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

Okay. We'll go ahead and get started. Thank you for coming out for the Keysight presentation session today. We're very pleased to have Keysight with us, global leader in electronic test and measurement and a key cornerstone in our test and measurement coverage. I'm Rob Mason, senior analyst at Baird, covering the advanced industrial equipment sector. Very pleased to have Neil Dougherty, CFO from Keysight with us, as well as Giampaolo Tardioli, who is the VP of Network and Operator Solutions in Keysight's Communications business. Neil is going to start off with just a couple of brief comments, kind of level set us, and then we'll jump into Q&A. Again, if you have any questions, feel free to send those up to me via the iPad, and we'll work those into the conversation. Neil?

Great. Thanks, Rob. So yes, as Rob mentioned, we're the #1 player in the electronic test and measurement or design and test markets. We essentially bring solutions to market across a wide range of industries to help our customers in their R&D labs get technology products into the marketplace with a focus on being first to market in the lab so that they can be first to market in the markets in which they serve. The biggest business is in the commercial communications space. We service commercial communications across both wireless and wireline ecosystems. Historic strength has been in the physical layer, but through acquisition, we've moved up into the protocol layers of both wireless and wireline technologies. So we now service an entire ecosystem end-to-end and up and down the protocol stack. We have a sizable aerospace defense business, again, focused on basically communications technologies for the defense industry, semi business, auto and energy -- or excuse me, in auto and energy, but primarily focuses on auto being EV and autonomous driving and then a general electronics business that really serves a wide range of industries kind of to put into those other buckets. I think there's a number of secular drivers across our industries right now that are very favorable. In the commercial communications space, it's 5G currently being deployed, but we're already starting to work on 6G, in the early phases of 6G. On the wireline side, 400 gigabits being deployed, 800 gigabit and terabit are in R&D. In auto, it's AV, it's EV. In semi, it's the move to smaller process architectures as well as the assurance of supply, general supply shortages that are driving demand today. So a number of secular themes across a wide range of industries that really bode well for Keysight as we look forward.

Super. Again, we tend to be a forward-looking group. But maybe just real quickly, I wanted to think back. We came into this year with a certain mindset around growth, 6% to 7% and EPS and operating leverage, maybe not quite. Expenses coming back into the P&L, so maybe a little bit below plan. I don't know if it's played out that way. You've been able to raise guidance. As we've gone through the year, supply chain despite it arguably getting worse through the year. China, throwing some wrinkles into the mix as well. So you've still been able to take both your revenue and EPS guidance up as we've gone through the year. Just maybe comment on what some of the factors are, how you've been able to execute in this environment and be able to do that.

Yes, it's certainly been a very challenging year. And I don't disagree with your characterization that the margin supply chain is maybe treading water. It's certainly not getting better, could be getting worse. I mean I think we had originally been expecting some improvement in the second half. We don't see any material signs of that at this point in time. But as you know, Keysight has continued to execute well. I think we benefit from the fact that we have a vertically integrated supply chain. A lot of the customized chips that we use in our product we make in-house and our own fab in Santa Rosa, California. And so that provides some protection from these challenges. I think we are -- while we're not seeing dramatic improvement in the supply chain, we are starting to see some benefits from some of the actions that we've taken. Notably, we pivoted some of our R&D resources 6, 9 months ago to work with our manufacturing teams to redesign boards or whatever it took to find alternate sources or design out long lead time parts to get in design in parts that are more readily available. Those things are starting to pay some dividends and help us as we look to get product out the door, I think. By and large, I think we're still doing a pretty good job of getting product into the hands of our customers on time lines that are acceptable to them. And so that's been great. And then from -- and as a result, we've taken up our revenue guidance for the year. Demand remains strong as measured by orders. That hasn't been the problem. There's still been this pretty significant disconnect between what we're seeing on the order side driven by strong demand and what we're able to deliver on the revenue side, which is really more a function of what parts are showing up on the loading dock, right, and on what time line. And on the expense side, at the beginning of the year, we did talk about the return of some of the COVID costs, most notably in facilities and travel and those types of things. That didn't really materialize in the first half of our fiscal year as expected. Just as we were kind of saying that, the Omicron variant hit, and we once again delayed return to the office and kind of continued to delay travel. We are just now starting to see those things start to happen, right? We've set a clock for employees in the U.S. to get back into the office with dates following at other locations around the world based on location and -- current location or current status of the pandemic in those locations. People are getting back on planes as evidence that I'm here this week, I'm in Europe next week. And so that is starting to happen. But it didn't -- it was slower than we expected. And so as a result, not just of that, but of other things as well, we've taken up our earnings guidance as well for the year.

