Our article could begin like a Belgian story: "Two tax experts walk into a room, one French, the other Belgian". Unfortunately, the wind of reform blowing through the OECD since the start of its so-called "BEPS" project has silenced the laughter. Now is the time for analysis and anticipation, not jokes. Let's face it: there's no kidding around with tax adjustments that could bring down an entire business model, based on concepts now widely shared by tax authorities everywhere. Even the Cayman Islands are on board. What a joke!
The rapid publication of the OECD's initiatives, broken down into 15 concrete actions, has - it has to be said - created a real buzz. Everyone has their own opinion and interpretation. But all agree that international taxation, as it has been practiced for decades, needs to undergo a revolution. Revolution is the right word: in France, we're well placed to know that the phenomenon never goes smoothly, and that ignorance of the basic issues at stake is a breeding ground for antinomian versions of history.
Our Belgian neighbors, hitherto rather quiet on the subject, have acted far more wisely. In response to the ebullition of practitioners, the administration recently issued a draft circular with 11 chapters and no fewer than 50 pages. Published in the country's two official languages and made public on November 9, the draft aims to clarify the basic concepts of transfer pricing and related issues, such as permanent establishments and mutual agreement procedures. It also summarizes the OECD principles and offers a welcome insight into the methods and terminology often used by practitioners, such as "PLI", "Cost Plus" and other such gimmicks.
The initiative is to be welcomed. It was high time that a government department ventured to clearly state its position and interpretation of the workings of transfer pricing. Admittedly, the circular is only at the draft stage, and will undoubtedly undergo a number of modifications. Even when finalized, it will only have the value of doctrine, and as such will have a limited place in the hierarchy of norms and the legitimate quest for legal certainty. Admittedly, this circular will only be binding on the Belgian tax authorities, and as such should not lead us to draw any direct conclusions on our French transfer pricing regulations. Let's face it, the present article is limited in scope, if only to satisfy our desire to write.
However, it is interesting to note the positions taken by the Belgian tax authorities on points that are still the subject of debate in France, or that are fuelling a long-running dispute. We have no doubt that good ideas often lead to others, and it is likely that the best-informed opinions of the Belgian tax authorities will prompt us to revise or reinforce our positions on the same subjects.
"The best market is the most expensive
This old Belgian saying will soon be a thing of the past. Indeed, the draft circular immediately begins its first chapter by setting out and explaining the arm's length principle. To the OECD's sibylline formulas without punctuation, the Belgian administration strikes back with simplicity - a resolutely Belgian quality - and states: this concept "stipulates that in their financial and commercial transactions, related enterprises are required to act as if they were unrelated. If they agree to terms that deviate from the terms that unrelated companies could agree to, the transfer price can be adapted " 2.
After outlining the factors of comparability for assessing the relevance of independent transactions, there is one point that is of particular interest to legal experts. The draft circular states that "the determination of a transfer price is based on the analysis of existing contracts concluded between related companies concerning inter-company transactions. The contracts are supposed to express what the functional analysis finds. Otherwise, the behavior of the parties involved in the transaction takes precedence, and contractual provisions are disregarded" 3.
In so doing, the Belgian administration seems to be siding with Anglo-Saxon theories, which relegate contracts to the rank of mere presumption. It should be remembered that Belgian law, which has its roots in the same sources as our own, is first and foremost a written law. Contracts play a key role in the entire legal architecture, and the specificity of tax law is unmistakable.
Admittedly, in French law, the tax judge has long given the administration the right to requalify the intentions of the parties in order to restore the fair value of a transaction 4. But requalification is not the same as eviction, and the administration must therefore take into account the existence of the legal act. To disregard the contract, as suggested by its Belgian counterpart, the administration would still have to implement the special procedure for abuse of rights. This boundary between requalification and eviction seems to be blithely crossed in the draft circular, to give precedence to substance over form, or to the actual behavior of the parties over contractual appearance. Will the French position, steeped in written law, hold for much longer?
