Why India’s new rules for taxing buybacks might be unconstitutional (and the simple fix to save them)

An article I read online recently raises important questions concerning India’s new rules for taxing buybacks.[1] I think the new rules might be unconstitutional, but before I explain why and propose a simple fix to save them, I’d like to offer a few observations and engage with some of the key themes in the article.?

Income characterization.

The authors point out that the new rules disregard a basic principle of the Income Tax Act i.e., that income is taxable under five mutually exclusive ‘heads of income’, and that income brought to tax under a specific head may not again be brough to tax under another, different head. The authors say that because they (i) treat the entirety of the proceeds a shareholder would receive upon tendering his shares in a buyback (the “buyback proceeds”) as a dividend, and (ii) allow the shareholder to recognize a capital loss equivalent to cost basis in the shares tendered and bought back, the new rules effectively tax buybacks under two heads of income, deeming into existence a dividend taxable under the head “Income from Other Sources”, and a capital loss recognizable under the head “Capital Gains”.

This is one way of looking at it. A better way would be to recognize that were it not for one significant and unexplained design choice (I refer here to the legislature’s decision to sever the link between distributions deemed to be dividends under S.22 of the Income Tax Act and the existence of accumulated profits at the distributing company – more on this later), the new rules align the taxation of distributions of accumulated profit via a buyback with the taxation of such distributions via a capital reduction or corporate liquidation.

India follows the classical system of corporate taxation i.e., taxing corporate income once at the corporate level and again at the shareholder level when dividends are distributed. Understandably, the legislature has therefore often been preoccupied with the possibility that undistributed profits generated in corporate solutions might escape taxation in the hands of shareholders. Parliament has sought to address this concern through two primary methods: ?

• A ‘super-tax’ on undistributed corporate profits. Originally enacted as S.23A of the Indian Income Tax Act, 1922 (the “1922 Act”)[2], the super-tax rules taxed shareholders directly on their proportionate share of a closely held company’s accumulated or undistributed profits if the company accumulated profits beyond its reasonable needs or failed to distribute its profits in a manner that rendered them taxable to shareholders. The super-tax, which was continued under S.104 of the Income Tax Act, was effective when individual income tax rates exceeded corporate income tax rates. However, it lost relevance following the amendments to the Income Tax Act in 1985, which equalized the individual and corporate tax rates, and was eventually repealed in 1987.
• An expanded definition of dividends. Even when corporate and individual tax rates are equal, undistributed corporate profit can reduce tax revenues because taxation at the shareholder-level is avoided. The problem is exacerbated when companies are able to return accumulated profit to their shareholders through means other than a dividend distribution. This was a particularly pressing concern in the early 20th century when English tax law and its derivatives, including the 1922 Act, distinguished between revenue and capital receipts and taxed only the former. Taxpayers used the law to their advantage by having their closely held corporations accumulate profits, issue bonus debentures that represented the accumulated profit, and then redeem the issued debentures. Courts treated the issuance of bonus securities as a capitalization of accumulated profit, and the subsequent redemption of the issued securities as a (non-taxable!) return of capital to the shareholder.[3] Similarly, courts treated distributions of accumulated profit by companies in liquidation as indistinguishable from, and therefore equivalent to, a return of capital.[4]

To address this situation, S.2(6A) of the 1922 Act was enacted in 1939[5]. S.2(6A) defined dividends expansively to include (a) in-specie distributions of corporate assets, (b) distributions of bonus debentures, (c) distributions upon liquidation, and (d) distributions upon a reduction in share capital, in each case to the extent the distribution is attributable to the distributing company's accumulated profit. The new definition was intended to ensure that any distribution of accumulated profit by a company to its shareholders, regardless of the way in which it was undertaken, would be treated as a dividend and taxed accordingly.[6] S.2(6A) was reenacted as S.2(22) of the Income Tax Act.

The super tax and the expanded dividend concept are testament to a legislative desire to treat distributions of accumulated profit as distinct from a mere return of capital and ensure they are taxed both in the hands of the company and its shareholders. Although originally enacted at a time when capital receipts largely went untaxed, neither mechanic was repealed when the capital gains tax was enacted (first in 1947, and then again in 1956). Thus, even when disguised as consideration payable for the ‘transfer’ of a capital asset – the gain component of which would otherwise receive capital gains treatment – the legislature intended for distributions of accumulated profit to continue to receive dividend treatment. This distinction is significant because capital gains are generally taxed at a lower beneficial rate than ordinary dividend income. ?This legislative intent becomes more compelling when the following examples are considered.

