The postponement of Vodafone's Supreme Court appeal to July 2011 has compounded the uncertainty regarding capital gains taxation of foreign investors in India. There are two dates which are key to the issue. First, the date when the Supreme Court's decision is given which we now know will be no earlier than the second half of this year. Secondly, and more certainly, there is April 1 2012, which is the date when the Direct Taxes Code Bill 2010 (DTC) will come into force. Section 5 of the DTC effectively contains the statutory enactment of the Vodafone tax charge, or, in the eyes of the Indian tax authorities, its re-enactment as the charge on non-residents already exists according to their view of current law under the Income Tax Act 1961 (ITA). The net result is that while there is scope for planning in anticipation of the DTC, those investors affected by the Vodafone proceedings will have to wait before knowing whether they will be faced with final substantial tax bills. It would be nice to think that the position post-April 2012 will be clear-cut and that the many uncertainties inherent in the Vodafone judgment will disappear. But, as explained later in this article, it would be wrong to think that the impact of Vodafone itself is short-lived and will only affect those who have made or will make transfers of shares in foreign companies owning Indian assets before April 1 2012. Techniques to mitigate the Vodafone impact following the Bombay High Court decision will remain valid in many cases.
Bombay High Court decision
Much has been written about the Vodafone case and it is not the purpose of this article to repeat the well-known background and history of the proceedings. The basic transaction effectively involved the sale by a Cayman company of shares in another Cayman company to a Dutch subsidiary of Vodafone. At the bottom of the vertical chain of companies sold was a controlling interest in a large Indian mobile telecoms operator. Perhaps the strangest, and sometimes forgotten, aspect of the proceedings is that they had nothing directly to do with the primary Indian capital gains tax (CGT) liability of the Cayman seller at all, but with the trickier question as to whether the Dutch buyer was subject to India's general fiscal obligation on payers to withhold tax when making payments to non-residents including payments of consideration which could represent capital gains. Before the challenge was made by the Indian tax authorities, existing fiscal jurisprudence suggested there was an insufficient connection with India for CGT to be applicable. Section 9 of the ITA applies to tax "all income accruing or arising, whether directly or indirectly, through the transfer of a capital asset situate in India". Even though the word "indirectly" is used, this refers to the accrual of income as opposed to the transfer of the asset. So, if the transfer is of an asset situated outside India, there should be no tax charge under this provision, and therefore nothing for the purchaser to do regarding withholding Indian income tax from the consideration.
It is worth noting that Vodafone was not a complete victory for the Indian tax authorities. The Bombay High Court decision delivered by a two-judge bench runs to 196 pages. Of particular interest was what the court thought of inward investment structures using intermediate holding companies offshore to hold Indian assets. Where these structures could be used to avoid CGT by selling shares in an offshore holding company instead of the underlying Indian asset directly, could they be attacked on anti-avoidance or similar grounds? Some of the salient general points of Indian tax law from the judgment are as follows:
It is perfectly open to a taxpayer to plan its affairs within the framework of the law to reduce the tax burden.
Where this is done, a valid legal act cannot be disregarded simply because it secures a tax advantage. In interpreting a fiscal statute, the court's duty is to follow the plain and unambiguous language of the statutory provision without expressing any moral judgment in relation to a taxpayer's conduct.
Similarly, the legal effect of a transaction cannot be displaced by probing into the substance of the transaction.
However, a transaction which is a sham or a colourable device with no legitimate underpinning can be ignored.
A shareholding represents a composite bundle of rights of a member in a company and, absent a sham transaction, shares cannot be looked through or recharacterised to represent direct interests in underlying assets.
A controlling shareholding is no different to any other shareholding in law. The fact that it gives the shareholder greater powers than a minority shareholder is merely a consequence of the rights attached to the holding: it does not turn the shareholding into a different capital asset.
Where income is taxable in more than one jurisdiction, judicial precedent (going back to transactions within and outside British India) recognises the need for apportionment.
The obligation to deduct income tax from payments to a non-resident extends to non-resident payers such as if it is not confined to persons subject to Indian taxation.
Until the reader got to page 148 of the judgment (which is where the discussion of these points stopped), it seemed as if the applicable principles would mean that the transaction was a bona fide sale of shares in an offshore company and that there was no basis for piercing the various corporate veils in the structure to say that there had been a sale of Indian assets. However, the court then applied their general principles to the facts and found as follows:
While none of the anti-avoidance principles applied to the facts, the overall transaction did not involve the simple sale of a share in a Cayman company.
The true nature of the transaction as evidenced by the transactional documents was that it consisted of the transfer of a composite mixture of assets and rights which enabled the controlling interest in the Indian telecoms business to change hands.
As evidenced in the disclosure to the Foreign Investment Promotion Board, the consideration paid was for the acquisition of "a panoply of rights including a control premium, use and rights to the Hutch brand in India, a non-compete agreement with the Hutch group, the value of non-voting non convertible preference shares, various loan obligations and the entitlement to acquire subject to Indian foreign investment rules, a further 15% indirect interest in [the telecoms operator]".
