiction that allows a company to deduct interest payments from their taxable income. “

An example of earnings stripping would be where a company resident in country X sells semi-finished goods to a related party resident in a tax haven (country Y) at a lower selling price when compared to the same product being sold to an independent party. This is to ensure that there will be no or a very small profit in country X. When the product is sold in country Y, at a market related price, the profit will be higher when compared to the same product being sold by an independent party. This is due to the fact that the independent party would have had to buy the semi-finished good at a higher market related input price whereas the related company in country Y did not. The higher profit that is achieved in country Y will be taxed at a much lower tax rate as it is situated in a tax haven compared to country X. This will lead to a tax saving for the group as a whole.

I hope the above made sense so far and if not please feel free to ask many questions in the comment box below. I will continue with the basics of transfer pricing in my next post. Until then, happy reading.