AI Is Supplying Blowout Earnings Again
By Maksym Misichenko · Nasdaq ·
By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
Panelists agree that while Cisco and Lumentum are benefiting from AI-driven orders, their high valuations and cyclical nature pose significant risks. The sustainability of margins and potential demand snap in 2027-28 are key concerns.
Risk: Demand snap in 2027-28 and potential margin compression due to open optics and white-box standards.
Opportunity: Multi-year supply agreements and extended demand for efficient optics due to power and cooling bottlenecks.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
In this episode of Motley Fool Hidden Gems Investing, Motley Fool contributors Tyler Crowe, Matt Frankel, and Jon Quast discuss:
To catch full episodes of all The Motley Fool's free podcasts, check out our podcast center. When you're ready to invest, check out this top 10 list of stocks to buy.
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
A full transcript is below.
Before you buy stock in Lumentum, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Lumentum wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $483,476! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,362,941!
Now, it’s worth noting Stock Advisor’s total average return is 998% — a market-crushing outperformance compared to 207% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
**Stock Advisor returns as of May 19, 2026. *
This podcast was recorded on May 14, 2026.
Tyler Crowe: We're talking earnings and investor questions on Motley Fool Hidden Gems Investing. Welcome to Motley Fool Hidden Gems Investing. I'm your host, Tyler Crowe, and today I'm joined by longtime contributors, Matt Frankel and Jon Quast. Today, we're getting lots of questions from members, and we thought this would be a really good time to combine some listener questions as well as earnings reports that we've seen coming out in the past couple of days, as a little bit of a marriage of good ideas. We're going to talk about Cisco's earnings. We're going to talk about Lumentum's earnings, and also we'll get into, hey, maybe we should talk about non-AI things for a little bit from our investor mailbag.
But as I said, we're going to start with Cisco. We got a question a little while ago from one of our listeners. I hope I get the name right, Halish Shankar, and was asking about Cisco and what are our thoughts on it. I thought it would be a great time to start the conversation today because Cisco reported earnings, and the stock is up 13.8% as we were recording because numbers were pretty good. Revenue growth was up about 12% for the year earnings was up, and obviously guidance was looking pretty good. Jon, why don't you run us through the numbers and what you guys saw in this particular earnings that I would say defied expectations of what Cisco has been for a while.
Jon Quast: It's so surprising to be talking about Cisco, one of the poster children of the dot-com bubble over 20 years ago, but really, the business is booming unlike ever before. You look at the most recent quarter, 12% top-line growth. Really, all of the growth is coming from one part of the business. Cisco has various components, but there's one part of the business that's driving everything, and that is networking. The company reported 25% year-over-year growth in the networking side of the business. Everything else is either down or basically flat.
Essentially what is happening here? As the AI infrastructure build-out marches on, all of these GPUs, the clusters, even the data centers themselves need to be connected, and this really plays to Cisco's strengths. It's getting a ton of demand, in particular, from the hyperscaler businesses. You think of the public cloud giants, the tech giants, the Magnificent Seven, these are the companies that are needing network solutions, such as the ones that Cisco provides. What is fascinating here, when you look at the orders to the hyperscalers last fiscal year, about two billion total for Cisco. Going into this fiscal year, it was expecting five billion total, and that was quite ambitious of a projection, more than doubling year over year. But we're three quarters into its fiscal year now, we're already surpassed that projection, and now management is saying, we're expecting nine billion of orders in this fiscal year from the hyperscalers. Incredible year-over-year growth there.
It's not just hyperscalers, it's also the neo-Cloud businesses. We just got report from Nebius today signed a $27 billion deal with Meta. Its business is growing like crazy. It needs networking solutions as well, but all in all, good quarter for Cisco.
Matt Frankel: Yeah, I wanted to lean into something Jon just mentioned and emphasize the word orders. Product orders to hyperscalers are growing at a triple-digit rate. AI infrastructure orders from hyperscalers were $1.9 billion in the third fiscal quarter. That's up from 600 million a year ago, so more than tripling. Overall, Cisco's orders were up by 35%. Jon mentioned the top line only grew by 12% in the quarter, but overall product orders, which are very indicative of future revenue, were up 35%. Even excluding hyperscalers were up 19%, we're seeing really strong demand across the business. Networking orders were up 50% year over year, as Jon mentioned, that was the strongest segment, and hyperscalers are the real story here.
