I. The case, the result, and the unfinished business

The decision in brief. Valeo Bayen, a French company, held two Indian subsidiaries: Valeo India Pvt Ltd (“VIPL”), which was wholly owned, and Valeo Service India Auto Parts Pvt Ltd (“VSIAPL”), originally a joint venture with an Indian group, which became wholly owned by Valeo Bayen after it bought out its joint-venture partner’s 40% stake in 2018. In December 2019 Valeo Bayen sold its entire VSIAPL shareholding to VIPL for Rs 64.78 crore, after which VSIAPL was merged into VIPL. Valeo Bayen claimed the sale was not a “transfer” for the purpose of the capital gains charge under Section 45 of the Income Tax Act, 1961, being covered by the exclusion in Section 47(iv) for a transfer from a company (Valeo Bayen) to its wholly-owned Indian subsidiary (VIPL); and, VIPL having deducted tax at source from the consideration, it claimed a refund of that tax in its India return. The Assessing Officer rejected the claim, treating the shares as held with trading intent so as to deny the Section 47(iv) exemption and bringing the entire consideration to tax under the residuary head of income from other sources, after an enquiry into the commercial substance of the transaction and the route not taken. The Dispute Resolution Panel confirmed the addition. Allowing the appeal, the Indian Income Tax Appellate Tribunal [TS-676-ITAT-2026(CHNY)] held the shares to be a capital asset on a holistic view of the holding, declined to question the assessee’s choice of merger route, found no defect identified in the DCF valuation, and held the transfer squarely exempt under Section 47(iv). The result is plainly correct.

Figure 1. The holding structure across the reorganisation. The French parent held VIPL (wholly owned) and VSIAPL (a joint venture until it bought out its partner in 2018) as sister companies. The sale of 23 December 2019 placed VSIAPL beneath VIPL, the relationship to which Section 47(iv) applies, and the fast-track merger then dissolved VSIAPL into VIPL. The two relationships never coexisted: the sale is what converted the sisters into a parent and subsidiary. 

The result in Valeo Bayen is plainly correct, and the reasoning is sound on each of its main analytical moves: the holistic test for asset characterisation, the armchair-of-a-businessman principle on the choice of merger route, and the refusal to permit rejection of a Discounted Cash Flow valuation without identification of specific defects. The ruling will be cited, and properly so, by multinational groups planning Indian sub-group consolidations. It is not important merely because it allows Section 47(iv). It is important because it shows what happens when assessment orders deploy characterisation tactically, and when anti-avoidance instincts are exercised outside the statutory discipline of GAAR. 

What is more interesting is the question the case did not reach. The Assessing Officer’s objection, reduced to essentials, was that Valeo Bayen had taken cash out of India shortly before the merger and paid no tax on it, when it could instead have merged the two companies directly and received nothing in hand. There is a real point buried there, though not the one the Officer pressed: under a direct merger the parent would have taken fresh shares in the merged company in exchange for its VSIAPL shares, and the gain, rolled into those shares, would have been chargeable whenever they were later sold, so the merger route would have carried a deferred tax cost that the cash sale did not. One can see why the Revenue might have read this as a textbook invitation to invoke GAAR: a tax benefit on one view, coupled with a sale between a parent and its wholly-owned subsidiary that could be cast as commercially unusual and tax-driven, the very combination Chapter X-A is meant to catch. The suspicion was understandable, but a suspicion that is understandable to entertain is not the same as a case that is made out, and if the asymmetry was to be tested as avoidance at all, the only instrument that authorised the testing was the General Anti-Avoidance Rule in Chapter X-A, which the Revenue never invoked. The order proceeds instead as a free-standing enquiry into commercial substance and alternative routing, a jurisdiction this commentary argues the Assessing Officer did not possess, and one which, as Section VII explains, would in any event have failed even through its proper channel, because the architecture of Indian GAAR would not, on these facts, have yielded the conclusion the Revenue needed.

And yet, on a closer reading, the case contained the materials to settle rather more than it eventually did. The Tribunal’s adjudicatory powers extend well beyond the grounds of appeal, allowing it to deal with issues not raised in the grounds provided both parties are heard on the point and the appellant is left no worse off for having appealed; but as a matter of practice the Tribunal, more often than not, decides only what is put to it, and here the focus of argument was, understandably, the application of Section 47(iv) to the undisputed facts, on which the appeal succeeded resoundingly. The assessment order, read carefully, raised at least four further questions that could fruitfully have been pressed, and that, had they been pressed, would have permitted the Bench to settle a much wider range of disputes for the future. The order contained a logical contradiction that, if isolated as a freestanding ground, could have disposed of the appeal even before any enquiry into Section 47(iv) became necessary. That contradiction, as Section IV shows, carried a further consequence the order never confronted: on the Assessing Officer’s own premise that Valeo Bayen was trading in shares, the gain was business profits, and the business profits of a non-resident without a permanent establishment in India are not taxable here at all. The relationship between the statutory reliefs that take a transaction out of charge, whether the exemptions of Chapter III or the exclusions built into the charge such as Section 47, and the General Anti-Avoidance Rules in Chapter X-A, the question that will shape Indian anti-avoidance jurisprudence for the next decade, was implicit in the structure of the assessment but was not raised, and understandably so. It is contested ground, where the Revenue’s reading and the academic view diverge and the answer is far from settled; and a taxpayer already assured of victory under Section 47(iv) had little reason to invite the focus of the litigation to shift onto that terrain. 

There is a wider wisdom in such restraint. Once a Bench is drawn from the narrow facts of a case onto foundational ground, the stakes escalate for both sides, and a litigant who might have won quietly on the facts instead stakes the result on a contested proposition of system-wide consequence. To have raised the Chapter X-A question would have been to invite precisely that contest, on terrain the Revenue would have defended to the hilt, in exchange for nothing the assessee needed. The discipline of winning well on the facts, rather than reaching for the landmark, is what counsel exercised here. The treaty ground, raised in the appeal memorandum but not argued with any depth, fell away for the same practical reason once the domestic exemption succeeded. 

