Abstract

Four decisions upon a single stock option plan have produced three irreconcilable answers, ranging from full taxation as salary to no taxation at all. The profession has read that spread as a disagreement about the character of a receipt, capital or revenue, salary or capital gains. This article suggests that it is nothing of the kind. What flows from a stock option is, in principle, a fruit of employment, and challenging its income character is likely to meet stiff conceptual resistance; what is in issue is the point in time at which it may be charged. Parliament declined to tax the grant, deferred the charge, and then named a single event upon which it would rest, the exercise of the option, with a measure tied to that event. It did not choose conversion into money at large, and it did not provide for an issuer repurchasing an option before exercise, or writing cheques to holders who exercise nothing and surrender nothing. Where the money arrives by those routes, the perquisite charge comes to rest in no man’s land. The article works the provisions to show that this is so, states the case for the Revenue at its highest, and then answers it, not by denying its premises, most of which are accepted, but upon the older principle that a casus omissus cannot be supplied by a court. That principle does not choose sides: an invitation to supply an omission was refused when it came from an assessee, and it cannot be accepted because it now comes from the Revenue. To tax the compensation payment as the deferred fruit of employment, a court would have to supply a taxable event, a measure and a class of persons, which is not the ironing of a crease but the weaving of new cloth. The gap is real, and it is Parliament’s to fill; the article suggests how, and cautions against doing so retrospectively.

I. The story, and the four answers

1. A stock option begins as a promise, becomes a right, ripens into a shareholding, and ends as money. Our law taxes neither the promise nor the right, however confident the expectation or secure the entitlement; it waits, and what it waits for is not realisation at large. It waits for two events which it has named with some care: the exercise of the option, when the employee ceases to hold a right to acquire shares and comes to hold the shares themselves, and the eventual sale of those shares. That choice of moment was neither obvious nor inevitable, as the divided House of Lords in Abbott v. Philbin [1961] AC 352, [1960] 2 All ER 763 (HL) was to demonstrate, and a choice of moment, once made, leaves edges. The four decisions on the Flipkart stock option plan are what happens when the money arrives without either event occurring, and on one plan and substantially one set of facts they have produced three irreconcilable answers, ranging from full taxation as salary to no taxation at all. What these cases show, to my mind, is what happens when the fruits of a stock option are reached without the statutory triggers being pulled, and both sides, taxpayers and Revenue alike, are driven to stretch those triggers to fit events for which they were never designed. The instinct of the profession has been to read that spread as a disagreement about the character of a receipt, capital or revenue, salary or capital gains, and to take sides accordingly. It is, to my mind, nothing of the kind: the four judgments were differing less about the answer than about something anterior to it, which none of them was invited to name, and which, once named, changes both what one may fairly expect of the judicial forum and what one must ask of Parliament. And that takes me to a larger question: what should a court do when a tax may be entirely justified as a matter of policy and yet, for want of a legislative framework, cannot be said to have the authority of law that Article 265 requires? That question has become pressing because of the most recent of these decisions. The decision of the Bangalore Bench of the Income Tax Appellate Tribunal in Pramod Kumar Jain v. DCIT [TS-1163-ITAT-2026(Bang)] was the first to be confronted with a transaction in which the option was actually surrendered rather than merely held, and its answer, that the consideration was taxable but as capital gains, is different in kind from the answers given in the other three. Read together, the four now cover the whole field of possibilities, and what emerges from reading them together is not visible in any one of them alone.

2. The facts, from a height, are these. Flipkart Private Limited, Singapore, framed a stock option plan in 2012 and granted options under it to employees across the group, including employees of its Indian subsidiaries. The options vested in the ordinary way, but the shares were unlisted, and an employee who wanted to turn his options into money had nowhere to sell what an exercise would give him. Money nevertheless reached the option holders twice, by two quite different routes. The first was a buyback: in August and September 2019 the Singapore company offered to repurchase vested options for a price, and those who accepted gave up their options in exchange for the money, the options being extinguished. Whatever else it was, this was an acceleration, for it allowed employees to reap the fruits of the plan without waiting for the shares which the plan promised, and indeed without those shares ever coming into existence. The second was of a different kind altogether. In December 2022, the group announced the divestment of its PhonePe business, from which a substantial part of the value of the Flipkart shares was derived, and the value of the options fell with it, from about USD 189 to about USD 166 per option, on the company’s own reckoning. In April 2023 the company told all option holders that although there was no legal or contractual right under the plan to compensation for the loss in value, or for the future accretion which would now not come, the board had decided of its own accord to pay USD 43.67 for each option held. It was paid on vested and unvested options alike, to those still in service and those long departed, and, with the plan extending beyond employees, to others as well. Nothing was given up: every holder retained every option he had before, and the communication said so in terms. The fine print of the arrangements is not in the public domain, and the position is stated as it emerges from the judgments; but so stated, this was a company doing considerably more than the letter of its scheme required, for reasons which were no doubt commercial and which nobody has suggested were anything other than genuine.

3.In the first case an asset moved against money. In the second, nothing moved except money. In neither did anybody exercise an option or receive a share; that is, in neither did the event on which our law fixes its charge ever occur.