And just speak to the -- you mentioned orders outpacing sales. So just speak to the backlog. If we look at the backlog, it's about 2x, maybe historically what the level would suggest based on what your revenue is right now. How do you see that based on your order funnel, based on how you see supply trending? How do you see your backlog flowing over the next?

Eventually adjusting, yes.

And just to clarify, when you book orders, those are orders that are shippable over the next 6 months, 12 months? The time frame?

That's right, yes. So if you set aside kind of the deferred revenue, the software and services buckets that are sold on a time basis, our order acceptance policy is that we don't put an order on the books unless we believe at the time of order, it's shippable within a 6-month period of time. So we have a relatively short duration on our backlog, which I think gets the quality of that backlog. We believe it's a very high-quality backlog. I think as we think about a migration from what is middle today or an elevated level of backlog to a more normalized level, that is going to take place over several quarters, at least a year, likely beyond that once the supply chain situation starts to resolve itself. I mean as we think about adding capacity, which we are doing, we're not doing it in a way to put in capacity that would allow us to flush backlog over a short period of time. We're more thinking about what's a sustainable level of demand in our business, and we'll work the backlog down over a longer period of time. And again, I think that's okay in the eyes of our customers because we really haven't seen that long of an extension of our lead times. It's within kind of acceptable range and so we can work that backlog down over a longer period of time.

And just your ability to manage pricing around that backlog. Do you have the ability to reprice? Is there a need to do that based on what you're seeing from the inflationary standpoint?

Well, we're certainly actively working on pricing. We've increased the frequency of our price increases. When we do a price increase, we do not reprice our backlog. And so if you think about our backlog today, it does have layered historical price increases in it, which we're now realizing. And the price increases that we're executing today will not be realized on the revenue line for several months out. So that does create a bit of a timing challenge. It does seem like the supply side price increases are more immediate. You want to get delivery, pay the new price is kind of the way it feels. And so there can be a little bit of a mismatch. But I think we're doing a pretty good job of managing through it as evidenced by gross margins that have been more or less stable over the last year.

Yes, yes. There's been a deliberate effort, I think, really over the last 5, 6 years to focus more on the R&D part of the value chain versus production test, you've migrated the mix of business there. As we think about maybe a more uncertain economic environment going forward, what does the history suggest that R&D test volatility is versus the production test volatility?

Yes. So I certainly know that the cyclicality of those R&D revenues for us or budgets in the -- from coming at it from a customer perspective are far more stable, right? What you tend to see on the manufacturing side is people invest in burst -- in capacity, then they absorb that capacity. And certainly in the cyclical situation, they tend to have excess capacity and this capacity purchase is a slower stop. Whereas our customers, our technology customers want to invest through the cycle on the R&D line so that they have new products, they're hitting market windows as you emerge from any macro pressures. And so if you look at -- I always say to folks when I ask this question was like show me the point in time when our customers' R&D budgets contracted, right? Not that it can't happen, but they tend to be far more stable over time, and we see that in the investments they make with Keysight. So today, Keysight revenues are about 60% R&D in total, about 30% Manufacturing, about 10% kind of post deployment optimization. And GP's business on the commercial communications side, it's actually closer to 70% R&D. So we definitely have this bias towards R&D solutions.

