Nvidia authorises another 150 billion dollars in buybacks: what does it really mean for shareholders?

Nvidia has just announced a number that is hard to ignore: 150 billion dollars. This is not new investment in data centres, acquisitions or spending to develop the next generation of GPUs. The board has authorised a 150-billion increase in the programme under which the company can repurchase its own shares. Adding this new authorisation to the amount still available gives Nvidia 235 billion dollars of remaining capacity, which it expects to use by the end of fiscal year 2028.

It is an enormous figure, but precisely for that reason it is worth avoiding two overly simple conclusions. The first is that Nvidia is about to buy 235 billion of its own shares immediately: that is not the case, because an authorisation is not an order already executed. The second is that a buyback of this size automatically makes the shares more valuable or prevents them from falling. Repurchases can create value for shareholders, but only under certain conditions. To understand this, we need to start with what actually happens when a business buys itself.

What a buyback really is

A share buyback, or repurchase of a company's own shares, occurs when a company uses its financial resources to purchase in the market some of the shares it previously issued. The company's capital does not increase: some of its resources are used to return value indirectly to shareholders. Repurchased shares may be cancelled or held as treasury shares, depending on the transaction's structure and the applicable rules.

Economically, the central development is the possible reduction in the effective number of shares outstanding. If total earnings do not change but are divided among fewer shares, each remaining share represents a slightly larger portion of the company and its earnings. This is why buybacks are often described using the cake metaphor: the cake does not get bigger, but it is divided into fewer slices.

In the example, the business has not produced one extra dollar of profit, but EPS, or earnings per share, has risen by approximately 8.7%. This effect is real, but does not by itself demonstrate that capital has been used well: we must also consider how much the company paid to repurchase the shares.

An authorisation is not a buyback

When we read that Nvidia has “announced a 235-billion buyback”, precision matters. The new measure adds 150 billion to the existing authorisation and brings the total still authorised to 235 billion. Nvidia said it expects to execute the remaining programme by fiscal year 2028, but this does not mean the entire amount will be spent tomorrow, next month or according to a linear schedule.

Repurchase programmes normally give management some flexibility. Actual decisions may depend on the share price, market conditions, operating needs, available liquidity and alternative investment opportunities. An authorisation therefore establishes the maximum capacity within which the company can act; it guarantees neither the pace of purchases nor the average price paid. Only repurchases actually executed can reduce shares outstanding.

How can Nvidia afford figures like these?

The scale of the programme becomes clearer when looking at the company's figures. In the second quarter of fiscal year 2027, ended 26 July 2026, Nvidia recorded 96.2 billion dollars in revenue and GAAP net income of around 59.7 billion. In the first six months of the fiscal year, net income reached 118 billion. At the end of July, the company also held 22.4 billion in cash and equivalents and 34.1 billion in marketable debt securities, totalling approximately 56.6 billion across the two components.

The buyback therefore stems from exceptional cash-generating capacity, although this must not be confused with liquidity held at a single point in time. Nvidia simultaneously continues investing in its technology platform, production capacity and the artificial-intelligence ecosystem. In the official announcement, Jensen Huang explicitly linked the authorisation to the ability to fund both technological transformation and capital returns to shareholders. This is management's assessment, not proof that every future repurchase will be attractive, but it indicates that the company regards both uses of funds as sustainable.

The per-share benefit is not automatic

A buyback can increase EPS without necessarily creating value. Imagine a share's reasonable economic value is 100 dollars. If the company uses 10 billion to buy its own shares when the market values them at 70, it acquires portions of the business below their hypothetical intrinsic value. For shareholders who remain, the transaction can be highly efficient.

If, instead, the same company buys at 150 dollars shares economically worth 100, it uses capital less advantageously. The number of shares falls and EPS may still rise, but an excessive price was paid to achieve that reduction. Rising earnings per share and poor capital deployment can therefore occur simultaneously. This is why a buyback announcement alone does not demonstrate that a share is undervalued.

The proper comparison also includes opportunity cost. Every dollar used to buy back shares cannot simultaneously be allocated to research, production capacity, acquisitions, debt reduction or dividends. A repurchase creates value when it is one of the best available uses of that capital, not simply when it makes EPS look better.

What can happen to the price-to-earnings ratio

A buyback can also influence the P/E, the ratio of share price to earnings per share. If a share costs 100 dollars and EPS is 5 dollars, the P/E is 20. If reducing the share count takes EPS to 5.50 dollars and the price hypothetically remains unchanged, the multiple falls to approximately 18.2. The share therefore appears cheaper relative to its earnings.

The market price does not necessarily stand still, however. Investors may react to the programme, change their expectations and revalue the shares. Moreover, in Nvidia's case, growth or decline in operating earnings may affect EPS much more than the reduction in share count. A buyback is one factor influencing the per-share figure, not a mathematical machine that automatically makes a valuation attractive.

Is there really a “parachute effect”?

