Right, I think we're on time. I know we've got a few minutes.
Few minutes left.
Yeah, yeah. Guys, when the lights go down, I assume it's.
That's right.
You've got your logo up here, which is good. Have you got any slides to talk to now?
No slides.
Okay, okay.
No slides.
You managed to get the—usually they have the Deutsche Bank logo.
All right, let's get started. Welcome back to this final session of day one of Deutsche Bank's technology conference. I'm delighted to welcome Neil Dougherty from Keysight, CFO. I'm Rob Saunders, European tech hardware analyst. You know, Keysight, for those of you who don't know, is a leading test and measurement equipment company, but also shifting to software. We're going to discuss the demand environment, tariffs. If you have any questions along the way, please raise your hand. Without further ado, maybe Neil, maybe just start by giving a little bit more of an intro than I just gave, just to the company for those who don't know. Then we can get into more of the questions.
Yeah, so as you suggested, Keysight is the leading provider in the electronic design and test space. We have a broad portfolio of hardware and software solutions, which we essentially bring to market targeted at a number of different end markets, most notably commercial communications, focused on both wireless and wireline communications, aerospace defense, where we're benefiting now from increased commitments in terms of spending as a percentage of GDP from the NATO allies and now multiple years of increased U.S. aerospace defense budgets as well. We have an industrial business that is focused on kind of next-generation automotive, including electrification of the drivetrain, move to more of a software-defined vehicle, which would include everything from autonomous driving to various levels of autonomous driving, software-defined vehicles, in-car infotainment systems, and really all the in-car electronics and in-car networks that are part of a modern automobile.
We have a semiconductor business that is really now being driven by move to smaller process architectures, things like silicon photonics, high bandwidth memory, are drivers that kind of link back directly to the AI side of things. Then we have a kind of a catch-all business in general electronics, which, you know, is kind of everything else that has electronic content in it. There are some submarkets in there, which we tack separately. We pay attention to the education, advanced education, research end markets, med tech. Those areas are embedded in that as well. Again, hardware and software, about approaching 40% of the revenues these days are coming from software and services. The balance coming from hardware and that mix continues to shift in the direction of software and services. Maybe just lastly, the business is pretty heavily levered to our customers' R&D labs versus their manufacturing lines.
You can think of it as high 50% R&D, maybe 30% manufacturing, with the balance being kind of post-deployment, installation, maintenance types of applications.
Great, you had your Q3 results last week.
Yes.
You did have a beaten race, which is congrats on that. You raised your sales growth to 7% for the fiscal year 2025, from I think closer to 5%. Can you pass out the delta in your improved outlook between AI spend, perhaps upside to aerospace defense, and anything else?
Yeah, absolutely. As we entered this year, keep in mind 2024 is a down year for us, a cyclical down year, but we'd started to see our end markets inflect, driven most notably by wireline and the AI buildout. We realized that all of these end markets, which I just walked through, were not in phase with one another. Some of them were recovering or set to recover. Others we felt were still likely to be a headwind for us as we entered FY 2025. Our base case was at the low end of our long-term growth target of 5%- 7%. We were forecasting growth this year of 5%. We've actually been able to increase that full-year expectation each of the last two quarter ends. At the end of Q2, we took it from 5% to 6%. At the end of Q3, we took it from 6% to 7%.
If you think about what has outperformed and what's kind of more performed in line with expectation this year, first of all, I'd say what's leading the growth charge for us is, in fact, our wireline business driven by AI. I would say that growth has been largely in line with what we expected. We expected it to be strong. It has, in fact, been strong and, again, largely in line with expectation. I think we've seen a little bit of upside on the aerospace defense side, driven, as I've mentioned, by budgets.
I really think that opportunity is more in front of us than it is being realized currently because there is a long lag that occurs between the time a politician stands on television and makes a commitment to increasing defense spending to the point that it's actually in budgets, assigned to programs, and then RFPs and other things have been conducted and it actually results in orders on a company like Keysight. Semiconductor, we expected to rebound and, in fact, we've seen a strong rebound, maybe slightly higher than expected. I think there is some upside that we've seen in the semi space. The auto business, we expected to be challenging and, in fact, has been challenging. I think that's in line with expectation. Probably the single area that has the single biggest area of outperformance for us has been in wireless.
Wireless, our base case coming into the year was for our wireless business, which is a large business for us, well north of $1 billion , was for that business to be flat, stable this year. In fact, we have seen some growth, some nice growth in that business. It's not growing as fast as it was a couple of years ago, but it's a big enough business and we've seen enough growth that it's a significant contributor to us being able to take up our growth rate for the year for the entire portfolio. It's been a good year. I think we've done a good job executing and converting that revenue growth to profitability. We've actually been able to take up our EPS growth expectations, even as we've absorbed significant tariff headwinds. It's been a good year so far with one quarter to go.
