Will This Dividend Hold? A Five-Step Check
Abbott has paid a dividend for 43 consecutive years. General Mills has paid for 43. PPL has paid for 42. On the one number almost every dividend list publishes, those are the same record.
They are not the same dividend. One of the three has raised its payment in 12 consecutive years, one in 6, and one in 3. Two generated more cash per share than they paid out across their last three completed financial years. One averaged negative free cash flow per share over the same window and kept paying anyway.
This is the check that separates them, in the order the evidence actually arrives.
Why does a 43-year record tell you so little?
Two different numbers get called a streak, and lists tend to publish the flattering one.
Years paid counts uninterrupted payment. Years increased counts consecutive raises. A company that froze its dividend for a decade and then resumed raising it keeps the first number and loses the second. Which of the two a site chooses to print is one of the reasons dividend growth figures disagree between sites.
The gap between them is not a rare accounting curiosity. Of the 582 payers on DividendAtlas that publish both figures, 417 have paid for more years than they have raised. For 199 of those, the gap crosses the ten-year line that most "dividend achiever" screens are built on: paid for at least ten years, raised for fewer than ten. The median gap, among the companies where the two disagree, is 17 years.
So a 43-year record is a fact about survival, not about strength. It tells you the dividend has never been cancelled. It tells you nothing about whether the company can still afford it, which is a question about this year's cash.
What should you check first?
Cash coverage, before anything else.
A dividend is paid in cash. Accounting profit is not cash, and the two separate for entirely ordinary reasons: depreciation, working capital swings, the timing of a large capital project. A company can report a healthy profit and still be funding its dividend from borrowings.
The comparison to make is the dividend per share against free cash flow per share, which is operating cash flow after capital spending. If the second number is smaller than the first, the shortfall is coming from somewhere. Debt, asset sales and cash reserves are all finite.
We treat this as the first check rather than the most sophisticated one because it fails loudly. A company whose earnings payout ratio is comfortable and whose cash coverage is not has already told you which of the two figures to trust. The full comparison of the two payout ratios sets out where they diverge and why the cash version moves first.
Does a high payout ratio settle it?
Less often than people expect.
The earnings payout ratio is the dividend divided by earnings per share. Above 100%, a company paid out more than it earned. That sounds decisive and frequently is not, because a single accounting charge can produce it without touching the cash position. Litigation provisions, goodwill write-downs and restructuring charges all land in earnings and none of them is a cash payment in the year it is booked.
What matters more is the direction. A ratio at 90% that was 55% four years ago is a different situation from a ratio that has sat near 90% for a decade. The first is a trend; the second is a policy. A ratio above 100% is worth reading carefully rather than reacting to, and the window you measure over changes the answer as much as the level does.
Which signals arrive early, and which arrive late?
The evidence for a cut does not arrive all at once, and ordering it is most of the skill.
- Cash coverage weakens first. It is a measurement of this year's operations and it moves before anything a board decides.
- Leverage rises next, because the gap has to be funded, and debt is the usual funding.
- Dividend growth stalls before it reverses. A board that is worried holds the payment flat for a year or two before cutting it. A frozen dividend after a decade of raises is a decision, not an oversight.
- The yield spikes late, once the market has drawn its own conclusion.
- Management language shifts last, or close to it, and by then the information is in the price.
A freeze is the most useful of these in practice, because it is unambiguous and public. Nobody freezes a dividend by accident. It is also a weaker signal than a cut, and the two look identical in a growth streak: we measured how far apart they really are. The seven signals, ranked by how much warning each one gives, covers the full sequence.
Is a high yield evidence of anything?
It is evidence that the price fell. Whether that matters depends entirely on why.
A yield is the dividend divided by the price, so it rises when either number moves. A company that has raised its dividend 8% a year for a decade and a company whose shares have halved can arrive at the same 6% yield from opposite directions.
The question to ask is not whether 6% is high. It is whether 6% is high for this company, against its own history and its own sector. A utility at 6% and a software company at 6% are not the same claim. The yield-trap checklist works through how to tell a repriced business from a cheap one.
Running the check on three companies
Abbott, General Mills and PPL are all judged on the same terms by our score: all
three carry the Defensive profile, so the comparison below is like for like.
Their records are close to identical. Almost nothing else is.
| Abbott | General Mills | PPL | |
|---|---|---|---|
| Years paid | 43 | 43 | 42 |
| Years increased | 12 | 6 | 3 |
| Dividend growth, 5y | +10.4% a year | +3.9% a year | -8.0% a year |
| Dividend per share | 2.52 USD | 2.44 USD | 1.14 USD |
| Free cash flow per share, 3y average | 3.59 USD | 3.82 USD | -1.12 USD |
| Earnings payout ratio | 51.5% | 88.6% | 90.3% |
| Yield | 2.20% | 6.07% | 3.26% |
| Dividend Health Score | 86, very_safe | 65, safe | 42, borderline |
All figures as of 27 August 2026.
Abbott Laboratories (ABT.US)Data as of 2026-08-27- Price
- $112.04
- Dividend yield
- 2.20%
- Annual dividend
- $2.52
Key statistics
Abbott passes every check in this article. It pays 2.52 USD against 3.59 USD of
free cash flow per share, its earnings payout ratio is barely above half, and and it
has raised the payment for 12 straight years, at 10.4% a year over the last
five. Its score of 86 puts it in the very_safe bucket, with high confidence.
