Analog Devices Q3 (19 August 2026): 59.3% of Every New Sales Dollar Became Profit Against a Guide of 50% — and the Q4 Guide Implies 81%

Analog Devices reported fiscal third-quarter revenue of $4.02 billion before the New York open on Wednesday 19 August, above the high end of its own guidance range, with an adjusted operating margin of 50.0% and adjusted earnings of $3.45 a share. The number worth having is the one underneath: 59.3 cents of every additional sales dollar arrived as adjusted operating income, against roughly 50 cents implied by the guide and 72.5 cents delivered a quarter earlier. The deceleration is real, it came entirely from the gross margin line rather than from spending — and the fourth-quarter guide implies the rate jumps back to about 81%.

What actually happened

ADI released third-quarter results for the three months to 1 August 2026 at 7:00 a.m. Eastern, with the call at 10:00 a.m. Every figure below is from the company's filed earnings release.

Revenue cleared the high end of the range. Adjusted operating margin landed at exactly 50.0% — the top of a range guided at 49.0% ±100 basis points. Adjusted earnings landed at exactly $3.45 — the top of a range guided at $3.30 ±$0.15. Consensus compiled before the print, near $3.93 billion and $3.33, was cleared on both lines.

The drop-through question, answered

The preview argued that revenue was the least interesting number in the release, and that drop-through — how much of each incremental dollar of revenue survives to operating income — was the measure that would tell you whether ADI was growing by filling capacity it already pays for or by adding cost. Here is the full sequence, with the guide it was measured against.

Two things stand out. The sequential rate decelerated from 72.5% to 59.3%, so the guide was directionally right about the slowdown and badly wrong about its size. And the year-on-year rate did not decelerate at all: 69.6% this quarter against 69.7% last quarter. Over a twelve-month window, the conversion of new revenue into profit is holding almost exactly steady while the base compounds — which is the more durable reading, because a single sequential step is sensitive to what happened to be shipped in a thirteen-week window.

Why the rate fell — and it was not spending

The preview set out three candidate explanations for a slowdown: conservatism, deliberate reinvestment, or a structural mix shift. The release lets you separate them arithmetically, because drop-through decomposes cleanly into what the gross margin line delivered and what operating expenses took back.

Operating expenses are the exonerated variable. Adjusted operating expenses rose $35.3 million sequentially — barely more than half the prior quarter's increase — and fell to 22.5% of revenue from 24.1%, and from 27.1% a year earlier. The operating-leverage mechanism that drove the original story got better, not worse. The reinvestment explanation is simply not in the numbers.

What changed is the gross line. Incremental adjusted gross margin fell from 85.4% to 68.1%, and adjusted gross margin as a percentage of revenue slipped from 73.0% to 72.5% — the first sequential decline in this run. When the incremental margin on new business (68.1%) sits below the average margin on the existing book (73.0%), the average has to fall, and that is what happened. That is the mix-shift explanation, in a mild form: the revenue arriving now carries slightly different economics from the revenue already in the base.

On a reported basis the leverage looks larger again: gross profit rose $917.8 million year on year on $1,141.6 million of extra revenue, an incremental gross margin of 80.4%, and reported operating margin expanded 1,170 basis points against 780 for the adjusted figure. That 390-basis-point gap is not performance. Acquisition-related expenses — mostly amortisation left over from the Maxim Integrated deal that closed in 2021 — were $397.8 million this quarter against $392.8 million a year earlier. Nearly identical in dollars, but 9.9% of revenue now versus 13.6% then. The charge did not shrink; the denominator grew. One further caveat on the reported line: special charges were a net credit of $24.2 million this quarter, worth about 60 basis points of margin, against a $4.3 million charge a year ago.

Automotive finally moved

The preview named automotive as the one line to watch, on the grounds that it was the only large end market with no AI in it and had been flat for five quarters. It moved.

Automotive had printed $856.1 million, $850.6 million and $871.6 million across the preceding stretch — roughly 2% of annual growth. At $998.2 million it is up 16% year on year and 14.5% sequentially, and it supplied $126.6 million of the quarter's $398.4 million revenue increase, the second-largest contribution after industrial. A line that had been standing still while the mix moved around it is now growing in its own right.

