Over six years PepsiCo's dividend grew from $5.5 billion to $7.6 billion. Over the same six years, free cash flow did not grow at all.
Cover fell from 1.20 times to exactly 1.00. Not 1.01. One.
Nobody is going to cut this dividend. There are more than fifty consecutive years of increases behind it and a board that understands what breaking that would cost. What has ended is something quieter and, for an owner, more consequential: the ability to raise it in any way that matters.
Where the growth came from
| Six years ago | At our report | Change | |
|---|---|---|---|
| Dividend paid | $5.5bn | $7.6bn | +38% |
| Free cash flow | $7–8bn | $7–8bn | Flat |
| Cover | 1.20× | 1.00× | The cushion is gone |
That is the whole mechanism. The growth was real and it was not earned. It came out of the gap between what the business produced and what it paid out — and that gap is now zero.
From here, every further raise has to come from borrowing, from selling something, or from cash flow that has not yet appeared.
A dividend covered exactly once is perfectly safe and completely finished. Both halves of that sentence are true, and only one of them gets written about.
A snack company carrying a drinks business
The usual story — "mature consumer business, slow growth" — misses what our report actually found. This is not one slow company. It is one excellent business attached to one that has collapsed.
- Frito-Lay earns a 22% operating margin and roughly 45% of divisional profit. Crisps, not cola, are the company.
- North American beverages have fallen to a 3.86% margin. That is not a mature business; it is a business barely making money on each unit sold.
- Snack prices were cut by up to 15% and volumes stayed flat. This is the most alarming single fact in the report — the pricing power that funded the dividend for decades did not produce a volume response when it was given away.
A company with pricing power cuts prices and sells more. A company without it cuts prices and sells the same amount for less money.
If the price cut does not come back as volume, the margin is simply gone — and the free cash flow that was already flat has to absorb it. That is how a cover ratio goes from 1.00 to below 1.00 without anything dramatic happening.
A high yield standing still, against a low one moving
And the crossover is further out than people expect
At our reports, PepsiCo yielded 4.1% near its 52-week low and Coca-Cola 2.4%. The obvious move is the bigger number. Work it through before you make it.
A 4.1% yield growing 0% against a 2.4% yield growing 5% a year: the smaller yield does not overtake until about year eleven. For a decade, the high yield pays you more every single year. After that, the growing one pulls ahead and never looks back.
So this is not a question about which company is better. It is a question about how long you intend to own it — and that is a question about you, not about them.
It is happening elsewhere, in plain sight
NextEra built its reputation on roughly 10% annual dividend growth. Buried in the guidance is a step down to 6% from the end of 2026 — disclosed rather than announced, and almost nowhere written about.
Same shape, different industry. The dividend is not in danger. The thing people actually bought it for has been reduced by 40%, and the share price is still quoted against a yield that describes the past.
Three checks, and one habit
Free cash flow divided by dividends paid, five years in a column. A rising dividend on flat cash flow is a falling line, and the falling line is the story — the dividend chart on its own shows a perfect staircase right up to the day it stops.
Fifty years of increases tells you about the past fifty boards. Three consecutive raises of 2% tells you what the current one thinks it can afford, and that is the only forward-looking information in the whole record.
Find an instance where the company changed price and look at what volumes did. A price cut that does not buy volume, or a price rise that loses it, is the moat being measured — and it decides whether cover recovers or keeps sliding.
A fifty-year streak is fifty decisions that have already been made. Not one of them binds the next board, and the only one that affects your return from here is the one that has not happened yet.
Treat the record as evidence about the culture — which is genuinely useful — and never as a forecast. The companies in this article will almost certainly keep paying. The question was never whether they pay. It is whether the payment grows, and for both of them the honest answer has changed.
One covered exactly once, one yielding less and still growing, and one that just cut its growth rate by 40%.
Coverage, margins and pricing power, worked through for each.
Frequently asked
Is PepsiCo's dividend safe?
At our August 2026 report, yes — and that is not the same as healthy. Free cash flow had been flat at $7–8 billion for six years while the dividend grew from $5.5 billion to $7.6 billion, taking cover from 1.20 times down to exactly 1.00. A dividend covered precisely once is paid in full and leaves nothing behind it, so the payment is secure while the increases have run out of room.
What does it mean when dividend cover reaches 1.0?
That every dollar the business generates is going out of the door. The dividend is still funded, which is why these companies rarely cut — but any future raise has to come from somewhere else: borrowing, asset sales, or cash flow that has not yet appeared. The streak can continue for years on token increases while the real payment stands still.
Is a high yield with no growth better than a low yield that grows?
It depends entirely on your horizon, and the crossover is further out than most people expect. A 4.1% yield that does not grow stays ahead of a 2.4% yield growing 5% a year until about year eleven. Before that the high yield wins; after it, the growing one wins and keeps winning. Neither is right in the abstract.
Why do companies keep raising a dividend they cannot afford to raise?
Because the streak itself has become the asset. Fifty consecutive years of increases attracts a particular shareholder base and sits in index definitions, so breaking it is costly in a way that has nothing to do with the business. The usual result is a token raise — enough to preserve the record, too small to matter to the owner.