And along those same lines, there's a software and service element mix that also builds in some resilience, we think, as we go forward?

Yes, absolutely.

Whether it's in the Communications business or more broadly, I think in the communications, we've talked about some of the solutions being 40% to 50% software, maybe software is 20%-ish overall. Not every piece of your business can get to 40% to 50%, probably software mix, but how do you see that trending over time? And where are some of the areas where software mix can come up?

Yes. Well, why don't I start, and I'll let GP make some comments. So yes, you're right, software today is about 20%, software-plus services about 1/3 of our total revenues. But as you know, some of our 5G solutions are 40% software content, something like O-RAN is closer to 50% software content. And so there are opportunities across our portfolio, particularly as newer technologies develop for higher levels of software penetration. And then in addition to that, Keysight using M&A has entered in kind of some pure software markets, things like our Eggplant acquisition which is software to test software or even the visibility business within Ixia. You're giving IT managers a chance to monitor data traffic on their networks, provide yet another opportunity for us to increase the overall level of software content across the Keysight portfolio.

Yes. So as Neil said, the recent acquisition really gave us an opportunity to move up in the value chain when it comes to software capabilities. A perfect example of that is the fact that we are serving the prodigal space of the commercial communications side. And Open RAN is a very good example that. When you look at Open RAN, you have a market like the network infrastructure market that traditionally has been fairly monolithic and driven by few vendors. And now you have 2 fundamental trends. You have the virtualization, so the capabilities of the network is now delivered through software. And you have the standardizations. These 2 factors are driving quite a bit of innovation in this market. And that's why you see in these areas that some of the solutions that we offer have a capability of software ratio in excess of actually 50%, in some of the solution, it's 60%. We see this actually being a trend. And going forward, more and more of our solutions will have more and more of the software capabilities because the complexity that needs to be solved needs to be done through tools like AI and machine learning capabilities that inherently are software.

And we want to go -- I want to go little deeper into some of your end markets. Maybe I'll just stop there and see if there are any questions from the audience, the iPad is kind of malfunctioning here. There's 1 question there.

Can you talk a little about looking at [ as you referred throughout ] some of the research [ on ] quantum and it seems like there's a bit of logic and [indiscernible]. It's a very complicated process, very low yields right now since still much of your work is in R&D., are you working with companies in pharma, in virology, neuroscience [indiscernible] full benefit from a much more predictable [ human ] development process. What are your thoughts there and how would you expect it [indiscernible]?

Yes, I'll make a few high-level comments. And then I'll give GP a chance to add. So obviously, we are active in the quantum space. I think Keysight had some tools that could be used for the folks that are working on quantum. We've done some small acquisitions in the space to bring in talent, get market access, working with the universities, the governments that are funding quantum research and I'll let you...

Yes. So great questions, by the way. So yes, we have made some acquisitions and 3 acquisitions in the course of the last few years. Interesting enough, you think about quantum, you think about quantum compute that is big gear and all golden looking. The majority of the contributions are in the software space. The reason is because of error corrections. Cubits are inherently very unstable. And in order to have a 1 cubit, a larger cubit, you might have maybe 100 physical cubits. And how you reduce that ratio is through error corrections. Error corrections, essentially, it's another way to call it, it's calibration, which is something that Keysight knows how to do very well. So we are very active in that area. We provide advanced quantum control systems. They basically control these cubits to a level of granularity and resolutions that can maintain the coherence of the cubit while at the same time, we can layer on top, control software and calibration software that maintains the error under certain limits.

Maybe we'll flip GP to the commercial communications business, just discuss that a minute. It seems like every discussion starts with 5G. But maybe we start on the wireline side or data center side, that's been very strong. You mentioned a number of drivers there, 400 gig being rolled out and then R&D on some of the next generation. So just where are we in those phases with those different generations? And what kind of time line can we expect?