Just to put this in perspective, Jon correctly mentioned that now Cisco expects $9 billion in orders this fiscal year for AI infrastructure. That's compared to just $4 billion in expected recognized AI revenue. More than double what they're recognizing in AI infrastructure revenue, they're expecting for future revenue because they're getting these orders in. This is really just a long way to say that the reaction to Cisco's quarter isn't necessarily about revenue. No one's that excited about 12% year over year top line growth or the earnings that they just reported on a per-share basis. It's as much about the orders it now has on its books that will be recognized in the future periods, and the anticipated acceleration in that number over time.
Tyler Crowe: Getting back to the question that our listener Halish asked related to, it's like, is this a good idea? This is one of the things I've been struggling with Cisco, and we'll get into it when we talk about Lumentum in the next section, as well, is that these are businesses that have been notoriously cyclical for pretty much all of their life as publicly traded companies. Cisco, like to Jon's point, was the poster child of massive build-out during the dot-com boom. If everybody had to have Cisco systems equipment, and then everyone was like, maybe we don't. Maybe we can use other stuff, and it was OK. It took decades for investors to see the highs of Cisco stock again. Looking back over the past 10 years, revenue has been up and down. Operating cash flow for this company is more or less what it was 10 years ago, and so this is where it's been a little bit of a struggle for me. The company is doing much better right now, but is this just a short-term catalyst of a typically cyclical business? Or is this something that's fundamentally different about the business, and we as investors should look at it differently?
Matt Frankel: Yeah, Cisco is at an all-time high after this earnings report. It's nearly doubled over the past year. The AI business has nearly doubled their expectations, as well. I would argue that not only is it a move that's justified, but this is a fundamentally different time. This isn't just cyclicality right now. I don't think we've seen an AI cycle over the past couple of decades. This is something that's new. It's something that wasn't really a big market opportunity. No one was talking about AI infrastructure a few years ago. Shares trade for about 26 times forward earnings right now. There's a solid case to be made that revenue growth will accelerate in the 2027 fiscal year, which starts very soon. I'd actually be comfortable opening a small position in Cisco at this level, even at an all-time high, and then adding incrementally. That's just my take on it.
Jon Quast: For me, Cisco just bores me. I'm sorry to all the Cisco shareholders out there. I just want to be honest about that up front. I wanted to just move on from this. However, given the question, I really took a honest, hard look at it, and I think I need to agree with Matt here that there may be a case for owning Cisco stock here at this price. Listen, it's still quite the value compared to some of its competitors in the space. Growth is accelerating. We look at the next quarter's projections, projecting 19% growth up from what was it 12-13 this quarter. That's an acceleration, that's a good thing. Operating margin recently went from 23%-25%. That's a good thing. You look at what Cisco's products it provides, it does seem like they are starting to take some market share here and if that continues, I don't think this is a terrible stock today.
Now, I think that I would still prefer Arista Networks, that's ANET, for its debt-free balance sheet. I like nice clean balance sheets, especially in the face of uncertainty. Cisco's isn't as clean as that, but I don't think it's crazy to own Cisco stock here. Hearing both your response and thinking about it myself, I've always looked at it in the same sense of these are all how bullish are you on AI build-out? You can look at the rates of spending and all the studies that are going out related to this. If you are a wholesale believer in what is being published and what is being projected for AI spending, then absolutely, these are 4-, 5-, 6-year catalysts that are going to be hard to avoid as investors; it all comes down to how much you believe it. We've discussed it many times a year before, and I feel bad like a broken record saying it all over again. But it really does come down to how much of a believer in this AI infrastructure build-out you are. Coming up next, we're going to even talk more about this with earnings related to Lumentum.
ADVERTISEMENT: Dell PCs with Intel inside are built for the moments that matter for the moments you plan and the ones you don't. Built for the busy days that turn into all night study sessions, the moment you're working from a cafe and realize every outlet is taken. The times you're deep in your flow, and the absolute last thing you need is an auto update throwing off your momentum. That's why Dell builds tech that adapts to the way you actually work, built with a long lasting battery, so you're not scrambling for the closest outlet and built in intelligence that makes updates around your schedule, not in the middle of. They don't build tech for tech sake. They build it for you. Fine technology built for the way you work at dell.com/dellpcs built for you.
