Keysight Technologies, Inc. (KEYS) Earnings Call Transcript & Summary

Okay. Great. Thank you, everybody, for joining us on the fourth and final day of our Communacopia and Technology Conference. My name is Mark Delaney. I have the pleasure of covering Keysight. And yet again, this year, I'm really happy to have Keysight with us at the conference. With us from Keysight. We have Satish Dhanasekaran, Keysight's President and CEO; and Neil Dougherty, the CFO. Thank you both for joining us.

Thanks for having us. Thank you.

Well, Keysight is a really fascinating company, provide design, emulation and test solutions across both hardware and software. And a large percent of the company is over $7 billion of annual revenue is tied to customer R&D applications. Maybe Satish, you could start. Talk a bit about what's allowed Keysight to be so successful and be a key part of customer workflows in end markets that include communications, aerospace, defense, auto and industrial.

Well, thank you, Mark. Again, this is a case of a business that has been around for 80 years from HP and Agilent, which really got reinvented as we were able to spin out and and be independent and invest in our future. And the core of the strategy has always been how do we take a business that was largely hardware-oriented business and a product business and transform it into a solutions business. Not an easy thing to do, but that's the journey we've been on. And that involved protecting the core, right to win and right to play for your customer base, which is no new for the core measurement tools and the metrology, but also pursuing a smart diversification strategy to add into applications that are much more mission-critical and more valuable for our customers and that invariably involve adding more software and solutions content and so going higher in the stack that you described from physical layer to application layer and -- to protocol layer and application layer really allows us to provide that cohesive portfolio and where we really benefit from with the breadth we have is the ability to take that R&D investment that we make towards the portfolio and monetize it over multiple end markets because ultimately, we serve engineers -- electrical engineers are our customers around the world.

We've seen evidence of the success the company is having with the last few earnings reports. Last 2 quarters, orders [indiscernible] a bit over $2 billion in each quarter. Companies that [indiscernible] order to grow sequentially again next quarter? What's driving that growth?

Yes. I mean we're seeing a confluence of end markets that are inflecting simultaneously. We start to see that first with the wireline parts of our market that was starting to grow with AI, defense technologies becoming a bigger priority in today's geopolitical environment, Europe investing in its own sovereign technology as a trend. And then our industrial end markets, which tend to, sort of, have a bigger exposure to manufacturing and PMI is also doing very well with all of the build-outs that are happening globally. So we're benefiting from a number of tailwinds. We invested to have the portfolio and the differentiation, and we're benefiting from having the right portfolio meeting great markets at this moment.

Neil, maybe one for you on this topic. As you think about the guidance you gave into the fourth quarter, some sequential growth, given how strong demand is, and we've been hearing all week around the data center market, particularly being quite robust, and you guys have been [indiscernible], is there a reason that order should grow more than normal sequential seasonality in 4Q?

I guess my first response to that is you're already seeing that strength in the $2 billion quarters that we've just put up already in the year over growth -- year-over-year growth that we put up through the first 3 quarters of the year. So I think that strength continues. Do we see a further acceleration at this point? Maybe. But right now, our base case is that you're going to see the strength continue, which should result in a normal seasonal uplift as we move from Q3 to Q4.

And Neil, staying with you, as you think about modeling revenue, right? You had orders of $2 billion, that's a nicely positive book-to-bill. Should investors be anticipating at some point next year, revenue hits that $2 billion level as well?

Yes. I think certainly that's the trajectory we're on. We're a little bit supply chain constrained at the current environment. We guided Q4 to $1.940 billion of revenue, which is up almost $100 million from where we were in Q3. So that's clearly the trajectory. Given the supply chain situation, things are a little bit nonlinear. We expect them to be a little bit nonlinear, so exactly when we burst through that $2 billion limit or $2 billion revenue per quarter threshold is a little uncertain, but certainly, that's the trajectory that we're on.