4.Out of these facts came four decisions and three answers. On the compensation payment, the Hon’ble Delhi High Court held in Sanjay Baweja v. DCIT [TS-5269-HC-2024(Delhi)-O], in the only Division Bench ruling in the field, that it was a capital receipt chargeable nowhere. Two months later, a single judge of the Hon’ble Madras High Court touched a different chord, declined to follow the Delhi route, and held in Nishithkumar Mukeshkumar Mehta v. DCIT [TS-6630-HC-2024(Madras)-O] that the receipt was a perquisite taxable in full as salary. A year afterwards, with both judgments before him, a single judge of the Hon’ble Karnataka High Court preferred the Delhi view and followed it in Manjeet Singh Chawla v. DCIT [TS-5466-HC-2025(Karnataka)-O]. On the taxability of the consideration for buying back vested options, a question which arose on a different transaction altogether, the Bangalore Bench of the Tribunal held in Pramod Kumar Jain (supra) that the consideration was taxable, but as long-term capital gains upon the transfer of a capital asset.

5. One plan, one workforce, substantially one set of facts and, for the most part, one set of counsel; and answers ranging from everything to nothing, the last of them upon the compensation payment arrived at with the conflict in full view. When so much is held constant and the outcomes still differ so widely, the interest lies not in the quality of the reading but in the nature of what was being read. Nor is this a boutique controversy. In an economy where public listings are the exception rather than the rule, employee liquidity arrives through buybacks and restructuring payouts, in programmes running to thousands of employees and thousands of crores. Whatever finally emerges from this litigation will govern the default currency of talent in the start-up economy, and it will do so for a class of taxpayers who have no say in how the question is framed and who will, for the most part, simply do what their employer’s Form 16 tells them.

6.It is striking that although each of the four decisions notices, in one form or another, that the statutory event never occurred, none of them asks what follows from that non-occurrence for a court called upon to decide the case, as distinct from what follows for the taxpayer; nor, so far as I have seen, does the debate which has followed them.

II. The proposition, in four steps

7. Let me state the problem in the way I think it should be stated, because the debate has been conducted in a currency which does not quite fit it. Let me break it down into four steps.

8. Step one: what flows to an employee under a stock option plan is, in principle, a fruit of employment. On this there is a remarkable convergence, spanning jurisdictions and generations. In Abbott v. Philbin (supra), where the House of Lords were divided sharply upon the moment of charge, they did not divide at all upon this: the value of the option when granted, said Viscount Simonds for the majority, "is nothing else than the reward for services rendered or, it may be, an incentive to future services"; and Lord Denning, from the other side, described the option as granted by the employer "as a reward or return for his services". The Commentaries on the OECD and UN Model Conventions proceed upon the same footing when they treat a stock option benefit as remuneration attributable to the period of employment which earned it. Our own law has never suggested otherwise. The definition of income in section 2(24) is inclusive and not exhaustive, and the word itself is one of the widest amplitude, to be given its natural meaning unconfined by the enumerated sub-clauses, as the Hon'ble Supreme Court held in CIT v. G.R. Karthikeyan [TS-2-SC-1993-O]; the expression, it has been said in an earlier case, is of elastic import, and its ambit is limited only by what the Act itself provides. It would, in my view, be quite a challenging, and perhaps an uphill, task, to successfully canvass the proposition that what a stock option plan yields is not of an income nature at all; the idea would meet stiff conceptual resistance from every direction at once. There are, however, some nuances on this aspect, and the foregoing statement is not without exception, which I will come to a little later. Once its income character is accepted, two further questions arise, and they are distinct: under which head such an income falls to be charged, whether as salary under section 15 read with section 17 or, failing that, under the residuary head; and, no less importantly, at what point of time it is to be charged. The second of these is the subject of this article, and it is the one which our legislature has answered in a particular and, as will appear, a rather exacting way.

9. Let me point out that two things are easily conflated at this stage, and it is worth separating them. The nature of an asset, whether capital or not, does not determine the character of a receipt, in either direction. If an employer gifts a house to an employee, the house is a capital asset in the employee's hands and the receipt is a revenue receipt in the same breath; and the converse holds equally, for the employment origin of a holding does not stamp every later payment referable to it as remuneration. Which receipt is under consideration, and what it was for, are questions that have to be asked separately each time.

10. It is equally important to bear in mind the fact that the earnings from a stock option have two components. The first is the reward for employment: the advantage of acquiring shares at a price fixed when the option was granted. The second is the return upon holding, the accretion in the value of what the employee comes to own after he owns it. The first does not lend itself to measurement while the option is merely held, for its worth is the difference between the price at which the option permits him to buy and the price the security commands, and that difference goes on changing with every day that passes; the employee may exercise tomorrow, or in five years, or never, and each course yields a different figure, so that until he acts there is no final sum to speak of. The second is easily computed, being the difference between the price at which the share is acquired and the price at which it is sold. Our statute recognises both and treats them differently. Section 17(2)(vi), read with Expl (c), charges the first as a perquisite, measured as the fair market value of the security upon the date of exercise reduced by the amount paid; section 49(2AA) then carries that same value forward as the cost, so that section 45 charges the second, upon sale, as capital gains, and nothing is counted twice. The line between them is drawn neither at vesting nor by the passage of time. It is drawn at the exercise: everything accruing up to that moment is remuneration by statutory fiat, whatever its economic character, and everything after it is capital. The two are relatable to the same instrument and must not be conflated, for they answer to different heads, different measures and different moments.

11. With those confusions cleared away, I return to the sequence. Step two: one view, and on the face of it the natural one, is that the fruit should be taxed when it is received, which is at grant. On that view, what happens to the value of the option afterwards is a separate matter altogether, and any movement in it falls to be dealt with on its own footing. It is worth asking what the first principles have to say, and Abbott (supra) repays a second look, because the division there was upon the very question with which this article is concerned. The majority held the option itself to be the emolument, to be valued at the date of grant, with everything thereafter falling on the taxpayer's own side of the line as the growth of his property; the minority would have waited for the exercise of the option, holding that until the shares are taken up the employee has nothing in hand. I dwell on the division because it makes a point easily lost in the present debate: the moment at which an equity-linked reward is taxed is a matter of choice and not of nature, and judges of the first rank have differed upon it. Our own law has settled where the minority would have placed it, though not, as will appear, for the minority's reasons.