This commentary works through three things in turn. It identifies what Valeo Bayen decides and what it does not. It then turns to the assessment order itself, where a four-layered error of characterisation invites analysis. And it asks the question the case raised but on which the Bench was not invited to pronounce: whether, after the introduction of Chapter X-A on 1 April 2017, the Assessing Officer's enquiry into commercial substance was jurisdictionally available to him at all. 

II. The four things Valeo Bayen actually holds 

Stripped to essentials, the decision holds four things. 

First, the conditions of Section 47(iv), that the parent (with or without its nominees) hold the whole of the share capital of the subsidiary, and that the subsidiary be an Indian company, are mechanical conditions. Where they are met on undisputed facts, as they were here, the statutory exemption operates of its own force. The transfer is, by legislative deeming, not a “transfer” for the purposes of Section 45, and no chargeable gain arises. 

Second, the characterisation of shares as a capital asset or as stock-in-trade is to be undertaken holistically, by reference to the entire history of the assessee's ownership rather than by isolating a particular acquisition. Where an assessee has held the majority tranche of an investment for several years as a long-term holding, the subsequent acquisition of a smaller tranche from a joint-venture partner pursuant to the partner's exit does not, of itself, transform the character of the overall investment into a trading asset. 

Third, the Revenue cannot question the commercial expediency of a taxpayer's choice between two lawfully available routes to a corporate reorganisation. Where Parliament has provided a fast-track merger procedure under Section 233 of the Companies Act, 2013 for wholly-owned group restructurings, faulting the taxpayer for not pursuing the longer NCLT-supervised route under Sections 230 to 232 is an intrusion that the Revenue is not entitled to make. The armchair-of-a-businessman objection is well-grounded in Indian jurisprudence and was correctly invoked. 

Fourth, on valuation, the Assessing Officer's preference for Net Asset Value over Discounted Cash Flow cannot be sustained on a mere comparison of the two outputs. The taxpayer has a statutory right to elect between the methods, and the Officer's power is to examine the assumptions underlying the chosen method and identify specific defects, not to substitute his own preferred methodology by reason of the result.

Each of these holdings is, in its own terms, unobjectionable. The result follows from them ineluctably. 

III. What the decision leaves untouched 

What the decision does not decide, and what it leaves carefully untouched, is the more interesting territory.

The treaty ground. The assessee’s appeal memorandum raised a ground under the India-France Double Taxation Avoidance Agreement: that the deeming fiction in Section 9(1)(i), which treats capital gains arising from the transfer of shares of an Indian company as income accruing in India, cannot be imported into the treaty framework, and that the applicable capital-gains article of the treaty would in any event allocate the taxing right elsewhere. The Tribunal did not address this ground, and on the principle of judicial economy the omission is understandable: once the appeal succeeds on the domestic exemption, the treaty ground becomes academic. Had it been reached, however, it would have raised questions of some difficulty rather than offered an easy alternative route to the same result. Its first limb, that the deeming fiction in Section 9(1)(i) cannot be imported into the treaty framework, was put more broadly than it needed to be. The fiction does its work within the Act, creating the charge, while the treaty sits above it and, through Section 90(2), allows the assessee the benefit of whichever regime is more favourable; the narrower proposition, that the treaty’s allocation prevails where it is more beneficial, would have rested on surer ground than the wider claim that the domestic fiction is disabled at the treaty’s threshold. The second limb, that the capital-gains article would in any event allocate the taxing right away from India, would have had to contend with the structure of Article 14 of the India-France treaty, which allocates gains on the alienation of shares of an Indian company to India where the shares represent a participation of at least 10 per cent in a resident company, the residence-only rule in the final paragraph reaching only those gains not already covered. How that framework applies to a transfer of the kind in issue is not free from difficulty, and it is better examined in a case where the point is fully argued than resolved in passing here. It is enough for present purposes that the assessee’s success under Section 47(iv) made the treaty ground unnecessary, and that nothing turned on its not being pressed. 

The GAAR question. Chapter X-A of the Act came into force on 1 April 2017. By the time of the assessment proceedings in Valeo Bayen, the General Anti-Avoidance Rules had been in operation for several years. The Revenue did not invoke them, and the assessee, quite properly given the way the case was put against it, did not raise the question of what the consequences of that non-invocation might be. The Tribunal's order, accordingly, addresses the Assessing Officer's enquiry into commercial substance, alternative routing, and the purpose of the transaction as if it were a self-contained exercise of ordinary assessment jurisdiction. Whether it could have been otherwise, whether the absence of a GAAR reference was in fact dispositive of the AO's jurisdictional competence to undertake any enquiry of that kind, is the question this commentary returns to in Section VII below. 

The head-of-income contradiction. The Assessing Officer's order treats the shares as stock-in-trade for the purpose of denying the Section 47(iv) carve-out, but treats the consideration as income from other sources for the purpose of assessment. These two characterisations cannot stand together. The point was raised by the assessee, but in the course of the wider argument on Section 47(iv) rather than as a freestanding ground capable of disposing of the appeal on its own. It is to that contradiction that we now turn. 

IV. The four-layered error in the assessment order 

The Assessing Officer's order, distilled, proceeds on the following logic. The shares of VSIAPL held by the assessee were not capital assets but were held with a trading intent, that intent being evidenced by the fact that the 40% tranche acquired from the joint-venture partner in 2018 was sold within 18 months in 2019. Because the shares were not capital assets, Section 47(iv), which is a carve-out from the charging provision applicable to capital assets, could not apply. The consideration was, accordingly, brought to tax as income from other sources under Section 56. 

This reasoning contains four errors, layered on top of each other, each independently fatal. 

The first layer: a wrong substantive characterisation 

The shares were not stock-in-trade. The assessee had held 60% of the equity of VSIAPL since 2012, for over seven years as at the date of sale, and had treated the holding in its financial statements as a long-term investment, not as trading stock. The acquisition of the residual 40% in 2018 was not a trading transaction; it was the buy-out of a joint-venture partner's exit, undertaken to consolidate an existing long-term investment. The Tribunal's holistic approach to this question, that the dominant character of the overall holding must prevail, is the correct one, and the analysis on this point in the order is sound. 