Yes, it's great questions. And again, we look at wireless and wireline as kind of part of the same ecosystem. You need both in order to deliver the promise of 5G. And when you look at the wireline side, I would like to mention maybe a couple of comments. One is that with the acquisitions of Ixia, now we have been able to move beyond the physical layer. So we have a product layer that is very active that allows us to extend our capability to offer solutions. But when you look at the 400G, the 800G, I would say 400G today, it's in the early phase of manufacturing. When you look at deployment, maybe less than 5% it's -- of transceivers in the world now are 400G. Actually, less than 25% are 100G. So that gives you a sense of how long it takes for this to move. So basically more than 70% is less than 100G. So 400G, early phase of manufacturing, we expect additional manufacturing capacity to be deployed over the next few years. 800G, it's in this active heavy R&D. And we see actually quite a bit of activity in terabits type of terra-g type of solutions, both on the transport layer and also on the data center. So again, we see these waves kind of overlapping and you cannot talk about 1 wave separate from the other, but there is a continuous overlap level of research, development and validation.

And you mentioned Ixia. Along with Ixia came security and visibility of the business. Just update us there on what you're seeing in that aspect.

Well, one of the -- at the end of the day, all these generations of communication systems have the goal of give us something that is infinite amount of capacity or throughput and 0 latency. That's the end goal at the end of the day. That translates to an infinite amount of data that needs to be dealt with. So when you look at visibility, our solutions and visibility have the capability to make sense out of this huge amount of data and understand how to filter it in a way that can be used by the operators, for instance, to understand how to monitor the network. So in general, we believe that those additional capabilities that we acquired through Ixia really gives us the capability to address the future challenges of the communication market. Think about -- I mentioned before, artificial intelligence, machine learning, they need to have the capability to digest an infinite amount of data in order to create this advanced models. But you need to do it in a way that is controlled and some of the visibility tools allows us to do that.

If we turn to 5G, the inevitable question is what stage or what inning are we in? So I'll go ahead and ask that and you can provide your take on that. But I'm just curious as well what the -- as the demand curve evolves, how your deployment of hardware versus software versus physical layer versus application layer? How you see those demand curves intersect?

Do you want me to start?

You want to start, yes?

Yes, maybe I'll just start with a little bit on the kind of where are we. And I think we still believe we're very early days, right? We're working primarily on rev 16 of the standard. Rev 17 is -- and 18 are to follow. And so each one of these revs of the standards is going to cause additional waves of research and investment. On the deployment side, still a very small percentage of the world's population has access to 5G signal. And so -- and those 5G networks that are being deployed are still on the lower end of the frequency bands, the sub-6 gigahertz frequency bands. And from our perspective, there's still a lot of investment that is currently ongoing in millimeter wave. And so it's really a question -- more of a question of when rather than if on those higher frequencies because we're also starting to engage on 6G and the frequencies don't go down, they only keep going up. And so this investment that is happening in higher frequency communications is a trend that is going to be with us for a while.

Yes. You said it's just the early phase. I mean, I don't know about you, but my personal experience with 5G is not great so far. If you look at your personal phone, I'm pretty sure that you don't see a huge improvement. And the reason is because some of the best bands have not been deployed yet. C-band has just been deployed starting in January. I think Verizon AT&T is still looking at that. T-Mobile had a 2.5 gigahertz bandwidth, that was a little bit ahead. So depending on your operator, you have different experiences. But when you look at 5G and the promises of 5G, we are largely still operating a Release 15 network. Release 16 is coming, is being deployed. That's where you have for the first time 5G stand-alone core, which gives you the capability to have multiple services. And then you have Release 17 is coming, which is now in R&D active phase, which will bring the industrial IoT. So all of the promises of 5G are still ahead of us. And then you have Release 18, which is a couple of years out here. And that's what you can see almost like a bridge to 6G, 5.5G, if you will, they call it 5G Advanced. So the key point here is that we are in the early phase of 5G and some of the complexity and technology of 5G is still being actively developed.