This expression often appears in material circulating on social media: with 235 billion available, Nvidia supposedly has a kind of parachute beneath its share price. The reasoning starts from a real factor. If the company regularly enters the market as a buyer of its own shares, it generates additional demand and can absorb some supply. After the announcement, the shares rose by more than 2%, indicating that the market welcomed the decision in the short term.

A buyback programme, however, does not constitute a guaranteed price floor. If earnings expectations deteriorate, artificial-intelligence growth slows, competition increases or the market applies lower multiples, the shares could fall even while Nvidia repurchases them. It is therefore more accurate to describe a potentially significant source of demand, whose strength will depend on how much, when and at what prices the company executes repurchases.

235 billion is enormous, but Nvidia is even larger

A number can look gigantic in absolute terms and take on a different scale when measured against the company. At the time of the announcement, Nvidia's market value was approximately 5,500 billion dollars. The authorised 235 billion therefore represented a little over 4% of market capitalisation. The comparison is merely illustrative, because the price, market value and pace of purchases will change over time, but it avoids giving the impression that the group is about to retire an enormous proportion of its shares.

The record also needs to be interpreted correctly. Nvidia describes the 150-billion increment as the largest increase in a repurchase authorisation ever announced. Reuters compared it with the previous record authorisation of 110 billion announced by Apple in 2024. This signals the scale the company has reached, but cannot tell us in advance how many shares will actually be cancelled: that will depend mainly on the average price paid.

Nvidia is already returning substantial capital

The new plan does not come out of nowhere. In the second quarter of fiscal 2027 alone, Nvidia returned approximately 26 billion dollars to shareholders through repurchases and dividends. At the end of July, approximately 99 billion remained under the previous authorisation; the addition announced in September subsequently brought the remaining total to 235 billion. The figures therefore show that the company has not merely approved a theoretical ceiling, but is already using buybacks on a large scale.

For shareholders, however, the programme's final outcome matters more than the gross amount spent. The decisive variable is the trend in the diluted number of shares outstanding. A company can repurchase shares in the market while, in the same period, issuing or allocating others to employees and executives through share-based compensation plans. In that case, some of the buyback merely offsets dilution, and the increase in remaining shareholders' economic stake is smaller than the announced spending suggests.

Buybacks and dividends are not the same

Both return capital to shareholders, but work differently. With a dividend, every shareholder receives money in proportion to their holdings. With a buyback, money goes to those selling while the company purchases; those who remain shareholders may benefit from the reduced share count and a larger percentage stake in the business.

Imagine owning 100 shares out of 10,000, or 1% of the company. If the business repurchases and cancels 1,000 shares and we do not sell, 9,000 shares remain and our 100 represent approximately 1.11%. Our stake increases without buying new shares. The tax consequences of dividends and capital gains, however, vary by country, investor type and how the shares are held, so this economic comparison does not replace a specific tax analysis.

Why do it now?

The authorisation comes while Nvidia continues growing at extraordinary rates, but the market is questioning the duration of the artificial-intelligence spending cycle, customers' capital needs and competitive pressure. One possible interpretation is that management views its own shares as an attractive use of funds at current prices. The official announcement indeed expresses confidence in long-term opportunities.

This nevertheless remains the assessment of those running the business, not a certificate of undervaluation. An exceptional company can have an overpriced share; a less impressive business can become an attractive investment if bought at a sufficiently low valuation. The announcement provides information about management's confidence and Nvidia's financial strength, but does not resolve the fundamental question of price.

The real signal to watch

The most significant aspect is not the buyback's mechanical effect on the share price over the coming weeks. It is what a decision of this kind says about the financial capacity the company has achieved. A business that can authorise another 150 billion in repurchases while continuing to invest heavily in its technology is signalling that it believes exceptional capital generation is sustainable. The latest results make this claim plausible, but do not eliminate risks related to competition, the AI investment cycle and the stock's valuation.

To assess the programme over time, I would therefore look at four factors: Nvidia's average purchase price, the net reduction in diluted shares, operating earnings growth and the alternative opportunities the company forgoes by using capital for buybacks. These figures, more than the announced amount, reveal whether remaining shareholders are receiving real value.

So is it good news?

It is certainly significant news. It demonstrates that Nvidia has extraordinary cash-generating capacity and that management is willing to allocate a substantial portion of capital to shareholders. If repurchases are made at reasonable valuations, if the number of shares actually outstanding falls and if earnings continue growing, the effect on per-share value could be substantial.

I would not, however, call the 235 billion a parachute capable of protecting the share price from falls, nor would I regard the programme as sufficient reason to buy Nvidia. The question remains the same: at what price will the company repurchase its shares, and what return could it have earned by allocating that capital to other alternatives? The largest increase in a buyback authorisation ever announced does not automatically make Nvidia an attractive investment. It does, however, make even clearer how much capital the company of 2026 can generate and distribute.

Note: this article is for information and educational purposes only. It is not a recommendation to buy, sell or hold Nvidia shares, nor does it constitute personalised financial advice.