The Great Indian Transfer Pricing Circus – a critical view of Indian TP provisions

The Great Indian Transfer Pricing Circus – a critical view of Indian TP provisions

by 

Vikram Vijayaraghavan, Advocate

M/s Subbaraya Aiyar, Padmanabhan & Ramamani (SAPR) Advocates, Chennai

Author's note & disclaimer: This Article appeared in the Third Quarter 2012  Volume 7 Issue 3 Page 7 of the Tax Justice Focus magazine available from this link. The content of the article relates to October 2012 and the TP provisions applicable at that time

The year 1991 was a watershed in modern India’s economic history; it was the year in which the Indian economy was `opened up’ or liberalized by the then widely-hailed Finance Minister, Dr.Manmohan Singh, now the Indian Prime Minister. The economic reforms of 1991 were far-reaching and opened up India for international trade and investment, taxation reforms, deregulation and privatization. These reforms caused huge cash flows into and out of India in the following decades.

The more liberal international trading regime established by Singh highlighted the issue of Transfer Pricing (TP).The Indian Government, like most others, is heavily dependent on tax revenue and simply cannot ignore the scope for tax avoidance created by transfers between subsidiaries of multinational companies. So it stepped up in 2001 and amended the Indian Income Tax Act of 1961 (via the Finance Act of 2001) and added a new chapter titled “Chapter X : Special Provisions Relating to Avoidance of Tax” and introduced Section 92 in Chapter X containing sub-sections 92A to 92F and Income Tax Rules (Rule 10A-10E) laying out specific TP provisions for the first time. In other words, the TP circus had begun in India.

Indian TP provisions were fairly OECD-like in the sense they based the TP regime on the arm’s-length principle or ALP (defined in Sec.92) of international transactions (Sec.92B) between associate enterprises (Sec.92A). The computation of ALP was laid down via five methods namely Comparable Uncontrolled Pricing (CUP), Cost-Plus (CPM), Resale Price Method (RPM), Transactional Net Margin Method (TNMM) and the Profit-split Method (recently a new sixth method has been prescribed by the Central Board of Direct Taxes, the tax administering body). Comprehensive documentation requirements are laid out in Sec.92D and Sec.92F contains all the definitions of the terms.

However, there are certain important differences between Indian provisions and the OECD TP guidelines. The definition of “Associate Enterprise” is quite broad under Indian TP compared to OECD; multiple-year data of the Financial results of comparable companies is not allowed in Indian TP except under certain circumstances unlike OECD Guidelines; arithmetic mean of comparables is used in Indian TP and not inter-quartile ranges; Indian TP has stringent documentation guidelines while lacking guidelines for intra-group set-offs, thin capitalization, intangibles etc. all of which are in contrast to the OECD guidelines.

There are two basic and serious problems with the current Indian TP provisions - firstly they are too general and vague to be useful laws; secondly their implementation has left very much to be desired. One worries that the first problem is intractable because TP by its nature seems an economic issue not lending itself to precise definitions which is essential for legalese and that the second problem is unsolvable because the ambiguity inherent in TP in general and the Indian TP provisions in general create opportunities for misinterpretation, over-reach and overzealous action.

Let us look at typical scenario, using the fictional multinational ABC to illustrate it. ABC (India) Pvt. Ltd. a subsidiary of American company ABC Inc., provides software development services to its foreign parent, a financial analytics application software firm. ABC (India) Pvt. Ltd. will likely choose cost-plus (CPM) and add a markup (say, 15%) to its costs. Given there are no benchmark figures for comparing the mark-ups in various sectors, the Revenue Department will normally not accept the taxpayers markup % and hold it as being too low adopting TNMM instead. It has become customary for the Department to try and apply TNMM to all varieties of International Transactions given that TNMM method is easy to apply without requiring too much precision. The Revenue Department will come up with a set of comparables to ABC (India) Pvt. Ltd. And therein lies the rub.

The comparable list of the Revenue Department will typically contain companies from the giants such as Infosys™ and TCS™ as well as small companies in unrelated software verticals (travel, healthcare etc.). These comparables make no sense whatsoever in the context of ABC; can a software megalith with more than 100k employees be compared to a 200 people firm? Can unrelated software verticals be compared at all? These comparables are there simply because proper comparables are incredibly hard to find. This is more so in a developing country like India where industry is still evolving, new industries are being opened up and the market hasn’t matured completely.

So due to this paucity of comparables, the only option in such a case is to take the existing comparables and perform ‘adjustments’. These adjustments are not enshrined anywhere and the TP provisions are delightfully vague on them! For example Rule 10B(e)(iii) on TNMM states “(iii)  the net profit margin referred to in sub-clause (ii) arising in comparable uncontrolled transactions is adjusted to take into account the differences, if any, between the international transaction and the comparable uncontrolled transactions, or between the enterprises entering into such transactions, which could materially affect the amount of net profit margin in the open market” . What `differences’ are to be accounted for and what will `materially affect’ the net profit margin is left to the taxpayer to substantiate and it is no surprise that the Department almost always disagrees. Many of the adjustments required such as risk, working capital, depreciation, idle capacity etc. are totally subjective and result in a lot of disputes between the taxpayer and Department. Furthermore, there are many “filters” applied for rejecting comparables such as those having export sales less than 25% of total revenue, those companies that make persistent operating losses and those companies that have high turnover and super-profits – well, what is super-profit? What is high turnover? Surprise, surprise – there are no quantifications for these filters and the result is the entire exercise often devolves into something farcical where the Department cherry-picks its comparables justifying a high profit % and the assessee picks its own set of comparables with a % close to its profit % and then both start bartering i.e., give and take of comparables. The difficulty is compounded by the fact that the comparables are selected on the basis of the published financials, which as we all know only display only the required minimum and so do not provide an adequate basis for comparing companies.

