By Pilar Barriguete & Edland Graci, Duff & Phelps, Spain
On March 6, the Spanish Court of Appeal reached a final ruling on transfer pricing adjustments arising from the Spanish tax authority’s audit of IKEA Distribution Services, a wholesale distribution company that sells products to related party retail companies.
The adjustments related to disagreements that the Spanish tax authority had with the IKEA subsidiary’s implementation of a transactional net margin method (TNMM) analysis for the 2007 fiscal year. The total adjustments at issue were 37 million Euros.
The ruling touched upon two main points, both of which were based on the court’s interpretation of the relevant OECD Transfer Pricing Guidelines. The first finding confirmed the appropriate use of a multiyear analysis for purposes of determining the arm’s length range of profitability in a TNMM analysis while emphasizing the need to make the comparison of the taxpayer’s single-year results relative to the year under review.
Secondly, the Spanish appeals court supported the taxpayer’s position that an adjustment to the median should be made only were comparability defects exist. The court found no comparability defects in IKEA Distribution Services’ TNMM comparable analysis and therefore found that any adjustment should have been made to the lower quartile of the arm’s length range rather than to the median.
The paragraphs below present a more thorough view of the main technical conclusions.
IKEA’s transfer pricing analysis
IKEA Distribution Services, whose activity is the wholesale of furniture and household appliances, conducted a benchmark analysis to identify an arm’s length range of operating margins for companies dealing with similar activities.
In doing so, the taxpayers applied the TNMM as the most appropriate method to analyze its transfer pricing policies and selected the return on sales as the best profit level indicator to compare its profitability.
The benchmark included financial statements of the comparable companies for the years 2003 to 2005. The interquartile range identified was the following:
|
Arm’s length range |
Return on sales 2003–2005 |
|
Lower quartile |
2.1% |
|
Median |
4.1% |
|
Upper quartile |
7.6% |
|
Average |
6.1% |
The taxpayer’s reported return on sales for the years 2006-2008 is presented below:
|
IKEA Distribution Services |
Return on sales |
|
2006 |
3.84% |
|
2007 |
0.42% |
|
2008 |
2.42% |
|
Average |
2.22% |
The identified market interquartile range extended from 2.1% to 7.6%, therefore, if the return on sales of IKEA Distribution Services laid within those margins, it should be considered within the market parameters. However, for the tax authorities, this was not the case regarding FY 2007 since the taxpayer’s return on sales was 0.42%, falling below the lower quartile of 2.1%.
Although the taxpayer’s margin was not within the identified arm’s length range during 2007, the company maintained that the multi-year approach for testing should be used for IKEA Distribution Services’ operations, leading to a comparison of the average return on sales of 2.22% for the 2006–2008 period in total to the identified arm’s length range, the result of which was within the interquartile range of arm’s length margins identified through the benchmark.
Hence, the taxpayer argued that no adjustment was necessary for 2007 (nor for 2006 or 2008).
The Spanish tax authorities argued that the comparison to the interquartile range should be made for 2007 on a standalone basis. On this basis, IKEA Distribution Services fell outside the interquartile range. Hence, the tax authorities applied a transfer pricing adjustment to the median value (4.1%) or equivalent to 37 million euros.
Moreover, they adjusted to the median, also, the taxpayer’s return on sales in 2008 (2.42%), since it was situated below such point, albeit within the arm’s length range.
Multiple year analysis
The Spanish tax authorities referenced the OECD guidelines when asserting the appropriateness of analyzing tax years on a single-year basis.
They stated that while the guidelines do consider it reasonable, in certain cases, to use of multiple years to analyze controlled transactions under the TNMM, the use of multiple years should apply only to the interquartile range construction from the comparability analysis. They took the position that individual years should be compared to the interquartile range for determining the amount of any adjustments amounts applicable.
The Spanish tax authorities also argued against the multi-year range advanced by the taxpayer, insisting that IKEA Distribution Services had not justified the reasoning behind the selection of the number of years (i.e. three years) for the calculation of the average and its comparison with the interquartile range.
