What is a share buyback?

A buyback runs in one direction. The company spends its own cash to acquire its shares, and those shares are then cancelled, not held as an asset and not re-sold. Because the shares cease to exist, the total count of shares outstanding drops, and the same company is now divided into fewer pieces. Every share that remains represents a slightly larger claim on the identical pool of assets and future profit. Nothing about the business changes; only the number of slices does.

One detail is specific to India and worth stating, because it differs across markets. Here the bought-back shares must be extinguished, physically cancelled within a short window, not parked as treasury stock to be quietly re-sold later. Some other markets let a company hold repurchased shares in treasury and reissue them, which blurs the effect; the Indian rule is cleaner, because a buyback genuinely and permanently reduces the count. That permanence is part of why a buyback is a real capital-allocation decision rather than a temporary manoeuvre: once the cash is out and the shares are cancelled, neither is coming back.

Two balance-sheet facts follow and are worth holding onto. A buyback reduces both cash and net worth, because money the company owned has left it and gone to the sellers. The face value of each share, its nominal or par value, is unchanged, which is one way a buyback differs from a stock split. What moved is the number of shares and the size of the reserves, not the denomination printed on each one. The diagram below traces a single round of it, from a thousand shares to nine hundred.

So far the buyback looks like a tidy piece of arithmetic: fewer shares, a bigger slice each, cash returned. Everything in this section is true and none of it tells you whether the buyback was a good idea. For that, the number that matters is not the share count at all. It is the price the company paid to shrink it. The rest of this guide is really one question asked from several angles: was that price above or below what the shares were genuinely worth.

Here is the point that most coverage skips, and it is the heart of a buyback. Shrinking the count raises the per-share slice mechanically, but a bigger slice of the pie is only worth more if the company did not overpay to carve it. A buyback creates no profit and builds nothing; it only moves cash from the company to the sellers and rearranges who owns what is left. Whether the shareholders who stay are better off depends on one comparison alone: the price paid versus the shares' intrinsic worth.

The logic is a transfer between the leavers and the stayers. When the company buys a share for less than it is truly worth, the seller has left value on the table, and that value accrues to everyone who held on. When the company buys a share for more than it is worth, the seller walks away with value that used to belong to the continuing holders. The buyback is the same mechanical act either way. Only the direction of the transfer changes, and the price sets the direction.

The figure computes both cases from one setup. Suppose each share is truly worth ₹200, the company has 1,000 shares, and it retires 100 of them, leaving 900. In the first case it pays ₹150 a share, a fifth below intrinsic value; in the second it pays ₹250, a fifth above. Same company, same 100 shares retired, same 900 left. The only difference is the price, and it flips the outcome for the holder who stays.

It helps to name the transfer in rupees. In the cheap case the company paid ₹150 for something worth ₹200, so each of the 100 sellers left ₹50 on the table, ₹5,000 in all, and that ₹5,000 is now spread across the 900 shares that stayed, which is the ₹5.56 a share they gained. In the expensive case the arithmetic runs backwards: the sellers extracted ₹50 above worth on each of 100 shares, and the continuing holders funded every rupee of it. The buyback did not create the ₹5,000 or destroy it. It only decided which side of the trade received it, and the price set the side.

A buyback creates no profit. It only decides who owns the profit that already exists, and at what price the leavers are bought out.

Notice too that the price in the figure is the price the company actually pays, which in the real world is the market price it faces, not the intrinsic value it hopes for. The two are different things, and the entire opportunity in a buyback lives in the gap between them. When the market price sits below intrinsic worth, the company can buy in cheaply and hand the difference to the holders who stay; when the market has bid the shares above worth, every share retired locks in a loss for them. A buyback does not set the market price. It only chooses whether to spend the company's cash into it, and that choice is where the judgement lives.

The honest difficulty is that intrinsic value is itself an estimate. Nobody rings a bell to confirm a share is worth ₹200; you have to judge it from the business, and reasonable people disagree. That is precisely why the price paid is the tell, and why a buyback announcement on its own proves nothing. A company convinced its shares are cheap should be glad to buy them; a company overpaying to shrink the count is spending shareholders' cash to move a per-share number. Reading which is which is a capital-allocation judgement, and one way to sharpen it is to compare the buyback price against a sober estimate of worth, the same discipline behind a price-to-earnings ratio read against real earnings power rather than a headline.