The treatment of subsidies and tax credits in the calculation of transfer prices
The other element that caught our attention is discreetly tucked away in a "Miscellaneous" section at the very end of the chapter. This deals with the treatment of "subsidies received" and their inclusion in the calculation of remuneration, the arm's length nature of which must be tested. These "subsidies" include both grants and tax credits or reductions. The subject is sure to arouse the curiosity of French practitioners, so reminiscent is it of the stir caused by the "Philipps" decision handed down by the Versailles Court of Appeal 5 and upheld by the Conseil d'Etat 6.
In this case, Philips France carried on a research activity for which it received subsidies from the French government's Fonds de Compétitivité des Entreprises, as well as research tax credits. Philips France entered into a "General Services Agreement" with its parent company, under which it undertook to transfer to the latter ownership of the unpatentable intangible rights resulting from its aforementioned research activity, at a price equal to the cost price of the corresponding operations, plus 10%. During the audit, the French tax authorities noted that, for the purposes of applying the contract, Philips France had deducted the amount of subsidies received from the French government and the research tax credit from which it benefited, in order to determine the cost price of the intangible assets transferred to its parent company, before applying the 10% mark-up to arrive at the price invoiced to the latter. Taking the view that this deduction led to an indirect transfer of profits abroad within the meaning of Article 57 of the French General Tax Code, the tax authorities increased the company's income by the amount of these subsidies and tax credits added back to the cost price used to determine the sale price.
The question was therefore whether the company could validly subtract from the base on which the margin was calculated the sums received under the preferential schemes applicable in France. In so doing, the method led to a significant reduction in the cost base, and thus to the 10% margin being relegated to a tiny fraction. The company's main argument - skilfully defended, it must be admitted - was that the benefits received were a reward for the investment and research effort attributed to any taxpayer falling within the scope, and should therefore not tarnish the subject of transfer pricing, which is in fact disconnected from these regimes. In order to draw the consequences from one subject to the other, the tax authorities would have had to demonstrate that independent third parties were passing on the effects of the preferential treatment in their pricing. The argument was successful, and the tax judge of last resort cut short the administration's attempts.
The subject would have deserved a better fate, had the administration's defense been different and divorced from the eternal burden of proof. In any case, the Belgian administration has shown fewer qualms, by directly ruling on the subject: "Subsidies are deducted from the cost/sales base, if there is a direct link between the subsidy and the production/sales of the product or the provision of the service. If there is no direct link between the subsidy and the production/sales of the product or the provision of the service, the subsidies are not deducted from the cost base. Tax deductions are not deducted from the cost base. An investment allowance, for example, is not deducted from the cost base" 7.
In so doing, the Belgian tax authorities have adopted a clear and unequivocal stance, at odds with the French tax judge. The service can always console itself with the thought that the decision would have been in its favor had the case been decided in the land of Jacques Brel.
France and Belgium share the same view of transfer pricing methods
The second chapter of the draft circular is devoted to transfer pricing methods. The document provides a clear explanation of each of them. These developments may be supplemented by the guide prepared by the DVNI in its day 8, whose illustrations still enlighten the layman today, even more than 10 years after its publication.
In describing the cost-plus pricing method, the Belgian authorities take a position on the cost base subject to remuneration, and more specifically on variations in this base. Indeed, it is common in the life of a company to experience sometimes significant fluctuations between its initial projections and the actual landing of its performance. In this respect, the draft circular offers a double perspective.
Firstly, the document points out that, in practice, transfer prices are generally predetermined on the basis of budgeted costs. The Belgian administration therefore indicates that it will monitor the impact on the transaction of the use of budgeted costs, as well as the difference between these costs and actual costs, and the sustainability over time of the type of costs (budgeted or actual) chosen. It specifies: "adjustments will be possible if the actual expenses associated with the transactions are systematically higher than the budgeted costs, or if the use of budgeted costs does not enable the cost base to be brought into line with an arm's length cost base " 9.