12. Step three: our legislature, in its wisdom, declined to tax at grant. The grant, therefore, is not a taxable event at all, whatever the value of what is granted. The charge was deferred, and one can see why, for the value of an option at grant is a matter of estimation upon which reasonable people may differ, and estimation is an uncomfortable foundation for a charge; and having deferred it, the legislature was obliged to say to what moment. It did not choose conversion into money at large. It fixed upon a defined event, and said so in section 17(2)(vi) read with Explanation (c): the event is the exercise of the option, and the measure is the fair market value of the security on that date as reduced by the amount paid or recovered. Section 49(2AA) completes the architecture by carrying that same value forward as the cost of the shares, so the eventual sale is charged under section 45 on the subsequent appreciation alone, and nothing is taxed twice. It is a complete and coherent scheme, and it provides for exercise and for nothing else. What happens, then, when the fruits of the grant are reaped otherwise than by exercising the option, is a question to which the provision returns no answer.

13. Step four, and here is the whole difficulty: in specifying that moment, Parliament contemplated exactly one route by which the employee would come to hold the shares and, in due course, their proceeds, namely exercise, then allotment, then sale. It did not contemplate an issuer buying the option back before exercise, and it did not contemplate an issuer writing cheques to holders who exercise nothing and surrender nothing. Consider the position which results. When the employee receives the grant, no charge arises, for the legislature has deliberately declined to tax at that point, and the door to taxability in that year is closed. When he afterwards reaps the fruits of the grant otherwise than by exercising the option, no charge arises under clause (vi) either, for the event upon which that provision depends has not occurred, and the door in that year is closed as well. The perquisite charge, in such a case, comes to rest in no man's land. There is, however, a distinction to be kept in view between income by way of perquisite, which is the fruit of employment, and the subsequent accretion in the value of what the employee comes to hold, which is capital in character; I return to this a little later.

Figure 1: The route Parliament provided, and the route by which the money actually reached the employee.

14. Before coming to that history, a word about the decision which preceded it. Before 1999, there was no specific provision at all, and taxability rested on the general concept of a perquisite; that is how the question came before the Hon'ble Supreme Court in CIT v. Infosys Technologies Ltd. [TS-62-SC-2008-O]. The Court held that an option is a right without obligation to buy, so that no perquisite could be said to accrue when the warrants were granted, and none when the options vested; and that upon exercise, the shares being subject to a lock-in, stamped non-transferable and incapable of being pledged, there was no cash inflow to the employees, and the benefit was only a notional one whose value was unascertainable. It is right to say at once what that decision does not do for the present question. It was a case in which the statutory event had occurred, for the options were exercised and the shares allotted, and the charge nevertheless failed because the value of what the employee received could not be ascertained. That reasoning can have no application where the employee has received money, for money is its own measure and no question of ascertainment arises; and to the extent that reliance is placed upon Infosys (supra) as authority for non-taxability in the cases before us, the reliance is, to say the least, questionable.

15. What the decision does establish, and what does apply, is the proposition upon which the Court rested at the end. Proceeding upon the footing that there was a benefit, it asked whether every benefit received by a person is taxable as income, and answered that it is not; that unless the benefit is made taxable it cannot be regarded as income; and that in the absence of legislative mandate a potential benefit could not be treated as the income of an employee chargeable under the head salaries. The two propositions are halves of a single requirement. A charge needs an event, and it needs a measure. Infosys (supra) was a case in which the event was present, and the measure was wanting; the cases before us are the converse, and the converse fails for the same reason. The point may be tested simply. Suppose an option lapses unexercised and the employer, out of goodwill, pays the employee a sum equal to what he would have made had he exercised it. The receipt is money and its quantum is beyond dispute, yet nobody would suggest that clause (vi) applies; and the reason is not that the sum cannot be valued but that there is no specified security, no allotment or transfer, and no date of exercise from which Explanation (c) may take its measure. The existence of a receipt does not supply the occurrence of an event.

16.The history which followed is the best evidence of what the legislature was doing. Until 1999 there was no specific provision at all, and the charge rested upon clause (iii) of section 17(2), which reaches only specified employees, namely a director-employee, an employee having a substantial interest in the company, and an employee whose income under the head salaries exceeds the prescribed limit; the executive filled the space as best it could through Circular 710 of 1995, which valued the benefit by reference to the date upon which the employee accepted the offer. It was in this backdrop, and for the years preceding these amendments, that Infosys (supra) was decided. The Finance Act, 1999 then inserted clause (iiia), brought all employees within the charge, and fixed the spread upon exercise as the measure. The Finance Act, 2000 withdrew it and moved qualified plans to taxation at the sale stage alone. The Finance Act, 2007 shifted the burden altogether to the employer under fringe benefit tax at vesting-cum-allotment, with section 49(2AB) supplying the cost. The Finance Act, 2009 abolished fringe benefit tax and inserted the present clause (vi) with effect from 1 April 2010, restoring exercise and allotment as the employee-level event. Four interventions in a single decade, and on not one occasion has the option as such been taxed; the earliest charge in the employee's hands has always been fixed at exercise or allotment. Whatever else the legislature has done in this field, and it has done a good deal, it has consistently thought in terms of a defined event upon which the charge is to rest, and not in terms of taxing value whenever it happens to be realised.

