4. Buybacks and dividends: when they create value and when they destroy it

Level 5: Competition and duration

The Theory course in this academy already explained what a buyback is and what it does mechanically to earnings per share, so that ground is covered. Here we go to the place where it is really settled whether the shareholder gains or loses, and which almost nobody looks at: the price paid. A buyback is neither good nor bad by nature. It is a purchase, and like any purchase it depends on how much you pay for what you get.

A buyback is just another investment decision

When the board puts cash into buying back its own shares, it is choosing one investment project among several. It could open a factory, buy a company, pay down debt or pay a dividend. It chooses to buy a piece of itself. The right question is identical to the one in the previous module: what return does it earn on that money? And the answer depends entirely on whether it is buying below or above what the business is worth.

Put another way: buying back shares is the best investment available when the share is cheap, and one of the worst when it is expensive. There is no third category.

The arithmetic, with concrete numbers

Take a company with 1,000 shares and an estimated intrinsic value of 100,000 euros. Each share is therefore worth 100 euros. Nothing that follows changes the business: same sales, same margins, same customers.

• Buying cheap. The share trades at 50 and the company spends 10,000 euros, retiring 200 shares with it. That leaves 800 shares and a value of 90,000. Each share is now worth 112.5 euros. Those who stayed are 12.5% better off without anyone having sold one extra product: they have bought for 10,000 something that was worth 20,000, and that difference is shared out among them.
• Buying dear. The share trades at 200 and the company buys back 50 shares. It spends 10,000 euros. That leaves 950 shares and a value of 90,000. Each share is now worth 94.7 euros. Notice that the outlay was the same in both cases, 10,000 euros, and that in both a value of 90,000 stays inside the company: the only thing that has changed is how many shares it has been spread across. Those who stayed have lost more than 5%: the company paid 10,000 for something worth 5,000, and that gift went to the ones who sold.

Look at what happens in the second case. The share count falls, earnings per share rise and the headline will say the company is returning value to shareholders. And yet the shareholder who stays is poorer. A falling share count proves nothing on its own.

The buyback that only papers over dilution

There is a variant that deserves a paragraph of its own because it is systematically misrepresented. Plenty of companies pay part of their employees' salary in shares. That issues new shares every year and dilutes the existing owners. To keep the total from rising, the company buys back on the market roughly the same amount it has issued.

When that happens, the press release will say that several hundred million has been returned to shareholders. It is false in the sense that matters. That money did not go to the shareholder: it paid in cash a payroll that had been booked in shares. The shareholder owns no larger a percentage of the company than before, has received nothing, and the cash went out all the same.

Spotting it is simple and needs nothing sophisticated:

• Look at the number of shares outstanding at the start and at the end of the year, not the amount bought back.
• If the amount bought back is enormous and the share count barely falls, the buyback was offsetting issuance.
• Compare the cash spent on buybacks with the year's share-based pay. If the two are similar, you already know what was going on.

Dividend versus buyback

The dividend has a quiet advantage: it is honest by construction. It hands out cash whatever the price and does not depend on the board getting the valuation of its own share right. Every shareholder receives the same amount per share and the payout shows up in a single line of the accounts. A buyback can be better than a dividend, far better, but only if it is done below value: that is where it concentrates the gain in whoever stays instead of spreading it across everybody.

The dividend also has a well-known drawback. It creates an expectation that is painful to cut later, because the market reads a cut as a sign of trouble, and that rigidity has pushed more than one company into keeping the payout going in precisely the years when it needed the money inside. A buyback, committing to nothing, is flexible. That flexibility is a management advantage and, at the same time, the door through which the opportunistic use you have just seen slips in.

A real case: buying back at the peak and asking for money in the trough

A well-documented pattern at the large US airlines during the good years before 2020: they put a very high share of their free cash flow, of the order of tens of billions of dollars between them over that stretch, into buying back shares with the price near record highs. When the bad year came and traffic collapsed, the sector needed public support and fresh capital, in some cases issuing shares at prices far below the ones it had paid to buy them back.

At the other extreme sits the classic case studied for decades: Henry Singleton at Teledyne, who over the seventies and the early eighties retired an enormous portion of the company's shares, buying them when they traded on very low multiples, and issued shares when they were expensive. The exact figures vary with the source, but the direction of the behavior is unmistakable, and it is exactly the opposite of the airlines'.

Keep the criterion, not the anecdotes: a buyback creates value when less is paid than the thing is worth and when the company does not need that cash to survive its own cycle. Everything else is presentation. This is educational material on how to assess capital decisions, not a recommendation about any company.