And you mentioned Open RAN earlier. Why is Open RAN important? Or what's impactful to the test and measurement business coming from Open RAN?

I'll take that?

Yes, go ahead.

I would say, I mentioned before, virtualization, so that means that things that in the past are used to be done through hardware, proprietary hardware, now are done through software using off-the-shelf computing capabilities, so that's one. The other one is standardization. So before, you had a monolithic base station that you buy from Ericsson or Nokia as the radio, the investment unit and the computing unit altogether. Now you have 3 different vendors that are providing that, and there are standards. So standards are good for companies like Keysight because we provide the interoperability capabilities. We provide a conformance testing capabilities. That's number one. Number two, the virtualization means that, as I mentioned, software is the way you deliver innovation. When you deliver innovation through software, the pace of innovation accelerates because that's much faster. And so the innovation cycle increases, more complexity and therefore, there is more opportunity for us to enable the design flow.

I think if you step back at a high level, we took a market that was 3 or 4 customers. And now there's dozens, if not hundreds of companies that are out there trying to participate in this market, provide a portion of that network access. And it also creates more handoff points, right, the interoperability between these different companies, they're doing a portion of the solution, all requires test. And so it's been a significant market opportunity for us versus a market that was previously defined by less than a handful of customers.

There are new customers.

And maybe this gets into my next question. If you think about -- Keysight clearly had a first-mover advantage in 5G when you think about these handoffs, is there an imperative that the customers need to standardize across some of their test gears so those handoffs are more seamless? Or just how should we think about your ability to protect that first-mover advantage versus customers blending the test and measurement vendors?

Yes. I mean, I think switching costs are high, right? It takes a long time to qualify a new vendor. And so I think from our perspective, our focus on the R&D lab, being first to market, as long as we continue to execute on behalf of our customers and they're largely committed to our platform at this point, right, there's significant cost to them, not only in terms of financial costs, but time to market cost to switch directions. And so in addition to that, there's an information advantage, right? Because for our largest customers, we have engineers that are badged sent to their labs every day, we're working hand-in-hand with them. And so there's great information flow about what their R&D priorities are that we can use to inform our own product development engine going forward. That being said, if you go into the labs of our customers, you're going to see some tools from our competitors in there as well. But I think if you look across the 5G ecosystem, we've seen pretty broad adoption of the Keysight platform, if you will, and we believe that's a defensible position.

Let's shift gears to the automotive space. That's another area that's been growing rapidly. It's not -- historically, it's not a huge legacy test and measurement market. Again, we're still trying to get our arms around maybe what the size of that is. I think for you, it's still less than 10% of revenue, automotive and energy. How do you differentiate in a greenfield space where there's not really an incumbent?

Yes, I think it's a great point. It is largely a greenfield space. I mean if you step back at the highest level and say what does Keysight do, we make tools and solutions for electrical engineers, and the internal combustion vehicle was a mechanical engineering device. And so we've very much served the legacy auto industry around the edges, right? But now if you think forward to an autonomous driving, electric vehicle, just the direction the industry is headed, it is wholesale electrical engineering problems that the industry is facing. I think how do we differentiate? Well, Keysight has the broadest portfolio of tools and solutions for electrical engineers. And we make the fundamental tools for electrical engineers to do their work, and it doesn't really matter what industry they're in. And all we then had to do is to say, how do we take these tools and solutions that we have, look at the specific problems that the auto industry is trying to solve and help them -- through use case identification, how are they going to use these tools and instruments, add software, add the things around the edges that are specific to the use model for auto, right? And so in the case of EV, it's a lot about power, right, reducing power, extending range, right, extending the battery life. Not just a single charge, but the charge/discharge cycles. So we have a power business that we've had for decades, right, that we could pivot that core technology to this new use case for the auto industry to help them build better and more efficient battery and charging infrastructure. On the AV side of things, how is autonomous driving going to be accomplished? Well, through a broad range of communications technologies, right? It's everything from cellular standards, 4G and 5G to Bluetooth and WiFi and radar and LiDAR. And we had solutions around all of those communications technologies. The question we had to answer is how do we piece those together in a way that's useful to the auto industry, but the core capabilities already existed and we're already differentiated within Keysight.