Jon Quast: Similar to the stock move that we saw with Cisco earlier today, shares of Lumentum were up as high as 21% on recent earnings reports. This is a company that we got a question about from one of our listeners, Nathan Holtstein. It's also a prominent member of several scorecards on the Hidden Gem side of various Motley Fool Investing Services. It's been recommended a couple of times by CEO Tom Gardner and some of the other in some of our recommendations services. It is something that's probably on a lot of investors' minds today. We want to get into earnings and try to get to Nathan's question as much as possible. But this isn't probably one people have heard about a lot. It's really one of those behind-the-scenes businesses. Matt, before you dig into what was actually in the report, give us the too long, didn't read of the 10-K for Lumentum as what it actually does.
Matt Frankel: Yeah, Lumentum has been around for a long time. It produces optical components for things like 3D mapping for high-speed communications, think fiber optics. For the longest time, this was a company that got most of its business from the rollout of 5G technology and telecommunications and things like that. But the key thing to know now is that the components Lumentum makes are extremely valuable parts of datacenter infrastructure, specifically when it comes to interconnections and other forms of AI infrastructure. That's what's really driving the business today.
The recent quarterly report was fantastic, even by Lumentum standards, which has produced a string of blowout earnings reports over the past year. Revenue was up 90% year over year, and 22% sequentially, that's an acceleration over the previous level. Earnings came in even better than expected. Adjusted operating margin expanded by 21 percentage points year over year to 32%. Beyond those headlines, Nvidia just made a $2 billion direct investment in Lumentum. Lumentum announced a brand new facility that it's building that should be online in 2028 that is directly related to its partnership with Nvidia, they're going to be providing the demand for it. It's like building a factory and you already have someone who's buying all the products. That bolstered its already cash-rich balance sheet. Lumentum now has over $3 billion of cash just sitting around. It's solidified that partnership between Nvidia and Lumentum, which is honestly its most important customer. Its guidance calls for 22% sequential growth in the current quarter, so things aren't slowing down just yet. This was a very strong report, and I completely see why the market reacted the way it did.
Jon Quast: When we were talking about Cisco, you're talking about a historically cyclical business that has this catalyst that is driving things. The Lumentum story is not any different. I think it's actually probably just the Cisco story on steroids right now because it's an even more extreme. Right now, over the past year, shares of Lumentum are up 1,200%, and currently shares trade for about 189 times earnings. Definitely not the more value-oriented 26, 27 times earnings that we were talking about with Cisco here. This is a high-flying stock that has a lot of very lofty expectations built into it almost entirely related to AI build-out, because, as to your point, Matt, previously, this was a business that came and went with the deployment of whatever wireless protocol was the new hot thing for telecom. When 3G was coming out, they had a lot of work. When 4G came out, and those periods in between, it was pretty dead. Again, when we're looking at Lumentum and those amazing numbers that we just saw, how do you reconci
Four leading AI models discuss this article
"Lumentum’s valuation leaves no room for execution misses in its Nvidia-dependent AI ramp."
Cisco and Lumentum both posted strong AI-driven order growth, with CSCO guiding hyperscaler orders to $9B this fiscal year and LITE reporting 90% YoY revenue growth plus a fresh Nvidia facility commitment. Yet both remain classic cyclical equipment names whose prior cycles collapsed once buildouts peaked. LITE’s 189x forward earnings and single-customer concentration with Nvidia embed assumptions of sustained triple-digit AI capex that history shows rarely last beyond 3-4 years. Cisco trades at a more reasonable 26x but still faces the same end-market risk if hyperscaler budgets flatten in 2027.
If Nvidia’s $2B direct investment and the $27B Nebius-Meta deal prove durable, LITE’s capacity expansion could compound for years without the usual telecom-style bust, justifying the premium.
"Lumentum's 189x P/E assumes hyperscaler capex intensity stays at peak levels indefinitely; history suggests cyclical networking/optical businesses face brutal margin compression when the build phase ends."
The article conflates *orders* with *revenue*, a critical distinction the panelists themselves acknowledge but don't stress enough. Cisco's $9B AI infrastructure orders guidance versus $4B recognized revenue tells you the backlog is real—but also that recognition is backend-loaded and subject to execution risk. Lumentum at 189x forward earnings (versus Cisco's 26x) has priced in not just AI buildout, but perfection: zero competition, zero margin compression, and sustained hyperscaler capex at current levels. The 1,200% one-year move is a valuation reset, not a fundamental one. Neither company has proven it can sustain these margins through a full cycle. The article never asks: what happens to networking/optical demand if hyperscaler capex normalizes in 2027-28?
If AI infrastructure spending truly is a multi-year, multi-trillion-dollar cycle (not a bubble), then current valuations may actually be cheap relative to TAM expansion, and Lumentum's margin profile could stick if it locks in long-term contracts with Nvidia and hyperscalers.
