Gross Margin Over Three Years: What a One-Point Slide Costs
Take a company with a 40% gross margin and a 15% operating margin. Let the gross margin slip one point to 39% on the same sales. Operating profit does not fall by one point out of 40, or 2.5%. It falls from 15 to 14, which is 6.7%. That gap between a “small” slide and a real hit to earnings is the reason I check gross margin before I look at almost anything else in an income statement.
Gross margin is the share of each sales dollar left after the direct cost of making or delivering the product. It is the first line where pricing power, input costs and mix show up together. I want three years of it, not one quarter, because a single quarter can be noise from a freight bill or a write-down, while a three-year drift usually says something about the business.
What follows is how I read it, using ten large US companies whose annual gross margins I pulled from my own data set as of September 18, 2026.
The arithmetic of a one-point slide
Start with the sales side. A company that moves from a 40% to a 39% gross margin keeps 39 cents of every dollar instead of 40. To earn the same gross profit dollars it needs sales to rise by 40 divided by 39, or about 2.6%. So a one-point slide means the company must grow just to stand still.
Then look at what sits below gross profit. Operating expenses, meaning salaries, marketing and research, mostly do not shrink when the gross margin does. Take $100 of sales, a 40% gross margin ($40) and $25 of operating costs, which leaves $15 of operating profit. Slip to a 39% margin and you keep $39, still pay $25, and end with $14. One point, gone from a $15 base.
The higher the operating margin, the smaller that effect. The lower it is, the more brutal. A business earning a 5% operating margin that loses a point of gross margin loses a fifth of its operating profit. This is why thin-margin retailers and distributors watch this line the way a pilot watches a fuel gauge.
Ten companies, three years
Here is what the data says for fiscal 2022 against fiscal 2025. I picked names across sectors on purpose, so the table shows how different the numbers look from business to business.
| Company | Gross margin FY2022 | Gross margin FY2025 | Change (points) |
|---|---|---|---|
| Newmont | 27.4% | 53.2% | +25.8 |
| Southern Company | 37.0% | 48.5% | +11.5 |
| Ecolab | 37.8% | 44.5% | +6.7 |
| Apple | 43.3% | 46.9% | +3.6 |
| Waste Management | 37.6% | 40.4% | +2.9 |
| Costco | 12.1% | 12.8% | +0.7 |
| Microsoft | 68.4% | 68.8% | +0.4 |
| Cigna | 12.4% | 9.0% | -3.5 |
| Mondelez | 35.9% | 28.4% | -7.5 |
| Nucor | 30.1% | 11.9% | -18.2 |
Read the extremes first. Newmont more than doubled its margin in three years, and Nucor lost about three fifths of its own. Neither is a story about management quality alone. Gold and steel are commodities, and the margin of a commodity producer is mostly the spread between a price it does not set and costs it only partly controls. Newmont’s annual reports on SEC EDGAR are where I would go to separate the gold price from the effect of its acquisitions, and I have not done that split here. Nucor’s 10-K filings would show the same for steel pricing and mix.
Starting-year traps in the table
I would not trust the table until I have moved the start date. Here is what happens when I do.
Nucor’s fiscal 2022 margin was 30.1%. Its 2020 margin was 11.1%. So the “slide” to 11.9% in 2025 is really a return to where the company stood before the steel boom, which peaked at 30.2% in 2021. Anyone who wrote in 2022 that Nucor had a 30% gross margin was describing a peak, and anyone who writes in 2025 that the margin collapsed is describing the same peak from the other side.
Southern Company shows it in the other direction. Its 2022 margin of 37.0% was a dip. In 2021 it was 43.5%, and it reached 49.9% in 2024 before easing to 48.5%. Measured from 2021, the three-year change is not +11.5 points but about +5 to the current level. Same company, same data, and the answer halves depending on where I start the clock.
That is the reason I always look at the whole series, not only the endpoints. Waste Management is a good example of a series that behaves: 38.6% in 2020, 38.0% in 2021, 37.6% in 2022, then 38.3%, 39.3% and 40.4%. The dip in 2022 and the climb after it fit a story in which prices outran costs once inflation settled, and the fact that the climb held for three straight years is what convinces me, not the endpoint gain.