Yes. Satish, maybe talk a little bit more around the supply chain constraints. It was a topic on the last earnings call, but any color you can share on that?

Yes. I think we're executing very well is the headline. I do think that every time we plan for a certain supply level, the demand keeps coming in harder and that's a good problem to have. But when you look at the macro supply environment today, whether it's for PCBs or components, you're often in environments where larger players are sucking up big chunks of capacity from the supply chain. And so the flexibility to respond in the short term is just is not there. So we're planning ahead. That's why I said on the earnings call, we're taking an 18-month view of supply and making some investments with our supply chain and in our own internal supply chain to be able to meet the demand across these markets. We feel really good about the portfolio as well. In through the 2023 and '24, we made some -- continue to make investments in R&D towards this future that we saw with wider bandwidths and more complex systems. And so the portfolio that we have now starting to launch is meeting great customer demand. So the demand for our new introductions have been a lot stronger right off the gate, which also compounds the ability to ship revenue in the near term. But we're working through it, and I'm confident we'll -- in a couple of quarters, we'll get through it. And I hope the demand continues to be more challenging in a couple of quarters, and we'll continue to work on the supply side. It's a good problem to have.

Yes. Absolutely. Yes. So hope over the next few quarters, and then maybe one for you, Neil. Any way to quantify how much revenue maybe you're not able to meet at the moment given the supply situation?

Yes. So first and foremost, again, I think given our differentiation in the marketplace, what we're really talking about is delaying the recognition of revenue or our ability to ship, not that we're walking away from business, given the supply chain constraints. I think -- just as a way to think about it, we busted through the $2 billion order level in Q2 with -- approximately $2.050 billion. I think $1 million or $2 million above or below that and generally, we have a 6-month order acceptance policy. So normally, [indiscernible] it leads you to expect that $2 billion plus of orders to ship by Q4, we've guided to $1.940 billion. You'd say, hey, maybe there's $100 million in there that's pushing out beyond our normal shipment window. Practicality is probably less than that because we have deferred revenue from software, some longer-dated programs, but I think $100 million is a way to, kind of, bound it at this point in time.

Okay. I know double ordering is hard to know when that's happening. I remember one CEO said to me at CES. He said, it would be great if -- when orders came in, if there was a little memo saying this is a double order. And I know it doesn't really work that way. But I mean, how does Keysight assess the risk that some of this $2 billion is because of supply chain constraints and double ordering?

I mean it's hard to -- as you said, it's hard to really know, but we -- from the -- to the extent that we see -- we have line of sight to what our customers are needing our products for. And as I mentioned on the call as well, in some cases, we're also having line of sight to what the entire ecosystems need are from likes of NVIDIA to their supply chain to what the hyperscalers are ordering. So we feel like all of this is towards real demand that's going into the AI data center. So there's not a lot of risk. [indiscernible] In history in this business would also suggest that we have had very low cancellations historically. Our customers are needing our products out of the gate, and they will take them if we could ship more right now.

Yes. If I could add to that. We're not seeing broad supply chain problems across a broad section of the portfolio. It's pretty concentrated. 20, 30 SKUs focused on this data center ecosystem. And if you look at those 20 or 30 SKUs, they're very highly differentiated in the marketplace. So in some cases, there isn't really another equivalent alternative in the marketplace for them to double book if they wanted to.

Okay. Very helpful context. Well, some of this demand, of course, coming from market, but a lot of it has to do with everything that Keysight has been working on. And you mentioned, Satish, this transition to being a full solutions provider in so many ways you're helping customers solve their engineering problems. You see that in the financials. So software and services, 33% of revenue last quarter, recurring revenue, you guys estimated, was just under 1/4 of total sales. Do you think longer term, where could those ratios get to?