17. That, then, is not a doctrinal disagreement at all. One view, and that is the view this analysis takes, is that it is a legislative vacuum, and once that is the basis to be proceeded with, the four judgments look quite different. The fora were not at variance upon what the statute says. The difference between them lay in what a court is to do where the statute says nothing, and each of them approached that difficulty differently. The Hon'ble Madras High Court read the words of the deeming provision as wide enough to reach a case for which they were not written. The Hon'ble Delhi High Court and the Hon'ble Karnataka High Court declined to read them so, with the consequence that the receipt fell outside the net altogether. The Bangalore Bench of the Hon'ble Tribunal found a route that fitted, because on its facts an asset had genuinely moved. Seen this way, the real question in the field is not definitional. It is the oldest question in statutory interpretation: what is a court to do when it is satisfied that the legislature would have provided for a case had it thought of it, and equally satisfied that it did not think of it? The other view, of course, is that the deferral embedded in the scheme of section 17(2)(vi) postpones taxability only until the point at which the benefit becomes measurable in monetary terms, and that a provision framed to defer a charge cannot be construed so as to convert it into an exemption; a construction which draws some support from the very indeterminacy I have described, for once a precise sum has reached the employee the reason for waiting has spent itself. That is the implicit, though not fully articulated, conceptual foundation of the judgment of the Hon'ble Madras High Court, and I deal with it, and with its sustainability in law, a little later in this analysis.

III. Working the provisions

18.Before coming to what a court may do about a gap, it is necessary to establish that there is one, and that requires the provisions to be worked rather than asserted. I take clause (vi) first, then the arguments which have been or may be built upon it, then the two transactions separately, and finally the other heads under which the Revenue may seek shelter if clause (vi) fails.

19. Clause (vi) of section 17(2) charges, as a perquisite, "the value of any specified security or sweat equity share allotted or transferred, directly or indirectly, by the employer or former employer, free of cost or at a concessional rate, to the assessee". Explanation (a) tells us what a specified security is, borrowing the meaning of securities from section 2(h) of the Securities Contracts (Regulation) Act, 1956 and adding that where an employee's stock option has been granted under a plan, the expression includes the securities offered under such plan. Whether the option itself answers that description is a question I take up separately, for it is upon that single word, offered, that the contrary view is built. Explanation (b) defines sweat equity shares, which these plainly are not. The critical provision is Explanation (c), which supplies both the measure and the timing of the charge: the value of the specified security is its fair market value on the date the option is exercised, reduced by the amount actually paid by or recovered from the assessee. Explanation (d) defines fair market value, and Explanation (e) defines an option as a right, but not an obligation, to apply for the specified security at a predetermined price. The charge, therefore, has four ingredients: a specified security; its allotment or transfer to the assessee; the absence of full consideration; and a value ascertainable in the manner Explanation (c) prescribes.

20. Let us turn now to the Pramod Kumar Jain (supra) decision, and see the practical application of these tests. That was a case in which the option holder gave up his options and received a price. No security was allotted or transferred to him, whatever passed from him, and there was no exercise date from which Explanation (c) might take a value. In paragraphs 15 and 16, the Tribunal applies these tests one by one and forms the considered view that it is "only after the option granted to the employees under the stock options plan is exercised that the incidence of taxability under section 17(2)(vi) arises". The inquiry did not stop there. The Tribunal put a question to itself, namely that "since the consideration received by the assessee upon repurchase of vested options by the FKS is not taxable under the head 'Salaries', the question arises under which head of income the same is taxable", and answered it by holding that the option rights were capital assets and that "the repurchase (i.e. buyback) falls within definition of 'transfer' under section 2(47)", so that the consideration for the buyback of the options was chargeable as capital gains. The approach was clinical and the outcome, in my humble understanding, unexceptionable. It is worth noticing, too, that this was the only one of the four matters to come by way of a regular appeal upon a completed assessment, with the scheme documents and the letters of offer upon the record; the others came as writ petitions at the withholding stage, and the difference in the quality of the material available shows.

21. The rationale of that analysis casts light upon the compensation transaction as well, with which the Hon'ble Delhi, Madras and Karnataka High Courts were dealing. The employees held rights to subscribe to shares at a specified price, and the monetary advantage inherent in doing so; and upon Flipkart's divestment of PhonePe, the value of those rights was considerably reduced. There cannot be, and there is no, dispute that these rights had all the attributes of a capital asset. The receipt of USD 43.67 for each option was thus plainly referable to a capital asset. Now, a capital receipt enters the definition of income only through section 2(24)(vi), as a capital gain chargeable under section 45, and section 45 comes into play only where gains arise from the transfer of a capital asset. No transfer occurred upon the compensation payment. It is the transfer which triggers the charge, and section 2(47) requires the sale, exchange or relinquishment of the asset, or the extinguishment of any rights in it; a diminution in the value of an asset which the holder continues to hold is neither. Test that payment against the same four ingredients, and the result is the same as before. No security of any description was allotted to the holder or transferred to him; he held the same options after the payment as before, and the communication said so in terms. What he received was money, and money paid by an employer is not the value of a specified security allotted or transferred to him; it is money. Nor can Explanation (c) be worked, for there is no date of exercise from which a fair market value may be taken, and no amount paid by the assessee to be deducted from it. Three of the four ingredients fail, and the third, the absence of consideration, is the only one satisfied. What follows is that a receipt of money referable to stock options does not, by that fact alone, answer the description of a perquisite under section 17(2)(vi); and that a capital receipt, without a transfer under section 2(47), does not attract section 45 and so does not enter section 2(24)(vi) either. Where there is no transfer, the same analysis which sustained the charge in Pramod Kumar Jain (supra) yields the opposite result, and the charge fails at the threshold.