The reason a buyback is so easily misread is that it reliably lifts one of the most watched numbers on the market, earnings per share, without touching the business behind it. Earnings per share is simply profit divided by the number of shares. A buyback leaves the profit alone and shrinks the denominator, so the ratio rises on its own. It looks like progress. It is division.

Follow it across several rounds and the point becomes stark. Hold profit fixed at an illustrative ₹10,000 and cut the share count from 1,000 to 900 to 800 to 700. Earnings per share climbs from ₹10.00 to ₹11.11 to ₹12.50 to ₹14.29, a rise of about 43 percent, while the company sells not one extra unit and earns not one extra rupee. The whole gain is the denominator shrinking by 30 percent. The chart draws the two lines that matter: the business, flat, and the ratio, rising.

There is a sharper way to say this that professionals use. A buyback is nearly always accretive to earnings per share, because a smaller denominator lifts the ratio almost automatically, yet it is only accretive to value when the price paid is below intrinsic worth. The two phrases sound alike and mean opposite things. A cash-rich company can even make the EPS lift look larger by funding the buyback from idle cash that was earning very little, so the profit given up is negligible while the share count drops, which flatters the ratio further without adding a rupee of real worth. Accretion to a ratio is not the test. Accretion to value is, and value is set by price.

A related real-world wrinkle is that many buybacks do not actually shrink the share count at all; they merely stop it from growing. Companies that pay staff in stock options and grants issue new shares every year, which dilutes existing holders, and a buyback of matching size simply mops those up so the count stays roughly flat. That can be a legitimate use of cash, but it is not the count-shrinking, EPS-lifting story the headline implies, and it is easy to miss if you look only at the announced buyback value and not at how many shares are being created on the other side. The number that matters is the change in the fully diluted count, after new issuance, not the buyback taken in isolation.

This mechanical lift is exactly what makes a buyback a tempting cosmetic. Executive bonuses, stock options and headline ratios are often tied to per-share numbers, so a buyback can be used to hit a target rather than to reward owners. It also feeds straight into the most quoted valuation ratio: at an unchanged price, a higher earnings-per-share figure mechanically lowers the price-to-earnings ratio, which can make a stock look cheaper without a single thing improving in the business. Anyone reading a buyback should separate the two questions the ratio blurs: did the company become more profitable, or did it simply divide the same profit among fewer shares. The tie-in to the value section is the punchline of the hero figure: in that example the earnings-per-share lift was identical in the value-creating and the value-destroying case. The EPS line could not tell them apart. Only the price could. That single observation, that an identical earnings-per-share lift can sit on top of either a real gain or a real loss for the continuing holder, is the reason this guide treats the price paid as the whole game and the per-share figure as a distraction dressed up as progress.

A company can repurchase its shares in two structurally different ways, and the difference decides who is treated how. The distinction is not cosmetic: it changes the price, the certainty and who ends up selling. In a tender offer, the company offers to buy a fixed number of shares at a fixed price, almost always a premium to the prevailing market price, and publishes a letter of offer with a defined window during which holders may tender. If shareholders offer more shares than the company wants, acceptance is proportionate: each tendering holder has the same fraction of their offered shares taken up, with a portion reserved for small shareholders. The price is known in advance; how many of your shares are actually bought is not.

In an open-market buyback, the company buys its own shares on the exchange over time, up to an announced maximum amount and maximum price, at whatever prices the market offers. There is no fixed acceptance and no premium tender price; the company simply becomes a buyer alongside everyone else until it fills its cap or the window closes. It is slower, less certain to complete in full, and the sellers are self-selected rather than treated proportionately. The schematic and the table below set the two mechanics side by side.

The acceptance ratio in a tender is worth working through, because it decides how much of your holding is actually bought. Suppose a tender seeks 100 shares but holders collectively offer 200. The offer is twice over-subscribed, so on the general portion roughly half of what each holder tenders is accepted and the rest is returned. Tender 40 shares into that and about 20 come back to you, still held, with only 20 sold. Two features soften this for smaller investors: a slice of the buyback is reserved for small shareholders, who are defined by a modest holding value, and acceptance within each category is proportionate. The practical lesson is that in an over-subscribed tender you cannot assume your whole lot will be taken, so plan for a partial acceptance and a residual holding rather than a clean exit.

