The Funniest Thing Companies File Under 'Transaction in Own Shares'

A public company buys its own stock. That is a momentous thing to do. You are taking real cash, sending it into the market, and converting it into equity that you then cancel. Fewer shares. More earnings per share for the people who are left. The capital structure shifts.

Yet if you are listed on the London Stock Exchange and someone is writing that story down for regulators, the announcement is filed under the headline "Transaction in Own Shares."

That phrase — "Transaction in Own Shares" — is the most boring label in finance, attached to one of the most consequential choices a company's board can make. It is the regulatory equivalent of calling a heart transplant "procedure involving chest cavity." Technically accurate. Emotionally inert. It hides the incentives.

The basic point is this: when you see "Transaction in Own Shares" on a filing, what you are actually reading is a company's capital-allocation confession. Management is telling you that, after considering capital expenditure, acquisitions, dividends, debt reduction, and everything else, they decided the best thing to do with the company's cash was buy the company itself. The label is dry because the regulators want it that way. But the plumbing inside is anything but.

"Transaction in Own Shares" is a specific disclosure category under the UK's Disclosure and Transparency Rules. It covers purchases, sales, redemptions, cancellations, and transfers of a company's own shares. Any listed company that does one of these things has to file an announcement with that headline. It is the same category whether you bought 20,000 shares or 2 million.

For the reader, the filing is basically a receipt. It tells you how many shares, at what price, on what date, and on which exchange. You'll see entries like "XPAR" (Euronext Paris), "CEUX" (Cboe Europe), "AQEU" (a multilateral trading facility), and so on — the venues where the buying happened. The detail tells you something about the company's effort to hide its footprint: spreading purchases across multiple platforms and multiple days is a way to avoid moving the market against yourself while you're still buying.

If you want to see the machine in action, look at TotalEnergies in July 2026. The French energy company filed "Transaction in Own Shares" announcements for two separate five-day buying windows. Between July 20 and 24, it repurchased 1.45 million shares for €107.5 million, across four European exchanges. Then between July 27 and 31, it did it again: another 1.67 million shares for roughly €125 million. That is about €230 million in buybacks over ten trading days.

The purchase prices climbed steadily through that first window — from €70.97 on July 20 to €76.07 on July 24, a 7 percent increase in five days. TotalEnergies kept buying even as the price rose. That is not accidental. If management believes the stock is worth more than the market is paying, rising prices during a buyback are not a reason to stop. They're a sign you're not buying enough.

But here is the weirder layer, and it's the one that actually matters for understanding the machine.

A lot of buybacks don't happen the way "Transaction in Own Shares" makes them sound. They don't happen as a company's treasury department clicking "buy" on a terminal at 10 a.m. and being done with the day. For the big ones, companies use something called an Accelerated Share Repurchase, or ASR.

An ASR is a buyback wrapped inside a bank contract. Here is how it works. A company agrees to pay, say, $2 billion to an investment bank. The bank immediately delivers most of the shares — typically 80 to 85 percent of the expected total — by borrowing them from institutional investors (pension funds, mutual funds, whatever) and handing them to the company. The company retires those shares right away. The share count drops on day one. Earnings per share jump on day one.

Then the bank goes into the open market and buys the shares it needs to return to the original lenders, working through a three-to-six-month settlement window. The final number of shares the company gets is determined by the volume-weighted average price of the stock over that period. If the stock went up during the window, the company gets fewer shares than the initial tranche implied. If it went down, the company gets more. The bank prices this risk into the deal upfront through a discount built into the VWAP calculation.

In other words, the ASR is a company paying cash today for shares it might not know the exact count of for three months. The bank is short the stock, hedging it gradually in the market, and collecting a fee for absorbing all the execution risk. The company gets an immediate EPS bump and a 10b5-1 safe harbor against insider-trading claims, because the transaction follows a pre-set contract rather than management's real-time decisions.

It is old finance in new costume. This is basically the same mechanic as a tender offer — a company making a large, immediate commitment to buy back stock — but routed through a bank's derivatives desk so the accounting looks cleaner and the settlement is private.

Boston Scientific used this exact structure in May 2026, announcing a $2 billion ASR with JPMorgan. Sallie Mae did the same with a $200 million ASR in March. They're not unusual anymore. They're plumbing.

Now the picture gets more interesting because of who is buying back the most.

Berkshire Hathaway resumed buying its own shares in March 2026 under new CEO Greg Abel. In Q1, the amount repurchased was $235 million — which is a rounding error for a company with a market cap above $1 trillion and a cash pile of roughly $397 billion. But then Q2 hit. Between mid-April and mid-July, the Class A share count fell by about 11,000 shares. At an estimated average price of roughly $721,000 per share, that works out to as much as $11 billion in buybacks for the quarter. That would be the largest buyback in Berkshire's history, beating the previous record of $9 billion in Q4 2020.

Berkshire shares were down about 3 percent year-to-date and roughly 10 percent from their May 2025 high when the program restarted. The stock was trading at around 1.5 times book value. Berkshire's own stated policy is to buy back only when the price is below conservatively estimated intrinsic value — and the buyback requires consultation between the CEO and Chairman Warren Buffett.

When the world's largest value fund starts buying itself at scale, it is a signal. Not a cute one. A structural one. The person whose entire career has been built on identifying mispriced assets is now saying his own company's shares are mispriced and pouring cash into them.

Abel personally bought $15 million of Berkshire stock — equal to his after-tax annual salary — and committed to doing the same every year for as long as he leads the company. He said he expects to lead for 20 years. That is not theater. That is the kind of personal alignment that was routine at elite firms before the 2010s and has become a genuine outlier again.

The simple model here is this: Berkshire has a mountain of cash, few attractive acquisition targets at current prices, and a share price below what its leadership believes the business is worth. So instead of letting the cash sit or chasing deals at inflated multiples, it buys itself. The share count drops. The book value per share rises. The remaining shareholders own a larger piece of a machine that keeps printing insurance float and compounding its portfolio, which approaches $360 billion in marketable equities.

It is financial engineering, sure. But it is financial engineering performed by a company whose entire reputation is built on not doing financial engineering. That's the oddness.

So what is "Transaction in Own Shares" really doing?

On the surface, it is a disclosure rule. Underneath, it is a window into how companies think about the value of their own equity relative to every other use of cash. When TotalEnergies files one of these notices, it is telling you that after weighing the energy transition, debt reduction, capital expenditure across 120 countries, and a €3.40 annual dividend per share, the board still had money left over and decided the best buyer was the company itself.

When Berkshire files an equivalent disclosure through the SEC, it is telling you that the person who spent five decades looking for undervalued businesses concluded that the undervalued business in the room was the one he runs.

The boring label matters because it flattens the significance. "Transaction in Own Shares" sounds like routine administration, the financial equivalent of a memo about office supplies. But it is actually a company voting with its balance sheet that its shares are the best investment it can find. Not the best investment in the world, necessarily — but the best one it can execute today, at today's price, with today's cash.

That is not a trivial thing. It is the closest thing a public company has to saying, out loud and with a check attached, "We think we are worth more than you do."

The machine only works if you keep buying at prices below intrinsic value. If you keep buying at prices above it, you are not returning capital to shareholders — you are redistributing it from long-term holders to short-term sellers, which is a different sort of transaction entirely. The label doesn't distinguish between the two. The reader has to do that work.

Anyway, the economic point is simple. When you see "Transaction in Own Shares," don't read it as paperwork. Read it as a confession. Management looked at every other option and decided the company itself was the bargain. The question is not whether that is a nice gesture. The question is whether management is right about the price.