• Corporate liquidations. Section 46(2) of the Income Tax Act, which was enacted in 1961, deals specifically with the taxation of distributions made by companies during liquidation. The legislature could have designed S.46 to give capital gains treatment to the amount by which a liquidating distribution exceeds a shareholder’s basis in his shares. Instead, S. 46(2) provides that liquidating distributions will, to the extent they are attributable to accumulated profit, be treated as dividends under S.2(22)(c) and taxed accordingly. ?Only liquidating distributions made in excess of a company’s accumulated profits will be considered to determine the gain, if any, to be taxed as capital gain under Section 46(2).
• Capital reductions. The Supreme Court of India too, seems to appreciate the distinction. In CIT v. G. Narasimhan[7], the court addressed the taxation of amounts distributed to shareholders pursuant to a reduction of share capital. It held that under S.2(22)(d), any distribution attributable to a company’s accumulated profits (whether capitalized or not) is taxable as a dividend in the hands of the shareholder. Only amounts distributed beyond the accumulated profits constitute a capital receipt and are subject to capital gains tax under Section 45.

When used as a mechanism to return profit to shareholders, for income tax purposes, there is not much that distinguishes a capital reduction from a share buyback – they are fundamentally identical. It is therefore odd that up until the enactment of the new buyback rules, a taxpayer could simply circumvent parliamentary intent to tax distributions of accumulated profits as dividends at ordinary income tax rate by undertaking a buyback instead of a capital reduction. This kind of disparate treatment of what is essentially the same activity is undesirable, and the new buyback rules should be viewed as bringing some much-needed parity.

This approach is not inconsistent with global practice. Take, for example, Sections 302 and 304 of the United States Internal Revenue Code:

• Section 302: Distributions in Redemption of Stock. This section governs the tax treatment of distributions made by a corporation when it redeems its stock from shareholders. A redemption is treated as a sale or exchange (and receives capital gains treatment) if it meets certain conditions such as:

(a) Substantially disproportionate redemption: The shareholder's ownership percentage is significantly reduced.

(b) Complete termination of interest: The shareholder completely relinquishes their interest in the corporation.

(c) Partial liquidation or corporate business purpose: Redemption aligns with specific business needs.

If these conditions are not met, the redemption is treated as a dividend, taxable as ordinary income, to the extent of the corporation's earnings and profits.[8]

• Section 304: Redemption through use of related corporations. This section applies to stock transactions between related corporations or affiliated parties to prevent tax avoidance through internal transfers.

(a) When one corporation (the “acquiring corporation”) acquires stock of another related corporation (the “issuing corporation”) from persons in control of both corporations (the “seller”), the transaction is treated as a redemption rather than a simple sale. The seller is treated as contributing stock of the issuing corporation to the acquiring corporation in exchange for stock of the acquiring corporation, and the acquiring corporation is treated as redeeming the stock so issued for the sale consideration.

(b) Whether the redemption proceeds receive sale or exchange treatment or are characterized as a dividend depends on whether the conditions under Section 302 are met with respect to issuing corporation. If the conditions are not met, then the IRC recharacterizes the transaction as a dividend to the seller (up to the extent of earnings and profits of both the acquiring and issuing corporations).[9]

(c) The aim is to prevent taxpayers from circumventing dividend taxation by structuring transactions as stock sales.

Clearly, ensuring distributions of accumulated profit, however affected, are taxed appropriately, is of prime concern to tax administrations globally. The new buyback rules reflect this shared concern.

Interplay with tax treaties

The authors highlight that because (a) S.90 of the Income Tax Act, which empowers the central government to enter into tax treaties, provides that “in relation to the assessee to whom such agreement applies, the provisions of this Act shall apply to the extent they are more beneficial to that assessee”, and (b) distributions subject to the new buyback rules are capable of? being characterized both as dividends (under Article 10 of the OECD Model Tax Convention) and capital gains (under Article 13 of the OECD Model Tax Convention), the taxpayer may have some agency in how the income is characterized for Indian tax purposes.