This is a simplified example and one can imagine much more complex scenarios in real life. Consider the case of software startups – their very USP is being incomparable to other companies; consider the cases of firms developing and/or trading in intangibles– the absence of clear guidelines makes it impossible to engage in a fruitful function, risks and assets (F.A.R) analysis as prescribed by TP provisions.

These issues with adjustments are not restricted to TNMM alone. Even when we use other methods, such as CUP, adjustments maybe required for both internal & external CUP causing very much the same confusion. Let us take a simple example: a toy company EFG India Ltd. exporting toys to its owner EFG PLC, UK on bulk contract basis as well as selling in the domestic retail market in India, has to make adjustments for the domestic vs. UK export market, wholesale vs. retail (i.e., volume discount) etc. in order to apply internal CUP.

In short, the TP provisions are not simple and practical enough to apply in reality and they more often than not lead to Pyrrhic battle for the taxpayer.

And it gets worse. Leaving the conceptual issues related to TP such as comparables, filters and adjustments, there are numerous practical problems in the implementation Indian TP provisions. The Transfer Pricing Officer (TPO) sometimes uses “customs data” blindly, uses powers to obtain information directly from third-party firms (under Sec.133(6)) without sharing the same or proceeding on the basis of cursory information obtained, applyies TNMM incorrectly, discards loss-making companies outright etc. In a recent case the TPO grabbed customs data directly from the Indian Custom authorities on coal imports without providing detailed information to the taxpayer; further investigation by the taxpayer revealed that the data compared coal of completely different calorific values and was unsuitable in a number of other respects..

These kinds of practical issues are common in Indian TP practice and combined with the theoretical flaws with TP it is no surprise that there has been a huge rise in litigation in Indian Courts on TP issues.

All these TP cases wind up in the traffic jam that is the judiciary (and quasi-judiciary) represented by the Commissioner (Appeals), the Income Tax Appellate Tribunal (ITAT), the High Court (HC) and the Supreme Court (SC) . The judiciary, for all its defects, is the place of last and often best resort for the taxpayer and it has slowly but surely pushed for a reasonable interpretation of TP provisions. However the entire process is uncertain and takes too long a time for the taxpayer. To make things worse, we now see an aggressive Revenue Department which does not wait for outcome of judicial proceedings but proceeds to attach properties & bank accounts unless some portion of the tax demand is paid. So, what is a taxpayer to do? The answer depressingly seems to be nothing, except to write articles such as these and hope for the best.

However, every cloud has a silver lining and the uncertainty of TP provisions and the Indian taxation regime in general has caused tremendous investor and public backlash in recent times. The Vodafone™ case relating to capital gains taxation is much publicized but it is symptomatic of the general discontent and angst at the terrible uncertainty, fickle nature and occasionally over-rigorous implementation of taxation provisions in India. This has led to tremendous pushback from those that matter i.e., those with money and the leviathan that is the Indian Government is slowly waking up. We recently saw some effort in the right directions by the Government in the recent formation of a Committee to study safe-harbor rules, in introducing a new placeholder section for such rules (Sec.92CB), by the introduction of Advance Pricing Arrangements (Sec. 92) and in prescribing a new TP method (Rule 10AB), though defined ambiguously, which basically allows use of any method (such as quotes, valuations etc.) for purposes of TP.

Frustratingly, we still see retrograde measures being introduced in tandem such as specified domestic transactions (Sec.92BA) now being brought under TP; retrospective amendments relating to definition of international transaction and to restrict the arm’s-length range, introduction of general anti-avoidance rules (GAAR). This last was so heavily criticized that it has been postponed.

What is the solution? This author feels that a three-pronged approach may work – first is to introduce sector-wide safe harbors which one believes the Government is working on and is great news for the industry. This would work by prescribing profit %’s for each sector tied to a published industry-wide index. Another solution is to consider the use of Formulary Apportionment (FA) wherever possible in addition to ALP – FA is an intuitive formulaic sharing approach splitting up the profits amongst the group companies across the globe. The third solution is of course to streamline the current provisions and make them more practical and applicable – some concrete suggestions are to use multiple year data, use inter-quartile ranges, to avoid cherry-picking comparables, to not discard loss making comparables outright, to provide clear and precise guidance on adjustments and filters, and to allow select technical expert references for comparability analysis.

Whether all this will happen is anybody’s guess. Unfortunately, from past experience one can say that the only certainty in Indian taxation is that there will always be uncertainty. Combined with the underlying ambiguity and hollowness of TP viz. a viz. the arm’s-length principle itself what we have here is a perfect storm.

In conclusion, it is clear that TP itself needs a fundamental re-think and that the ALP arm’s-length principle while good in theory does not pan out well in practice. It is time alternative systems in TP are thought through and brought to the fore. It is also clear that with respect to Indian taxation, the TP provisions are the most important in terms of tax revenue as well as one of the most controversial and highly litigated tax provisions and the pressing need of the hour is to reform them and make their language and implementation certain and practical for the taxpayer.