All right, I guess we can get started. Thank you all for coming. My name is C.J. Muse with Evercore ISI, semiconductor equipment analyst. Very pleased to have KLA Corporation. We have Bren Higgins, CFO and EVP of Global Operations, and we have Oreste Donzella, EVP of EPC, and Kevin Kessel is here as well. Welcome.
Yep. Thank you for having us.
Thank you.
I guess maybe first question really just talk about kind of the current environment, which I'm sure you haven't been asked recently. This is really a cycle unlike any cycle I've seen, right? Where end demand trends are softening, yet you guys can't keep up with demand, and your backlog is growing incredibly, and I think you expect it to continue to grow. You know, give us the lay of the landscape and how are you thinking about, you know, things. Do we get a soft landing, hard landing? You know, what are your thoughts?
It's a great question, and you're right. It's certainly been top of mind for a lot of investors. Before I get started, I just wanna make the obligatory safe harbor statement. I'll make some forward-looking statements, and those statements are subject to risk. You can find our risk factors in our SEC filings, which you can access via our website. You know, what's funny, I was talking to some of my CFO peers and certainly with the macro backdrop, it seems like what's interesting about this time is with every negative macro data point we seem to get, I get $100 million in orders. I think it's reflective of a few things.
It is different, but I think it also shows the level of demand relative to the ability to supply and how much under demand we've been, the industry has been able to supply. I think we've done a pretty good job as a company this year, and we talked about this a fair amount in our Investor Day in terms of how we manage our supply chain, and so on. We've seen sequential growth through the year. We had a tick down in March from December related to Omicron and some of the shutdowns associated with that. But Jun e was a good result. September, we guided a sequential increase and are comfortable with our perspectives on September shaping up the way that we thought.
And we do see December as a sequential growth opportunity as well. One of the other things I said at earnings was that I thought that as we exit the December quarter, that the expectations around output for the company, that there's sustainability in that, as we move forward. You have the usual normal kind of customer churn that you see related to slots and slot planning. The demand and pressure to ship today in aggregate is as strong as it's ever been. I'd say supply chain challenges are probably not much better. I mean, they've continued to be difficult. They continue to be difficult now. A lot of work done by our supply chain teams. I run operations and supply chain at the company, and so I'm living this daily.
Different issues, but lots of issues that continue to pop up. So a lot of pressure to continue to ship. So, we feel pretty good about what's in front of us. It sets up for a pretty good year for the company from an outperform point of view. We expect the company to be somewhere in the mid-20s overall in terms of growth. On top of the last couple of years, we've seen a nice trajectory of growth, be pretty close to doubling the size of the company since 2019 from a revenue point of view. I think it sets up pretty well.
Now look, at the end of the day, as we look at, you know, some of the macro backdrop and some of the customer comments that are out there in terms of some of the weakness in some of the consumer markets, particularly around mobile and PCs, how that affects memory, we're certainly aware of that, and we'll see how that plays out in terms of how customers are planning as we move into 2023. Not gonna guide 2023 today. I think we're a little bit a ways from there. Certainly what you're hearing from those customers on the memory side is that there'd be reductions in their plans for next year. Although memory tends to be, because so much of incremental bit supply comes from new capacity, memory tends to move quickly, kind of both directions.
They're big bets. In the old days, when it was all from shrinks and adding bit supply through shrinks, you know, wafers weren't necessarily growing. It was an easier thing for our customers to manage. Now they have to make much bigger bets. Sometimes what they say early in the year or in the previous year about what they're gonna do the next year can vary, and it's all obviously very market dependent. But we're so far, and what I would expect to see is, given the strength of demand in logic and foundry, that we would continue to see whatever, if slots became available, they get filled very quickly by those other customers. I feel pretty good about that. I feel pretty good about the backlog we have.
