Diageo

We held this report back deliberately. Diageo's financial year ends on 30 June, and on Thursday 6 August it published its full-year results alongside a Capital Markets Day — the first full year and the first strategy under a chief executive who has been in the seat seven months. Publishing the day before would have been publishing into the dark.

Here is what came out. The full-year dividend is 50 US cents. Not 51. Not 55. Fifty — exactly the minimum floor the board had promised in February when it halved the interim payment. A company that lands precisely on its own floor is telling you something, and it is not that things are going better than expected.

For an American holding the ADR, the story needs no interpretation at all, because it is denominated in dollars with no currency argument available:

A company halved its dividend, wrote off $2.4 billion, reported falling sales — and the shares rose ten percent in two sessions. That is not a market reacting to the results. That is a market that had already priced something worse, and was relieved.

So what are you actually looking at? Diageo owns Johnnie Walker, Guinness, Smirnoff, Captain Morgan, Baileys, Tanqueray, Don Julio, Casamigos and Crown Royal, plus 34% of Moët Hennessy. It is the largest producer of international premium spirits by net sales — though not, as is often written, the largest drinks company in the world; Kweichow Moutai is bigger by both revenue and market value.

And underneath the brands sits the thing that makes this business genuinely unusual, which we will spend Part V on: $8.7 billion of spirit lying in casks doing nothing — 81% of the entire inventory — of which $5.7 billion is Scotch. That is a barrier no amount of money can cross, because it is denominated in years. It is also, as we shall see, the reason the balance sheet is under strain.

The shape of the story is worth holding on to. For roughly two decades to 2021, Diageo was one of the great British compounders — mid-single-digit organic growth, operating margins near 30%, a rising dividend every year, and a premium market rating to match. The share price peaked at $203.23 on 4 January 2022. Even after last week's rally it remains 53% below that level.

What broke was not the brands. It was the assumption that premiumisation would continue indefinitely, and the credibility of a management team that had been trusted on its numbers for twenty years and then missed badly.

Our house rule is that any dividend-paying company gets its payout examined properly — coverage on free cash flow rather than earnings, the trend over time, and what would have to happen for a cut. Diageo makes this the central section, because the cut has already happened and the question is whether it was enough.

First, a correction to the story almost everybody is telling. The popular line is that Diageo has just cut its dividend for the first time since the company was created. That is not quite right, and the real sequence is more revealing.

So the streak ended in 2025, with a freeze — and the cut in 2026 is the second signal, not the first. Put the whole sequence together and you get three escalating retreats in eighteen months: the buyback was suspended, then the dividend was frozen, then the dividend was halved. Managements do not do those things in that order by accident.

Here is the same story in the only currency that admits no argument — the actual cash an American ADR holder received:

An American who owned one Diageo ADR received two dollars and fifty-two cents in December, and eighty cents in June. Note also that the FY2024 and FY2025 payments were identical to the cent — the freeze, visible in dollars, with no exchange-rate excuse available.

Coverage collapsed from 1.84 times to about 1.1 times in three years and stayed there. A 1.1 times cover on free cash flow is not "covered" in any meaningful sense — it is technically covered, with no cushion whatsoever for a bad year, and 2023 through 2025 were bad years.

But the binding constraint was actually the earnings test, not the cash test. In fiscal 2025 Diageo earned 164.2 cents per share before exceptional items and paid out 103.48 — a payout ratio of exactly 63.0%, which was precisely the ceiling of its own stated policy. The policy was not breached. It was maxed out, with earnings falling. There was nowhere left to go but down.

The verdict: the rebased dividend is safe, and it is no longer a reason to own the shares. At 50 cents a year — $2.00 gross per ADR, a little less after the depositary fee — the forward yield on $96.35 is about 2.1%, against a trailing 3.4% that straddles the cut. Anyone buying Diageo today because the screen shows a 3.4% yield is buying a dividend that has already been abolished. ★ And note that last week's rally has made this worse, not better: the shares went up, so the forward yield went down.

★ And notice the elegance of what management actually did, because it deserves credit. The roughly $1.2 billion a year freed by the cut, plus the roughly $2.3 billion coming from the sale of the East African Breweries stake to Asahi, is close to the $3 billion of debt reduction needed to bring leverage from above 3 times down inside the 2.5–3.0 times target by fiscal 2028. The cut is not a panic. It is a calculation. Sir Dave Lewis said it himself in February: the reduction "will accelerate the strengthening of our balance sheet."

When we analysed British American Tobacco we found a genuinely high-quality franchise inside a category everybody agrees is dying, and a board that chose to keep paying its shareholders and deleverage slowly. Diageo's board looked at a similar picture and did the opposite.

Which one was right depends entirely on the answer to a single question: is spirits actually in structural decline, or merely in a cyclical trough? If it is cyclical, Diageo cut a dividend it did not need to cut and the shares are genuinely cheap. If spirits is the new tobacco, Diageo moved first and it is British American Tobacco that has a problem. That is the question the rest of this report is really about.