Great. You mentioned tariffs. There was a little bit of disappointment I'm sensing last week around some tariff discussion. Maybe you could just clarify how you're seeing the tariff situation impacting your product cost and margin, and where do you see the biggest impact today?
Yeah, so, obviously we had now kind of two rounds of tariffs. There was the initial round of tariff announcements back in April, which was kind of baselined at 10% for a lot of countries. We saw those tariff rates increase most recently here in August. I think right now we're likely to settle in somewhere around the $40 million a quarter in terms of total tariff headwind as a result of the two waves of tariff announcements. The good news is we feel like we largely have it handled. We've taken a set of actions to mitigate this tariff exposure. It is going to take some time to fully be realized in our P&L. It's a mix of cost action and operational efficiency actions and another set of actions that are focused on passing the remaining costs onto our customers via price increase.
It's those latter increases that really do take time, right? We have not made an attempt to reprice backlog. We've honored outstanding quotations to our customers largely. Given our relatively large backlog, $2.4 million of backlog, it does take time for new quotes that include the pricing changes to turn into orders and then ultimately turn into revenue. What we've said is we would have the April round of tariffs fully offset within Q1 and a second set of actions here in August. We would expect to have the August tariff increases offset before the end of Q2. I feel pretty good about where we're at and feel like we've got a really robust plan to ultimately make these things earnings neutral to our P&L.
Okay, let's dive deeper a little bit on the end markets. Let's start with wireline. Obviously, I think investors, I'm guessing, seem to be looking at you as a potential AI play, but it's quite hard to parse out how much of your wireline story is to do with hyperscalers and how much is just general demand. Is there any way you can sort of quantify the AI contribution to that wireline growth that you're seeing?
Let me start by quantifying our wireline business. We have a commercial communications business that is in the vicinity of $2.5 billion and approaching 50% of that. We're getting very close on a quarterly run- rate of having 50% of that business come from the wireline side of things. Obviously, the major underlying driver of our wireline growth over the last 18 to 24 months has been this rapid acceleration of investment in AI. I think the good news, there's a couple of bits of good news here. First, we're very bullish on the opportunity as we look forward into FY 2026 and beyond, right?
What we are seeing is a shortening of the technology cycles as the industry looks to move to faster speeds from 100, 400, largely being 400- 800, which is where most of the activity is right now, ultimately moving to 1.6 Tb, then 3.2 Tb 6.4Tb, and even talking about 12.8 Tb at this point in time. As those cycles shrink, you essentially need to redesign all of the equipment that makes these networks work. The network silicon, the connectivity, the optical transceivers, the network equipment, and then ultimately, we have touchpoints with all of those aspects of the hardware buildout itself. We actually have touchpoints with the hyperscalers themselves that are building out these data centers.
You have a combination of multiple touchpoints between Keysight Technologies' physical and protocol layer solutions to service this ecosystem, and shortening design cycles as the industry works through various technology inflection points that give us confidence in the opportunity as we move forward. This is very exciting and, again, a significant driver of growth for us as we look both backward over the last, say, 12 to 18 months and forward into the coming year.
Great. Would you call out anything in particular? It sounded like it's mainly interconnect driven. Is it Ethernet? Is it PCIe? Is there perhaps a trend towards scaling out versus scaling up that you're seeing that could be an inflection for you guys? Is it just, just as you say, a shortening of the whole cycle overall that's driving your business?
Yeah, I think it's an interesting question. I think we have the opportunity to benefit both from scale out and from scale up. As I think of those two opportunities, and keep in mind, as I've said, our business is significantly more levered to R&D. That's true at the Keysight level, and it's even more true within our wireline business, which is probably 70% R&D versus manufacturing. Given that emphasis in this space on R&D, I think of the scale out play as more of a volume play, whereas the scale up play is more of a technology play.
As I think about those two things, I tend to think of the scale up play as the bigger long-term opportunity for us because not just the move through various speeds, but if you think about the multiple developing alternatives to the NVIDIA ecosystem, whether that's Ethernet or UALink or PCIe-based solutions, there's a complete set of tools and equipment that need to be built out in support of each of these standards. We have touchpoints at silicon, connectivity, network appliance, and with the hyperscalers who will ultimately deploy these technologies. It's kind of a little bit of a more the merrier standpoint when you're selling into the R&D lab. The more technologies that they're trying to bring to market, we're agnostic ultimately to winners and losers, but it creates a tremendous opportunity in the interim.