We're a technology enabler at KLA, so most of what we tend to do better in the front end of R&D and in a ramping of a facility, less around the incremental wafer starts at a mature production node. A lot of our backlog, particularly around some of our high-end systems, is related to customers pursuing their roadmaps. I feel pretty good about what we have and how that translates into output moving forward. I'm trying to get a lot more out the door than I'm able to still, and I don't see it letting up anytime soon.
If you think about, I think you've talked about what, $13 billion+ in backlog exiting June, and I imagine that you know will grow September, December. Do you worry at all about that being at risk? I mean, I think that you know from what I hear you know customers are scariest of canceling on you and ASML kind of the most and reluctant to get out of line. I guess if things get terribly worse. I mean, is that something where you know you would see risk or no?
You know.
Based on the mix you see in the.
It's a question I continue to poke at, obviously, internally as well. As I said, since we play this role in enabling roadmap progression, a big chunk of our backlog tends to be for products that have long lead times. Customers usually make investments in those areas irrespective of what's happening with top lines. They don't delay their roadmap. They may delay the size of a project. If they delay a whole project, that could change the equation a little bit. If they're delaying the size of something within a particular timeframe, it's less of an impact. I feel pretty good about what we have from that point of view. Now we do have some capacity-centric products. Those products could be impacted if you saw a slip like that.
And I think finally, there's probably some issues related to facilities, which sometimes causes delays in timing of shipments where it's a brand-new greenfield fab, and if the facility is delayed for usual construction reasons, you could also see slips there where it's not practical necessarily to take a shipment if your project's delayed six months. There's always those kinds of issues, but at least where we stand today, I think customers would be very reluctant to get out of line as it relates to a lot of our high-value systems. The lead times on those systems can be, you know, as long as two years right now. I think we feel pretty good about that roadmap.
Capacity-centric products, there could be some exposure in those areas if you saw some slowing or some big delays, but that's a smaller part of our overall mix.
Makes sense. Maybe we could talk about kind of the moving parts of 2023, and not you know, asking for kind of formal guidance, but maybe, you know, what's a positive? What might be a concern? If we started with leading edge, or real bleeding edge, you know, I would think that, you know, Intel would be a real positive for you, as they are going more and more merchant, you know, process control where your share should be rising nicely. TSMC pushing hard at three and five, but you know, maybe some less capacity depending on, you know, Intel's decision, in terms of, you know, their GPU tile. Do they go three or five? Is there some kind of squishiness there? And then I guess Samsung.
How, if you add up all three, how would you kind of think about that?
Yeah. I think the competitive dynamics of leading edge. I don't wanna get into the specific customer dynamics, but the competitive dynamics of leading edge are strong and they have been very public about the plans in terms of future opportunities. Frankly, the diversification of end demand at the leading edge has also been very good. I think that we would see a continuation to your point, right? If you start to see and it makes sense, right? As you're doing development, you tend to oversample. You do a lot of very kind of rigorous process monitoring. You're trying to figure out systematic defectivity. You're monitoring it every step. Then as you're ramping a facility, you introduce, okay, volume to that process, and how does volume change the maturity?
Then you get up to production and you're like, "Okay, I've got a fairly predictable process. My yields are where they want to be, where I want them to be." Your sampling strategy tends to drop off a bit. We're participating more in that part of the business than we used to because of the proliferation of designs at the leading edge and the amount of process flows that are flowing through. But it's less of a part of a business for KLA or less impactful than what we see in what I'll call the development in the ramp. If you see a wafer start reduction, then obviously there's an impact to that from a timing point of view. But I think we're gonna participate more on the front end of that.
On the legacy side, we're just kind of talking through the markets, think about 28 nm and above. Process control intensity tends to be fairly modest in legacy. The reason for that is if you've got a customer who's been making a part for a long period of time and now has demand and has to add new capacity, unless something's changing in the specs, which happens in automotive, but generally unless the specs are changing, the process control intensity is gonna be fairly modest. On the legacy side, we haven't really shipped a lot of that business yet. We've got, you know, some in the second half of this year, and we'd expect to see some into next year. If you look at those markets, automotive looks pretty good.