Semantics are important, as they call for a certain degree of flexibility on the part of inspectors during control operations. Calling for the systematic calibration of transfer prices on the basis of actual data supplied by accounting firms not only requires an additional effort on the part of companies, but also raises the question of the nature and timing of the adjustment to be made. Every year, at the end of the financial year, a microcosm of accountants, auditors and financiers is stirred up to decide how to deal with this adjustment, whether to correct the accounts for the financial year, or enter a credit note or additional invoice for the following year (without raising the VAT and inventory value issues associated with these gesticulations!) The position currently adopted by the Belgian tax authorities means that this issue can be ignored, as long as it remains marginal and not systematic.
Secondly and more importantly, the circular offers an interesting perspective on the slippage observed in cost bases. It explains that "It is generally accepted that higher costs due to inefficiencies are borne by the company supplying the goods or services, since it is this company that is in a position to influence the magnitude of the costs in question. In this situation, an independent buyer will not accept any price adjustments".
By making each player in the value chain accountable, the document deals a serious blow to the tendency observed in the administration (particularly in France) to consider that as soon as the tested entity takes on a routine profile, it should be protected against the risk of losses and systematically generate a margin on these costs. As if, in the real economic world, an independent player would accept that his supplier let his costs slip, only to charge him back shamelessly.
This common-sense stance is reminiscent of an unfortunately little-known decision of principle handed down by the Versailles Court of Appeal, which we have already commented on in another life. The decision, handed down following the adjustments claimed from the Unilever 10 group, led to the same consequences, by enshrining the fact that the costs of inefficiency should not be borne by the party who did not cause them.
In this case, Astra Calvé, which later became Unilever France, produced margarines in its four factories throughout France. Sales were made to a Belgian group company (an astonishing coincidence) on the basis of the price-plus-10% method. This margin was applied to the capital invested in production, as calculated on the basis of a normal production rate observed in other similar plants. However, the margarine plant was experiencing serious efficiency problems, due in particular to the obsolescence of certain equipment, but also to other intrinsic factors. In practice, the implementation of these factors was causing production costs to soar well above projected expenses that the apparently comfortable 10% margin was not enough to offset operating losses. The tax authorities, anxious to ensure that the negotiated margin prevailed, considered it appropriate to include all costs incurred, including those linked to the plant's inefficiency, in the cost base. The tax judge took a different stance, rejecting the department's claims and returning the plant to its responsibilities of good management.
Along the way, it is interesting to note that this intellectual approach, although rational, had to wait for the recent work of the OECD to remind us that every economic player must bear the consequences of the risks he incurs, provided he concentrates the material, human and financial resources to control these risks. In this respect, the draft Belgian circular simply embodies this concept, while laying down a common-sense principle that the administration will henceforth have to adhere to. Let's hope that it transpires beyond the border!
Comparability analysis in focus
The third chapter of the draft circular focuses on comparability analysis, which is placed at the heart of any demonstration of the arm's length nature of transfer pricing. The document cuts short any hesitation right from the introduction: "Comparability analysis is important for all transfer pricing methods used to assess whether related transactions comply with the arm's length principle and therefore with market conditions. It forms the basis for any justification of the transfer prices used " 11.
Let's take a moment to consider this trend, which aims to position the search for comparables as the keystone of the dialectic of proof in transfer pricing matters. On the French side, this approach seems to be the result of a cross-fertilization between the French tax authorities and a section of the legal profession, which, by dint of its focus on the economic sphere, has forgotten that the demonstration of an abnormality is first and foremost a matter of legalism, which the specificity of tax law cannot erase. Let's put it bluntly: transfer pricing is a legal issue. Admittedly, it is based on concepts borrowed from economics and finance. However, it cannot avoid the most basic concepts of our law, particularly civil law.