"The market is dangerously over-extrapolating short-term AI infrastructure order growth into permanent, high-multiple terminal value for historically cyclical hardware providers."
Cisco (CSCO) and Lumentum (LITE) are currently riding a massive wave of AI-driven capital expenditure, but investors must distinguish between order book growth and sustainable free cash flow. Cisco's shift toward hyperscaler networking is impressive, yet its legacy business remains a drag on margins. Lumentum’s 189x P/E ratio is a massive red flag, pricing in perfection that assumes the current AI build-out cycle will never face a digestion period. While the order growth is undeniable, the market is conflating 'demand' with 'long-term competitive moat.' I am wary of the cyclical trap; these companies are essentially becoming high-beta proxies for Nvidia's (NVDA) success rather than independent compounders.
If the AI infrastructure build-out is a multi-year secular shift rather than a standard hardware cycle, current valuations may actually be justified by the sheer scale of the total addressable market expansion.
"AI infrastructure demand could drive multi-year upside, but cyclical risk and ultra-high valuations keep the risk/reward far from one-sided."
These earnings echo a broader AI-infrastructure theme: Cisco's networking fueled by hyperscaler orders and Lumentum's optics tied to Nvidia's AI hub, with Nvidia investing directly in Lumentum and a new capacity buildout planned. The bulls argue for a durable, multi-year capex cycle for AI data centers that could lift growth and margins. The bears warn this is a cyclical, capital-intensive space where demand can snap, and markets are already pricing extreme multiples (CSCO ~26x forward earnings; LITE far higher). The piece glosses over fundamentals like cash flow, margin sustainability, and competition, plus potential supplier churn among hyperscalers.
The case rests on a lasting AI capex cycle; if hyperscalers pause or accelerate away from certain suppliers, these 'orders' may not crystallize into durable earnings, and the high valuations could compress.
"Power and cooling bottlenecks could extend rather than curtail optical demand past 2027."
Claude highlights the backlog recognition lag at Cisco, but this underplays the potential for Lumentum's Nvidia-backed facility to lock in multi-year supply agreements that smooth revenue beyond typical cycles. Power and cooling bottlenecks in hyperscale builds, unmentioned so far, could actually extend demand for efficient optics rather than trigger the feared 2027 snap. This shifts the risk profile from pure demand destruction to supply-chain execution challenges that favor incumbents with direct ties like LITE.
"Nvidia's dual role as customer and investor in Lumentum creates hidden concentration risk that a supply-chain execution narrative doesn't address."
Grok's power-and-cooling bottleneck thesis is speculative—I'd need capex data showing hyperscalers are *constrained* by optics efficiency rather than chip availability or real estate. More critically: Nvidia's $2B Lumentum investment and direct facility commitment create potential conflict-of-interest dynamics. If Nvidia later vertically integrates optics or shifts suppliers, LITE loses both customer and anchor investor simultaneously. That dual dependency risk hasn't been quantified.
"The shift toward open-compute standards and commoditized optics will compress Lumentum's margins, rendering its current valuation a trap."
Claude is correct to highlight the dual-dependency risk, but everyone is missing the elephant in the room: the commoditization of optical transceivers. As hyperscalers like Meta and Google push for open-compute standards, Lumentum’s proprietary moat will erode regardless of Nvidia’s current investment. We are seeing a shift toward 'white-box' optics where the software-defined network matters more than the hardware vendor. This makes LITE’s 189x P/E fundamentally unsustainable as margins inevitably compress toward hardware-commodity levels.
"Margin compression from open optics/white-box standards could drive LITE's multiple down despite Nvidia ties."
Gemini's commoditization concern is valid, but the bigger risk is margin compression from open optics and white-box standards. If hyperscalers accelerate standardization, LITE’s pricing power could erode even with Nvidia ties, and long-term contracts may not fully offset mix effects. The debate should quantify embedded margin exposure and capex hedges; otherwise the 189x P/E could re-rate meaningfully lower in 2–3 years if open standards gain traction.
Panelists agree that while Cisco and Lumentum are benefiting from AI-driven orders, their high valuations and cyclical nature pose significant risks. The sustainability of margins and potential demand snap in 2027-28 are key concerns.
Multi-year supply agreements and extended demand for efficient optics due to power and cooling bottlenecks.
Demand snap in 2027-28 and potential margin compression due to open optics and white-box standards.