I mean, if you go look at lots today, you don't see a lot of cars out there. I think that there's very real memories of what happened just a couple of years ago when supply chains were shut down. That business looks pretty good. Activities in native China look like, given the backlog that we have and given the investment that's happening in legacy nodes, but also in memory, but also infrastructure, wafer capacity, reticle capacity, that I would expect China to be relatively consistent next year with this year. I don't see necessarily a fall off there. We'll have a lot more to say as we go through it.
We get some a little bit more input from customers as we get closer to the end of the year to give a clearer picture on what we expect for the year.
Makes sense. Maybe we could move to process control intensity. Has outgrown WFE for the last two years. I guess would like to kind of parse through how much of that was kind of the mix shift foundry logic from 50%-65% of overall spend, and how much of that is some of the new kind of innovations like gate-all-around and other 3D dynamics driving that?
Yeah. It was one of the things we wanted to talk about at Investor Day was with the introduction of new technology that we have seen really end market embrace of the nodes from a design point of view. It's always been good for KLA that when you have a robust design environment and customers have to manage a lot of designs that test design rules in different ways and have to deliver to pretty tight market windows, that has been good for overall process control. We've seen that, and we've also seen that because of the introduction of these technologies, that because of that demand, that the ability to take the previous node capacity, if they didn't have the follow-on demand from other customers, and try to migrate it to the next node.
When you had the delays in EUV, as an example, scaling was slowing, there wasn't the end demand, and there wasn't necessarily a technical driver to keep customers from being able to pursue the capital efficiency of trying to reuse a lot of their equipment. As we moved over the last few years or so, as we've seen designs really start to pick up at 7 nm, and we went through some data at Investor Day about how that had changed relative to the previous couple of nodes. We saw that pick up, and we've seen that, okay, customers have had to invest more, but also the inability to migrate that capacity, less reuse. Those have been two big factors that have driven the business, but then driven that mix.
The mix shift tends to be a very favorable thing for us. Logic foundry process control intensity tends to be almost twice as high as in memory overall, particularly at the leading edge. You could argue at the lagging edge, it's probably more like memory, but at the leading edge. We'll call it sort of mid- to high-teens% of wafer fab equipment, where memory is 9% or 10% as a percent. It does give us a nice tailwind from mix. But I think the mix change has been a byproduct of this change in design and how that's affected how customers sample and monitor, but also how they manage their overall capacity.
Makes sense. Why don't we hit on the competitive landscape? I don't know, Oreste, if you wanna chime in on this?
Yeah.
Whether you just wanna hit on EPC, but I know you know all this cold as well. I guess, you know, what are the biggest battlegrounds that you see over the next three to five years, if at all?
You wanna start, Oreste?
Sure. First of all, I mean, semiconductor process control, as you know, we hold a very strong market share position. We are 4-5x bigger, larger in market share in the cumulative process control for semiconductor, so that is comprising the segments we serve, like optical with respect to metrology serving, and then we don't serve, like, CD-SEM, and we still hold 4-5x larger share than the nearest competitor. What we see, we take competition very seriously. In KLA, we used to have a strategic planning. We do strategic planning every year. That is a very important pillar of our KLA operating model. So we just started this strategic planning process right now.
We take the competition so seriously that we play the role of competitors, and we try to figure it out how our competitors are going to behave in terms of spending in process technology, capacity expansion, and so on. And this is what we do every year. So what we see in the landscape, of course, we see very, very strong position, as you know, in the optical wafer inspection. This is the segment that is outgrowing the process control and the WFE overall, so we have incredible position in that segment, and we continue to grow either in size and in share. Of course, we see also challenges from other competitors there.
The only way we have to compete effectively in the market is to make sure that we listen to our customers, we see what they do, 'cause sometimes listening is not enough. We need to be in the fab and understand how they deploy our tools, our competitive tools, and then we are ahead of our competitors in developing differentiated technology. This is part of what I call the strategic planning process. Of course, there are a lot of questions about reticle inspection. There are a lot of questions from investors, from other analysts and so on. In reticle inspection, the market is extremely complicated because reticle inspection, we compete in the mask shop, where we need to inspect the reticles that are coming out of the mask shop and also in the wafer fab.