Great. If anyone has any questions, please do raise your hand. Let's switch to wireless. Obviously, I think if you think historically it was one of your larger businesses, it sounds like wireline is now bigger than wireless at the moment. How is AI influencing your test product development in wireless? I'm thinking of features like RF performing, using AI to software innovation. How are you seeing that and how can you monetize that?
Yeah, so a couple of comments. First of all, just to correct the record, wireless is still bigger, but the gap is closing and we're very close to 50/50. Wireless is still the larger business. We'll see when that transition happens and ultimately wireline becomes bigger, which appears to be the trajectory that we're on. I think as it relates to AI, what we're seeing is within our customer set, primarily on the network side, and we're in the very early stages of this. You can think of it as more research rather than development at this point in time. They're looking at how do they use AI to optimize these networks? How do they optimize for power efficiency? How do they optimize for network traffic?
How do they model the behavior of the actual user themselves so that they can better predict and ultimately shape the form of the beam to provide better connectivity for a mobile user? What Keysight does to enable this AI learning within this wireless network ecosystem is we essentially have simulation and emulation tools that can simulate very high volume network traffic that can be used for these models to learn. We can simulate the traffic, they can create a model that can learn and ultimately apply these AI algorithms to, again, model power efficiency or call scheduling, these types of things. Once the model has been built, you can use these same tools to test the efficacy of the actual AI-driven model itself. It's a highly software-driven focus to actually enable the AI modeling and the learning of these AI algorithms for these ecosystems.
Got it. Sorry, there's a question. I can't see. Oh, hi. Yeah, go ahead.
I don't necessarily know specifically about the transaction, but certainly non-terrestrial networks are a significant driver of investment in our ecosystems today. One of the areas where I talked about earlier that surprised to the upside was growth in wireless. A lot of that investment, one of the areas where we're seeing it, is this investment in non-terrestrial networks. When you think about these non-terrestrial networks, the connectivity between an earth-based device and a satellite at the physical layer is not that much different than a terrestrial wireless connection. We use similar tools, but you have added complexity, right? The satellite itself is moving at, what is it, 27,000 kph or something like that. You have this speed, you have the distance, and you have the complexity of handoffs from one satellite to the next that need to be modeled.
Similarly to other industries, we have multiple touchpoints within these ecosystems. Folks that are making components that go either into the satellite or to the ground station, subassemblies, things like antenna arrays that need to be developed and tested, and then the actual satellite itself, as well as the ground station and the communication channel, earth to satellite communication channel. We can provide, in this case, simulation capability to simulate these data links to test the ground station, to test the satellite, and then a whole slew of both physical and protocol layer tools to test the components and subassemblies that make up these complex networks. This is exciting for us because of the multiple touchpoints.
Right, yeah. How is your positioning within the satellite space, I mean, relative to your position in the wireless, in the wireless sort of classic wireless operator area? Do you have a comparable share?
Yeah, I mean, I think the thing that is unique about Keysight relative to other players in the broader test space is the breadth of our portfolio. We've been in this business a very long time. We have by far the most complete set of tools and we have for a long time. Even if you go back and look since the launch of Keysight and the spin-off from Agilent, at that time we were very much focused on physical layer test. What we did post-spin is we wanted to expand our touchpoints with these marketplaces and move up the protocol stack. We moved into these protocol layers. We bought a company in 2015 called Anite, which got us protocol layer solutions for wireless. In 2017, we bought a company called Ixia, which got us protocol level solutions for wireline.
As a result, we have a pretty unique portfolio, again, a complete set of physical layer tools coupled with protocol layer solutions for both wireless and wireline that's unique in the marketplace. What that allows us to do for our customers, instead of just selling them tools and allowing them together to cobble together a solution by buying tools from multiple vendors, we have a better opportunity to understand the specific challenges of the industries we're facing and provide them the complete solution with hardware, with software, with services that addresses that and essentially accelerates their time to market, which is why the focus on R&D. What are we really selling? You're selling time to market to our customers. It's a breadth of our portfolio that I think uniquely positions us in the marketplace to do that.
Got it. Let's switch to Spirent, a company I used to cover. You're awaiting regulatory approval to acquire Spirent. You've had to make some concessions along the way. It sounds like you're going to close the deal by the end of October, but there's still some approvals to go. Maybe you can just remind us where you are in terms of the approval process and which part of the portfolio you'll be acquiring and you're looking forward to having part of.