IKEA Distribution Services had argued that volatility caused by market risks and reductions in raw material prices over the period warranted the consideration of multiple years. The tax authorities believed this position was not correct because they did not believe it was consistent with the facts presented.
In the first court hearing on the matter, the Central Administrative Economic Court (known in Spanish as “TEAC”) agreed with the tax authorities decision. According to the TEAC: “The multiple-year consideration referred to in the Guidelines is that related to the comparability study, in which the median considered is that which has resulted from considering the three years to which the study refers. However, once this is determined, its comparability should be determined on an individual basis with respect to each of the individual [year’s] results reported”.
The Court of Appeals agreed with the tax authorities and the TEAC’s conclusions regarding the appropriate comparison of taxpayer’s individual year results to the interquartile range, but also agreed that the use of a multi-year range was appropriate based on the OECD guidelines stating that, “examining multiple year data is often useful in a comparability analysis, but it is not a systematic requirement”.
One point in the arm’s length range
As anticipated, the Spanish tax authority applied an adjustment to the return on sales of the entity for the FY 2007 and 2008 by increasing it to the level of the median (4.1%) of the return on sales identified by the taxpayer in its benchmark analysis.
The tax authority considered the median result as the most representative point in the range, arguing there were significant points of non-comparability because of differences in sales volumes. In such cases, the Spanish tax authorities argued that the median was most appropriate because it is a measure of central tendency. Therefore, the tax authorities decided to adjust the margins of the companies to the median of the arm’s length range.
The TEAC agreed with the decision of the Spanish tax authorities for the FY 2007 since the taxpayer’s margin was below the interquartile range but did not accept the adjustment for the FY 2008 since the mark up of the company (2.42%) was within the interquartile range, although below the median.
According to the court, a difference in sales volume is not enough to make comparability adjustments requiring the taxpayer to realize results at the median of the interquartile range.
Moreover, the fact that the benchmark was accepted as a whole in the comparability analysis and that the entity under examination occupies a leading position within its sector for its sales volume does not in itself cause a lack of comparability. Consequently, the court recalled paragraph 3.60 of the OECD guidelines recognizing that “if the relevant condition of the controlled transaction (e.g. price or margin) is within the arm’s length range, no adjustment should be made.”
Finally, when the case came to the Court of Appeal, its reasoning was similar to the TEAC’s, admitting that to apply an adjustment to the median, it is necessary that there are “defects of comparability”.
Contrary to the decision of the first instance court, the Court of Appeal argued that if it was agreed that comparability defects did not exist for 2008 for the same reason these did not exist for 2007. Therefore, the court upheld the plea ruling that the adjustment should have been made to the lower quartile (2.1%) in 2007, not the median (4.1%), and agreed that no adjustments should have been made for the FY 2008 since the taxpayer’s return on sales falls within the arm´s length interquartile range.
Legal certainty
Like other current transfer pricing cases reaching trial, the court relied upon the OECD guidelines.
The decision of the Spanish Court of Appeal together with the first instance court provided some legal certainty and security for taxpayers with regards to two transfer pricing areas that had been previously open to arbitrariness.
These two fields relate to the appropriateness of using the multiple-year analysis approach and the selection of a point in the arms’ length range.
The multiple-year analysis
Regarding the first area of concern, the selection of a multiple-year analysis, the court concluded that while the market interquartile range can be calculated on a multiple year basis, the taxpayer must compare the former to its financial position within the individual year subject to adjustment on a year-by-year basis.
Selection of a point in the range
With regard to the issue of selection of a point in the range, the Spanish Court of Appeal determined (in line with the OECD guidelines) that where the range comprises results of relatively equal and high reliability, any point in the range can satisfy the arm’s length principle, whereas, if comparability defects remain, the use of measures of central tendency can be more appropriate.
Additionally, the court emphasized the importance of consistency in that if the TEAC did not identify defects during 2008, neither could be attributed to the year 2007 for the same benchmark and interquartile range. Hence, the adjustment was made to the lower quartile instead of to the median.
Be the first to comment