In mask shop, if you look at right now, 90+% , I would say 95% of the reticles are inspected with the KLA 193-nanometer optical reticle inspection tools still. There is a small fraction of reticles, EUV reticles, the advanced reticles, that are now inspected on, and this is more on the path finding and not really applicable to the production of the masks or the reticle that are inspected with the actinic reticle inspection from another competitor of ours. But it's still in the path finding. We believe that whenever the industry is going to need another wavelength that we call actinic reticle inspection for EUV masks is going to be mostly in the time that ASML is going to release the High-NA EUV. Why?
Because the geometry on the reticle is going to shrink in such a way that you need now an EUV reticle inspection to inspect EUV masks or EUV reticles. In that case, we will have a perfectly on time solution, and we believe we will be very well positioned to pretty much capture all this market. This is what we do in mask shop. In wafer fabs, this is a smaller market of reticle inspection. This is what we call reticle re-qualification. In general, we attack this market from two different fronts. We have a proper reticle inspection tool that is competing against Lasertec, but we also have the so-called print check application. This is very, very important because we are the only supplier that can inspect reticles after you print the reticle on the wafer with our most advanced wafer optical inspection.
People don't count this in the market share of reticle inspection per se, but in reality, this is incremental dollars we get on the optical wafer inspection to serve a reticle inspection market. I would say this is probably the one that is capturing more, most interest in terms of a competition landscape in the near future. I don't see any big changes in the other segments where we hold a very, very strong position and will continue to hold based on differentiated products in our portfolio.
Your reticle inspection share inside the fabs is what?
It's if you consider only the reticle inspection per se, excluding the wafer optical inspection that is still qualifying reticles on wafer, I would say it's 50/50 right now.
In the fab. If you look at what we'll call production, absent sort of R&D pathfinding and production, it's probably closer to 70% if you take production and mask shop. I think what Oreste is, you know, one of the things he's getting at is if you look at our shares, about 54.5%, as reported by Gartner, it's gone up a little bit over the last year or so. In the markets that we participate, it tends to be meaningfully higher than that. I think that what he was articulating there, and it really is how we go at most markets, is it definitely is much more of a portfolio approach. That what we're trying to do is meet our customers' technical requirements, but also their economic ones.
A lot of our competitors are point products, so they come to a competitive situation with a certain technology or a certain product, and then it's how does that product do everything in all the different use cases? What we're able to do, I think at KLA, is how they leverage the portfolio to address technical issues, but at the same time, try to leverage other capabilities to drive the economic motives. The biggest part about that is what makes sense in a production environment. There's R&D, what we call pathfinding, defect discovery, and then there's production. And so I think one of the things that we've been able to do is allow our customers to kind of mix and match and somewhat agnostically from our perspective, across our solutions to be able to to meet those requirements.
In process control, about nine markets that we're in, so broad markets, we're number one in seven of nine of those. Then the other two, we think are opportunities around e-beam, uses of e-beam, both from review, which is reviewing defects that are found by the optical inspectors. Inspection, which is used to do defect discovery early on, and then metrology, which is using it to measure overlay, and calibrate overlay targets, as you're doing, basically building a chip. In all those areas, we have offerings in the market that drive the relevancy of our optical solutions, that drives 80%+ of the dollars into the optical solutions, and increases the relevancy of that.
I think increasingly as a result, then also helps us from an overall process control intensity point of view. The thing I left out earlier about we talked about outgrowing WFE, and I talked a little bit about the market dynamics, but also the share dynamics are part of that. Our share of WFE was about 6.2% last year, and in Investor Day in 2021 Investor Day, we talked about driving it up to about 7.25 and above, you know, maybe as high as 7.5 Kind of percent, overall. That's gonna come from these assumptions on the overall market, but also some additional share gain opportunities we think exist as we move forward.