However, in my opinion, the domestic characterization of an item of income generally should not have to bend to how that same item of income might be characterized under a treaty, particularly when it comes to dividends. The fact that items of income that are considered fees for technical services under the Income Tax Act are nevertheless often not characterized as fees for technical services under Article 12 of the OECD Model Tax Convention (and are instead characterized as business profits under Article 7) has little bearing when it comes to whether a profit distribution should be characterized as a dividend under the treaty because the definition of dividend is (intentionally) far more expansive and designed specifically to encompass distributions which are “subjected to the same taxation treatment” as dividends under a countries domestic law.[10] In fact, paragraph 28 of the commentary on Article 10 of the OECD Model Tax Convention provides that:

“Payments regarded as dividends may include not only distributions of profits decided by annual general meetings of shareholders, but also other benefits in money or money’s worth, such as bonus shares, bonuses, profits on a liquidation or redemption of shares (see paragraph 31 of the Commentary on Article 13) and disguised distributions of profits.”

Paragraph 31 of the commentary on Article 13 of the OECD Model Tax Convention similarly provides that:

“If shares are alienated by a shareholder in connection with the liquidation of the issuing company or the redemption of shares or reduction of paid-up capital of that company, the difference between the proceeds obtained by the shareholder and the par value of the shares may be treated in the State of which the company is a resident as a distribution of accumulated profits and not as a capital gain. The Article does not prevent the State of residence of the company from taxing such distributions at the rates provided for in Article 10: such taxation is permitted because such difference is covered by the definition of the term “dividends” contained in paragraph 3 of Article 10 and interpreted in paragraph 28 of the Commentary relating thereto, to the extent that the domestic law of that State treats that difference as income from shares."

The United States takes a similar approach under its tax treaties.[11]

Accordingly, whether or not a taxpayer chooses to be taxed under a treaty or domestic law should be determined assuming that the buyback proceeds will be taxed under Article 10. Article 13 will not apply. This is not just because the definition of dividend under Article 10 is wide enough to encompass the buyback proceeds as characterized under domestic law, but also because the new proviso to Section 46A operates to ensure that there is no gain that would fall to be addressed under the capital gains article.

In and of themselves, the new buyback tax rules also shouldn’t increase the risk of international double taxation or denial of foreign tax credits. The OECD commentary is quite clear and requires the residence country to provide a tax credit as long as taxation in the source country is in accordance with the treaty (i.e., the treaty accords the source country a right to tax) even if the residence jurisdiction differs with the source country as to the characterization of an item of income.[12] Thus, even if the residence country would treat the buyback proceeds as capital gains, because India has the right to tax capital gains arising from a disposal of shares of an Indian company under the vast majority of its treaties, the residence country is obliged to grant a credit for the taxes paid in India.

Why the new buyback tax rule is probably unconstitutional

The authors highlight that the new buyback rules disregard the fundamental principle that dividends represent a distribution of accumulated profit and treat the entirety of the buyback proceeds as dividends, ignoring shareholder basis in the redeemed shares. I agree. Longstanding legislative practice, since the enactment of the expanded definition of dividends, has been to treat profit distributions as dividends to the extent of the distributing company’s accumulated profits. The new buyback rules sever this link, and it is not at all clear why.[13]

But there is a bigger problem. Can the portion of the buyback proceeds equivalent to a taxpayer’s basis in the shares even be considered income? I believe the answer is no, and therefore, at least to the extent that they treat a return of basis as dividend income, the new rules are probably unconstitutional and liable to be invalidated and read down.

• The constitutional power to tax. Parliament derives its power to tax income under Article 246 of the Constitution read with Entry 82 (Taxes on Income other than agricultural income) of the Union List in the Seventh Schedule. The meaning of the word income as used in Entry 82 was the subject of interpretation in Navinchandra Mafatlal v the Comm’r of Income Tax[14], where a constitution bench of the Supreme Court upheld the capital gains tax as being intra-vires Parliament’s power to tax income. The Supreme Court held:

“What, then, is the ordinary, natural and grammatical meaning of the word "income"? [I]n the United States of America and in Australia both of which also are English speaking countries the word "income" is understood in a wide sense so as to include a capital gain. Reference may be made to Eisner v. Macomber 252 USR 509, Merchants' Loan & Trust Co. v. Smietanka 255 USR 509 and United States of America v. Stewart 31 USR 60 and Resch v. Federal Commissioner of Taxation 66 CLR 198. In each of these cases a very wide meaning was ascribed to the word "income" as its natural meaning. The relevant observations of learned Judges deciding those cases which have been quoted in the judgment of Tendolkar, J., quite clearly indicate that such wide meaning was put upon the word "income" not because of any particular legislative practice either in the United States or in the Commonwealth of Australia but because such was the normal concept and connotation of the ordinary English word "income". Its natural meaning embraces any profit or gain which is actually received. This is in consonance with the observations of Lord Wright to which reference has already been made. Mr. Kolah concedes that the word "income" is understood in the United States and Australia in the wide sense contended for by the learned Attorney-General but he maintains that the law in England is different and, therefore, entry 54 which occurs in a Parliamentary statute should be construed according to the law of England. We are again brought back to the same argument as to the word having acquired a restricted meaning by reason of what has been called the legislative practice in England - an argument which we have already discarded. The argument founded on an assumed legislative practice being thus out of the way, there can be no difficulty in applying its natural and grammatical meaning to the ordinary English word "income". As already observed, the word should be given its widest connotation in view of the fact that it occurs in a legislative head conferring legislative power.”