With this premise in mind, the fourth and final paragraph of Article 57 of the French General Tax Code stipulates that only "in the absence of precise information for making the required adjustments" is taxable income determined by comparison with that of similar businesses operating normally. In other words, comparability analysis is merely an alternative method for the tax authorities, who are initially invited to make adjustments by any means. Take, for example, a contract between two companies belonging to the same group, providing for remuneration that exceeds the levels produced by comparable companies. Unless it can be shown that there is a significant imbalance between the parties (another legal concept borrowed from contract law, by the way!), the contract remains enforceable against the tax authorities, who should therefore be able to demand performance of the contract independently of any economic analysis. It should be noted that this subsidiarity of comparable research was still reflected, albeit in a very prudish way, in the former article L13AA of the Livre des procédures fiscales, which required the production of such studies "only if the method so requires" (in other words, the method used to determine transfer prices).
Today, the quasi-automaticity of the search for comparables is reiterated by the tax judge, who quotes over and over again his now widely-repeated recital: "When it finds that the prices charged by a company established in France to a related foreign company are lower than those charged by similar companies operating normally, i.e. at arm's length, the tax authorities must be considered to have established the existence of an advantage which they are entitled to reintegrate into the results of the French company, unless the latter can justify that this advantage had at least equivalent counterparts for it. In the absence of such a comparison, however, the tax authority is not entitled to invoke the presumption of profit transfer thus established, but must, in order to demonstrate that a company has granted a liberality by invoicing services at an insufficient price, establish the existence of an unjustified difference between the agreed price and the market value of the asset transferred or the service rendered " 12.
That's all it took for the French administration to follow suit. In its recent instruction on the new vintage of the documentary obligation, the French tax authorities unashamedly affirm the imperative need to produce a comparability analysis. It states unequivocally that "A detailed comparability analysis and functional analysis of the company and its associates should be prepared for each category of transaction, reflecting any changes from previous years. For each category of transactions, the comparability analysis describes the company's remuneration conditions, justifying any differences with those of independent companies " 13.
So, of course, the very principle of arm's length requires us to behave like third-party companies. It's a shortcut to consider that comparability analysis is at the heart of any transfer pricing study. But the desire to compare everything erases the specificities or factors, exogenous or endogenous, which act as vectors of value. As surely as Narcissus met his doom, the search for one's own reflection is not always virtuous.
A Vademecum of comparability analysis
These considerations aside, the draft circular outlines what could be a reliable search for comparables from the Belgian point of view. The approach is commendable, if we consider that it pursues the objective of legal certainty by limiting errant interpretations and disputes between technicians. However, it fails to address the specificities of each situation, which may call for alternative approaches. Above all, by asserting a Belgian model, it leads to the de facto exclusion of research carried out abroad (particularly if the tested party is located outside Belgian territory) and which could be based on other criteria.
For example, the draft circular adds to the OECD's recommendations on the use of multi-year data. The guidelines state: "The examination of multi-year data is often useful in a comparability analysis, but it is not a systematic requirement. Multi-year data should be used where it improves the transfer pricing analysis. There is no need to set standards for the number of years to be covered by multi-year analyses " 14. The Belgian administration, however, goes further and considers that "at least 3 years should be considered in the analysis of comparables " 15. In practice, this also corresponds to the approach commonly applied by the French administration, which tends to favor the use of a three-year period preceding the year of the audited financial year.
But as is often the case, the devil is in the detail, and the key point is hidden in the second part of the sentence. The project continues: "On the other hand, in the part to be tested, the data is logically limited to that relating to the transaction examined for a given year".
This assertion seems surprising to us, even if we have to admit that it echoes the usual practices of the French administration too. Indeed, how can we reconcile the principle of the independence of fiscal years and the right of recovery (of three years in both countries, barring exceptions) with this necessarily arbitrary approach? The approach would not capture the exogenous economic effects potentially observable in a particular year, the effects of which would be smoothed out (and therefore attenuated) if a range of years had to be systematically employed.
If the French authorities were to take such a step, we believe that a 2009 decision by the Versailles Administrative Court could prevent them from doing so. In the "Man camions et bus" case, the tax judge rejected the Minister's claims, ruling that by neglecting the preponderance of the presence of an economic player like Renault on the French market, and by focusing on pan-European references, the administration had failed in the dialectic of proof incumbent on it16. By extension, it can be considered that by not sufficiently capturing the economic factors affecting the market in a particular year, the comparability analysis would necessarily be flawed.
