Are Stock Buybacks Good for Shareholders? Singleton's Teledyne Test

Stock buybacks are good for shareholders when a company pays less for its own shares than the business is worth, and bad for them when it pays more. That is the whole answer. A buyback is not a gift to owners, a vote of confidence, or a polite substitute for a dividend. It is a purchase, and a purchase is only as good as its price.

Nobody demonstrated that more completely than Henry Singleton, the engineer who founded Teledyne in 1960¹ and ran it for roughly three decades. He is the rare executive who did both halves of the trade at scale — selling his company's stock when the market paid handsomely for it, then buying back the great majority of it when the market would barely pay anything at all. What follows is a historical illustration of that decision, not a recommendation to buy or sell any security, and not a view on any company carrying the Teledyne name today.

The decade when Teledyne's own stock was the expensive thing to sell

Singleton was not a financier by training. He held a doctorate in electrical engineering from MIT and spent 1954–1960² running a division at Litton Industries, one of the original conglomerates. When he and a colleague started Teledyne, the plan was to assemble a diversified industrial and technology company by acquisition, and the currency for those acquisitions would be Teledyne's own shares.

The 1960s were kind to that plan. Investors of the era prized conglomerates — companies built from many unrelated businesses under one listing — and Teledyne's stock traded at a rich multiple of its earnings, meaning the market was willing to pay a high price for each dollar the company earned. Singleton treated that high price for what it was: an opportunity to sell something expensive. Over the decade Teledyne acquired roughly 130 businesses,³ paying for most of them with newly issued stock. Each deal swapped a small slice of a richly valued company for a whole operating business valued modestly. He was using an expensive currency to buy cheap assets.

Then, near the end of the decade, he stopped. Enthusiasm for conglomerates faded, Teledyne's own multiple compressed, the prices asked by sellers rose, and paying with stock no longer favoured the buyer. Singleton ended the acquisition program and dismissed the team that ran it. He did not chase a few more marginal deals to keep the growth story alive for the financial press.

That halt matters to the buyback story, because it shows he was never sentimental about his share price. He did not believe Teledyne stock was intrinsically worth issuing or intrinsically worth hoarding. He believed it was worth issuing at one price and worth buying at another, and he organised the company's capital around the difference. Founders who have sold equity in a private company know this from the other side of the table: the price at which ownership leaves the building is not a detail of the deal. It is the deal.

With the issuance window closed, cash began to accumulate. The businesses Teledyne already owned were durable and cash-generative, producing more than they needed to reinvest. The question became what to do with the surplus.

Eight tender offers, and nine-tenths of the company retired

The reversal came in 1972. The market that had once paid richly for conglomerate earnings now wanted nothing to do with them. Measured against the cash its businesses were throwing off, Teledyne's shares had fallen to a fraction of what they appeared to be worth. Singleton concluded that the cheapest good business available to Teledyne was Teledyne.

So he bought it. Between 1972 and 1984 the company ran eight tender offers⁴ — a tender offer being a public invitation to shareholders to sell their shares back to the company at a stated price by a stated date — and, with open-market purchases alongside, retired roughly 90 percent of its outstanding shares.⁴

The first tender is instructive. Teledyne offered to buy about one million shares. Nearly nine million were offered up by holders⁴ eager to sell into a brutal bear market, and Singleton took them all. The stock, rather than rising on the news, drifted lower — so he tendered again. Each time an offer closed and the quotation sagged, another wave of willing sellers appeared.

It would flatter the story to say every purchase was a steal. A programme run over twelve years is never that uniform. The offers were not all identical in structure — some exchanged company bonds for stock rather than sending cash out the door — and the later ones were struck at higher prices as the shares recovered. The discipline was not that every purchase was cheap. It was that price governed the decision every time, measured against what the operating businesses were actually producing, rather than against a schedule set in advance.

It is worth pausing on how contrarian this was. In the 1970s, buying in one's own shares was unusual and faintly disreputable. Conventional wisdom held that a company with surplus cash should pay a dividend or make an acquisition, and shrinking the share count read as a confession that management had run out of ideas. Teledyne paid no cash dividend through the heart of the programme, beginning one only late in Singleton's tenure.⁵ His reasoning was consistent: a dividend is taxed on the way out and hands the same value to every holder regardless of price, whereas buying stock at a low price gives every remaining shareholder a larger piece of the company for the same cash. Holders who wanted income could sell a few shares into the tenders. Those who stayed owned a steadily larger share of everything Teledyne produced. This publication has made the broader case that free cash flow matters more than yield for owners who now live on a portfolio; Singleton made it in practice, a tender at a time, for more than a decade.

Warren Buffett has described Singleton's record of running businesses and deploying their capital as the best in American business⁶ — a judgment that rests less on any single year than on the coherence of the whole. He issued when the stock was dear and bought when it was cheap, and did nothing in between that undercut either move. Teledyne itself was broken into more focused pieces in the 1990s,⁷ which is its own lesson about the limits of the conglomerate form, but it does not disturb the arithmetic of the years in between.

Why a repurchase is a trade between the owners who sell and the owners who stay

The mechanism behind Singleton's result is simple enough to hold in one hand, and once held, most of the public argument about buybacks resolves itself.

A share is a claim on a business's future free cash flow — the cash left after paying for everything needed to keep the business running and competitive. When a company buys shares and retires them, the total cash the business generates does not change, but the number of claims on it shrinks. Each remaining share represents a larger fraction of every dollar the business will ever produce. And because long-term returns are driven by the growth of that cash stream over time, a cheap repurchase has a quiet compounding effect: the growth runs through a smaller base of owners. The buyback does not invent the cash. It decides who owns the next dollar of it.