Definitely higher. And one of the things that we think about is how do we make our portfolio more valuable to our customers and differentiate it, and I even use the word unique in some cases. But when we have that type of portfolio and we can connect that and offer solutions to a customer, you're really helping them with your time to market. And when you have that dynamic, it allows you to also have a conversation about the value you're bringing to the equation. And so you start to see our gross margins go up sequentially also this year to record levels of 68%, which we're very happy with. We still see more opportunity to grow that. And equally, as we sell solutions, we're also able to service the customers over a longer period of time. And that's where the services attach rates are important and that's recurring revenue for us and the software parts of the portfolio are important. And what we're seeing this year is a little bit of more manufacturing business that we are winning in the AI data center so that numbers of the recurring revenue you quoted reflects the mix shift that you're seeing. But I think over the long term, we remain confident that we'll continue to grow that ARR and services as our solutions get more adopted.

On the theme of AI, I know it's driving a lot of the demand for your products, but do you see risk of AI disrupting any of the products you provide?

Not really. I think AI has really been a tailwind for us from a market -- AI data center market perspective. But even when we look broader, I think customers are changing the way they're thinking about the engineering workflow. And anytime there's change, I think there's an opportunity for us to be a consultant and partner with them and help them rethink how things get done, add more content to our own portfolio, which we are in the process of doing to make it easier so that it fits into their workflow pretty well. The simulation parts of our tools that we have are highly differentiated and they're very focused on providing the most complex physics capabilities, whether it's in RF or high-speed digital or photonics. And so we're not involved in any of the areas where we feel like that's a risk because of AI.

Okay. I think software is over 20% of your revenue overall. To what extent is that software running on equipment of other test companies?

Most of our software runs on our hardware, right. And I would, sort of, characterize it as there's a software that's an app that really enables our customers to drive insights in our physical layer and the emulation platforms that we have, have a higher software mix and they allow the customers to emulate real-world conditions in a lab, and that's another differentiated capability. And then we have the simulation business, which is all software. So I suppose it can run on any server.

Okay. Very helpful. I wanted to talk around your exposure to R&D relative to production test. Keysight's always been more on the R&D test side. But talk about where you are today, how that ratio may vary by end market?

Yes. We still think the majority or 50% plus of our business historically has been in the R&D labs, and that's been an [indiscernible] intentional a part of the strategy. We've also added some operational piece to the portfolio with the acquisitions of Ixia and some of the Spirent businesses that we have, building more differentiation as we go through our strategy. I think what we're seeing now with the AI data center ramping, simultaneous capacity ramps that are occurring in our prime contractors in our defense business, and also the China Plus One strategy that that's occurring in Southeast Asia, there's a lot more manufacturing activity, and we're winning there. It's a good margin business. We're winning. But we recognize that some of that is episodic, but over the long term, we continue to focus on growing that R&D parts of our portfolio because it builds more sustainability to our revenue stream.

As you were saying, you have some operational exposure, too. I mean, as manufacturing parts of these industries grow, I mean, you still participate. At least selectively, where you can get the right profit margins?

Absolutely. Yes, high-value manufacturing.

We spoke on emulation and simulation already. And I think those have been strategic initiatives to get even bigger there as you described at the last Investor Day. Maybe talk a little bit more around how those parts of the portfolio, perhaps, could grow in the longer term?

Yes. I mean, we're quite excited by -- even as we're winning in the physical layer of AI as infrastructure is getting rolled out, people are testing the physics associated with electrical and optical signals, and we're agnostic to that and from a portfolio, we're working to characterize how these AI data centers would behave under a range of conditions under different LLMs and other things And those kinds of use cases are growing in today's world where customers are using a very heterogeneous stack of hardware and software capabilities. So the ability to emulate how those would behave in the real world, how the data will flow, where the bottlenecks are. I think those are the kinds of things that we're working on right now, and we feel good about that part of the portfolio. And as 6G which is, again, the next generation of communication technology and wireless is going to ramp, that's going to be another tailwind to our emulation part of our portfolio. On the simulation side of the business, we historically had just a small presence in the RF part of the portfolio. We've then built out high-speed digital simulations. On top of it, we've added multiphysics capabilities with acquisitions of ESI and most recently, with the Optical, we've really built a nice set of portfolio of tools for our entire simulation business, which will exceed $0.5 billion this year for the first time and will be a high recurring revenue from that part of the portfolio. We feel good about it. And as we bring those assets together, our ability to solve many more challenges for our customers just goes up because we now have the full stack there.

