156 ITPJ JULY/AUGUST 2000 Case Notes GERMANY Tax Court Ruling on Transfer Prices and External Gross Margin Analyses! Dr Heinz-Klaus Kroppen2 and Axel Eigelshoven3 I. INTRODUCTION lated ing in the same industry. T The determination of arm's length transfer prices and the comparables were, however, not disclosed to the tax- OF SECRET COMPARABLES before the ne amount of the construc- le price method, with DSIRE)20 der vogele, have not been literature. T ther analysed tax o between th tributor on the
JULYIAUGUST 2000 ITPJ 157 the appropriate gross margin being determined based on the taxpayer was deprived of the possibility to put forward an external comparison of four distributors operating in its own arguments in support of its position. the same market as the taxpayer. Initially, comparable Ger database i as wel as industry codes and company profiles. were deleted. The Court also rejected this proposal available on a limited basis in the database. After identify f comparability and the tax office could influence the ing the comparable companies, the tax auditors obtained selection of comparable companies to the disadvantage of more detailed inform ffices which was then further analysed the taxpayer The efforts of the tax office to detemine appropriate trans- C. Position of the tax authorities and the OECD ate comparables are, in principle, to be welcomed In prac- In the opinion of the tax authorities, comparable com- n operating margins, even if the taxpayer has fully dis- anies can be used in court proceedings without disclosing the identity of the companies concerned. The tax author- closed all required information to the tax office, so that an ities. hence, follow the established practice of the Federal estimate is not admissible. The operating margins are not Tax court. l4 However, the Ministry of Finance deems the are often based on(inadmissible) rules of thumb. In the right to be heard before the courts to be violated if the files maintained regarding comparable companies are only dis- authors expenence, arm's length prices have so far hardly closed to the lax Court rather than to all participants in the ever been determined by tax auditors using a theoretically Court proceeding. Therefore, the ruling in the Dusseldorf databases). The main reason for this may be the lack of the Federal ministry of Finance. Yet, that ruling goes one (financial statements and balance sheets are currently only when used to determine transfer prices in Tax Court pro- available for 10,000 to 15,000 companies). In addition, both a lack of resources to the tax authorities and insuffi cient experience on the part of tax auditors in performing The oECd does not analyse in detail the issue of secret sis. However, in the case at hand the main concen is that Pricing Guidelines(hereinafter: OECD Guidelines ) the the tax auditors'determination is based on data that could oeCd discusses the net margin method, clearly rejectin not be disclosed to the taxpayer because of tax secrecy le he use of data that are not available to the taxpayer. It is gislation pressly made clear that these rules should be applicable to all transfer pricing methods, which would include the B. Holding of the Lower Tax Court of Dusseldorf resale price method. The OECD justifies its objection by The Dusseldorf Tax Court first stated that a disclosure of defang that the taxpayer must have the possibility to courts would otherwise not be ensured. The oeCd. like data of comparable companies to the taxpayer in the court the Lower Tax Court of Dusseldorf, therefore also goes proceedings would lead to a violation of the tax secrecy of the comparable companies. The Court noted that the tax- payer has the right to be heard before the Court, 0 and this ncludes the right to access to all evidence and facts rel- 6.. See Administrative Principles on Transfer Pricing of 23 February 1983 evant to the case. In the opinion of the Court, however, of the text, see Heinz-Klaus Kroppen and Axel Eigelshoven, cl there is no conflict between the taxpayer's right to access in The Tax Treatment of Transfer Pricing (Amsterdam: IBFD, 1987),1 all information and the strict tax secrecy legislation. Tipke and Heinrich Kruse,(Koln: Dr. Otto Schmidt Verlag, 1996), Sec. 162 AC crecy or the disclosure of data of the comparables, mey r d uers eurn hof keun: tr tko, schn ida verlig. 9a ess respectively, cannot be justified by the obligation of the AStG, note861 ourt to grant the taxpayer the right to access the Court' s 7. See also wolf-Dieter Hoffmann and edgar Schnitzer, "Globale file Kroppen, in Jahrbuch der fachanwdlte fir Steuerrecht 1999(Herne/Berlin The Court also rejected an exclusive disclosure to the Tax Neue Wirtschafts-Briefe, 2000), at 51 s y gl at 191 for an example. Court, which was suggested by the tax office. 