22. One of the pleas taken by the taxpayers in the compensation cases is, however, worth setting out, for it carries a risk of its own. It was urged that the payment was a one-time voluntary payment unconnected with the employment. If that is so, it must equally be unconnected with the options, for the options themselves came to the holder by reason of his employment and by no other route; and a payment unconnected with the options can hardly be characterised as compensation referable to a capital asset. The nexus between the options and the employment is unmissable, these options being the fruits and rewards of employment, and a plea which denies all connection with employment therefore proves too much: pressed to its conclusion, it unsettles the very characterisation it was deployed to support. And there is more. A receipt described as voluntary and made without obligation invites the residuary head under section 56(1), if the receipt is income in nature, and section 56(2)(x) whether it is or not, that provision charging sums of money received without consideration by deeming and not by description. Both of those propositions require an answer, and I come to them a little later. What the episode illustrates, and it is an illustration to which I shall return, is how a vocabulary adopted for one purpose can defeat another.

23. It is right to add a word about the reasoning of the Hon’ble Karnataka High Court, which travels, on the face of it, somewhat further than the conclusion required, particularly upon the ascertainability of cost; but the conclusion is in any event justifiable upon narrower ground, and I do not think it necessary to examine that aspect any further.

24. The position at this stage may be summarised. Under the law as it presently stands, consideration received upon the surrender or repurchase of a vested option may be charged as capital gains, because an existing capital asset has been transferred. A payment made merely because the economic value of an option has diminished, while the option itself remains wholly intact, presents a different case altogether. Section 17(2)(vi) neither identifies a taxable event nor supplies a workable measure for such a payment; section 45 fails for want of a transfer; and neither section 56(1) nor section 56(2)(x) makes good the deficiency, for the reasons I come to presently, though I record that neither has been tested in these cases. The economic case for taxing such a receipt may be a strong one. But an economic justification cannot supply the elements of a charge which Parliament has omitted to provide.

IV. The case for the Revenue, stated at its highest

25. Stated compactly, then, the Revenue's case rests upon five propositions- whatever be their sustainability in law, and it is worth setting them down in the order in which they must be established, for each depends upon the one before it. First, that the option is not a free-standing piece of property but an incident of the employment or service relationship, so that everything which flows from it flows, in the last analysis, from that relationship: the definition in section 2(37) of the Companies Act, 2013 says as much, both sides in Abbott (supra) treated the option as a reward for services, the model commentaries treat the benefit as employment remuneration, and section 2(24) is of the widest amplitude. Second, that an unexercised option is not a capital asset, being no more than a contractual right to receive property later, subject to vesting, cancellation and exercise, so that the capital receipt characterisation is unavailable and the door to sections 45 and 2(24)(vi) is locked before section 17 is reached; and that the line of authority upon compensation for the sterilisation of a profit-making apparatus, from Kettlewell Bullen to Oberoi Hotel and Saurashtra Cement, does not assist, for that line concerns a subsisting source of income, and an option which has never become a share is not such a source. Third, that clause (vi) nevertheless reaches the payment, because Explanation (a) says includes and not means, because Parliament wrote offered and not allotted, and because the charge extends to benefits conferred directly or indirectly, a cash payment which restores the impaired value of an offered security being an indirect conferment of that value. Fourth, that Explanation (c) is the ordinary measure for the ordinary case and not a condition precedent, so that its inapplicability to an extraordinary cash payment does not defeat a charge where the value is already known to the rupee; that Infosys, so far from assisting the taxpayer, is against him, for there the difficulty was that a benefit under lock-in could not be valued, whereas money is its own measure; that CIT v. B.C. Srinivasa Setty [TS-2-SC-1981-O] is a capital gains authority which has no application where there is neither a capital asset nor a transfer; and that the arithmetic, crude as it may appear, follows from the facts, for the holder paid nothing for the options and gave up nothing to receive the money, so that there is nothing to be netted off and the entire receipt is the benefit. And fifth, that if clause (vi) should fail, the receipt does not thereby escape, for it remains a profit of employment under section 15, or income under the residuary head in section 56(1), or a sum of money received without consideration within section 56(2)(x)- as the assessee, as recorded in the arguments raised before the Hon’ble High Court, claimed it to be a gratuitous payment unconnected with the employment.

26. Three further observations complete the case, and they are not the least of it. The first is that the taxpayer's own pleading assists the Revenue at several points. A payment said to be voluntary, made without obligation, and unconnected with the employment is a payment which invites the deeming in section 56(2)(x); and a payment truly unconnected with the employment cannot easily be connected with the options either, the options having come by no other route. The second is that the payment was made upon unvested options as well as vested ones. That is very difficult to describe as compensation for the impairment of a perfected capital asset, for an unvested option is not yet even an enforceable right; it is far more readily described as a distribution to those who were upon the plan by reason of their relationship with the group, which is to say, as a benefit of that relationship. The third is procedural. Three of these matters arose upon applications for a certificate of nil deduction, where the question is not whether the receipt is ultimately chargeable but whether the payer may safely be told that it is not. Where a genuine controversy exists as to character, the Revenue may say, a certificate of nil deduction is not a matter of right; and the Hon'ble Madras High Court, it will be noticed, affirmed the refusal of the certificate as a corollary to its conclusion upon the merits, while holding in terms that the basis upon which the impugned order had proceeded was itself flawed.

27. That is the case against the taxpayer, put as high as I can put it. I have set it out at this length for two reasons. The first is that it has never been assembled in this form, either by the Revenue in these proceedings or in the debate which has followed them, and an argument which has not been assembled cannot be answered. The second is that its answer does not lie in denying its premises, most of which I accept. It lies in a distinction which the case, for all its force, never confronts: the distinction between what a statute was meant to achieve and what a court may do to achieve it when the statute has not said so. To that I now turn.