But the substantive characterisation error is only the first layer. 

The second layer: the wrong head of income even on the AO's own logic 

Assume, for the sake of argument, that the AO was right that the shares were stock-in-trade. The consequence does not lead to Section 56. The Income Tax Act organises chargeable income into five exclusive heads under Section 14: salaries, income from house property, profits and gains of business or profession, capital gains, and income from other sources. The heads are mutually exclusive. Section 56(1) operates as a residuary provision, charging income only where it does not fall within any of the heads A to E. 

If shares are held as stock-in-trade, the gain on their sale is unambiguously business income chargeable under Section 28. It is computed under Sections 29 to 44DB, with the deductions, set-offs, depreciation allowances, and special provisions applicable to that head. It cannot also be income from other sources. The residuary head is residuary precisely because the other heads have already been canvassed and excluded. 

The AO's order makes no attempt to explain why, having held the shares to be stock-in-trade, he did not charge the consideration under Section 28. It asserts both characterisations at once, stock-in-trade for the purpose of denying the Section 47(iv) carve-out and not-business-asset for the purpose of bringing the consideration under the residuary head, without acknowledging that the two cannot be sustained together. 

The error is not merely formal. The choice of head determines the computational mechanism, the available deductions, the rate of tax, the period-of-holding rules, the loss set-off and carry-forward consequences, and, critically, the treaty article that would have applied had the assessee been entitled to treaty protection. To treat the same receipt as falling under one head for one analytical purpose and under another head for another is to treat the head of income as a matter of revenue convenience rather than legal classification. That is not a permissible approach under the Act.

The third layer: denial of cost of acquisition even within Section 56 

Even if one were to indulge both of the AO's earlier choices, accepting the stock-in-trade characterisation and permitting the recourse to Section 56, the assessment would still be defective at the computational stage. Income under Section 56 is, like income under any other head, charged on a net basis. Section 57 expressly provides for deductions, including expenditure laid out wholly and exclusively for the purpose of earning such income. 

In the case of a sale of shares, the cost of acquisition is precisely the kind of expenditure that Section 57 contemplates. The AO did not allow it. The entire Rs 64,78,01,960 of sale consideration was brought to tax as if it were a gratuitous accretion to the assessee's wealth rather than the proceeds of a sale against which a cost had to be set off. The assessee raised this as ground 6 of the appeal memorandum, that “the Ld. DCIT and Ld. DRP failed to allow the cost of acquisition of shares as a deduction under Section 57 of the Act”, but the point does not feature in the Tribunal's operative reasoning either, having been rendered moot by the success of the Section 47(iv) ground. 

The fourth layer: no charge at all on the AO’s own premise 

There is a further consequence the order never confronted, and it is the most striking of all because it follows from the Assessing Officer’s own characterisation. If the shares truly were stock-in-trade, then Valeo Bayen was, on the Officer’s premise, carrying on a business of trading in shares, and the gain was business income. But the business profits of a non-resident are taxable in India only to the extent attributable to a permanent establishment here. Under Article 7 of the India-France treaty, in the absence of a permanent establishment there is no charge on business profits at all; and the order records no permanent establishment, nor is one suggested. On the Assessing Officer’s own characterisation, therefore, the consideration was not taxable in India by yet a further route, independent of Section 47(iv) and independent of the head-of-income contradiction. 

This is, of course, an alternative and additional answer rather than the primary one. The primary position remains that the shares were a capital asset and the transfer exempt under Section 47(iv). The fourth layer bites only on the premise the Assessing Officer himself adopted, and it shows how comprehensively that premise defeats the addition he made: characterise the shares as a trading asset and the gain becomes business profits that, for a non-resident without a permanent establishment, India cannot tax. The premise the Officer needed in order to deny the exemption is the very premise that, carried to its conclusion, removes the charge altogether. 

What the layering reveals 

These errors are not isolated. Read together, they expose the assessment order's organising logic. The order is not the product of a doctrinal enquiry that began with the question “under which head of income does this receipt fall?” and proceeded to an answer through the application of legal principle. It is, on the contrary, the product of an enquiry that began with the result, that the entire consideration should be taxed at the highest possible rate without deductions, and worked backwards to whatever doctrinal labels could plausibly support that outcome. 

That is why the order can simultaneously hold the shares to be stock-in-trade (to deny Section 47(iv)) and treat them as not-business-asset (to bring the consideration under Section 56). The two characterisations are deployed instrumentally, each to defeat a different statutory provision standing between the AO and his preferred outcome. The head of income, in this approach, ceases to be a legal classification and becomes a tactical choice. 

The Tribunal, dealing with the case as it was argued, did not need to express the matter in those terms; the result on Section 47(iv) was independently sufficient to dispose of the appeal. But the observation is worth making in commentary, because the same approach is observable in a wider class of orders coming up before benches across the country. The residuary head under Section 56 has, in recent years, become something of a default destination for receipts that the assessing officer wishes to tax without the constraints of deductions, set-offs, and exemptions that the proper head would impose. Correct characterisation is a discipline the Act requires, and the value of seeing it affirmed at the assessment stage, rather than corrected several years later in second appeal, is not small. 

V. The contradiction and the limits of what the Tribunal was called upon to decide 

The four-layered error described above is, in one sense, the assessing officer's mistake. There is a further question worth asking, however: why did the appellate machinery, the Dispute Resolution Panel and then the Tribunal, not pause longer over the head-of-income inconsistency? The answer, on a careful reading of the record, lies less in any shortcoming of the appellate forums and more in how the case was put before them. 

The Dispute Resolution Panel's confirmation 

The Dispute Resolution Panel, constituted under Section 144C, is meant to be a forum of substantive review by a panel of three Principal Commissioners or Commissioners, with the power to confirm, modify, or set aside variations proposed by the Assessing Officer. Its directions are binding on the AO under Section 144C(10). The DRP is not a rubber-stamp; in the legislative scheme, it is the principal mechanism by which transfer-pricing and international taxation variations are tested before they crystallise into final assessment orders. 