12 The Tax 3. However, see Harald kuckhoff and rof schreiber, errechnungspreise in Court reasoned that the tax secrecy was, on the one hand also binding on the Court and, on the other, the exclusive 10 'Tax Court Code, Sec.96, Para.2. disclosure to the Court would again violate the taxpayer's 11. Tax Court Code, Sec. 78 right to access the files. moreover, the duty to grant the 12. This view differs from the Federal lesteuerblatt(hereinafter:BSiBL)I taxpayer the right to be heard before the Court was Court dated 18 December 1984, Bund (1986),at229 designed to put the parties in an equitable position in the 13. Letter of the Federal Ministry of Finance dated 7 April 1986, BSIBL I proceedings. This right would be infringed if the Tax (1986), at 128 Court had to rule on the comparability of companies and 14. See Federal Tax Court Opinion dated 24 September 1976, BSiBL n(1977), at 196; Federal Tax Court Opinion dated 21 May 1982, BStBL I (1982), at 604 eral Tax Court Opinion dated 18 December 1984, BStBL. I (1986), at 226 C 2000 IBFD Publications BV
ITPJ JULY/AUGUST 2000 further than the German tax authorities by generally which information deficits are inherent. The determination rejecting secret comparables made by the tax auditors must, therefore, be based exclu sively on the same information that is available to the tax D. Position of the german courts and tax literature ayer at the time the contract is concluded. This is also acknowledged by the OECD, because a range of prices rather than one price are deemed to be appropriate. 9 ceedings, the prevailing opinion in German tax literature is Assuming that prio re formed in perfect markets(where assured that the data do not allow any reference to the tax- would be no evident justification for a range of prices, e that only anonymous data, if any, may be used. '5 It must be there is by definition only one equilibrium price),ther payer who is protected through tax secrecy legislation. 6 For these reasons, with respect to the dusseldorf Tax The right to be heard before the courts and the taxpayer's Court case neither the tax auditors nor the tax office could right to access facts and evidence of the court can by no use information that was not available to the taxpayer at means justify a violation of the tax secrecy. 7 An exclusive the time the contract was concluded. In addition. the disclosure to the court would lead to a violation of the tax- OECD and the Dusseldorf Tax Court are correct in stating payer's right to be heard before the court and would be a that if secret comparables are used, there is a risk that the reason for an I under Section 119 No. 3 of the Tax tax office will only use information that is favourable for Court Code. 8 Therefore, the only controversial issue is the tax authorities. 20 Some German tax auditors. however whether anonymous data can be used as a basis for an hold in the tax literature that in the Dusseldorf Tax Court case an arbitrary selection had been excluded because the es comparable companies were objectively identified. The in principle not admissible. However, this does not apply selection modes could have been verified by the tax to anonymous comparative data in the public domain (e.g. payer:. The selection process is, however, no means the official list of average rent prices( MietspiegeD, offi- objective. This is supported by experiences in the United cial value surveys for land(Bodenrichtwertsatze), surveys often focus on differences in opinion regarding which on industry averages, and lists of managing directors salaries). These data are publicly available in Germany companies should properly be included as comparables.In and are sometimes even generated by state institutions. addition it is to be remembered that if an identification of such data are not protected by tax secrecy legislation. the"secret comparables"could be made on the basis of publicly available databases, this would mean a loss of the The difference between the data sources can be illustrated tax secrecy of the comparable companies by the following example: A limited liability company sells a company car to its parent company abroad. The The use of anonymous data would therefore require that DEM 15,000 on the basis of the so-called <ar to be the data be altered or that specific information not be dis- managing director determines the value of the closed. If the courts would allow such a procedure, this List",which is often used in Germany by car dealers for would leave t determining prices. Other information is not available to possible for both the Tax court and the taxpayer to verify the managing director. The tax auditors are aware of other the comparability standards and the need for adjustments car sales by another company under