Another point of attention is the origin of the references selected as a result of the search for comparables. Here, the Belgian administration is more courageous than its French counterpart, taking a clear stand. Thus, when looking for so-called "external" comparables, the draft circular states: "The Administration accepts pan-European studies, preferably based on the 15 EU countries that were part of it before its enlargement in 2014" 17.
Clearly, in accordance with Community law, the tax authorities cannot display a national preference by demanding the use of comparables that are necessarily domestic. It would have been an outrage to see the Belgian administration sanctioned by the Brussels judges! But the use of comparables from the Europe of 27 does not seem relevant to us either, given the significant market differences between member countries. For our part, we subscribe to this position, subject to taking into account any specificities of each sector, as highlighted by the "Man" ruling mentioned above.
Haro on losses!
In the same chapter, the tax authorities state bluntly: "With regard to comparables in the intervening period, the tax authorities will not accept any company that has had two or more loss-making years " 18. While this postulate is clearly in line with a trend widely shared by auditors in both countries, it seems to us no less contrary to the most elementary economic and fiscal concepts.
It's worth remembering that the vagaries of economic cycles can unfortunately lead to chronic losses. Our administrations know this better than anyone else: what can we say about countries such as Belgium and France, which have been in deficit for so long that they almost seem to wallow in it? In this environment, it seems absurd to ask economic players to adopt a financial stance that governments have long since renounced.
What's more, certain economic sectors or market strategies require major investments to be made in the hope of producing more virtuous returns in the future. This is the very essence of entrepreneurship. Suggesting the rejection of loss-making companies therefore runs the risk of capturing only players situated on a potentially different level of the economic and commercial cycle than the audited taxpayer. In so doing, the approach ignores a fundamental element of comparability analysis, namely corporate strategy, which is one of the OECD's five basic criteria19. What's more, this approach would send back to the taxpayer a smooth, inaccurate and therefore distorted reflection of a company that is necessarily virtuous and shielded from the vagaries of the market. Needless to say, this state of grace has no basis in reality.
From a legal point of view, it should be remembered that nowhere in positive law is there any reference to a prohibition on making losses for two years or more. On the contrary, our carry-forward mechanisms implicitly but necessarily validate the possibility of a taxpayer making a loss. So it's a delicate matter to deny under transfer pricing what tax law validates in other circumstances.
Above all, the exclusion of loss-making companies from comparable panels would in practice mean depriving the audited taxpayer of a freedom offered to independent third-party companies. In so doing, the very principle of arm's length competition would be shattered, given the obvious difference in treatment that the situation would entail.
This scent of iniquity is reminiscent of previous, well-founded discussions which, on the subject of so-called "implicit" guarantees arising from membership of a group, led to the remuneration of a transaction that was simply induced and that third-party companies could not possibly know about. The French judge cut short the discussion in a landmark ruling by the Bordeaux Administrative Court 20, the recital of which was taken up shortly afterwards by the Conseil d'Etat 21. In these rulings dealing with the "halo effect", the solution was unequivocal: one cannot set up against a related taxpayer what one disregards in an arm's length environment.
Last but not least, there is now a well-established body of French case law to the effect that operating losses are not in themselves sufficient to demonstrate an indirect transfer of profits for transfer pricing purposes 22.
For once, then, law and economics are on the same wavelength, arguing in favor of a strong mitigation, if not outright rejection, of the Belgian administration's approach of rejecting comparable companies with two years of losses. On the French side, we believe that the taxpayer has the necessary weapons at his disposal to counter an identical stance, even if adopted by the auditors.
Welcome explanations of today's controversial issues
A welcome initiative: the draft circular sets out to explain recent OECD developments in the most contemporary and scrutinized areas, namely intra-group services, intangible assets and financial relationships between affiliated companies. These three topics alone take up almost half the report.