The Tribunal's account of the DRP proceedings in Valeo Bayen is austere. The order states, at paragraph 3, that “aggrieved by the draft assessment order passed by the A.O, the assessee filed objections before the Disputes Resolution Panel (DRP), who confirmed the addition made by the A.O.” That single sentence is the entirety of the Tribunal's record of what the DRP did. 

A DRP that confirms an assessment without engaging at length with each strand of the AO's reasoning is unusual but not unheard of. It is conceivable that the DRP did examine the head-of-income question and recorded reasoned conclusions simply not reproduced in the Tribunal's order, which would naturally pick up only those parts of the DRP's directions bearing on the grounds of appeal as framed. What can be said with confidence is that the Tribunal's record does not show the DRP engaging in detail with the doctrinal architecture of the assessment, and that the DRP's directions did not draw the Tribunal's attention to the head-of-income inconsistency as a freestanding issue. That is itself a matter of institutional interest, given the role Section 144C was enacted to perform; but it is also entirely consistent with how DRPs have, in practice, tended to operate. 

The Tribunal's recital and its operative reasoning

 The Tribunal noticed the head-of-income contradiction. It is recorded at paragraph 6, in the recital of the assessee's arguments, that the AO “has treated the consideration received by the assessee as income from other sources without considering the fact that the shares of VSIPL held by the assessee are in the nature of capital asset the transfer of which if at all is taxable need to be brought to tax under the head ‘capital gains’”, and further that the AO “has held the same as adventure in the nature of trade though he proceeded to make the addition under head ‘income from other sources’.” 

This is the Tribunal correctly summarising the submission, which identifies the substance of the contradiction discussed in Section IV above. But the recital is in paragraph 6, where the assessee's arguments are set out, while the operative reasoning is in paragraphs 8 to 12; and the contradiction does not appear there as a freestanding ground of decision. The reason is straightforward. In the form in which it was put, the point was deployed in support of the wider argument that the receipt, if at all taxable, ought to fall under capital gains and be exempt under Section 47(iv); it was not pressed as an independent ground that the assessment was unsustainable on its own logic regardless of where the receipt fell. The Tribunal’s powers would have permitted it to take up the point of its own motion, on notice to both sides; but in practice it decides what is argued, and the contradiction, as argued, went to the characterisation of the receipt rather than to the validity of the assessment as a whole. The Tribunal, on the points squarely before it, reached the right conclusion on each. 

Had the contradiction been pressed as a freestanding ground, the appeal could have been disposed of on a short and self-contained basis, before any enquiry into Section 47(iv) became necessary. The argument would have run as follows. The assessment order, on its own logic, contains a contradiction that is dispositive: the Assessing Officer holds the shares to be stock-in-trade, yet charges the consideration under Section 56. Those two characterisations cannot stand together. If the shares are stock-in-trade, the consideration is business income under Section 28; if they are a capital asset, it is capital gains under Section 45, in respect of which Section 47(iv) disapplies the charge. On no reading is it income from other sources. On the Officer’s own characterisation, in other words, the assessment is unsustainable regardless of where the receipt is finally placed; the addition would fall on that ground alone, leaving the remaining issues to be addressed only for completeness. 

Disposing of the appeal in this way would have carried the additional benefit of settling, by way of clear ratio, the proposition that the head of income is a matter of legal classification rather than tactical choice, a proposition that would have closed off a line of assessment increasingly used in transfer-pricing and international-taxation cases. But this was not how the case was argued: the contradiction was deployed in support of the wider Section 47(iv) ground rather than pressed as an independent basis of decision, and a Bench ordinarily decides the case on the ground on which it is asked to decide it. The point survives, accordingly, for a future case in which it is pressed as a freestanding ground. 

What remains for the next case 

Two propositions therefore await future affirmation. The first is that an assessment order cannot, simultaneously and inconsistently, characterise the same receipt as stock-in-trade for one purpose and as residuary income for another. The second is that even within Section 56, the cost of acquisition must be allowed as a deduction under Section 57. Both are well-established in general terms; neither was specifically reaffirmed by Valeo Bayen, because the Section 47(iv) ground rendered them unnecessary to the decision. A future Bench, with these points squarely argued before it, will have the opportunity to set them down in clear ratio. 

There is no criticism in this observation. The Valeo Bayen Bench decided the case as it was put. The wider doctrinal points require to be raised by counsel, in a forum where they can be properly developed on adversarial argument, and there is no shortage of cases coming up in which they will be. 

VI. Two routes, both blessed: the proper analytical frame 

Before turning to the GAAR question, it is useful to set out what the transaction actually was, considered as a piece of corporate planning. 

The Valeo Group held two operating subsidiaries in India: VIPL, a manufacturing company wholly owned by the French parent, and VSIAPL, an aftermarket trading company in which the French parent held 60% from incorporation in 2012 and the balance 40% from June 2018 following the exit of a joint-venture partner. By December 2019, both companies were 100% subsidiaries of the French parent and operated in parallel in India. The group decided to consolidate the two into a single Indian operating entity. 

Indian law offered two principal routes to that consolidation. 

The first was a direct merger of VSIAPL into VIPL, accomplished either through the NCLT-supervised procedure under Sections 230 to 232 of the Companies Act, 2013 or, because both companies were wholly owned by the same parent, through the fast-track procedure under Section 233. In either case, the merger itself would have been tax-neutral. Section 47(vi) would have exempted the transfer of capital assets by VSIAPL to VIPL; Section 47(vii), read with the definition of “amalgamation” in Section 2(1B), would have exempted the extinguishment of the French parent's VSIAPL shares in exchange for new VIPL shares; Section 49(2) would have preserved the cost base; and Section 72A would have governed the carry-forward of losses. The French parent would have ended up with augmented shareholding in VIPL but no liquidity event. 