One route-specific fact matters for reading a buyback in India today. The regulator grew uneasy with the on-exchange open-market route because participation through the market is uneven, unlike a proportionate tender, and it has progressively curtailed and then withdrawn that route. As of 17 July 2026 the tender route dominates in practice, so any older explainer that presents the on-exchange open-market route as a routine, freely available option is describing a mechanism that has since been narrowed. Regulatory positions change, so confirm the current framework with SEBI before relying on it.

A company cannot buy back shares with money it does not have, and it cannot use just any money. Under Section 68 of the Companies Act 2013, a buyback may be financed only from the company's free reserves, its securities premium account, or the proceeds of a fresh issue of shares or other securities, and it may not be funded from the proceeds of an earlier issue of the same kind of shares. The cash paid out reduces those reserves, which is one reason a buyback lowers net worth: reserves are part of shareholders' funds, and paying cash to the leavers depletes them.

The size is capped, not open-ended. In any financial year a buyback cannot exceed 25 percent of the aggregate of paid-up capital and free reserves. The approval threshold scales with size: a buyback of up to 10 percent of paid-up capital and free reserves can be authorised by the board, while anything above that, up to the 25 percent ceiling, requires a shareholder special resolution. There is also a leverage guardrail: after the buyback, total debt must not exceed twice the paid-up capital plus free reserves, which is meant to stop a company from borrowing heavily just to shrink its share count. SEBI layers process rules on top, including how the offer is disclosed and run.

The debt guardrail deserves a second look, because it targets the most dangerous kind of buyback. A company that borrows to buy back its own shares can lift earnings per share twice over, once by shrinking the count and once by replacing equity with cheaper-looking debt, while quietly making the balance sheet more fragile. The two-to-one ceiling on post-buyback debt exists to stop that engineering from running unchecked. It does not make a debt-funded buyback wise; it only caps how far it can go. For a holder, a buyback that leans on new borrowing rather than genuine surplus cash is a reason to look harder, not to relax, because the per-share number has been bought with leverage rather than earned by the business.

The process itself follows a set path, which is worth knowing so an announcement is legible. The board approves the buyback, and above the ten percent threshold the shareholders pass a special resolution; the company then publishes a letter of offer setting the price, the size and the timetable, backs the promised cash with an escrow arrangement so the money is demonstrably available, opens the tendering window, accepts shares proportionately if it is over-subscribed, pays the accepted holders, and finally extinguishes the shares it has bought. Each step is disclosed, so a careful reader can see the price, the size relative to reserves and the funding source before deciding anything, rather than reacting to the headline alone.

There are three real reasons a company buys back its shares, and they are worth stating plainly because the first two can be legitimate and the third is where the mischief lives. The first is to return surplus cash: when a company generates more than it can reinvest at a worthwhile rate, handing it back is a rational alternative to letting it sit idle or funding weak projects. The second is to signal undervaluation: a company spending its own cash on its own shares is, in effect, saying it regards them as cheap, and because the signal is backed by money rather than words, the market often reads it as credible. The third is the mechanical lift to earnings per share, which improves a ratio without improving the business at all.

It also helps to place a buyback on the ladder of things a company can do with a rupee of surplus cash. The highest use is to reinvest in the business itself when the returns are good; below that sit sensible acquisitions and paying down expensive debt; and only when none of those clears the bar does returning cash, by dividend or by buyback, become the best remaining option. Read this way, a buyback is not a triumph but a quiet admission that the company cannot find a better internal use for the money, which is entirely reasonable for a mature, cash-generative firm and a warning sign for one that should still be investing to grow. Where a buyback sits on that ladder is part of judging it.

The trouble is that every one of these stated reasons has a bullish telling and an honest telling, and the honest one usually comes back to price. A buyback is a capital-allocation decision, and the discipline of judging a use of cash against its alternatives and its price is exactly the habit that the method we teach is built to install. The table separates the story from the substance.

The undervaluation signal deserves particular scepticism, because it is the easiest to stage and the most flattering to believe. A board that genuinely thinks its shares are cheap will usually commit a meaningful sum and will not be selling personally at the same time; a board using the announcement for effect can put up a token amount, generate the headline, and change little underneath. The size of the buyback relative to the company's cash and market value, and whether insiders are buying or selling alongside it, say far more than the wording of the press release. A signal backed by a small cheque is a small signal, however confident the language wrapped around it.