V. Casus omissus: a principle which does not choose sides

28. The answer to that case does not lie in disputing its premises. It lies in a principle older and sterner than any of the propositions upon either side of this debate, and I approach it with a certain diffidence, for I had occasion to consider it judicially many years ago, in Tata Tea Ltd. v. JCIT [TS-5438-ITAT-2002(Kolkata)-O]. The question there was whether an amendment extending a tax holiday from five years to ten could be read as reaching units whose five years had already run, and the invitation to supply the omission came from the assessee. It was refused. Two passages from that order carry the authorities, and they read as follows: Casus omissus, which broadly refers to the principle that a matter which has not been provided in the statute but should have been there, cannot be supplied by us, as, to do so will be clearly beyond the call and scope of our duty, which is only to interpret the law as it exists. Hon’ble Supreme Court, in the case of Smt. Tarulata Shyam v. CIT [TS-14-SC-1977-O] at page 356 has observed: "We have given anxious thought to the persuasive arguments..., (which) if accepted, will certainly soften the rigour of this extremely drastic provision and bring it more in conformity with logic and equity. But the language of sections... is clear and unambiguous. There is no scope for importing into the statute the words which are not there. Such interpretation would be, not to construe, but to amend the statute. Even if there be a casus omissus, the defect can be remedied only by legislation and not by judicial interpretation... It will be well to recall the words of Rowlatt, J. in Cape Brandy Syndicate v. Inland Revenue Commissioners [1921] 1 KB 64 (KB) at page 71, that: ‘...in a taxing Act one has to look at merely what is clearly said. There is no room for any intendment. There is no equity about a tax. There is no presumption as to a tax. Nothing is to be read in, nothing is to be implied. One can only look fairly at the language used.’" As for the Lord Denning’s observations in Seaford Court Estates Ltd. v. Asher [1949] 2 All ER 155 (CA), which have been heavily relied upon by the learned counsel, the House of Lords itself, in a later judgment in the matter of Magor & St. Mellons Rural District v. Newport Corpn. [1951] 2 All ER 839, did not approve the proposition advanced by Lord Denning... Lord Simonds, supporting the majority view..., unequivocally and categorically rejecting Lord Denning’s theory on the relevance of intent of Legislature: "...What the Legislature has not written, the court must write, and fill in the gaps. This proposition... cannot be supported... It appears to me to be naked usurpation of Legislative function in the thin guise of interpretation and it is less justifiable when it is guesswork with what material the Legislature would, if it had to discover the gap, have filled it in. If a gap is disclosed, the remedy lies in an amending Act..." Lord Morton observed that "These heroics are out of place"...

29. The authorities collected there have lost nothing in the intervening years. The proposition approved in CIT v. National Taj Traders [TS-5033-SC-1979-O], that a case not provided for in a statute is not to be dealt with merely because there seems no good reason why it should have been omitted, and that a casus omissus cannot be supplied except where the reason for it is found within the four corners of the statute itself, remains the rule. So does the holding in Petron Engineering Construction (P.) Ltd. v. CBDT [TS-5023-SC-1988-O] that an omission in respect of a matter whose provision may have been desirable, but which the legislature has not made, is not a defect which the mode of construction Lord Denning advocated can cure. And our own Supreme Court has echoed Lord Simonds more than once, in State of Kerala v. Mathai Verghese AIR 1987 SC 33, in Jumma Masjid v. Kodimaniandra Deviah AIR 1962 SC 847 and in Punjab Land and Development Corporation v. Presiding Officer (1990) 3 SCR 111.

30.  I set out that passage for a reason which goes beyond the authorities it collects. In Tata Tea (supra), as also in a number of cases cited therein, the invitation to supply the omission came from the assessee, and it was refused. Here the invitation comes from the Revenue. A rule of construction which varies with the identity of the party invoking it is no rule at all; and if a court may not read words into a taxing statute to relieve a hardship, it may not read them in to prevent an escape either. The traffic runs both ways. Where the law has missed something out, however logical its inclusion might have been, it is not for the judicial forums to supply the deficiency at the instance of either side.

31. On a different note, there is a subtle irony worth noting. Lord Denning was the most widely quoted English judge of his century, and he remains among the most quoted in our own courts; and yet, upon the questions which mattered most to him, he was very often in a minority among his own colleagues. It was his approach in Seaford Court Estates (supra) which the House declined to follow, even though he was in the coram, in Magor (supra). It was he, again, who was in the minority in Abbott (supra), nine years later, upon the very question with which this article is concerned. And it is his timing rule, not the majority’s, which our own Parliament has since adopted. There is a lesson in that sequence which is worth more than the anecdote: an answer may be right, and the method by which a court reaches it may still be beyond its office. That is the whole of what follows.

32. And Lord Denning’s metaphor contains its own limit, which is almost always quoted without it. A judge must not alter the material of which the Act is woven, but he can and should iron out the creases. The question in any given case is therefore not whether the judge may act, but whether what is asked of him is the ironing of a crease or the weaving of new cloth.