The second route, the route the group in fact chose, was a sale of the VSIAPL shares from the French parent to VIPL, followed by a fast-track merger of VSIAPL into VIPL. The sale fell within Section 47(iv); the subsequent merger fell within Sections 47(vi) and 47(vii). The end-state was the same as under the direct-merger route, a single consolidated Indian operating company wholly owned by the French parent, but with one significant difference: the French parent received Rs 64.78 crore in cash, a receipt that Section 47(iv) keeps outside the capital gains charge. 

It is to be noted that both routes are tax-neutral routes as expressly provided by Parliament. Sections 47(iv), 47(vi), and 47(vii) are parallel provisions in the same Chapter, addressing different but related forms of intra-group reorganisation. The legislature has not preferred one over the other; it has provided both, recognising that a parent may sometimes wish to consolidate while holding its investment as paper, and sometimes wish to use the consolidation as an occasion to repatriate capital. The Act accommodates both intentions. 

The Assessing Officer's complaint, reduced to its essentials, was that the group ought to have chosen the route that would not have produced cash in the French parent's hands. That is not a complaint about tax avoidance. It is a complaint about a commercial choice between two statutorily blessed alternatives, each with its own integrity. The Tribunal was right to refuse to entertain it.

Figure 2. Two routes to the same corporate end-state, with opposite tax outcomes. A direct merger would have swapped the parent’s VSIAPL shares for fresh VIPL shares, rolling the cost base over under Section 49(2) and deferring, not eliminating, the gain, which becomes chargeable when those shares are later sold. The route actually taken sold the VSIAPL shares to VIPL for Rs 64.78 crore, a transfer that Section 47(iv) deems not to be a transfer, so no capital gains charge arises at all. Both routes realise the India investment; only the merger route carries a tax cost on doing so. 

VII. GAAR was not the missing instrument; it was the wrong one 

We come now to the question the case truly raises, and that the Tribunal does not address because it was not asked to. 

The three possible answers 

Chapter X-A of the Income Tax Act, 1961, comprising Sections 95 to 102, came into force on 1 April 2017 after a long and considered legislative gestation. It empowers the tax authority to declare an arrangement to be an impermissible avoidance arrangement and, on such declaration, to determine the consequences in terms of the Act by, among other things, disregarding, combining, or recharacterising any step or part of the arrangement. The threshold conditions for invocation are substantive: the arrangement must have the obtaining of a tax benefit as its main purpose, and it must satisfy at least one of the four tainted-element tests of Section 96. The procedural protections are equally substantive: invocation requires a reference under Section 144BA to the Principal Commissioner, an opportunity to be heard, and review by the Approving Panel constituted under Section 144BA(15). 

Given that Chapter X-A is now in force, what is the proper role of an Assessing Officer confronted with a transaction that satisfies the conditions of an unconditional statutory relief, but which the Officer suspects of being motivated by tax considerations? The principle is not confined to the exemptions in Chapter III; it applies equally to the exclusions built into the charge itself, such as Section 47, which deems specified transactions not to be transfers so that no capital gain arises under Section 45. In Valeo Bayen the relief was of the latter kind, but nothing turns on the distinction: in each case Parliament has provided that, where the conditions are met, the receipt is not brought to charge, and in each case the only authorised route to displacing that result on avoidance grounds is Chapter X-A. If the Revenue wished to challenge the transaction as avoidance, GAAR was the only permissible route, and on these facts, even that route would have failed. Three answers are conceivable in principle. 

The first is that the AO has no role at all of the kind he undertook in Valeo Bayen. Where Chapter X-A occupies the field of anti-avoidance, the residual common-law and judicial doctrines that previously empowered an Officer to look behind the form of a bona fide transaction, the colourable-device doctrine, the substance-over-form principle in its anti-avoidance application, and the lingering shadow of McDowell in its broader reading, have been subsumed into the statutory scheme. The AO may suspect avoidance, but the only mechanism available to him to act on that suspicion is a reference under Section 144BA. He cannot, of his own motion, undertake an enquiry into commercial substance, business purpose, or alternative routing of a transaction that satisfies such a statutory relief on its terms. To do so would be to bypass the procedural and substantive safeguards that Chapter X-A erects, and to revive in the AO's hands a discretionary jurisdiction that Parliament has consciously taken away. The only qualification to this position, at best, concerns cases where GAAR is unavailable to the tax administration for reasons other than the merits. Where a transaction falls below the monetary threshold for GAAR's application, GAAR simply does not reach it, and whatever anti-avoidance law otherwise applies, including the residual judicial doctrine as confined by Azadi Bachao Andolan and Vodafone, applies undisturbed; on that there is little dispute. Grandfathering stands differently, and on a contested footing. One view treats grandfathering as no more than a choice-of-law rule that preserves the pre-GAAR position, JAAR included, so that a grandfathered transaction remains exposed to the judicial doctrine. The better view, in my respectful submission, and for the reasons developed later in this commentary, is that pre-commencement grandfathering protects the settled expectation of the investor against the new recharacterisation power that GAAR introduced; to allow JAAR to reach the same transaction would be to withdraw indirectly the very protection the grandfathering confers, and would in any event ask of the pre-GAAR doctrine more than, after Azadi Bachao and Vodafone, it was capable of delivering against a genuine transaction. 

The second answer is that Chapter X-A and the residual judicial doctrines coexist as parallel regimes, so that an Assessing Officer who cannot, or does not, proceed under Chapter X-A remains free to mount an ordinary anti-avoidance enquiry under the older doctrines. On this view the enquiry in Valeo Bayen would have been within jurisdiction, and the only question would be whether it was substantively right, which, on these facts, it was not. The difficulty with this answer, developed below, is that it treats the residual doctrine as a reservoir the Revenue may draw on whenever the statutory regime is unavailable or inconvenient, including where its thresholds are not met; and that is precisely the move that allows the Revenue to achieve indirectly what the statute does not permit it to achieve directly. 

