Ciena just posted the best quarter in its history, and the stock fell about 10% anyway. That gap between result and reaction is exactly where the investment question lives.
For fiscal third quarter 2026 (reported Thursday, September 3, 2026), revenue came in at $1.67 billion, up 37% year over year, with adjusted EPS of $2.11, a 215% increase versus the same period a year earlier. Against a consensus EPS estimate around $1.74, the company beat on both profit and sales, with EPS topping estimates by more than 20%. Adjusted operating margin rose to 22.5%, the highest in company history. These are not soft beats padded by one-time items. The underlying business accelerated.
And yet the sell-off is rational once you look at two numbers: 41.7 and 1.75.
Two Ciena customers together accounted for 41.7% of quarterly revenue, meaning the AI-driven demand is real but narrow. This concentration has been climbing. In fiscal second quarter 2026, three customers each contributed at least 10% of revenue and together represented 47.4% of total revenue. Now just two drive nearly the same share. The hyperscaler bet is becoming more concentrated, not less, even as the headline revenue number broadens.
The second number is the Q4 guide. Ciena guided its fiscal fourth-quarter revenue to $1.75 billion plus or minus $50 million, a midpoint that only aligns with the roughly $1.7 billion analyst consensus rather than clearing it. For a stock priced for relentless acceleration, matching expectations is effectively a miss. The market adjusted that instantly.
Here is the bull’s counter: the demand signal underneath this quarter is extraordinary. On the earnings call, management said that one month into the fourth quarter, Ciena had booked nearly as much demand as it received during the entire third quarter. Backlog rose $800 million sequentially to $8.5 billion, and management expects it to exceed $10 billion by fiscal year-end. Executive Advisor Scott McFeely said the vast majority of the expected $10 billion backlog carries customer-requested delivery dates in 2027, and that the backlog covers most of the company’s preliminary 2027 revenue outlook. That kind of visibility is rare in hardware.
Ciena has already issued a preliminary fiscal 2027 outlook calling for at least 30% revenue growth, pointing to $8.3 to $8.4 billion, with operating margins of 25% to 27%. If those numbers hold, today’s sell-off looks like noise. The company also pointed to several product ramps, including HyperRail, its next-generation line system expected to begin scaling in 2027, and Vesta co-packaged optics, expected to start contributing revenue in 2027 and ramp into 2028.
The bear argument is not that Ciena is struggling. It is that Ciena’s growth is rented. If one or both of those two dominant cloud customers slows its optical build, redirects spend toward a competitor like Credo Technology (which completed its acquisition of silicon photonics developer DustPhotonics in May), or decides to vertically integrate, the revenue curve bends sharply. More cautious analysts worry about customer concentration and potential technology disruption more than the current AI build suggests is warranted. That is a minority view today. It would not be a minority view if one of those two cloud customers showed up in next quarter’s disclosures at a reduced share.
Management said supply constraints are likely to persist through 2027 and possibly into 2028, which is both a risk and a form of protection: Ciena cannot easily be displaced from deployments already underway. Its incumbency in long-haul and data-center interconnect is structural, not accidental.
The concentration question is ultimately a valuation question. With shares around the low $300s after the post-earnings drop and a P/E multiple around 120, investors are paying for a multi-year compounding cycle. That multiple is defensible if the two large customers keep spending and the 2027 guide firms up. It becomes very difficult to defend if either customer moderates or the December Q4 report lands in-line again. Watch the customer disclosure table in the Q4 filing. That single data point will tell you more than any macro commentary about whether 37% growth is a Ciena story or a two-customer story.