That is the good news, and it is real. But it is only half the equation, because the company parted with cash to make it happen. The question is whether the fraction of the business acquired was worth more than the cash paid for it.

Buffett once put the arithmetic in three lines, and it is worth borrowing. Imagine three equal partners in a business worth $3,000; each stake is worth $1,000. The partnership buys one partner out for $900. The two who remain now share a business worth $2,100, so each holds $1,050 — an immediate gain of $50, though neither of them did anything. Buy the same partner out for $1,100 instead, and each remaining partner is immediately $50 poorer. Nothing about the appearance of the two transactions differs. The share count falls in both. Only the price differs, and it differs completely.

This is why a buyback is best understood as a transfer between two groups of shareholders — those who sell and those who stay — rather than a payment to shareholders as a class. When the price is below value, the stayers gain at the expense of the sellers. When the price is above value, the sellers gain at the expense of the stayers. Management's job is to be on the right side of that transfer on behalf of the owners who remain. Singleton was on the right side because the decision was never about the share price in itself; it was always about the share price relative to the cash underneath it.

Owners who have sold a business have run this calculation in private. Most have at some point bought out a co-founder, a departing executive or an early investor, and nobody in that room described the cheque as "returning capital." They called it a purchase and negotiated the price hard, because overpaying the departing party came directly out of the pockets of those who stayed. A public-company buyback is the same transaction with more zeros and a stock-exchange quote in the middle. The quote is not the value. It is the asking price. The gentler language survives mainly because the people negotiating the price — management — are usually not the people whose pocket it comes from.

A sceptical reader will object here, and the objection is fair. Fewer shares lift earnings per share — profit divided by the number of shares — even when the business is unchanged and the price paid was far too high. Per-share figures rise, bonus targets tied to them are met, and the remaining owners are poorer in economic substance while the reported numbers say the opposite. It is one more reason the headline profit figure and the cash a business actually produces deserve to be read separately. If a repurchase can manufacture a prettier number, perhaps it is cosmetics rather than capital allocation. Sometimes it is exactly that. The Teledyne test does not deny the abuse; it names the conditions under which the abuse is absent. And it is telling that repurchase announcements almost never state a price above which the company will stop buying — a discipline no management team would dream of omitting when bidding for an outside business.

Three ways a repurchase copies Teledyne's shape and none of its substance

If the principle is this simple, why do so many programmes end badly? Because the arithmetic is simple and the institutional discipline is rare. Three specific failures turn a programme that looks like Teledyne's on the surface into its opposite.

The first is buying dear. Companies have the most spare cash when business is good, and business is good when the stock is expensive. The result is that corporate repurchases in aggregate have historically been heaviest near market peaks and thinnest near troughs — precisely backwards from Singleton's pattern. Some of the largest American banks repurchased shares aggressively in the years before the 2008 financial crisis, then were forced to raise fresh capital by issuing new shares at far lower prices than they had just paid. They bought dear and sold cheap, and every dollar of that sequence moved value from the owners who stayed to the owners who left.

The second is borrowing to do it. Singleton paid with cash the operating companies had already earned. A company that borrows to retire shares is running a different transaction: it is swapping a flexible claim for a fixed obligation. That can be rational when the shares are genuinely cheap and the debt is small relative to the cash the business reliably produces. It is dangerous when neither condition holds, because the obligation does not adjust to the cycle the way an equity claim does. The share count falls either way; the resilience of the business does not.

The third is buying to offset rather than to retire. Where a company issues shares to employees and then repurchases roughly the same number in the market, the share count barely moves while the cash genuinely leaves. That is a way of paying compensation, and it may be perfectly sensible, but it should not be counted as capital allocation. The plainest check is also the least glamorous: read the change in shares outstanding over several years rather than the dollar figure in the announcement.

Underneath all three sits the objection heard loudest in political seasons — that cash spent on shares is cash taken from workers, factories and research. Sequence matters more than slogan. A management team that defers a necessary machine or cuts genuine research to fund a repurchase is not following Singleton; it is liquidating the business slowly in order to decorate the share count. Singleton's programme ran the other way around. The operating companies were funded first, and the cash spent on tenders was the surplus that remained after the businesses had what they needed to stay durable. When a company has already paid for its own future and still produces more cash than it can use, that cash has to go somewhere. Buying more of a good business at a good price is what any private owner would do with it.

What the announcement never says, and what the share count does

The practical consequence for anyone assessing a company — or listening to a manager explain why they own one — is that the announcement is not the event. Most repurchase programmes are authorisations: board permission to buy up to a stated amount over a stated period. Permission is not a purchase. Programmes are announced and quietly left unused; others are executed enthusiastically at the worst possible prices. The only record that tells the truth is the share count over time, set beside the price at which the buying actually happened and the cash the business was producing while it happened.

That gives owners a short set of questions that travel well beyond one conglomerate in the 1970s. Were the operating needs of the business funded first? Was the buying done when the shares were cheap relative to the cash underneath them, or when the cash was simply available? Did the share count actually fall? And did the company use borrowed money to do it? A programme that passes all four is a management team behaving like a partner. One that fails them is a gesture, paid for by the owners who stayed.

None of this makes repurchases good or bad as a category, which is why the category argument that resurfaces every earnings season tends to generate more heat than light. Buybacks do not create value; the growth of free cash flow does. A repurchase merely decides how many owners that growing stream has to be shared among — and at what cost. Bought cheaply, it concentrates decades of compounding into fewer hands. Bought dear, it is a slow leak dressed as a return of capital. The difference is price, which is the same thing every owner who has ever sold a business, or bought out a partner, already knows in their bones.

More of this thinking sits in the firm's capital allocation library, which collects the pieces on how cash gets deployed and what separates the deployment that compounds from the deployment that merely looks busy.