The third answer is that there is no residual common-law anti-avoidance jurisdiction in Indian tax law independent of Chapter X-A. The Supreme Court's decisions in Azadi Bachao Andolan and Vodafone International Holdings BV together close the door on judicial recharacterisation of bona fide transactions, and Chapter X-A is the first and only legislative mechanism that authorises such recharacterisation in defined circumstances. On this view, both the AO and the Tribunal were operating in a doctrinal territory that no longer exists. 

Why the first answer is correct 

The first answer is, in my respectful view, the correct one. 

The second answer is incoherent with the structure of Chapter X-A as enacted. Parliament does not build an elaborate substantive and procedural mechanism only to permit its bypass at the discretion of the very officer it constrains. The objection holds whichever trigger is relied on. If the residual doctrine may be summoned merely because the officer finds the statutory route inconvenient, the safeguards are not safeguards but suggestions, to be observed when convenient and ignored when not. And if it may be summoned because the statutory thresholds are not met, the objection is graver still: the officer would then be using the residual doctrine to reach precisely the result the statute, on its own terms, withholds, denying by an uncodified route a benefit the codified regime leaves undisturbed. That is to do indirectly what the statute does not permit directly, and it is the same vice, whether the protection withheld is the failure of a Section 96 threshold or, as in the grandfathering cases, an express exclusion of pre-commencement transactions from GAAR's reach. A regime enacted to discipline the outer limits of judicial anti-avoidance cannot coherently be read to leave those limits exactly where they were, available for use whenever the statutory regime does not deliver the Revenue's preferred outcome. That is not the legislative scheme. 

The third overstates the position by erasing the constitutional foundation on which judicial doctrines of substance over form rested before GAAR. Those doctrines were not free-standing inventions of tax authorities; they were exercises of judicial interpretive power, anchored in the Constitution and developed through case law over decades. They cannot be obliterated by a statute that does not purport to obliterate them. But the third answer's central insight, that GAAR is now the proper home of the substance-over-form enquiry in tax cases, is sound. My position, accordingly, is not that JAAR has been abolished but that it has been confined: it retains its place where GAAR is not in the field, for pre-commencement transactions and matters the statute leaves outside its scope, while ceasing to be available as a free-standing supplement within GAAR's domain. The distinction matters, and I develop it below. 

If the first answer is right, the question whether Valeo Bayen was a fit case for GAAR has a precise answer. It was not. The Revenue did not invoke Chapter X-A, and the AO's enquiry into commercial substance and alternative routing was therefore not merely substantively flawed but jurisdictionally unavailable. The order ought to have been set aside on that ground alone, with the Section 47(iv) analysis appearing only by way of confirmation that the exemption applied on its terms. 

The Tribunal did not reason this way. The order proceeds as though the AO's enquiry were jurisdictionally competent and the only question were whether it was substantively right. Again, that framing reflects how the case was argued: the assessee's defence on Section 47(iv) was, on the facts, more than sufficient to win the appeal, and the assessee had no occasion to put the deeper jurisdictional point. But it does mean the door is left open for the next AO to undertake the same kind of enquiry, secure in the knowledge that, on the Valeo Bayen approach as actually reasoned, the only available defence is a fact-bound demonstration of bona fides. A future case in which the jurisdictional point is pressed squarely is therefore awaited with some anticipation. 

Having said that, I should acknowledge that the question admits of other conceptual approaches; the view I have taken is one of them, and, to my mind, the right one. 

GAAR as sword and shield: why JAAR cannot fill the gap 

There is a further, and to my mind decisive, reason why the first answer must be correct. It draws on the considered architecture of the GAAR debate as it developed in the years preceding Chapter X-A's introduction, and on a structural feature of GAAR that is sometimes overlooked. GAAR is not only a sword in the Revenue's hands; it is also, by design, a shield in the taxpayer's. Once such a regime is in place, it is meant to do both. 

The point was made forcefully by Graham Aaronson KC in the United Kingdom's GAAR Report of 2011, which informed legislative thinking in several jurisdictions, including, indirectly, India. Aaronson observed that the temptation of the judiciary to stretch interpretation in order to achieve what is perceived as a sensible outcome creates uncertainty operating in both directions, sometimes favouring the taxpayer and sometimes harming. A targeted statutory GAAR was recommended precisely as a means of disciplining that stretched interpretation: once the legislature provides a defined anti-avoidance regime with substantive thresholds and procedural safeguards, the case for open-ended additions in the name of an uncodified judicial anti-avoidance rule becomes weak. The Shome Committee's recommendations on Indian GAAR in 2012 struck the same note, observing that GAAR “would also enable taxpayers to plan for a change in the anti avoidance regime that would allow legitimate tax planning... while eschewing dubious tax avoidance arrangements.” The dual function is foundational: GAAR curbs dubious schemes, and it equally allows and protects legitimate tax planning within its framework. 

Once that dual character is grasped, the implications for the present case follow with some force. What GAAR is designed to leave alone, legitimate tax planning that does not satisfy the impermissible-avoidance-arrangement thresholds of Section 96, cannot lightly be recharacterised as abusive by appealing to a more elastic and uncodified version of the judicial anti-avoidance rule (“JAAR”, as it has come to be called in tax commentary). The codification of GAAR was intended to bring predictability and to discipline the outer limits of the judicial doctrine. If the AO, finding that the GAAR thresholds cannot be met, or simply not invoking them, were free to fall back on JAAR to achieve the same result, GAAR's codification would be hollowed out from the day of its enactment. The shield would not exist; the sword would have been replaced by two. 

This is not a hypothetical concern. The point arose, in a different but analytically related form, in the Supreme Court's recent engagement with the Tiger Global line of cases, where one of the questions canvassed was whether JAAR could be deployed to override an express grandfathering of pre-GAAR transactions under Rule 10U. Without resting on any particular reading of where that line of authority finally settles, the structural objection can be stated at the level the present argument needs. The proposition that JAAR could be used to deny benefits that GAAR had been disapplied from withdrawing carries matters rather far: the Revenue would, in effect, be doing indirectly what the legislature has barred it from doing directly. Applied to the present setting, the same logic applies with even greater force. If GAAR cannot be circumvented by JAAR where the legislature has expressly grandfathered transactions out of GAAR's scope, then a fortiori it cannot be circumvented by JAAR, or by any free-standing commercial-substance enquiry of the kind the AO undertook here, where the Revenue has simply chosen not to invoke GAAR at all. 