Is that industry moved to define a platform yet where -- such as 5G, where they can standardize on the platform or is it still evolving and it's more bespoke solutions based on the vendor?

Yes, more of the latter. I mean, I think certainly, we like industries that have standardization because people test to the standard, right? I mean just at the simplest level. And you're not really seeing that broadly adopted in auto yet. And certainly, from where we sit, it makes a lot of sense. But again, we're very early. We look at the auto opportunity right now measured in decades, not in years because it takes that long to turn over the installed base. It takes a long time to get our new auto technology from the lab onto the roadways, given the life safety issues that are involved. And so they tend to be a little bit slower moving, but long technology cycles.

I want to touch real quick on your semiconductor business as well. You've referenced semiconductor being around 10% of revenue as well. I don't know if that touches all of your R&D capabilities or if that is more of the production area. But how do you think about the cyclicality of that business today versus what it's been in the past?

Yes, it's certainly -- it would appear at this point to be lower than it's been in the past. We've run this kind of prolonged semi upswing. And given the supply chain situation that we're in, given the -- there's kind of multiple drivers, right, there's the technology side moving to smaller process architectures. There is the flat-out demand side. There aren't enough chips, there's chip shortages we need to get capacity into place. And then there's this whole kind of nationalistic or regional element about assurance of supply, right? And so you're seeing re-onshoring of fab capacity into Europe and into the U.S. because people realize it's a national security -- or assurance of chip supply is really a national security issue. And so you -- we're getting positive demand signals from our semiconductor customers today out through, say, 2025. right, which is by historical standards, that is a long amount of visibility for that space.

Yes. Okay. Real quickly, just on the business model. Again, Keysight targets mid-single-digit type growth, 40% incremental margins at that level. You talked about some of these expenses eventually coming back to the P&L right now. Is that a dynamic that you think carries into '23, just this annualization of the expenses coming back? Or do you have levers, be it your variable compensation that modulate that if we still get good growth and we get better supply and backlog?

I mean, I think both of those things are true. So first of all, we are going to see costs return to the structure as our employees are trying to work as they return to airplanes. So there are going to be some real costs that return. Similarly, from a -- just from a margin perspective, I think the inflationary environment clearly makes things harder and not easier, right? And so that's clearly a pressure. And right now, we're maintaining gross margins. And even that is a challenge in this environment. But that being said, the backlog is there, the demand is there. If we start to see supply chain improvement, you should be able to put up nice revenue growth. And over the longer term, I think that the right way to think about Keysight is delivering that 40% incremental any time our growth is mid-single digits or better, right? So there could be some short-term challenges. But again, if the supply chain situation resolves itself, maybe you get a little bit more help on the top line, which allows you to kind of reabsorb some of these expenses that are coming back into the P&L.

And accordingly, free cash flow is about 100% target of net conversion on net income. CapEx is a little higher this year on some investments. What's the sustainability on the CapEx? Does that -- is this just kind of a 1-year dynamic or does that carry into '23?

Yes. So this is our second year of kind of elevated CapEx, and we have a couple of big programs that are ramping up. I would tell you that -- a couple of things. So one, we are also seeing increased investments in capacity. So that part of it is like going to continue even as some of these programs wrap up. And from a CapEx cash flow standpoint, it might extend a little bit. The commitments are basically all out there. But just like we're seeing in other places, that's taking longer to get things into the plant. And so the cash flows are a little bit delayed. So I would expect next years to be a little bit elevated relative to historic norms, but we'll have to see where that land.

Perfect. Perfect. We need to end it there. We're out of time. There is a breakout session, so feel free to join us outside at Astor B, but we want to thank Neil and GP for joining us.

Yes, thank you.

Thank you.