Keep two questions apart at all times, because conflating them is how both bulls and bears get Diageo wrong. "Can anyone take Johnnie Walker from Diageo?" and "will as many people drink whisky in 2045?" have completely different answers. This part is about the first. Part VI is about the second.

Here is the balance-sheet fact that most coverage of this company never mentions:

Four-fifths of Diageo's inventory is spirit sitting in a cask, doing nothing, for years. And $5.7 billion of it — 53% of everything on the inventory line — is Scotch alone.

The Scotch Whisky Regulations 2009 require Scotch to be matured in Scotland, in oak casks of no more than 700 litres, for "a period of not less than three years", at a minimum of 40% alcohol. But three years is only the legal floor. The expressions that actually carry the margin — Johnnie Walker Black at twelve years, Green at fifteen, the eighteen-year-olds, Blue with its rare aged stocks — represent capital committed twelve to twenty-five years ago.

You cannot decide today to sell 18-year-old Johnnie Walker. That decision was taken in 2008, by people who had to guess what 2026 would want. Diageo's premium Scotch sales this year were set by a capital-allocation decision made before the iPhone had an app store.

This is why a competitor with unlimited capital cannot close the gap. They can buy a distillery, buy grain, buy casks and hire the finest blender alive — and they still cannot have eighteen-year-old stock until 2044. There is no acquisition, no capital programme and no quantity of private equity that compresses the timeline. It is one of the very few barriers in consumer goods denominated in time rather than money. Compare a soft drink, a snack or a razor: all of them replicable inside eighteen months by a well-funded entrant.

An honest report has to say this immediately. Maturing stock rose 10.8% in a single year — and Scotch stock rose 16.4% — while net sales fell. That is not confidence. That is inventory building up because it cannot be un-built. Diageo laid down that Scotch against a demand forecast that has since been cut three times, and it cannot get the money back out for a decade.

A maturing-stock moat is a one-way bet on your own demand forecast. When volumes disappoint, the moat becomes a working-capital sink: cash goes in, nothing comes out for years, and the write-down risk sits in the out-years. It is a direct reason net debt rose while profit fell. The barrier that keeps competitors out also keeps Diageo's capital in. Both readings are true, and the analysis is better for holding them at once.

Diageo's core categories are protected from imitation by statute. Nobody, anywhere, with any amount of money, may sell a competing product called "Scotch" — it must be distilled and matured in Scotland. Tequila must be made from blue Weber agave in designated Mexican states, policed by the Consejo Regulador del Tequila. Cognac must come from the Charente.

Now notice the second edge of the same blade. The statute that makes the moat un-crossable makes the cost base un-relocatable. A tariff on Scotch cannot be engineered around: Diageo cannot open a facility in Kentucky and keep calling the product Scotch, because the law protecting it forbids exactly that. Its options collapse to three — absorb the margin hit, raise the price and lose volume, or sell less.

Compare a snack or soft-drink manufacturer facing the same tariff: it shifts a production line across the border and carries on. Diageo's moat forfeits that option by construction. One statute, two consequences, opposite signs.

⚠️ One calibration, because it is easy to overstate: on Diageo's own stated assumptions, Mexican tequila is currently exempt under the USMCA. In 2026 the trap is biting on Scotch and European liqueurs, not on the tequila portfolio.

Nobody is going to take Johnnie Walker from Diageo — not with capital, not with marketing, not with time, because time is precisely what they lack. The risk was never that a competitor would cross the moat. It is that the town inside it gets smaller. Buffett's castle-and-moat metaphor assumes the castle is worth besieging. Diageo's problem is not the moat's width; it is the castle's occupancy rate.

Four separate arguments get bundled together as "young people don't drink any more." They deserve separating, because they carry very different weights of evidence.

Guinness. A 267-year-old Irish stout — heavy, dark, slow to pour, and by any reasonable reckoning the least fashionable thing in the entire portfolio — is the standout grower in this company and has been for several years, recruiting precisely the young drinkers and the women that the bear case says have left the category. Growth has been led by Great Britain and Ireland, helped by the "splitting the G" social-media phenomenon and by the success of Guinness 0.0.

If the pond were simply draining, Guinness could not be doing what it is doing. Whatever is happening to alcohol demand, it is far more selective than "young people don't drink."

★ And at Thursday's Capital Markets Day, Diageo settled the question that had been hanging over it. Bloomberg had reported in May that the company was exploring a sale or spin-off of Guinness — valued at north of $10 billion — alongside a review of its 34% stake in Moët Hennessy. Diageo has ruled out selling either. We think that is the right call, and the irony was worth stating while it lasted: the asset management was reportedly considering selling was the only large one that was growing.

Read that table without the headline and something becomes clear: operationally, this was a decent year. Sales fell 2% and operating profit still rose 2%. The adjusted margin expanded by more than a point. Free cash flow rose by nearly half a billion dollars. Net debt came down. Europe, Latin America and Africa all grew. That is a company being gripped.