The structural point bears stating once. Where GAAR applies, GAAR is the instrument; where GAAR could apply but has not been invoked, the absence of invocation is itself a substantive choice with substantive consequences. JAAR, as a residual judicial doctrine, has a role only where GAAR is not in the field at all, in respect of transactions prior to GAAR's commencement, or in domains the statute clearly leaves outside its scope. It does not have a role as a free-floating supplement to be deployed whenever the AO finds the GAAR procedural route inconvenient. 

On this view, the question whether the AO's enquiry in Valeo Bayen was jurisdictionally available has a clear answer. It was not. GAAR was the appropriate instrument for an enquiry of that kind; the Revenue chose not to invoke it; and there is no residual JAAR jurisdiction the AO could have fallen back on to achieve the same result. The enquiry the AO undertook had no doctrinal home at all. The Tribunal's order, by resolving the appeal on Section 47(iv), reaches the correct destination by the most direct route open to it on the case as argued; but a future case in which the jurisdictional point is squarely raised will, I respectfully suggest, find that it has no answer to give the Revenue except the answer the legislature has already given through the architecture of Chapter X-A. 

The counterfactual: a GAAR enquiry that, in my view, would not have succeeded 

It is worth pausing on the counterfactual. Suppose the Revenue, instead of permitting the AO to apply a free-standing commercial-substance test, had made a reference under Section 144BA, and the Approving Panel had taken up the question whether the share-sale-then-merger structure was an impermissible avoidance arrangement. How would that enquiry have proceeded? 

Section 96 defines an impermissible avoidance arrangement as one whose main purpose is to obtain a tax benefit and which satisfies at least one of four tainted-element tests: it creates rights or obligations not ordinarily created between persons dealing at arm's length; it results in misuse or abuse of the provisions of the Act; it lacks commercial substance, in whole or in part; or it is entered into, or carried out, in a manner not ordinarily employed for bona fide purposes. On the Valeo Bayen facts, each limb is, in my assessment, doubtful at best. How the tainted-element tests apply to a given set of facts is, of course, a matter on which reasonable views can differ, and what follows is the view I have reached rather than a proposition I would put beyond argument; reasoning through the limbs in turn, however, I have come to hold it with some confidence.

The “main purpose” test would have required the Revenue to show that obtaining a tax benefit was the main purpose, not merely a purpose and not merely a foreseeable consequence, of the arrangement. The group's stated commercial objective was the consolidation of two parallel Indian operating subsidiaries into one, and that objective was achieved. The tax benefit, the application of Section 47(iv) to the share sale, was a feature of the route chosen; but it is hard to see how it could be the main purpose when the same end-state could have been reached, by a different route, also tax-neutrally. The group did not invent a tax-neutral structure for a transaction that would otherwise have been taxable; it chose between two statutorily provided tax-neutral structures on liquidity grounds. 

The arm's-length limb would have required a finding that the share sale created rights or obligations not ordinarily found in arm's-length dealings. A sale of shares between wholly-owned group entities is structurally a non-arm's-length transaction, but the question under Section 96(1)(a) is not whether the parties were at arm's length but whether the rights and obligations created were of a kind not ordinarily found between arm's-length parties. A sale of shares at a DCF valuation, with consideration paid in cash, withholding tax deducted at source, and the proceeds applied to fund downstream operations, is in no respect structurally unusual. 

The misuse-or-abuse limb would have required a finding that Section 47(iv) was being used for a purpose other than that for which it was enacted. But the provision was enacted precisely to facilitate the movement of capital assets from a parent to its wholly-owned Indian subsidiary, which is exactly what occurred. There is no plausible reading on which the use of the provision in Valeo Bayen could be characterised as a use beyond its intended scope. 

The commercial-substance limb is defined in Section 97 in deliberately precise terms: an arrangement is deemed to lack commercial substance if its substance is inconsistent with its form, if it involves round-trip financing, accommodating parties, elements that offset or cancel each other, or transactions conducted through one or more locations not employed for substantial commercial purposes. None of these applies: the share sale was a real transaction, with real consideration actually paid and remitted between real entities, exhibiting none of the offsetting, round-tripping, accommodating-party, or routing features the section identifies. 

The non-bona-fide-manner limb is the closest the Revenue might have come, but even there, what would the argument be? That the group chose Section 233 because the fast-track procedure was quicker? That is precisely why Section 233 exists. The legislature designed it for wholly-owned-group restructurings, and using it for that purpose is the opposite of non-bona-fide. 

My own assessment, accordingly, is that a GAAR reference on the Valeo Bayen facts would not have succeeded: it would in all likelihood have failed at the Approving Panel stage, and perhaps at the stage of the Section 144BA reference itself. I recognise that others, weighing the same limbs, might come out differently, particularly on the manner in which the arrangement was carried out; but, for the reasons given, I do not think the thresholds are met, and I hold that view with conviction. 

What Chapter X-A was actually enacted for 

The reason a GAAR reference would have failed is the reason the AO's free-standing enquiry should have failed too: the transaction was a legitimate exercise of statutorily provided alternatives by a multinational group consolidating its Indian operations, not the kind of artificial, multi-step, treaty-shopping or value-shifting structure that Chapter X-A was enacted to address. 

Chapter X-A is, by design, a residual instrument. The Shome Committee Report on GAAR (2012), the parliamentary debates around its eventual notification, and the CBDT's own clarifications have consistently emphasised that GAAR is intended to address arrangements that are artificial: arrangements that have the appearance of a commercial transaction but lack its substance, that are constructed primarily to extract a tax benefit not contemplated by Parliament when the underlying provisions were enacted, and that involve elements such as round-tripping, accommodating parties, or opaque routing through low-tax jurisdictions, characteristic of avoidance rather than ordinary commerce. 