Unofficial statistics derived from observation in the field tend to show that these types of transactions regularly occupy the attention of the auditing departments, and therefore naturally account for the bulk of the adjustments notified. As we are wont to say, although the subject matter is not particularly complex, transfer pricing services are often the cream of the crop when it comes to transfer pricing. How many groups fail to deduct shareholder costs from the base to be re-invoiced to subsidiaries, or to describe in detail the benefits provided to debtor subsidiaries?
On this point, the draft circular specifies how to handle so-called "low value-added" services, which should significantly lighten the documentary and supporting burden on taxpayers concerned by these flows.
Government attention to intangible asset and cash flow transactions is more a reflection of the times. The former echoes the nature of our modern economy, and the fact that key success factors are now almost exclusively embodied in intangible elements. The latter are the continuation of a political movement that began in the wake of the 2008 crisis, which pointed an accusing finger at finance and the abuses that can result from intelligently thought-out capital movements within integrated institutions such as business groups.
As far as intangible assets are concerned, once again Belgium is distancing itself from its culture of written law, in order to adopt the Anglo-Saxon theses that permeate the new OECD principles. Notably, with regard to the qualification of intangible assets, the administration announces: "It has been explicitly chosen not to adhere to the accounting standards that apply to intangible assets. An intangible asset therefore does not always have to be expressed in the entity's annual financial statements in order to be entitled to remuneration for its use or disposal. Nor does an intangible asset have to be protected by law in order to claim remuneration for its use or disposal" 23. In so doing, the Belgian administration is moving more towards the notion of intangible "element", which voluntarily no longer echoes a right, the contours of which are more certain and the qualification assumed. This thesis refers back to the famous "something of value" developed by American theorists, which escapes any accounting or legal definition, and which reflects the idea of an intangible element that carries value. In a way, this notion refers to the "goodwill" already familiar to appraisers.
For the time being, this step has not been taken by our French tax authorities, and even less so by tax judges. Our attachment to written law, from which stems the need to be able to qualify the flow in order to apply the appropriate tax treatment, permeates our legal system, even if it means, unfortunately, that our tax authorities have a difficult task to perform, to say the least.
In the same way, the Belgian administration is moving away from legal ownership of intangible assets to emphasize economic ownership, reflected in the performance and control of essential functions known as "DEMPE " 24. In this way, the contract, like the title deed, would be no more than a bundle of clues, which operational reality could lead us to disregard. Here again, if the same system were to be replicated in France, we might well wonder how it would fit in with our own article L64 of the Livre des procédures fiscales, which only allows a contract to be completely disregarded in the very limited context of an abuse of rights.
Another significant development is that the Belgian tax authorities have adopted the OECD's "ex-post" approach. This approach allows tax authorities to reassess the value of an intangible item when it has been the subject of a transaction, and the circumstances that led to the valuation have changed significantly in subsequent years. Here again, this approach runs counter to eminent concepts of written law, namely the intangibility and force of contract. In our civil law, the contract is formed and becomes immutable when there is agreement on the thing and on the price, unless of course it can be demonstrated after the fact that there has been a manifest error as to the cause or object, or an imbalance in the obligations of the parties.
Above all, this ex-post approach enables the tax authorities to extend the recovery period. The draft circular discreetly states: "In the context of the good faith execution of the relevant double taxation agreements, the tax administration will proceed, at the latest 7 years after the closing of the accounting period of the disposal of an intangible asset, to an adjustment of the transfer price in accordance with the principles described above " 25.
In France, the tax judge has ruled that the price cannot be re-evaluated by the administration years after the transaction, on the grounds that the value of the brand has changed due to market developments. 26 The very essence of the economy is random, and the taxpayer would be hard-pressed to anticipate its volatility! In the absence of a crystal ball, taxpayers cannot always predict the success or failure of their operations. The emergence of a new patent or a competing brand can undermine the value of these elements, and therefore the intrinsic value reflected in the sale or concession price. This approach, which may seem outdated in the light of OECD developments, in our view ensures a higher degree of legal certainty, by limiting the trade-in period to its ordinary law period, and by protecting the binding force of contracts.