VI. What the court is being asked to supply

33. Consider, then, what a court must supply in order to tax the compensation payment as the deferred fruit of employment. It must supply a taxable event, for the payment is not the monetisation of the option: the holder holds every option he held before, and the plan continues to bind him. It must supply a measure, for the statutory formula is inoperable where there is no exercise date and no price paid. And it must supply a class of persons, for the same payment upon the same terms went to those who had left the company years earlier and to others who were never its employees at all, who have no subsisting relationship of employment upon which a charge to salary might be hung. A court which supplies the event, the measure and the class has not ironed a crease. It has woven new cloth. The requirement of Article 265, with which I began, is that all three components of a charge, the subject, the person and the measure, must come from the statute; and the Hon’ble Supreme Court said so in terms in Govind Saran Ganga Saran v. CST [TS-5014-SC-1985-O] and in Mathuram Agrawal v. State of Madhya Pradesh (1999) 8 SCC 667, the former being the decision upon which Infosys (supra) itself proceeded.

34. There is a further difficulty with the pragmatic thesis, and it is fatal upon its own terms rather than upon any principle of restraint. That thesis, at its highest, is that the option was the emolument and that clause (vi) merely postpones the recognition of a charge which had already attached. If that is so, the charge attached at grant, and the year of charge is the grant year. Every assessment in these cases was made for the year of the cheque. Upon the thesis which alone supports them, therefore, those assessments are for the wrong years, and the right years closed long ago; a timing argument which proves the assessment year to be wrong cannot sustain the assessment, it destroys it. I put the point no wider, for the other routes to taxability, whether under section 56 or upon the construction favoured at Madras, would each operate in the year of receipt. It is the deferral thesis, and that thesis alone, which carries this consequence. Nor is the difficulty a new one. In Abbott (supra) itself Lord Radcliffe began by observing that it is a natural enough assumption for the tax gatherer that if a transaction does not attract tax in one year it must in another, and that he did not regard that as a good general principle upon which to found the construction of an income-tax code. The observation has lost nothing with age.

35. As indicated earlier, let me now turn to Explanation (c) as a deferral spent once the benefit becomes measurable. The construction has a certain symmetry, and I have not understated it. But it asks the reader to accept that a provision which names an event is really about the availability of a figure, and that once a figure exists the event may be dispensed with. That cannot be right, and the case of the lapsed option shows why. If an option lapses unexercised, no charge arises, though the value of the benefit foregone could be computed to the rupee upon the day it lapsed. Measurability was never the condition; exercise was. Explanation (c) does not say that the perquisite shall be valued when a value becomes available; it says that the value of the specified security shall be its fair market value upon the date on which the option is exercised. Take away the exercise, and there is no date, no security, and nothing to value. What is left is not a deferral that has spent itself, but a charge that has never arisen.

36. Let me now turn to the residuary provisions, and the answer turns upon a distinction which is easily missed. Emil Webber v. CIT [1993] 200 ITR 483 (SC) is authority that income retains its natural connotation, that a payment integrally connected with a taxable stream may be income though made by one who is not the employer, and that where the salary head fails because the payer is not the employer, the income falls to be charged under the residuary head. That is right, and it governs a particular situation: a receipt which is income by nature but which fails a specific head for want of a condition personal to that head, such as the identity of the payer. Such income is orphaned, and section 56 receives it. It is quite another thing when a receipt belongs by its nature to a head that has considered it and released it, i.e., when an income is taxable under a head but is not taxed under that head because the conditions precedent for such taxation are not satisfied. Such a receipt is not orphaned but released, and it cannot be recycled through the residuary head; that is the principle of Nalinikant Ambalal Mody v. S.A.L. Narayan Row [TS-1-SC-1966-O], of CIT v. D.P. Sandu Bros. Chembur (P.) Ltd. [TS-5-SC-2005-O] and of Cadell Weaving Mill Co. (P.) Ltd. v. CIT [TS-4-HC-2001(Bombay)-O]. A payment referable to the capital value of an asset, which the capital gains head declines to charge for want of a transfer, falls in the second class and not the first. The release I rely upon is the release by the capital-gains head. Clause (vi) never described this payment, and I do not treat it as having considered and released what it did not mention. The payment is referable to a capital asset which the holder continues to hold; section 45 looked at that class of receipt and declined to charge it in the absence of a transfer. That is why section 56(1) cannot be asked to finish the work.

37. As for section 56(2)(x), which charges a sum of money received without consideration, the answer is shorter, and I confine it to a single ground. The payment was not a gratuitous transfer of wealth. It was made upon a published formula, at a uniform rate for each option held, to several thousand counterparties, and it was made against the options as held and by reason of their impairment. A provision which descends from the gift-tax measures, and which is directed at bounty, does not reach a commercial payment of that character. I do not rest this upon the employment nexus, for to say that the money was consideration for services would be to reopen the head which has already been closed; the ground is simply that a payment computed upon and referable to the holder’s own asset is not a payment for nothing. I record, nonetheless, that the provision was not invoked in any of these matters and that the point is untested; and that the vocabulary of voluntariness, adopted by the taxpayers to defeat the salary head, is what makes the argument available at all.

VII. The limits of the policymaker’s perspective

38. What then of the policymaker’s perspective, upon which I have relied elsewhere and which the Revenue’s case may fairly invoke? I do not think the answer is to reject it. The answer is to notice its domain. Neil Brooks has argued, persuasively to my mind, that legislative intent in tax statutes is largely a fiction, that legislatures do not in any meaningful sense intend the detail of a taxing enactment, that the intent which exists is the policymaker’s, and that where two readings are open the reading consonant with the policymaker’s perspective should be preferred. That is a tie-breaker among available constructions. It presupposes that the words will bear both readings and asks which to choose. It is not, and was never offered as, a licence to construct a charge where the words bear no reading at all for the transaction in hand. The distinction between ambiguity and omission is the whole of the matter. Where there is ambiguity, the policymaker’s perspective is a legitimate and valuable guide. Where there is omission, as in this case, it is an argument for amendment addressed to the wrong forum.