The damage is all in the exceptional items — and they tell their own story about capital allocation. The $1.5 billion of impairments falls principally on Türkiye — the Mey İçki business acquired in 2011 for around $2 billion in the emerging-markets push — and on Don Papa, the Philippine rum bought in 2022 for up to €437.5 million. Last year's write-downs were the same story in a different costume: $458 million off Distill Ventures, the incubator built to find the next big brand, and $231 million off Aviation American Gin, bought from Ryan Reynolds for up to $610 million.

Note carefully what has NOT been impaired: Casamigos and Don Julio. The tequila deals everybody criticised are intact. What Diageo has actually written off is the machine it built to find the next Casamigos — and its emerging-market shopping.

That is a genuinely counter-intuitive finding and it changes the indictment. ★ It also cuts both ways on the moat: the fact that Diageo's own brand incubator failed is evidence that new spirits brands are easy to start and very hard to scale — bad news for Diageo's capital allocation, and quietly good news for its competitive position, because the challengers are not scaling either.

Lewis's framing of the cost programme was notably careful: "These savings will allow us to invest in the turnaround without needing to reduce operating profit." ★ Notice what is absent from that sentence, and from his three stated priorities — competitive category strategies, "customer, customer, customer", and a redesign of the operating framework. There is no mention of growth, of the consumer, or of the categories. This is the language of an operator fixing a business, not of a brand-builder defending a franchise. Given what Part V says about the moat, that may be exactly the right language.

The sequence is worth stating plainly, because it is unusual. Ivan Menezes, chief executive for a decade, died suddenly in June 2023 — announced on the 5th, one day before he was due to hand over. Debra Crew stepped up early and the Latin America warning arrived five months later; she left in July 2025 by mutual agreement after roughly two years. The chief financial officer then ran both jobs for six months. Sir Dave Lewis arrived on 1 January 2026. And the chairman has changed too — it is now Sir John Manzoni, not Javier Ferrán as many sources still report.

Diageo has replaced its chief executive, its chief financial officer and its chairman inside about eighteen months. There is a benign reading — a company that needed new eyes has got them, all at once, and the new team has moved decisively on the dividend, the leverage and the cost base. There is also the plain fact that nobody currently running this company has been running it long enough to be judged. We score management a 4: not for incompetence, but because there is almost no track record to score, sitting on top of a decade of capital allocation that has just been written down twice in two years.

The usual line is that Berkshire Hathaway has never owned Diageo. That is true, and it is also the least interesting way to put it.

★ In the early 1990s Berkshire Hathaway bought 31.2 million shares of Guinness plc — a stake of roughly $302 million, at around 644 pence a share. It is widely cited as Buffett's first significant investment in a company outside the United States. Guinness plc became Diageo in 1997. Buffett did not take the Diageo shares.

Berkshire's first foreign stock was the company that became Diageo — and Buffett chose not to own the thing it turned into. We would not over-read a decision taken nearly thirty years ago for reasons that were never fully explained. But it is a genuinely striking piece of history, and it sits oddly with how rarely it is mentioned.

★ There is a modern echo, too. Berkshire disclosed a stake in Constellation Brands in early 2025 — so its appetite for alcohol brands is intact. But Constellation is overwhelmingly a beer business: Modelo and Corona, sold to a growing American demographic. Berkshire bought exposure to alcohol and chose the beer end, not the aged-spirits end. That is a revealed preference worth noticing. And Buffett's own famous preference for Cherry Coke is a red herring here — Berkshire has held Guinness, Constellation and other drinks positions. The absence of Diageo is a judgement about the business, not squeamishness about the product.

The entire valuation argument reduces to one question: which earnings number is real? Diageo earned $6.61 per ADR before exceptional items and $3.12 after them. At 14.6 times the first figure this is a reasonable price for a franchise of this quality. At 30.9 times the second it is expensive.

★ But the more immediately useful fact is what last week did to the entry point. Before the results the shares were $87.83 and the analysts' average target of $99 implied 13% of upside. After a 10% rally that same target sits 2.8% above the price. The market has already taken the re-rating that a "cheap quality franchise" thesis was supposed to deliver — and it took it in two sessions, on a set of results that included a halved dividend and $2.4 billion of write-offs. Whatever opportunity existed here, a good part of it was available last Tuesday and is not available today.

Our reading, and we will state it plainly: this year's exceptional items are largely genuine write-downs of past acquisitions rather than recurring costs — Türkiye, Don Papa, the restructuring charge for an operating framework being rebuilt once. Unlike the amortisation question we examined at Thermo Fisher, these are not an annual feature of an ongoing acquisition programme; Diageo is not currently acquiring. So the pre-exceptional figure is the more honest guide to the earning power — with the important caveat that a company writing down acquisitions two years running has not finished the exercise.

★ One thing we are discarding, and we will say why. Our own discounted-cash-flow model returns a fair value above $400 a share — nearly 370% above the price. That number is not credible: the model is choking on a year of large one-off charges and extrapolating from a distorted base. We would rather show you a broken model and discard it than quietly publish the flattering output. No valuation of Diageo should rest on a DCF taken over a restructuring year.