audit, where a com- based on complete and unaltered data. parable car(same model, age, mileage, etc. was sold for a Of course, the tax office is free to determine the audit ice of DEM 18, 000 to an unrelated party. Therefore, the requirement through comparison with transactions that are tax auditors intend to assume a hidden profit distribution not disclosed to the taxpayer. For the determination of in the amount of dEm 3, 000 arm's length transfer prices, this means, for example, that The admissibility of determinations of an arm's length the licence database used by tax auditors to collect third price on the basis of anonymous comparative data (in the party royalty rates(which is not publicly accessible)can troversialin ple, the Schwacke List)is not likely to be con- only be used for sampling purposes. 2 It must not, how- above the case described above, since this is a com- ever, be used independently to determine or estimate an mon means of valuing cars in Germany. In addition. arm,'s length royalty nobody would suggest to analyse the comparable transac- tions on which the Schwacke List is based On the other hand, determinations made using comparable transactions that are known only to the tax auditors are inadmissible. Tipke and Heinrich Kruse (koln: Dr. otto Schmidt verlag, 1996) Sec. 30, note price differences. The reason for the price differences is 16. See Federal Tax Court dated 27 October 1993, BSIBL I(1994), at 210 the imperfectly competitive environment. due to the lack 17. Heinrich Kruse, in Abgabenordnung/Finanzgerichtsordnung, eds. Klaus of information, market participants cannot determine the "true market price". If it would be admissible for the tax 18. Federal Tax Court dated 18 December 1984, BStBL Il(1986)at 229 did, auditors to use more information than the normal market however, not show this opinion participant can use for its analysis, the tax auditors would 19. See OECD Guidelines, note 1.45. and Administrative Princip be in a better position than a market participant. This 2.1.9 US Treas. Reg. Sec. 1. 482-)(e) would mean that the price determined by the tax office 20. See OECD Guidelines, note 3 would have nothing to do with the" true market price, in /StR 16(1999), at 521 21. See Harald Kuckhoff and Rolf Schreiber, Quo vadis Fremdvergleich", 8 22. See Kroppen and Eigelshoven, supra note 6, at 50. C 2000 IBFD Publications B\
JULY/AUGUST 2000 ITPJ E. Determining appropriate gross margins based on It must then be clarified how comparability can be guaran blicly available industry averages teed without conflicting with the taxpayer's right to inspect the files or the taxpayer's right to be heard before Tax auditors have often used industry averages to estimate the courts. If the taxpayer is able to prove satisfactorily third party gross margins. In our opinion, industry average that the structure of its enterprise differs from that of the compar ables, a gross margin comparison will presumably admissible in determining arm s length gross margins, be completely unreliable(if it could even be determined because the data(e.g. the Schwacke List, the official list of Alternatively, a net margin comparison could be made average rent prices, and the official value surveys for land) because net margins are usually less dependent on differ are also accessible to the taxpayer Apart from admi ences in terms of functions and risks 8 Of course the tax bility, there would seem to be a question as to whether the payer would be free to furnish evidence of company spe- data are appropriate to be used in determining a specific cific reasons for deviating from industry average transfer price. In this context, a distinction must be made In the area of transfer pricing for multinational enterprises, between a determination of appropriate transfer prices and the estimation under Section 162 of the General Tax code it is, however, normally impracticable and inadmissible to or Section 1 Paragraph 3 of the Foreign Tax Code use average gross margins of an industry. Compared with smaller enterprises included in the collected comparative Market prices are to be determined according to the arm's data of the Federal Ministry of Finance ( e.g. for hair- length principle. Under this rule, statistical empirical data dressers and butcher shops), these enterprises regularly or average values in the public domain must only be used show a variety of special features. Particularly, functions if third parties also determine their prices based on such risks and capitalization differ widely. However, these fac- data. This applies in Germany, for example to the use of tors have an important influence on the determination of value surveys for land, the Schwacke List, or the official an appropriate gross margin. According to the established