The composite bundle of assets included items which were capital assets situated in India in their own right and therefore, their sale attracted CGT.
Although not a function of the court, the consideration would need to be apportioned between the various rights and assets sold to compute the CGT bill.
The curious position reached was that the court did not find any evidence of tax avoidance in the way in which the sale was structured, but having looked at the large number of transaction documents, concluded that it could not be right to say that what happened was simply the sale of a share in a Cayman company. Without any close analysis of how the ancillary documents to the share sale agreement could themselves have involved the transfer of other capital assets, the court concluded that this was a composite sale of a bundle of rights and assets-even though those rights and assets were not owned by the seller. It seems to follow from this that Vodafone is a very particular decision on its complex facts: the exception that proves the rule. The rule remains that it is not possible for the tax authorities to pierce the corporate veil when it suits them. Indeed, in general the veil cannot be pierced. But if the documentation suggests rather more than the transfer of shares, the court will follow that through to see what the real transaction between the parties is. There is nothing new in terms of jurisprudence in this. What was more novel was how the Bombay High Court applied that jurisprudence.
Some of the court's logic was based on an interpretation of the legal agreements which suggested there was some ambiguity between identifying what was the bargain struck between the parties (relevant for tax purposes) and the mechanics required to complete that bargain. There was no doubt that the sale of the Cayman company carried with it all the indirect underlying benefits relating to the Indian telecoms business. If a holding company is sold, it is hardly surprising that its value is derived from its investment in its subsidiaries, which in turn derive their value from their businesses. To suggest that the real sale in such a situation is a sale of the controlling interest in the subsidiaries' businesses is a somewhat startling proposition: it seems from the general statements made by the court earlier in the judgment that they would also agree with that sentiment. But where the Vodafone facts apparently diverged is that the transaction there required the putting into place of structural arrangements in India that would regulate the conduct of the business of the Indian company, the relationship between the shareholders of the company and the contractual entitlements accruing or arising in respect of agreements with the Indian partners. In other words, the manner in which the purchaser effectively obtained control of the Indian business was not simply by buying the share in the Cayman company.
It is likely that on a different and simpler set of facts where it is possible to distinguish more clearly between the sale of the shares and consequential matters including completion mechanics, the Indian courts would uphold the transaction as the sale of shares and nothing else. Clearly, great care needs to be taken in drafting the documentation right so that consequential steps are not confused with the subject-matter of the sale. But there seems no reason why, even today, after the Bombay High Court decision, offshore sales of shares in foreign companies with underlying Indian interests should give rise to CGT. The simplest position is where the sale really does not require anything to be done in India even as a completion or post-completion matter. An intra-group reorganisation of an offshore group is an example of this where the ultimate ownership does not pass to third parties. If that reorganisation involves hiving down shares in the foreign company to a new subsidiary, that does not have any Indian nexus. But even in third party sales, it would be wrong to think that all such sales would be caught by the Vodafone decision. Indeed, thinking this would be to do the Bombay High Court something of a disservice given the general principles which they so carefully set out in the first 148 pages of the judgment.
Treaty implications
In Vodafone, the effective seller was a Cayman company – an entity which was not entitled to the benefit of a tax treaty since none exists between India and the Cayman Islands. The question arises whether Vodafone would also affect a situation where the seller is treaty-protected as there are many inward investment structures established on this basis. Suppose a US company owns a Mauritian company (MCo 1), which sells shares in another Mauritian company (MCo 2) which has a controlling interest in an Indian company. Suppose also that both MCo1 and MCo2 hold Mauritian certificates of residence and are entitled to treaty protection based on the Supreme Court judgment in Union of India v. Azadi Andolan Bachao.
If the true analysis of the sale is that MCo1 sells shares in MCo2, then the Indian tax analysis is that this is not a transfer of a capital asset situated in India in any event: Vodafone is inapplicable. There is no need for MCo1 to rely on the India/Mauritius tax treaty. But suppose that one could somehow pierce the corporate veil and say MCo2 is in reality selling its controlling interest in the Indian company. If that is the case, then the treaty ought to apply to say that the transaction is still not taxable in India because MCo2's supposed transfer of shares in the Indian company is protected by the treaty exemption from capital gains. To pierce the corporate veil further and to suggest that the real transaction is a sale of its business by the Indian company, which is subject to Indian tax, seems far-fetched and unwarranted by Vodafone. One cannot simply ignore the fact that the Indian company continues with its business following the sale of the shares so how can it have transferred its business to anyone?
The position gets more interesting if there are completion matters to do with the sale of the shares which have an Indian connection. Suppose, like in Vodafone, there is a brand change in the business. If this means that a valuable brand right has been transferred, this still begs the question: "Who has transferred it?" If the answer is MCo2 (which owns the controlling interest directly in the Indian company), then the question is what is the nature of its right? In Vodafone, it was suggested that this right is a separate capital asset situated in India. On that basis, MCo2 could still rely on the treaty to claim CGT exemption on the sale of this right.
In other words, while the interaction between Vodafone and treaties is not completely clear, the result under current law should be that where someone is properly entitled to the CGT exemption under a treaty, that should prevail.
