Calix Just Posted Record Revenue and the Stock Still Has to Answer for Its Margins

# Calix Just Posted Record Revenue and the Stock Still Has to Answer for Its Margins

Here's a fun game: find the earnings report where everything looks great on the top line and something's quietly rotting three lines down. Calix's Q2 2026 print, dropped July 20, is a pretty clean specimen. Record revenue. A double-digit EPS beat. Record backlog. And a gross margin that's sliding like it owes someone money.

Let's start with the good stuff, because there's genuinely a lot of it. Calix put up $293.3 million in quarterly revenue, up 21% year-over-year and 5% sequentially — an actual record, not a "record if you squint" number. Non-GAAP EPS came in at $0.47 against a Street estimate of $0.41, a 17.5% surprise that's the kind of beat that gets a stock added to every momentum screen on the planet. Remaining performance obligations hit $386.4 million, also a record, which is the company's way of saying "we can see revenue coming before it arrives." Software and services revenue — the part of the business that actually deserves a growth multiple — hit $50.5 million, up 16% YoY and 7% sequentially. That's the Calix One / Agent Workforce Cloud story working as advertised: an AI-native platform pitch that broadband service providers are apparently buying into, not just nodding along with on earnings calls.

So why isn't this an uncomplicated "back up the truck" story? Because gross margin fell 240 basis points sequentially to 54.8%, and management is guiding for another 280 basis point drop to 52.0% in Q3. Stack those together and you're looking at roughly 480 basis points of margin erosion in two quarters. That's not noise. That's a trend, and it's happening at the exact moment the top line is accelerating — which tells you the growth is being driven disproportionately by appliance (hardware) revenue rather than the higher-margin software mix everyone wants to believe is the real story. Appliance revenue was $242.8 million of the $293.3 million total — that's about 83% of the business still riding on boxes, not platform. The software narrative is real, but it's still the minority shareholder in this revenue mix, and the hardware side is currently the one setting the margin trajectory.

There's also a small but telling wrinkle in the "beat" itself: depending on whose estimate you're using, Calix either beat revenue by 1.2% or missed a higher consensus figure around $295.7 million. That's not a scandal, it's just estimate dispersion — but it's a reminder that the "record revenue, big beat" headline is doing a bit more narrative heavy lifting than the underlying number strictly supports.

Here's my actual take: this is a stock caught in a tug-of-war between two legitimate stories, and the market hasn't decided which one wins yet. The bull case — broadband buildout tailwinds (BEAD funding and friends), a genuine software transition, record backlog de-risking near-term numbers, an AI-native platform pitch that's differentiated enough to matter — is not hand-waving. It's backed by real sequential software growth and a backlog number that gives you actual visibility, which is more than most small-cap industrials can offer. But the bear case isn't hand-waving either. Nearly 500 basis points of guided margin compression in two quarters is the kind of thing that, left unaddressed, turns a "record revenue" quarter into a "why did EPS growth stall" quarter down the road. If appliance revenue keeps outrunning software, the mix shift that's supposed to re-rate this stock toward a platform multiple just... doesn't happen. It becomes a hardware company with a software garnish, priced like something else entirely.

The next real answer comes with Q3 earnings, expected around late October. That print needs to do one specific thing: show the 52.0% gross margin guide holding — or better, show software revenue growth outpacing appliance growth enough to start bending the mix back in the right direction. If margins keep sliding past guidance, the "record revenue" headlines stop mattering and the multiple compresses to match the actual business, not the story about the business.

Until then, Calix gets to keep both narratives alive simultaneously, which is exactly why it's showing up on every quant screen with a pulse. Momentum and earnings surprise factors don't care about gross margin trajectories — they care about beats and records, and Calix just delivered both. But factor models aren't the ones who have to explain Q4 guidance if 52% becomes 50% becomes a pattern. Watch the margin line, not the headline. The revenue record is the easy part of this story. The margin math is the part that decides whether this rerates or reverts.