Yeah, absolutely. There's kind of one step left, and there's one big step left, and then there are some kind of formality steps that we'll also need to go through. The big remaining step is we're waiting on approval from China SAMR. All I would say is our conversations with SAMR have been and continue to be constructive, and we're highly optimistic that we're going to get this transaction closed by the end of our fourth quarter. We'll continue to work with them until those approvals are in hand and we can close the transaction. I think from a strategic viewpoint and why this transaction was important and why we're excited to have it be part of Keysight , there are a couple of SAM expansion portions of their business, which create a material expansion to our served addressable market. One is around precision location, kind of GPS precision location.
This is probably the crown jewel from our perspective. It is something that we believe is going to be critical to 6G. It's critical in the aerospace defense sections. I just talked about how they're using AI to improve the beam forming as they model the likely behavior of a user with a mobile user with a cell phone. The more precise location you can get, the better off you're going to be able to do in modeling those types of things. I think there's a gap in our portfolio. We believe it's going to be integral to success in 6G, and it's something that we're excited to have on board. The second thing would be more of a network monitoring thing. We do a lot of stuff, again, in the R&D labs, in manufacturing, helping bring these communications ecosystems to market and getting them deployed.
We didn't have a lot of touchpoints with those ecosystems once the networks were deployed. This gives us a further extension of touchpoints with the ecosystem into deployed networks and yet another revenue stream. Excited to have those things on board. Ultimately, we expect once we get the business integrated, it's going to be accretive to both gross and operating margins at the Keysight level. A great opportunity for us to create value for shareholders.
Great. That's a neat segue to aerospace defense, given their positioning business. I guess that could be relevant for robotaxis as well. Let's talk about aerospace defense. There's been a lot of different policy announcements. The NATO spending increased, the big beautiful bill. There's been DOGE. What have you seen at a net level? You said it sounds like it's more ahead of you.
Yeah.
There are, I guess, potential headwinds you might have seen from government efficiency measures. What have you seen so far?
Yeah, I think a couple of things. First of all, we believe net additive. Certainly net additive at the macro level. We have actually seen less detraction from things like DOGE than you might have expected because, again, where we tend to play, aerospace defense technology spending, and the DOGE was really more about efficiency gains rather than trying to reduce fundamental spending in core technologies as we move forward. I think one of the things that we're benefiting from in the U.S. is there does seem to be one of the few areas where there does seem to be some political alignment between the Democrats and the Republicans in terms of the need to invest not just in defense, but in defense technology. I think that's been helpful. We've seen increasing defense budgets across both the Trump 1, Biden, and now Trump 2 presidential administrations.
That has been additive and we would expect to continue to be additive. One of the big changes that we're starting to see now in terms of picking up activity is the increased commitment to spending as defense spending as a percentage of GDP among the NATO allies. This can take time to work its way through, but we are seeing increased activity, increased quoting activity, bidding activity on programs with our European defense customers. These tend to be long-term commitments. It's something that we're excited about and has us very bullish on the aerospace defense opportunity as we look forward.
What are the sort of priorities that excite you the most in the defense modernization? Is this just a modernization process of bringing this existing equipment up to date, or is it more these Germanies of this world that are stepping up for the first time?
I think it's both. I mean, I think as these NATO allies who maybe have been underspending for a period of time look to have dramatic increases in their spending, a certain percentage of that is going to go to things that we don't necessarily care about, boots and, you know, weapons and bullets. A certain percent is going to go to things that we very much do care about, technology investment. That technology investment primarily centers around communications, signal capture, signal jamming, understanding the communications signature of the battlefield, if you will, electromagnetic spectrum operations, radars. These types of things are where we have the biggest contributions to make.
I think it's highly likely that as the NATO allies increase spending, a not insignificant percentage of that business, particularly as it relates to Keysight with a $1 billion plus aerospace defense business, is going to create significant opportunities for us moving forward.
Got it. Let's switch to automotive in the interest of time. We've got this big change happening with both on the powertrain to EVs. There's a change from the West to China, which looks like it's going to dominate. At the same time, we've got this transition to software-defined vehicles, which China seems to be leading on. How are you positioned relatively given that China seems to be taking the lead in a lot of this innovation?
Yeah, I mean, I think I'd say two things. As it relates to the drivetrain and the move towards electric vehicles, China has been a significant disruptor, not unlike what happened in solar panels. There are lower cost Chinese batteries available. I think it creates real questions as to what level of investment is going to continue in the OEMs and the other big auto producing countries around the world. That's a question to be answered. If there is good news in there, it's that our EV business has been significantly de-risked at this point. There's not a lot of downside. We'll see if those markets recover. Where the real opportunity lies for us is in the software-defined vehicle space.