There is a pattern worth naming that cuts across all three reasons. Companies tend to buy back most heavily when they are flush with cash and their shares are already expensive, near the top of a good run, and to stop buying in downturns when cash is tight and the shares are genuinely cheap. That is precisely backwards from the value logic of this guide, which rewards buying below worth and penalises buying above it. It is not a rule that every buyback is mistimed, but it is a strong reason to check the price against value rather than assume management has timed it well, because the aggregate record suggests the temptation runs the wrong way.

Read the third column and a theme emerges: not one honest reading is settled by the announcement itself. Each turns on a fact the press release rarely leads with, and most of them reduce to the same question the value section put at the centre. Is the company buying its shares for less than they are worth, out of genuine surplus, or spending owners' cash to move a number. The corporate action is easy to describe. Reading whether it is sound is the skill. And the raw material for that reading, the price against a sober estimate of worth, the funding source, whether insiders are buying or selling, and the size relative to the company's cash, is all disclosed for anyone willing to look past the headline.

For years, buybacks in India carried a tax structure that quietly made them more attractive than dividends. The tax sat at the company level: under Section 115QA, the company paid a buyback distribution tax on the amount distributed, and the shareholder received the proceeds tax-free under Section 10(34A). Because the shareholder's slab rate never touched the money, a buyback was a tax-efficient way to return cash, especially for holders in high tax brackets, and it was widely used for that reason.

The Finance (No. 2) Act 2024, the July 2024 Union Budget, reversed the principle: the tax on a buyback moved from the company to the shareholder. As of 17 July 2026, for buybacks completed on or after 1 October 2024, the treatment works as the figure shows.

The consequence is direct. Because both a buyback and a dividend are now taxed in the shareholder's hands at the slab rate, the old tax edge of buybacks has largely disappeared, and the buyback-versus-dividend decision now turns on price, timing and the individual holder's tax position rather than on a structural tax saving. If you want the wider picture of how dividend income, capital gains and trading income are taxed in India, that is the subject of the guide on trading taxation in India.

One difference between the two survives the tax change, and it is about control of timing. A dividend is declared by the company and lands on every shareholder whether they want the income this year or not, triggering tax for all of them at once. A buyback is optional at the holder's end: only those who tender realise the proceeds and the tax, while those who hold defer both and simply own a little more. So even with the rates now aligned, a buyback still hands the individual holder more say over when the taxable event happens, which can matter for someone managing income across tax years.

For a holder, the practical effect depends on the slab. Before the change, a shareholder in a high tax bracket could receive buyback proceeds without the slab rate biting, which is what made the route so tax-efficient; after it, that same holder is taxed on the whole amount as a deemed dividend at their slab rate, so the after-tax proceeds are lower than they once would have been. The exact figures depend on the individual's total income and the rates in force, which is why this guide states the direction rather than a number: the tax outcome moved against the high-bracket holder and toward parity with a dividend. Confirm your own position before assuming either way.

Three corporate actions are constantly confused because all three touch shareholders, yet they do very different things to the company and to the share count. The clean way to tell them apart is to ask three questions: does cash leave the company, does the share count change, and who is taxed. A buyback is the near-perfect mirror of a bonus issue. A bonus adds shares and pays out no cash, so the count rises and each share is worth proportionately less while the total value is unchanged. A buyback removes shares and pays cash to the sellers, so the count falls and, unlike a bonus, it changes who owns the company. Set out side by side, the three stop being interchangeable.

The ownership row is the one to dwell on, because it is where a buyback truly parts company with a bonus. A bonus issue hands new shares to everyone in proportion, so nobody's slice of control changes. A buyback pays out the holders who tender and cancels their shares, so the holders who do not tender, which typically includes long-term owners and promoters who sit tight, end up owning a larger share of the company than before without spending a rupee themselves. A buyback therefore quietly redistributes ownership and can nudge control toward those who stay, an effect a dividend and a bonus both lack. That is worth watching whenever the non-tendering holders are the same people who run the company.

One more idea ties the buyback and the dividend together for a holder comparing them. The cash a company returns each year can be measured as a total payout yield, the dividend yield plus the buyback yield, where the buyback yield is the value of shares repurchased divided by the company's market value. Read that way, a buyback is simply the second channel through which a company hands cash back, and a firm that returns little as dividends may still be returning a great deal through repurchases. The combined figure is the honest measure of how much cash is actually coming back to owners, and it is the one to compare across companies rather than the dividend alone.