In Navnit Javeri v Ass’t Comm’r of Income Tax, another constitution bench of the Supreme Court upheld Section 2(22)(e) of the Income Tax Act as constitutionally valid. In this case, the court further refined its interpretation of the meaning of income under Entry 82. It held that:

What the entries in the List purport to do is to confer legislative powers on the respective legislatures in respect of areas or fields covered by the said entries; and it is an elementary rule of construction that the widest possible construction must be put upon their words. This doctrine does not, however, mean that Parliament can choose to tax as income an item which in no rational sense can be regarded as a citizen's income. The item taxed should rationally be capable of being considered as the income of a citizen. But in considering the question as to whether a particular item in the hands of a citizen can be regarded as his income or not, it would be inappropriate to apply the tests traditionally prescribed by the Income-tax Act as such.

[T]he question which now arises is, if the impugned section treats the loan received by a shareholder as a dividend paid to him by the company, has the legislature in enacting the section exceeded the limits of the legislative field prescribed by the present entry 82 in List I? As we have already noticed, the word "income" in the context must receive a wide interpretation; how wide it should be it is unnecessary to consider, because such an enquiry would be hypothetical. The question must be decided on the facts of each case. There must no doubt be some rational connection between the item taxed and the concept of income liberally construed. If the legislature realises that the private controlled companies generally adopt the device of making advances or giving loans to their shareholders with the object of evading the payment of tax, it can step in to meet this mischief, and in that connection, it has created a fiction by which the amount ostensibly and nominally advanced to a shareholder as a loan is treated in reality for tax purposes as the payment of dividend to him.

Navinchandra Mafatlal makes clear that the meaning of income under Entry 82 is expansive. However, the court still recognizes some inherent limits to what may properly be considered income within the meaning of Entry 82. Read together, Navinchandra Mafatlal and Navnit Javeri stand for the proposition that for an income tax to be considered constitutionally valid, (a) there must be some rational connection between the item taxed and the concept of income liberally construed, and (b) liberally construed, income means any profit or gain actually realized.

• The new buyback tax is not entirely a tax on income. As discussed, the new buyback rules disregard the fundamental principle that dividends represent a distribution of accumulated profit and treat the entirety of the buyback proceeds as dividends, ignoring shareholder basis in the redeemed shares. This is the problem. Even under the most expansive definition of income, it is plain that a return of basis / capital cannot be considered income.[15] The Income Tax Act itself recognizes this fact in the design of the capital gains tax – which taxes only the excess of the amount realized on the transfer of a capital asset over the asset’s cost of acquisition. The fact that the new proviso to S.46A allows the taxpayer to set off his basis against future capital gains cannot change the fact that the basis itself does not represent income and cannot legitimately be taxed. For these reasons, and to this extent, the buyback tax is constitutionally invalid and liable to be struck down or, at the very least, read down.

The simple fix.

The obvious and most straightforward way to address the possible unconstitutionality of the new buyback rules would be to simply amend the rules to reestablish the link between dividend treatment and the existence of accumulated profit, along the lines of S.46(2) (corporate liquidations) and the Supreme Court ruling in G. Narasimhan (capital reductions).

The unfair fix.

There is however, one other, more complicated possibility. Under Entry 86 of the Union List in the Seventh Schedule, Parliament is empowered to legislate with regard to “taxes on the capital value of the assets, exclusive of agricultural land, of individuals and companies; taxes on the capital of companies.” In Navinchandra Mafatlal the Supreme Court declined to find that the power to impose a tax on capital gains derived from Entry 86. However, Entry 86 was the basis for India’s now repealed wealth tax.[16] The entry basically allows for a tax on the capital value of an asset. What is a taxpayer’s basis in a share if not the capital value of the share? A court could very well find the new buyback tax rule constitutional under Parliament’s taxing power under Entry 86.