list of average rent prices. On the other hand, it is not in practice of the Federal Tax Court, the objective of an est- line with the arms length rule to determine appropriate however, be to determine the tax base that is gross margins based on industry averages. These average most probably correct. It seems to be highly doubtful that data normally cover too broad a range of gross margins. an average value of a whole industry could comply with are based on a very diversified group of enterprises with use of industry statistics efore has rightfully rejected the This is due to the fact that industry-related average values this rule. The OECd the quite difierent functions and risks, Independent parties Finally, it should be noted that the use of industry averages an appropniate gross margin on the basis of functions and for determining and estimating appropriate gross margins instead of industry averages 24 lated average values can at best be used for sampling purposes Another issue relevant to the use of industry averages is hether such averages can be used for an estimate under Section 162 of the General Tax Code and Section 1 Para. 3 of the Foreign Tax Code. First, it is to be noted that in Ger- man tax law an estimate of the tax base is only admissible if the taxpayer failed to comply with its duty to cooper ate." This limitation on use is mandatory, on account of the prohibition to tax based on suspicion(Verbot der ver- e: Richtsatz -Sammlung 1998(Herne/ office must first apply the standard methods, whereas an 24. See oEcD Guide lines, note. 2. 21 estimate on the basis of operating margins or return on 25. Administrative Principles, Sec. 93.1 assets is only admissible in exceptional cases. 26 nung/Finanzgerichtsordnung, eds. Klaus Tipke and Heinrich Kruse [Koln: Dr Oto Schmidt Verlag, 1996), Sec. 162 AO, note 6: Franz Wa Therefore, it is to be considered whether the use of indus- in Ai echt, eds. Hans Flick, Franz Wassmeyer and Hubertus try averages for the application of the resale price method Baumhoff (Koln: Dr. Otto Schmidt Verlag, 1996), Sec. 1 AStG, note861;Klaus is admissible. This could only be justifiable if the taxpayer wrirtscha is- Brief., ser set. AS (G, nte 254.g et a- (HermelBer and the comparable companies form a relatively homo-. 26. Franz wa in AuBensteuerrechf, eds. Hans Flick, Franz wass geneous group (i.e. the enterprises concerned have similar and Hubertus Baumhoff (Koln: Dr. Otto Schmidt Verlag, 1996), Sec. 1 ASIG, functions, risks, and capital structures). This type of est AuBensteuergesete, eds. Klaus Brezing et al. imate, however, already turms out to be difficult if the tax- (Heme/Berlin: Neue Wirtschafts-Briefc), 1991, Sec. 1 AStG, note 258 payer or the tax office claims that the cost, capital, or risk 28. See OECD Guidelines, note 3.2 structure differs from that of the comparables and that 29. Federal Tax Court opinion dated 18 June 1960, BSIBL. m(19 therefore an estimate based on these data would not lead to sB(1986),at2 an appropriate solution. For example, in a ruling of the ederal Tax Court dated 18 December 1984,the tax and 28 Heinrich Kruse(Koln: Dr. Otto Schmidt Verlag, 1996), Sec. 162 AO, notes 2 office argued that the restaurant that had been subject to an 30. OECD Guidelines, no also Harald Kuckhoff and rolf estimate realized lower food sales than the comparable Schreiber,"Quo 8 IStR 16(1999), at 519: Helmut company taken into account in the industry avera Becker. "Commen Handbuch internationale verrech nungspreise, eds Becker and Heinz-Klaus Kroppen (Koln: Dr. Ott 2000 lBFD Publications BV
ITPJ JULY/AUGUST 2000 II. COMPARABILITY ANALYSIS IN APPLYING gross margins to be applied to fashionable articles. The THE RESALE PRICE METHOD appropriate gross margin as determined by the tax auditors was therefore 28 per cent. According to the statement of A. Relationship between resale price method and the facts by the Court, this was justified by the tax auditors comparability analysis on the grounds that Apart from the formal aspects of the case, the Tax Court of the plaintiff sold fashionable wear which was predomi Dusseldorf also dealt with the substance of the arm's nantly manufactured in Italy and, hence, in a country which was used to quite different standard sizes because south length principle. For determining an appropriate transfer European people were normally smaller than central Euro- price, both the tax office and the Tax Court applied differ pean people. Also, it was to be taken into account that the ent approaches; however, these approaches were always manufacturer was located in a country which had the repu based on the resale price method. An analysis appears tation for delivering ordered goods either with defects, with interesting because the Court and the tax