You can think about autonomous driving as probably the really big opportunity to play out over at least a decade, maybe a couple of decades if you think about true level five autonomous driving. Also included in there would be infotainment systems, in-car networks, a lot of in-car electronics are also included in this space. I do think you're going to see the broader global automotive ecosystem continue to develop and invest. It goes beyond the car itself. You talk about vehicle to everything communications. You've got infrastructure elements and other things that are going to need to include these communications technologies. When you think about autonomous driving, how does it ultimately get accomplished? It's through a confluence of multiple different communications technologies, things like 5G and 6G, but also Wi-Fi, in-car LAN, radar, LIDAR systems. These are all areas where Keysight has core expertise.
I think we're uniquely positioned to help the industry, particularly as things standardize, right? We love industries that have standards. The auto industry has resisted that in a lot of ways. You can think of that about charging, right? The Tesla plug is different than the BMW plug as an example. If you talk about getting to true broad scale autonomous, there's going to have to be some standardizations. These industries are going to have to come together, which creates interoperability challenges and all sorts of other things, which is good for Keysight. As I said, where we are seeing strength in auto today is in that software-defined vehicle. This is an area where we're very bullish as we look forward.
If China does take the lead, do you see is there a risk in five to ten years that those guys switch from sort of best-in-class Western vendors to, you know, domestic vendors, even if they're inferior, maybe they're under pressure?
Right now I think that's unlikely. I just think, first of all, it's significantly a more challenging technological problem than the battery problem where there has been some significant Chinese disruption. I also just think it's unlikely that these major auto OEMs around the world, you think about the big names in Japan, Korea, and Europe, are just going to cede these markets, right? I just think that there is going to continue to be development across the board and ample opportunity for Keysight to continue to compete and win.
Got it. You've got your semiconductor general electronics business. Obviously, there's a lot of design activity around AI chips, perhaps not actually that much volume relative to the overall semiconductor industry, but if you're playing into the R&D side, then presumably you've seen quite a big uptick. How is that affecting you so far today? Where are we in the kind of baseball innings of that?
You're talking about the use of AI in chip design?
The AI-related semiconductor demand, for example, custom silicon from hyperscalers, obviously new potential customers. How have you seen that impact your?
Yeah, so I mean, if you think our semiconductor business is pretty, you know, we have a couple of highly differentiated positions. We're not a broad semiconductor tools player, but within our semiconductor business, we make a tool called the parametric tester. It's essentially a wafer level test at the end of the manufacturing process. We are definitely seeing AI-driven demand, right? You can point it back to high bandwidth memory. You can point it back to the demand back to the migration towards smaller process architectures and an increasing focus on silicon photonics. All of those things directly link back to AI. We tend to be more agnostic to what the actual functionality of the chips itself. We're more of a wafer level test on the production processes. We can tie the demand that we're seeing back to the underlying market drivers.
AI is clearly one of those drivers that is driving demand across the semi-ecosystem.
Got it. Last couple of questions, unless there's any from the audience. Just one more on China and one more on margins. On China, it's about high teens percent of revenue, as I understand it. How has the sort of overall geopolitical backdrop affected that market? What do you see going forward? You've touched a bit on in terms of auto, but what about for your whole business?
Yeah, you know, I think there are puts and takes as it relates to China. First of all, we've been dealing with now for six or seven years the ongoing evolution of trade sanctions on China. Obviously, it started with the restriction on selling at aerospace defense. Then it was Huawei. Then it was semiconductor. This has been an ongoing headwind, which we've largely been able to plow through and continue to grow our business. I think with the escalation of the tariff environment and just the prolonged nature of this, we're seeing an increased portion of the industry move to at least the China Plus one strategy, if not just an all-out move out of China. While we have seen some pullback in some areas of demand in China, it's been offset by a corresponding increase in demand in Southeast Asia.
It gives in one area and takes away in the other. I think that the net benefit is that we continue to see growth across the broader Asia region as a result.
Got it. Just a last one on margins. Obviously, you've been transitioning to a greater portion of recurring revenue, a greater portion of software. How much more margin upside can you push through from that transition from here?
Yeah, I mean, I continue to think about the high-level operating model, which we have laid out for investors, which essentially calls for our business to grow at, you know, 5% to 7% on a sustained basis over the cycle on average. When we grow at that level, we can deliver 40% operating leverage to the bottom line. You see that this year, we're actually slightly over-delivering to that metric, ex-tariffs, and the tariffs have obviously been a little bit of an exogenous shock to the system. As our business has continued to recover, we expect to continue to grow in line with those long-term averages and expect to continue to increase our overall profit margins by delivering to this strong incremental. That's the way we manage our business, and we believe we can continue to deliver to that model.
Thank you very much. I really appreciate the time. Let's go to the drinks.
All right, thank you very much.