Read together, the picture is coherent. A buyback and a dividend both send cash out; the buyback also shrinks the count and reshuffles ownership, while the dividend pays everyone in proportion and leaves ownership untouched. A bonus issue is the odd one out: no cash leaves, the count rises, and it is an accounting reallocation rather than a distribution. And on tax, the 2024 change means a buyback and a dividend are now taxed alike in your hands, which is the single fact that erased the buyback's old advantage.

Strip away the corporate-finance language and a buyback presents a shareholder with one of two positions, depending on whether you act. If you tender or sell into the buyback, you receive cash and exit part or all of your stake. In a tender offer, remember that acceptance is proportionate, so if the offer is over-subscribed you may get only a fraction of your tendered shares bought back and keep the rest. And since 1 October 2024, the cash you receive is taxed as a deemed dividend at your slab rate, while the cost of the accepted shares becomes a capital loss you can carry forward. The after-tax value of tendering is therefore very personal, and it depends on your bracket.

If you hold, you do nothing and your proportional ownership rises slightly as the total share count falls. But a larger slice is only worth more if the company did not overpay to shrink the count, and none of it is free value: the cash that lifted your ownership share also left the company. A buyback can be a sensible return of surplus capital at a fair price, and it can equally be a device to prop up earnings per share when the underlying business is not growing. The honest reading, the one this whole guide keeps returning to, is that a buyback is a capital-allocation decision to judge on its price and its purpose, never an automatic positive.

Doing nothing is itself a decision, and it should be a considered one rather than a default. The question a tender really asks you is whether the offered price is a good one at which to sell part of your stake, given what you think the business is worth and what you originally paid for it. If the offer sits comfortably above your own estimate of value, tendering to the extent you are allowed can be sensible; if it sits below, holding and letting your ownership share rise may serve you better, provided the company is not overpaying with your cash to get there. The offer price is information, not an instruction, and the buyback has handed you a choice rather than made one for you.

One practical mechanic sits underneath all of this: eligibility is set by a record date. Only shares held as of that cut-off can be tendered, and the entitlement to tender in a proportionate offer is calculated from the holding on that date, so buying in purely to participate is rarely the free lunch it can look like once the offer price is already reflected in the market price. When you weigh whether to tender or hold, the honest inputs are the same three this guide has returned to throughout: the price offered against your own estimate of worth, your tax position at your slab rate, and your view of the business you would be reducing or keeping. None of those is answered by the announcement, and all of them are yours to judge.

So the honest one-line answer to whether a buyback is good news is that it depends on the price, and the announcement rarely tells you. A buyback returns cash by shrinking the share count, which lifts per-share ownership and earnings per share as a matter of arithmetic. It creates value only when the company pays less than its shares are worth, handing the holders who stay what the sellers left behind, and it destroys value when it overpays. Everything else, the route, the tax, the EPS lift, the signalling, sits on top of that one test. Judge a buyback the way you would judge any use of a company's cash: against its alternatives, and above all against its price.

• Companies Act 2013, Section 68. Establishes the power to buy back securities, the ceiling of 25 percent of paid-up capital plus free reserves in a financial year, the board-versus-special-resolution thresholds at 10 percent, the free-reserves, securities-premium or fresh-issue funding sources, and the post-buyback debt-equity limit. Stated as of 17 July 2026; verify the current text at the source. mca.gov.in
• Finance (No. 2) Act 2024, buyback taxation. The July 2024 Union Budget shifted the tax on a buyback from the company to the shareholder, treating the amount as a deemed dividend taxed in the holder's hands for buybacks on or after 1 October 2024, with the cost of the shares becoming a capital loss. Cited at the level of principle; confirm the current provisions and rates with the Income Tax Department. incometaxindia.gov.in
• SEBI (Buy-Back of Securities) Regulations. Set the process rules for buybacks and, over recent years, progressively curtailed and then withdrew the on-exchange open-market route, leaving the tender route dominant in practice as of 17 July 2026. Confirm the current framework at the source. sebi.gov.in
• On the value logic of a buyback. The principle that a repurchase adds value for the continuing shareholders only when the shares are bought below their intrinsic worth, and destroys value when they are overpaid for, is standard in the corporate-finance treatment of payout policy. The worked examples in this guide follow that logic and are illustrative, not forecasts.