So the $0.5 billion number, that's just simulation or simulation and emulation?

It's just simulation.

Simulation, okay. Let's double click on the Communications Solutions Group. Your biggest segment as a percentage of revenue, a little under half the total company typically on an annual basis. On the 2Q call, the company has said that in the first half of the year, AI revenue already has surpassed all of what have been done in fiscal '26. So in that $500 million to $600 million range. And then last quarter, you talked about AI accelerating even more. What specifically, from an AI standpoint, is driving that kind of strength in the CSG segment?

Yes. I mean we, sort of, referred to our AI exposure in our wireline business. Historically, if you look at our commercial communications segment, 2 components, wireless and wireline, wireless was slightly bigger than wireline and for the first time this year, I think we said wireline has actually exceeded wireless, and that shows you the demand in the AI parts of the business is just going up. And the first half was strong. And by all indications, the second half could be stronger than the first half from everything I see today. And I would say the R&D part of that portfolio is growing nicely, equally, the manufacturing exposure that we have from all of the infrastructure build-outs ramping up and our portfolio being in a great place. So we're seeing more demand for both parts of the portfolio there for the AI data center.

As you think about the AI part of the business within CSG, can you help investors think about how much may be coming specifically from optical tests, so NPO and CPO type applications?

Yes, we have a broad exposure to different components and different customer types. So our breadth is really a source of strength and scale for the business. When we think about the interconnect part of the portfolio, if you -- it's a little over 1/4 of the total business from an interconnect point of view, and we do service both optical and electrical interconnects there. And what we're seeing is customers are working on some complex physics problems. And there's more scale to build for the optical side of things. Right now, copper is the predominant contributor to the business.

Okay. Well, I want to talk on the hyperscale part of the business, and Keysight's been growing its direct hyperscaler revenue stream. Can you share more what products Keysight is doing when it's going directly to hyperscalers?

Yes. I think it's some of the same set of tools and portfolio. I mean, one of the things we observed in the last 4, 5 years is an inflection in adoption of new technology driven by hyperscale scale that they're bringing to wireline networks. The adoption rate from going from 10 to 20, 40, 80 to 100, to 400 now and 800. I mean, you look at those curves and what we're starting to see is those design cycles compress, customers are going faster, the next generation waves that are overlapping, really creating a nice tailwind for us to grow our contributions in this marketplace and more silicon designs coming out from the hyperscalers is another tailwind for our business, more customers to serve and more [indiscernible] tools, which has also been another tailwind. I also think that even though our [indiscernible] resized AI exposure from a hyperscaler point of view, to be about 10% of the total mix this year, but I think that is a little bit underrepresented because of the influence downstream that the hyperscalers have with regard to their supply chain. So that number is actually a lot bigger. So we're participating meaningfully in the physical layer opportunity. We're working with a number of hyperscalers in the emulation opportunity feel good about the runway that exists as we look ahead.

Very helpful. Maybe we can turn to the wireless part of the business. Obviously, there was a downturn post some of the 5G build-out that the whole industry saw, but wireless has been growing. It grew pretty significantly on the last -- in the last quarter and contributed to that $2 billion total. Can you talk a bit more about what's driving the wireless orders? Because I think 6G is a little bit out in time.

Yes. I think we're excited by the early start we're seeing in the 6G business. So that was one of the contributors. We also see the non-terrestrial network opportunity to be potentially bigger as we look ahead. But again, another contributor to growth for our wireline business -- for our wireless business. And third, I think the supply chain of wireless, again, driven by the activity around AI is -- was a key contributor to growth in that part of the portfolio.