Finally, with regard to financial transactions, the draft circular provides a number of elements that the tax authorities will take into account when processing files involving such transactions. Chapter X deals successively with intra-group loans, the provision of intra-group guarantees for loans, and treasury agreements. The central notion of "rating" (or credit rating), which is involved in the analysis of all these flows, is also clarified in the very first sections. On this point, the Belgian tax authorities endorse the OECD's position, de facto rejecting the notion of "halo", which the French tax judge has already had occasion to deal with. The draft document confirms: "It should also be borne in mind that no compensation should be paid for the implicit guarantee " 27.
This chapter shows that the Belgian administration has adopted the OECD principles and encourages a sequential approach when determining the arm's length nature of a financial flow. In this context, access to specialized databases will be more than necessary. However, as the subjective element in these transactions is limited, the analysis should be facilitated by the same effect.
In conclusion
The Belgian tax authorities' approach is to be commended. Not only does it clarify a number of essential concepts, it also has the merit of clearly stating the tax authorities' intentions on a number of issues. Our French administration would be well advised to replicate the formula, rather than sowing administrative instructions over the years, the content of which is not always crystal-clear. The fact remains, however, that the circular's clear and marked attachment to the work of the OECD presents a new aspect of Belgian tax law, leaning more towards economic rationality than written law. The document is still at the draft stage. While adjustments will undoubtedly still be made, this ambitious first draft already highlights the future areas of control for both the Belgian and French administrations. Brel sang "quand on a que l'Amour". That was long before BEPS. From now on, transfer pricing taxation is clearly part of the dance!
1/ Article published in Nouvelles Fiscales (Lamy) n°1239, March 1, 2019, courtesy of Wolters Kluwer Group.
2/ Circulaire relative aux prix de transfert, p. 5.
3/ Idem.
4/ See for example CE, July 20, 2007, n°232004.
5/ CAA Versailles, Oct. 11, 2016, n° 14VE02651.
6/ CE, 8ème et 3ème ch. réunies, Sept. 19, 2018, n° 405779.
7/ Circulaire relative aux prix de transfert, p. 7 and 8.
8/ Prix de Transfert, Guide à l'usage des PME, Nov. 2006.
9/ Circulaire relative aux prix de transfert, p. 12.
10/ CAA Versailles, 6ème ch. 5 déc. 2011, n° 10VE02491.
11/ Circulaire relative aux prix de transfert, p. 12.
12/ See eg. CE, 8ème et 3ème ch. réunies, 19 sept. 2018, n° 405779.
13/ BOI-BIC-BASE-80-10-40-20180718, published 18 Jul. 2018, BIC - Base d'imposition - Transfert indirect de bénéfices entre entreprises dépendantes - Obligation documentaire permettant le contrôle des prix de transfert, §430.
14/ OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations, July 2017, §3.75.
15/ Circulaire relative aux prix de transfert, p. 18, §83.
16/ CAA Versailles, 3ème ch. 5 mai 2009, n°08VE02411.
17/ Circulaire relative aux prix de transfert, p. 21, §108.
18/ Circulaire relative aux prix de transfert, p. 21, §111.
19/ OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations, July 2017, §1.36.
20/ CAA de Bordeaux, 3ème ch. 2 sept. 2014, n° 12BX01182.
21/ CE 9eme et 10ème ch. réunies, 19 juin 2017, n° 392543.
22/ For a recent example, see TA Melun, 3ème ch. 14 juin 2018, n° 1502063.
23/ Circulaire relative aux prix de transfert, p. 21, §137.
24/ OECD acronym for Development, Enhancement, Maintenance, Protection and Operation.
25/ Circulaire relative aux prix de transfert, p. 21, §168.
26/ CE 9ème et 8ème sous-sections réunies, March 16, 1990, n°41059.
27/ Circulaire relative aux prix de transfert, p. 21, §264.