Finally, in the above example, there is the question whether one could somehow say that the real transaction is a sale by the top US company of Indian assets and therefore CGT would be payable. An attempt by the Indian tax authorities to pierce the corporate veil in this way and to deny treaty benefits was unsuccessful in the well-known ruling of the Advance Ruling Authority in E*Trade Mauritius Ltd. delivered last year. Absent an extremely aggressive set of facts verging on tax fraud, it is really quite difficult to see how such a challenge could be maintained. There would need to be at least a serious misrepresentation in the manner in which tax residence certificates had been obtained by the Mauritian companies before such an argument could get off the ground.
Vodafone and the DTC
Once the DTC comes into force on April 1 2012, the position will be governed by what is section 5 of the DTC. As is explained below, this does not mean that Vodafone will cease to be good law. Of course we do not know what the Supreme Court will decide, so any meaningful discussion at this stage necessarily presupposes that the Bombay High Court decision will be upheld on appeal.
One preliminary point needs to be made about the DTC.
Section 1 (3) states that:
"Save as otherwise provided in this Code, it shall come into force on 1st April 2012".
Section 5(1) (d) of the DTC states:
"The income shall be deemed to accrue in India if it accrues, whether directly or indirectly, through or from the transfer of a capital asset situated in India".
Taking the two provisions together, the natural reading is that the DTC will apply to transfers from April 1 2012. But this is not wholly clear so some care needs to be taken in transitional situations where something may need to occur after that date such as where a sale completed before that date provides for deferred consideration after that date. There are a number of ways of dealing with these situations but it is hoped that more detailed commencement provisions will be introduced later this year, making these added measures unnecessary.
Section 5 as set out above is similar to section 9 of the ITA. So, the Indian tax authorities would say that these words are no more than a re-enactment of current law. The reproduction of the words "directly or indirectly" from section 9 enables them to look beyond direct sales as explained in section 2 above.
Therefore, any limitation to the extent of this charge needs to be by way of express exception. That is provided by new words in section 5(4) which exclude from the CGT charge:
"Income from transfer outside India, of any share or interest in a foreign company, unless at any time in 12 months preceding the transfer, the fair market value of the assets in India, owned, directly or indirectly by the company, represent at least 50% of the fair market value of all assets owned by the company".
It should be remembered that in Vodafone, there was much talk about the fact that a controlling interest in an Indian company was being transferred. The new charge and exclusion has nothing to do with controlling interests at all, which is unexpected. For example, if a foreign company (FCo1) owns another foreign company (FCo2) which owns 10% of an Indian company and nothing else, a sale by FCo1 of FCo2 would not fall within the exception: the indirect interest in the 10% stake represents more than 50% of the assets owned. But if FCo2 owned a controlling interest in an Indian company and other non-Indian assets which are greater in value than the Indian interest, then the exclusion would apply. Where this does provide a sensible result is where an international trading group decides to sell a business division where India is just a small part of that division. If it sells an intermediate holding company offshore which indirectly owns the Indian business through a chain of companies as well as other companies outside India, the indirect sale of the Indian company would not be subject to Indian CGT. But it will still be important to make sure the sale occurs at the right level in the chain.
It should also be noted that the exclusion does not apply if "at least" 50% of the fair market value of the assets are Indian assets. This again is odd in that one might reasonably have expected this to say more than 50% in accordance with international tax principles. In other words, the charge should apply where the transfer relates to assets which are principally Indian assets. It may be that this particular provision is relaxed in the final version of the DTC.
Post 2012
Post-April 2012, considerable focus will be on the fair market value of assets held by offshore companies. One can reasonably expect a certain amount of reorganisation of shareholdings as well as before that date.
In a treaty situation, if a CGT exemption is available, the charge should not apply subject to two caveats. First, care needs to be taken that there is sufficient substance in the treaty seller to justify its entitlement to treaty relief. Section 291(5) of the DTC precludes treaty relief unless a tax residence certificate has been obtained. But this is not of itself conclusive: it is a minimum requirement. So one cannot simply assume post-DTC that such a certificate on its own will be sufficient in every claim for treaty relief. Secondly, as is well known, a new general anti-avoidance rule will be introduced which will have priority over treaties. One can speculate at length as to how this will apply where intermediate holding companies are established in treaty jurisdictions. The simple truth at the moment is that there is uncertainty which may or may not be resolved by the anticipated GAAR guidelines. In the meantime, it remains key to focus on corporate governance of such entities as well as their substance.
Finally, the question remains whether Vodafone will be defunct after April 2012. A simplistic view would be to say yes. However, the way in which the Bombay High Court chose to analyse the transaction and find assets and rights sold beyond the transfer of the Cayman share has nothing to do with the DTC. This is a question of judicial construction of documents to find the transaction between the parties. There is no reason why that approach, if upheld by the Supreme Court, will not continue. So even if the new exception in section 5(4) is available, it will still be necessary to ensure that the transaction both reflects the intention of the parties and is documented properly as a share sale.
Nikhil Mehta (nm@taxbar.com) is international tax counsel at Amarchand & Mangaldas and Gray's Inn Tax Chambers.
