Adding range to transfer pricing rules
Transfer pricing has been raking in significant tax controversies and has also emerged as one of the leading arms of the Indian tax law.
Transfer pricing has been raking in significant tax controversies and has also emerged as one of the leading arms of the Indian tax law contributing a significant share to tax collection. In order to align Indian transfer pricing regulations with global best practices, finance minister Arun Jaitley, in his maiden Budget speech of 2014, proposed introducing ‘range concept’ and the use of multiple-year data.
Range concept
In cases where MNCs enter into transactions with their Indian counterparts, such transactions are subjected to Indian transfer pricing regulations. These regulations mandate addition to income of Indian counterpart in cases where transactions are found not to be arm’s length price (ALP). The regulations prescribe use of specific methods to determine ALP. The regulations also prescribe use of arithmetic mean (with a tolerance range of 3%) in case where the methods suggest more than one ALP. This has resulted in skewed results in certain circumstances (say, in case the arithmetic mean could be skewed where one of the ALP was an outlier to the rest of the lot).
The draft rules suggest application of range concept in cases where ALP is determined using either the cost plus method (CPM), resale price method (RPM) or transactional net margin method (TNMM), and at least nine entities have been identified as comparable for the determination of ALP. In other cases, the originally prescribed arithmetic mean would continue to apply.
While applying the range concept, the numerator (profit or other such indicator) and the denominator (turnover or operating cost or other such denominator) are to be aggregated to derive data point for each of comparable entity. The data points lying between 40th and 60th percentile would constitute the range. If the ratio (margin) being tested falls outside the above range, then the median of the range would be considered as the arm’s length margin and transfer pricing adjustment will be determined accordingly.
While the introduction of range concept is a welcome move, some of the proposed rules may end up causing more hardship to taxpayers rather than providing relief. For instance, the condition of there being a minimum of nine comparable entities may seldom be satisfied in case of niche businesses. Further, the range is applicable only to three methods mentioned above and does not include commonly used other methods such as the comparable uncontrolled price (CUP) method. Another important aspect that requires consideration is that the draft rules use the term ‘a minimum of nine entities’. This could mean that the range would be applicable only on comparison with external comparable entities and not for comparable comparison of taxpayers’ transactions with third-party customers.
The most critical aspect, however, is the range of 40th to 60th percentile being proposed. The range may be considered too narrow, as only a few comparable entities within the comparable set would form the basis of the arm’s length range. Further, the proposed range does not appear to be in line with the government’s objective of alignment with global best practices. Some countries such as the US, having a more mature transfer pricing legislation, use the inter-quartile range, i.e. the 25th to 75th percentile, as an acceptable arm’s length range. In certain cases taxpayers may even be worse off than they would be by applying arithmetic mean for determining ALP.
Multiple-year data
One of the major challenges while deploying the prescribed methods was the availability of comparable data which is current or contemporaneous. In such cases, taxpayers were forced to use multiple-year data for determining ALP. During the course of assessments, the tax authorities have denied use of earlier years’ data, which was also upheld by the courts after considering the prevailing Indian transfer pricing regulations.
In a departure from this position adopted by the Indian tax authorities, the Board has proposed to amend the rules requiring taxpayers to use data for three years, i.e. for the year under consideration and two years immediately preceding the relevant year, to compute ALP. Use of three-year data is proposed to be mandated in cases where the ALP is determined using the above mentioned three methods—CPM, RPM and TNMM. In certain cases, use of two-year data may be permitted.
Allowing use of multiple-year data will considerably help improve comparability analysis as it will account for impacts of economic and product cycles, long-term arrangements, etc. This may also help bring about a level of standardisation across taxpayers and tax authorities for calculation of ALP. However, limiting the use to only specified methods may restrict taxpayers in a certain lifecycle to use single-year data to draw-up their ALP analysis.
The Board has taken a step in the right direction by putting up draft rules, having a closer and a more critical look at some of the proposals, and also by considering recommendations from various stakeholders. It would ensure implementing non-adversarial tax regime as promised by the finance minister to the business community.
The author is partner, Direct Tax, BDO India