General Mills Inc (GIS.US)Data as of 2026-08-27- Price
- $40.51
- Dividend yield
- 6.07%
- Annual dividend
- $2.44
Key statistics
General Mills is the interesting one. The 6.07% yield is nearly three times
Abbott's, and the payout ratio of 88.6% is high enough to look alarming beside
it. But cash coverage holds: 2.44 USD paid against 3.82 USD of free cash flow per
share. That is why it scores 65 and sits in safe rather than lower. Growth has
slowed to 3.9% a year and the streak of raises stands at 6 years against
Abbott's 12. That is a real weakening, and it is a different problem from an
unaffordable dividend.
PPL Corporation (PPL.US)Data as of 2026-08-27- Price
- $34.99
- Dividend yield
- 3.26%
- Annual dividend
- $1.14
Key statistics
PPL is where the 42-year record stops meaning what it appears to mean. Free cash
flow per share averaged -1.12 USD over its last three completed financial years,
against a dividend of 1.14 USD. The five-year dividend growth rate is negative at
-8.0% a year, so the payment today is smaller than it was five years ago, inside
a record that has never been interrupted. Its score of 42 lands in borderline.
The utility caveat matters here, and it cuts both ways. Regulated utilities routinely spend more on their rate base than they generate in operating cash, and doing so is how they grow the asset base their allowed returns are set on. Negative free cash flow is therefore normal for the sector and is not, by itself, a cut warning. What it does mean is that the dividend depends on continued access to capital markets, which is a different risk from an operating one.
Where does a score fit?
It does the sweep. It does not do the judgement.
Running five checks by hand on one company takes an afternoon. Running them across a catalogue of several thousand takes a system. That is what our Dividend Health Score does. It reads coverage, balance-sheet strength and payment record, then combines them into a number from 1 to 99 and one of five buckets. A confidence level comes with it, so a thin-data company cannot pass itself off as a well-documented one. How the score is built, and what each pillar reads, sets out the method. The related argument for why quality beats headline yield is the same case made from the portfolio side.
Calibration helps here, and it is the kind of thing a single company page cannot
give you. Of the 720 companies on DividendAtlas that currently carry a score,
190 sit in very_safe and 343 in safe. A further 148 are borderline, 38 are
risky, and one is very_risky. General Mills at 65 is therefore not a warning
sign. It is an ordinary result, and the work is in knowing which part of the
check pulled it down.
Use the score to narrow several thousand companies to 20. Then run this article on the 20.
The odds shift sharply as the yield rises, and we have
measured how far: 91% of payers yielding
under 2% are rated safe or better, against 23% of those above 6%. The record
is just as slippery on its own: 21 companies with an unbroken 25-year payment
history are paying less than they did five years
ago. And once you know a dividend
will hold, when the cash actually arrives
is a separate question with a lumpier answer than most expect. Property companies
run on a different measuring stick again, and
whether that makes them less safe has a
clearer answer than the debate suggests. And when a company carries
no score at all, this
article is what you run instead.
What none of this can tell you
Every check above reads filed financial statements, and filed statements are backward-looking.
A score cannot read a loan covenant that steps in at a leverage threshold. It cannot read a regulator's next rate decision, an acquisition announced this morning, or a new chief executive who wants to fund a buyback instead. It cannot see a controlling shareholder who needs the income and will defend the dividend past the point the numbers justify.
PPL is a live example of the limit rather than a failure of the method. Its negative free cash flow is a genuine finding, and reading it correctly needs a fact the statements do not contain: whether the regulator will keep allowing the returns that make the spending worthwhile. No ratio answers that.
Those are the cases where a dividend gets cut with the ratios looking acceptable, and where one gets defended long after the ratios stopped supporting it. The check narrows the field and tells you where to read the annual report. It does not replace reading it.
Frequently asked questions
- What is the single best predictor that a dividend will be cut?
- Cash coverage. A company can report accounting profits and still fail to generate the cash the dividend needs, and the cash figure moves first. Compare the dividend per share against free cash flow per share before looking at anything else.
- Does a long dividend record mean the dividend is safe?
- No. A record tells you what a company has done, not what it can still afford. Of the 582 payers on DividendAtlas that publish both figures, 417 have paid for more years than they have raised, and 199 have paid for at least ten years while raising for fewer than ten.
- Is a payout ratio above 100% always a warning?
- No. A one-off accounting charge can push the ratio above 100% without changing the cash position at all. The direction of travel over several years matters more than the level in any single year.
- Why does a high yield make a dividend look riskier?
- A yield rises when the price falls. A yield well above a company's own history usually means the market has repriced the shares, and the reason for the repricing is what you need to find.
- Can a dividend safety score replace this check?
- It can replace the sweep, not the judgement. A score reads coverage, the balance sheet and the record across thousands of companies in a way no one can do by hand, but it cannot read a contract, a regulator or a management team's intent.
DividendAtlas provides data and research for informational purposes only. Nothing here is investment advice or a recommendation to buy or sell any security. Always do your own research.