39. I hold to this even though I have argued elsewhere that judicial decisions have a normative effect which our system captures imperfectly. A ruling which no policymaker could have imagined and no taxpayer honestly expected will not be absorbed into policy; it will be reversed, and sometimes retrospectively. But the corollary matters here. If the system's answer to a decision which departs from the policymaker's perspective is legislative correction, then legislative correction is equally the system's answer to a gap which the policymaker never filled. The feedback loop runs in that direction by design, and the court is not meant to pre-empt it by supplying the amendment itself. When courts do venture into that space, the consequence is not certainty but its opposite, for what seems to one judge an obvious crease seems to another an alteration of the material, and taxpayers are left to guess which bench they will draw.

VIII. Where this lands

40. It remains to say something about where this is likely to go. The proximate hinge is procedural. On the judicial front, the Hon’ble Madras High Court judgment is in appeal before the Division Bench, and it will be interesting to see how things unfold in the courtroom. However, my deeper expectation is that the terminal event will be legislative either way, because a vacuum of this kind, once named, is an invitation, and because every escape of this shape in our history, self- generated goodwill, tenancy rights, has ended in a deeming amendment.

41. If Parliament is to write, let it write well, and the materials lie ready upon its own statute book. The clean rule is a deeming provision treating every option-linked realisation, whether by repurchase, surrender or cancellation, and, if the residual gap is to be closed, payments referable to option holding, as capital gains, with cost deemed nil save to the extent of amounts actually paid or already taxed as a perquisite. Section 46A does that work for a company’s purchase of its own specified securities: it deems a transfer and supplies a computation where the general charge would otherwise fail for want of an event. What is wanted here is that structure, and not that subject, close the narrow remaining vacuum, preserve the exercise-stage perquisite intact, and tax the whole of the embedded reward, since a nil cost leaves nothing untaxed. The rougher alternative, a widened perquisite entry taxing pre-exercise cash at slab rates, is a policy choice Parliament is entitled to make.

42. I would add one institutional observation, since I have been arguing elsewhere for a better mechanism of advance certainty. This controversy ought never to have required four rounds of litigation across three High Courts and a Tribunal. A single scheme, a single corporate event, thousands of identically placed taxpayers and a question of pure characterisation is the paradigm case for a prospective ruling or a clarificatory circular, obtained before the money moved rather than litigated for three years afterwards. That the system had no way of delivering that answer is a comment not upon the judges, who decided what they were asked to decide, but upon the architecture within which they were asked to decide it.

IX. The unfinished business

43. Strip this controversy to its studs, and the disagreement was never about whether a fruit of employment is income. It always was. The disagreement is about a gap. Parliament chose to tax the fruit only when it turns into money, described one way in which that happens, and left the other ways unaddressed. Two of the four fora filled the gap by construction, in opposite directions; one found a route that fitted, because upon its facts an asset had genuinely moved. The remaining space is not a puzzle for judges. It is an item of unfinished business, and the law has long known what to do with unfinished business. There is no equity about a tax, no presumption as to a tax, and no casus omissus for a court to supply, however inconvenient the silence may be. The charge, let it be said plainly, was never abolished. It was made to wait for an event; and where that event does not come, and cannot come, the trigger stands deferred into oblivion. That is a state of affairs which Parliament may repair whenever it chooses, and which a court may not.

44. There remains the question of retrospectivity, if and when an amendment comes, and I would enter a word of caution. Retrospective correction is at its most defensible where a court has read a provision otherwise than the policymaker intended, for there an intention existed and the amendment merely restores it. This is not such a case. Here the policymaker formed no intention about the transaction at all; the design assumed a single route by which an option turns into money, because that was the route everyone then knew, and the assumption, entirely reasonable when it was made, has been overtaken by the commercial practice of an unlisted economy. There is nothing for a retrospective amendment to restore. It is also material that the taxpayers concerned are salaried individuals who did what their employers told them and who had no part in framing the question.

45. That there is legislative competence to amend with retrospective effect is not in doubt, and Article 265 furnishes no answer to it, for such an amendment supplies the very authority of law which the Article requires. The constraint lies elsewhere. The Hon’ble Supreme Court has held, in Rai Ramkrishna v. State of Bihar [TS-5003-SC-1963-O] and in the line which follows it, that although the power to legislate retrospectively is undoubted, its exercise may in a given case be so unreasonable as to attract the constitutional guarantees; and among the considerations treated as relevant are the length of the period covered, whether the liability was one the taxpayer could have foreseen and provided against, and whether he was in a position to pass it on. There is also authority in this very field. In Infosys, the Hon’ble Supreme Court has held that clause (iiia), inserted in 1999, was not clarificatory and could not operate retrospectively, precisely because it introduced for the first time the mechanism by which cost, and therefore value, became ascertainable; and that a mechanism of that kind cannot be read retrospectively unless the legislature expressly says so. An amendment which, for the first time, supplies the event upon which a charge is to rest stands upon the same footing. A prospective amendment in this field would be unobjectionable and is probably overdue. A retrospective amendment styled as clarificatory would be a different matter, not only because it would shake taxpayer confidence, but also because the affected taxpayers are salaried individuals who, however well placed, belong to a different genus altogether when it comes to the ability to afford expensive litigation in the constitutional courts, which alone can decide its vires.

46.The majesty of the law is not served by collecting what the law has not imposed; a revenue foregone for want of a charge, however inconvenient to the exchequer, is a small price for the principle that the State takes nothing from the citizen except by authority of law. And where that authority is wanting, the remedy is to enact it, and not to infer it.