The paradigmatic cases that motivated the introduction of GAAR were Vodafone International Holdings (offshore transfer of an offshore holding company deriving its value from Indian assets), the treaty-shopping concerns that drove the Azadi Bachao litigation and its aftermath (use of intermediate jurisdictions to access favourable treaty rates), and the broader base-erosion and profit-shifting concerns the OECD's BEPS project was addressing at the international level. These are arrangements involving multiple steps across multiple jurisdictions, each step justified, if at all, only by its contribution to the overall tax outcome. 

An intra-group share sale between a French parent and its Indian wholly-owned subsidiary, governed by an unconditional statutory relief that Parliament has provided precisely for that situation, is not in the same conceptual universe. It is a single-step domestic transaction (consideration paid by an Indian company to a non-resident for shares of another Indian company) using a domestic provision (Section 47(iv) operates by reference to the Indian residency of the subsidiary). It implicates no tax treaty other than incidentally, no intermediate jurisdiction, no artificial entity, no value-shifting, and no base erosion. It is, in the ordinary sense of the word, ordinary. 

To suggest that GAAR is the appropriate instrument for examining such a transaction is to misunderstand what GAAR was enacted to do. It is not a free-standing anti-avoidance jurisdiction empowering tax authorities to second-guess any taxpayer choice that produces a more favourable result than some alternative. It is a targeted instrument for cases where the artificiality of the arrangement is the doorway to the enquiry. Where the arrangement is not artificial, where it consists of a single transaction using a single statutory provision in the way Parliament expressly contemplated, GAAR has no application, and the question of GAAR invocation does not arise. 

The corollary is the proposition that gives this Section its title. The AO's free-standing enquiry in Valeo Bayen was misconceived not because GAAR would have been a better procedural vehicle for the same enquiry, but because the enquiry itself was one that no instrument, neither GAAR nor any residual judicial doctrine, authorised in the first place. GAAR was not the missing instrument; it was the wrong one. The right instrument was the discipline of statutory construction, applied to a statutory relief that operates of its own force where its conditions are met.

VIII. What survives, what fails, and what waits for the next case 

The result in Valeo Bayen is correct. Section 47(iv) applies on its terms; the conditions are met on undisputed facts; the exemption is unconditional and does not require, as a matter of statutory construction, any further enquiry into commercial substance or business purpose. The Tribunal was right to allow the appeal, and its holdings on the holistic test for asset characterisation, on the armchair-of-a-businessman objection to alternative routing, and on the limits of the AO's power to reject a DCF valuation will each be cited and applied for years. 

What this commentary has tried to suggest is that the case had within it the materials to settle rather more than its reasoning eventually settles, but only if certain points had been put to the Bench more squarely than they were. A distinction is worth drawing here. Some of these were defensive grounds that could only have strengthened the assessee’s position or, at worst, been redundant: the four-layered error in the assessment order, including the head-of-income contradiction, the denial of cost of acquisition under Section 57, and the further point that on the Officer’s own trading premise the business profits of a non-resident without a permanent establishment were not taxable in India at all. These were raised in the appeal memorandum but not pressed as freestanding grounds; had they been, the Bench would have had the occasion to settle a class of issues that recur in transfer-pricing and international-taxation assessments across the country, and the assessee would have risked nothing by pressing them. The jurisdictional question, whether the AO could undertake a commercial-substance enquiry into a transaction protected by an unconditional statutory relief without a GAAR reference, stands on a different footing. It too was implicit in the structure of the assessment; had it been articulated as a ground, the case would have offered the perfect vehicle for a Bench to set down, with the authority of judicial pronouncement, the proposition that GAAR’s codification means what Aaronson and Shome said it would mean. But pressing it would have meant inviting the very GAAR enquiry the assessee had every reason to keep off the field, and a litigant already assured of success on Section 47(iv) cannot be faulted for declining that invitation. The treaty ground was raised but, having been overtaken by the assessee’s success on Section 47(iv), was understandably not pressed at length. Each of these, with respect, awaits another case, though not all of them awaited the same kind of choice. 

The deeper lesson, perhaps, is one for counsel as much as for any Bench. The doctrinal quality of a decision is shaped not only by the Bench that hears it but by the way the case is put before it. A Tribunal whose powers permit it to go beyond the grounds will, in practice, still decide what is argued, and a case argued narrowly will, however well decided, settle only what it is asked to settle. Where a case carries within it the materials to resolve doctrinal questions of wider importance, and where raising them costs the client nothing, there is, in my respectful view, much to be said for pressing them squarely, even when the narrower ground is sufficient for the appeal. The harder judgment is reserved for the question that cannot be raised without inviting a contest the client is winning by avoiding; there, restraint is not timidity but advocacy, and the line between the two is precisely what separates the head-of-income point from the call to invite GAAR. The Bench can decide only what it is given the occasion to decide; counsel’s art lies in choosing, case by case, which occasions are worth creating. 

Valeo Bayen will be cited, and cited for the right propositions. The unfinished business of which this commentary speaks is in no sense the fault of the order itself or of the Bench that pronounced it; it is, rather, a reflection of how the adversarial process plays out when a strong defence on the narrowest ground succeeds at the threshold. The next case in which the deeper questions are pressed will, one hopes, find the doctrinal materials laid out in such a way that the Bench can give the kind of considered judgment the legislative architecture deserves. 

IX. Untested, but not unavailable 

A final caution is worth sounding. That GAAR was not pressed in Valeo Bayen is a fact of this case, not a settled practice to be relied upon. Those who plan consolidations of this kind would be unwise to read a single non-invocation as a lasting dispensation; what one assessment leaves untested, another may not, and the prudent assumption is that the question will one day be asked. The assurance such a taxpayer should seek is not that GAAR will be left unused, but that, if it is invoked, the transaction will answer the impermissible-avoidance tests on their own terms, an assurance that, on facts of this kind, the architecture of Chapter X-A itself supplies.