That matters more than the headline growth rate, because it is the closest thing in this release to evidence that the upturn is broadening beyond two customer groups. The ISM manufacturing index at 55.6% in July — a seventh straight month of expansion — implied a manufacturing recovery that should eventually reach an analog supplier's broad-market and automotive lines. This is the first quarter in which it visibly did.

Consumer is the counterweight, and deserves a mention precisely because nobody will give it one: flat sequentially and up only 6% year on year, down to 10% of revenue from 13%. Not every end market is participating.

A new fixed charge arrives

There is a development in this release that was not in the preview, and it works directly against the arithmetic described above. ADI completed the $1.5 billion all-cash acquisition of Empower Semiconductor on 7 July 2026 — inside the reported quarter. Empower's integrated voltage regulator and silicon capacitor technology is about delivering power at the point of compute, which is the constraint that binds as AI rack density rises: the limiting factor has become power delivered per unit of area, not total watts available.

It appears in three places in the release. Cash paid for acquisitions was $1.50 billion. Transaction costs were $23.4 million. And the fourth-quarter outlook now embeds $405 million of acquisition-related expenses, against $397.8 million in the third quarter.

That third item is the mechanically interesting one. The reported-margin story of the past year has been a fixed charge diluting against a rising revenue base. Acquiring a company adds new intangibles to amortise, which tops the charge back up. The dilution continues — $405 million against a $4.3 billion base is 9.4% of revenue, still below this quarter's 9.9% — but it is now a race between a growing denominator and a numerator that has stopped being constant. The reported-to-adjusted earnings gap was $0.71 this quarter and is guided at $0.72 next quarter, which is the same point stated in per-share terms.

What the Q4 guide is actually saying

The fourth-quarter guide is where the information in this release concentrates: revenue of $4.3 billion ±$100 million, reported operating margin around 42.6% ±150 basis points, adjusted operating margin around 52.0% ±100 basis points, reported earnings per share of $3.14 ±$0.15 and adjusted earnings per share of $3.86 ±$0.15. The board also declared a quarterly dividend of $1.10 a share, payable 15 September.

Read mechanically, that midpoint asks for 6.9% more revenue and 200 basis points more adjusted operating margin — an implied drop-through of about 81%, above anything actually delivered in this cycle including the 72.5% quarter. Only two things can produce it. Either incremental gross margin recovers sharply from the 68.1% just posted, which would mean this quarter's mix effect was transitory; or operating expenses stay close to flat in dollars while revenue adds $278 million, extending the leverage that is already running at 22.5% of sales. Both are possible. Neither is the low bar the third-quarter guide turned out to be, which is the honest way to read a company that has just been shown to guide conservatively.

What would change the picture: adjusted gross margin continuing to slip while revenue grows would mean the mix effect is structural rather than a one-quarter step, and the 81% would not be reachable without operating expenses falling in absolute terms. Automotive returning to its flatline would put the broadening thesis back in doubt. Neither is knowable now, and both are checkable in three months.

Cash generation remains the backstop under all of it. Operating cash flow was $1.60 billion in the quarter and $5.55 billion on a trailing twelve-month basis, 40% of revenue; free cash flow was $1.46 billion and $4.94 billion, 36% of revenue, against capital expenditure of $146 million in the quarter. ADI returned $1.69 billion to shareholders in the quarter — $535 million of dividends and $1.16 billion of buybacks — and $5.17 billion over twelve months, while also paying $1.5 billion of cash for Empower.

What does not read across

The read-across from this print is to the analog and mixed-signal peer group and to the industrial supply chain, not to an index in any mechanical sense. The same day's other semiconductor report — Wolfspeed's fiscal fourth quarter — is the mirror image: a small, loss-making supplier trying to swap electric vehicle demand for data centre power, where ADI is a large, highly profitable supplier already drawing on both, and now buying more of the second. Applied Materials' quarter sits one layer further up again, in the equipment that builds the capacity.

For currencies, the discipline is to keep the categories apart. Capital spending cycles eventually reach the growth factor that the meter scores, but on 19 August the release that sits in the dollar's path is the FOMC minutes at 2:00 p.m. Eastern, not an analog chipmaker's gross margin. The two arrive on the same day and belong in different files.

Educational macro context only — not investment advice.