However, this would be unfair in the extreme. Even the wealth tax applied at a rate of 1%, only to individuals whose net wealth exceeded the specified threshold of INR 3,000,000, and accounted for debt owed on the assets included in the determination of a person’s net wealth. The buyback tax rules are subject to none of these safeguards. The tax applies at ordinary tax rates, there is no threshold a taxpayer must cross before the rules are applicable (unless taking into account the buyback proceeds, the taxpayer’s income falls within the zero bracket), and there is no scope to account for debt incurred to acquire the shares.

Were it up to me, I would reestablish the link with accumulated profit.

The views expressed in this article are solely my own and do not represent the views or opinions of my employer or any affiliated organizations.

[1] https://anagrampartners.in/archives/insights/taxing-buy-back-of-shares-creating-an-artificial-quagmire. I highly recommend reading the article by Anagram Partners first to get up to speed on the new buyback rules.

[2] Indian Income-tax (Amendment) Act, 1930, Section 4

[3] Bouch v Sproule, [1887] 12 AC 385; Comm’rs of Inland Revenue v Blott, [1921] 2 A.C. 171; Comm’rs of Inland Revenue v Fisher’s Executors, [1926] A.C. 395; Comm’r of Income tax v Mercantile Bank of India, (1936) 63 I.A. 457. C.f. Swan Brewery v Rex [1914] A.C. 231

[4] Comm’rs of Inland Revenue v Burrell, [1924] 2 K.B. 52; Comm’s of Income Tax v Muthukaruppan Chettiar, (1935) 62 I.A. 203; Comm’r of Income Tax v Siddha Gowder, I.L.R. 55, Madras 818 (1932)

[5] Indian Income-tax (Amendment) Act, 1939, S.2

[6] The Finance Act 1955 further expanded this definition to address avoidance of the corporate super tax. Companies would return accumulated profit to their shareholders by way of loans and advances, thus avoiding both the application of the super tax and dividend taxation in the hands of the shareholders. Under new sub-clause 2(6A)(e), loans and advances to shareholders are treated as dividends to the extent of the corporation’s accumulated profit.

[7] [1999] 102 Taxman 66 (SC)

[8] If the redemption is classified as a dividend, the redeemed stock's basis is not reduced or removed from the shareholder’s basis in their stock holdings. The basis of the redeemed shares is added to the basis of the remaining shares held by the shareholder (if any). However, if all shares are redeemed, and the redemption is treated as a dividend, the shareholder retains the entire basis of the redeemed stock. This basis remains available to offset gain or loss in future transactions but provides no immediate benefit because dividend income is taxed separately.

[9] When the proceeds of the transaction are treated as a dividend the basis of the stock sold or transferred to the acquiring corporation does not disappear. Instead, it is added to the basis of the seller’s stock in the acquiring corporation.

[10] Commentary on Article 10 of the 2017 OECD Model Tax Convention, paragraph 23

[11] U.S. Dep’t of the Treasury, Technical Explanation of the United States Model Income Tax Convention (1996), Commentary on paragraph 3 of Article 10, and Technical Explanation of the United States Model Income Tax Convention (2006), Commentary on paragraph 3 of Article 10

[12] Commentary to the 2017 OECD Model Tax Convention, paragraph 5 of the commentary on Article 13, and paragraphs 32.1 – 32.7 of the commentary on Articles 23A and 23B).

[13] Neither the Finance Minister’s budget speech, nor the notes to clauses, nor the memorandum explaining the provisions of the Finance Bill, 2024 shed any light on the rationale for this departure from previous longstanding legislative practice. One potential explanation is that Parliament wanted to avoid the complexity that would be inherent in determining how much of the company’s accumulated profit to attribute to a selective buyback. But similar complications exist in the case of capital reductions, so Parliament’s sudden reticence is perplexing.

[14] [1954] 26 SC 758 (SC)

[15] See Eisner v Macomber, 252 U.S. 509; C.f. Comm’r of Income tax v Glenshaw Glass, 348 U.S. 426. The United Supreme Court, in Glenshaw Glass, defined income as “instances of undeniable accessions to wealth, clearly realized, and over which the taxpayers have complete dominion.” No sane individual would argue that returning a taxpayer’s basis to him constitutes an ‘accession to wealth’.

[16] Banarasi Das v Wealth Tax Officer, [1965] 56 ITR 224 (SC); Sudhir Chandra Nawn v Wealth Tax Officer, [1968] 69 ITR 897 (SC)