auditors have a delay, or not at all. In addition, it was to be taken into been confronted with difficulties similar to those experi account in determining an appropriate margin that the plain enced by the taxpayer in applying the resale price method tiff was engaged in the trading of high-price fashionable The Tax Court of Dusseldorf case is typical, in so far as in wear which was only purchased by very few customers practice there are often only transactions with unrelated This meant that the plaintiff had only a limited growth potential because it could only deliver its goods to selected suppliers that are barely comparable. The problem of retailers.Also, enterprises distributing high-priced luxury barely comparable transactions draws attention to another articles were more likely to be more strongly affected by issue related to the determination of transfer prices political and economic crises than enterprises distributing namely the appropriate application of the comparability goods for everyday demand. 2 analysi Given the above circumstances the tax auditors deter The term comparability analysis was coined by the OECD. mined a five per cent premium on the gross margin of 23 It means that in comparing a transaction between related per cent as provided by the industry average, resulting in a and unrelated companies, all economically relevant differ- gross margin of 28 per cent ences are to be adjusted. 1 A direct comparison of the First, it is remarkable that this computation is not meant to transactions (i.e. without adjustment) is only possible in be an estimate but a determination of an appropriate trans exceptional cases This means that the comparability ana- fer price, considering that the administrative principles lysis is of crucial importance in determining the appropr require a transactio obas ed analysis . Therefore, the shift in two steps. In the first step, all differences between the must first be justified. This shift can be reasonable if the There may be differences in functions performed (e.g. tions are performed over time or differences in specific advertising, service or transport), risks (e.g. inventory, transactions can be ignored for reasons of practicality.34 price,currency,credit or collection risk), contractual terms Although the administrative principles generally do not (e. g. terms of payment or delivery), economic circum- allow an aggregation of transactions, it seems to be fair stances(e.g. property of trademark rights or geographic that an aggregation should be admissible in the Dusseldorf markets), and business strategies. In a second step, exist- Tax Court case. Thus, the question of comparability will ing differences must be quantified and adjusted accord- arise ngly As discussed above all differences must be identified and In the Dusseldorf Tax Court case, there were also consid uantified, although the OECD emphasizes that only those erable differences between the transactions to be com- differences that impact the margin are to be adjusted, 35The ared. For this reason, it is interesting to see how adjust ax auditors'analysis, however, also provides for the ments are to be computed in the opinion of the tax adjustment of differences that are of minor importance within the scope of the resale price method. For example, the tax auditors emphasized that on account of smaller B. Resale price method applied by the tax office standard sizes in Italy, the German distributor was likely to realize a higher gross margin than an average distributor The constructive dividends were determined by the tax However, an advantage of the resale price method(as office using an external gross margin comparison, follow- compared with the comparable uncontrolled pr nce metho ing two approaches. The first approach was based on that product differences play a much less significant role industry averages, while the second(as already described (if any at all). 3 Furthermore, the tax auditors did not take above)was based on gross margins realized by four into account the comparative group(i.e. German distribu wholesalers in the same indust tors)that probably distributes the same standard sizes In the first approach, the appropriate gross margin was determined on the basis of publicly available industry averages collected by the tax authorities. According to the 31. See OECD Guidelines, notes 1.15-1.35: US Treas Reg. Sec. 1.482-1(d) tax auditors' file of wholesalers in the textile industry 2. See sTrE20(1999),at788 appropriate gross margins range from 17 per cent to 23 per 33. See Administrative Principles, Sec. 2. 1.2. cent.The industry data, however, also allowed higher 35. See OECD Guidelines, note 1.15 36. See OECD Guidelines, note 2. 17: US Treas. Reg. Sec. 1. 482-3(c)(3(II(B) 2000 IBFD Publications BV