What I conclude from each pattern
I sort what I see into four cases, and the sorting matters more than the number.
A steady or rising margin alongside rising sales points to pricing power or better mix. Apple is my cleanest example in this set: gross margin from 43.3% in 2022 to 46.9% in 2025, with each year higher than the last. I wrote about why the services mix matters more than unit growth for the same company, and this series is the accounting side of that argument. It does not prove the mix shift is the cause, but it is what the shift would look like.
A falling margin alongside rising sales means the company is buying growth or cannot pass on costs. Mondelez is the name on my list that raises this question. Its margin dropped from 39.1% in 2024 to 28.4% in 2025, a fall of 10.7 points in one year, which is far too large to blame on mix. Cocoa prices are the obvious suspect for a chocolate maker, and I would check the earnings call before saying so with confidence. A fall that size, in one year, at a food company is the sort of thing that makes me stop everything else.
A falling margin alongside falling sales is the pattern I fear most, because both price and volume are weak. I did not find a clean example in this list of ten, which is itself a small piece of information about how big US companies behaved over these years.
The fourth case is a margin that barely moves. Microsoft’s ran 68.4%, 68.9%, 69.8%, 68.8% and 67.9% between 2022 and fiscal 2026 by my data, a range of under two points on a base near 69%. For a software and cloud company that stability is the point. I looked at what the price asks of that steadiness in my Microsoft piece, and the answer depends on the margin not drifting lower as cloud infrastructure becomes a bigger share of revenue.
When a low margin is the design
A low margin is not a bad margin. Costco earned 12.8% in 2025, up from 12.1% in 2022, and nobody would call that weak, because the model is to keep the markup thin and collect membership fees on top. I looked at that in what the membership fee does to the valuation argument. Reading Costco’s 12.8% against Microsoft’s 68.8% and concluding that Costco is the worse business would be a category error. What matters for Costco is the direction: a rise of 0.7 points in three years on a business this size is small in absolute terms, and I take it as a sign that pricing discipline held rather than as a growth driver.
Cigna belongs in this bucket too, with a margin of 9.0% in 2025 against 12.4% in 2022. Managed-care and pharmacy-services companies carry large pass-through revenue, so the margin is low by construction and small shifts in mix move it. A 3.5-point decline in a business with a nine percent margin is proportionally large, but I would want to know how much of it is mix inside the company before I read it as pricing pressure. I do not know that split from the data I used, and I would not guess.
Where this method could mislead me
Gross margin is not defined the same way across industries. Some companies put depreciation of plants into cost of sales; others put it below the line. Retailers count freight one way, software companies another. Banks and insurers do not report a gross margin at all. So I compare a company with itself over time and, at most, with two or three close peers, and I distrust any chart that ranks unrelated industries on one axis. My table above mixes miners, a utility, a steelmaker and a software firm, which is fine for showing the range, and useless for choosing among them.
The counter-case to my whole approach is that a falling margin can be the right choice. A company that cuts prices to win share, and earns more total profit for it, will look worse on this line and better on the next. If I see a slide paired with faster sales growth and a stable operating margin, I could be wrong to worry. There is a second limit. Gross margin says nothing about how much the company spends to get the sales. A firm can hold a 60% margin and burn all of it on marketing. That is the reason I read it together with the operating margin, and why I treat it as a signal to investigate rather than a verdict.
The threshold I would act on
When a company’s gross margin falls by more than about two points over three years while sales are flat or shrinking, I stop and read the last two earnings calls for the reason: input costs, product mix, price actions or a one-time item. The margin tells me that something changed. The call tells me what. If management gives a specific, measurable reason and the margin has stopped falling within two quarters, I stay calm. If the explanation is vague and the slide is still running, I treat it as a real problem and downgrade my view of the business.
For the ten names above, the ones I would look at first are Mondelez and Nucor. Mondelez because a ten-point drop in a year needs an explanation, and Nucor because I want to know whether 11.9% is a floor or a waypoint.
Analysis and opinion only, not investment advice. Figures come from annual financial statement data in my own database as of September 18, 2026, and company filings on SEC EDGAR are the place to confirm any of them; calculations were checked on September 22, 2026.