Makes sense. Maybe wanna frame, I guess, two areas of maybe debate. So the first one is on the actinic, or the move to EUV actinic. Basically, you know, you dominate the mask shops, and your view is that maybe having kind of the pathfinding first-mover advantage is not such an advantage, and that really the adoption's gonna happen with the move to High-NA, and the tool that you're working on internally today will be ready for that insertion, and you feel confident about securing leadership market share in that market.
For a couple of reasons. First of all, because we have been working the mask shop with a differentiated product for many, many years. We know how to do, for example, die-to-database. That is one of the approaches you use to find the defects on a reticle. We have done for many, many years. We don't need to do like a test run on this kind of algorithms and this kind of capability. We are very, very confident that whenever the market needs, we have the right tool because we have already had a long experience in all these algorithms, optics and sensors, many, many years in the mask shop. Also sometimes you gotta be at the right place at the right time.
Sometimes being too early in the market will create maybe some platform that you are compromising based on whatever you have available in terms of components now. Then when the market really needs, you are absent. What we are trying to do, we could have entered the actinic reticle inspection market five years ago, started development maybe seven, eight years ago. We didn't do it. We didn't do it because we knew that the market didn't need this tool in 2021 or 2022. Now we know because we talk, we work with imec, we work with our customers. We also work with and talk to ASML. We understand that the way how the pattern is going to shrink on the mask is going to need this actinic reticle inspection by the time the High-NA EUV is done.
We are tailoring our development tool to exactly what the market needs. That's been, frankly speaking, one of the differentiators of KLA, having the right solution at the right time, not too much earlier, not later, of course. I feel confident.
Makes sense.
Look, the pathfinding application we have, we have an E-beam solution that's in the marketplace today to help compete and address that. To Oreste's point, until you see the pitch shrink or the density of the pattern on the wafer, on the reticle shrink that you would expect with the High-NA EUV that creates more of an opportunity for actinic in production and even the use of pellicles, which is another complexity in the very complex subject. We think we have a solution in the market that addresses the needs today. Certainly in the EUV environment, we're getting, you know, a significant amount of the business, as Oreste said. We are investing in these next-generation solutions, and KLA tends to invest many years ahead. Customers always want optionality.
They'll always tell you what they. You know, "Yeah, sure, I'll try it. Yeah, do it." I think to Oreste's point earlier is our strategic planning process and how we drive the operating model to try to understand the requirements, which could be five to seven years down the road, is a very important part of what we do. You could argue the most expensive program in the company today, where we spend the most money, is in these areas. It'll take some time from a development point of view, but we think the market need is sort of tied into the introduction of High-NA and we'll be at the market at the right time.
I think that's investor FocusOne in the competitive landscape. I think number two is the optical e-beam debate. Obviously, you guys dominate optical, but at a certain point, you can't find the defects and you need to consider e-beam, but obviously very costly and very difficult to use in line. It sounds like, you know, your solution, your, you know, e-beam is more complementary with AMAT and that really the competition or theoretical competition is from ASML. I guess, do you agree with that assessment and maybe can you kind of expand on that?
Let me say that optical with respect, you know, it continues to grow and continues to surprise everybody, including myself. 23 years ago, I started working KLA. I was working in the fab before. I remember the guy who interviewed me at KLA asking me which division you want to work for. I said, "Of course, e-beam. e-beam is the future." It was 23 years ago, and I was totally wrong because look at what happened since then with the wafer optical inspection is growing like faster than any other segment in the process control. Because we found a way how to progress our technology I mean in optical wafer inspection, not only from optics sensor point of view, but also with AI. AI made the impossible possible in reality.
As you know now, detection algorithm, classification algorithm, all our tools are based on machine learning, neural network modeling. That means they will extend the wafer optical inspection forever. This is one. The second thing is, of course, we wanted to understand how to complement our wafer optical inspection with some E-beam solution. We see the E-beam solution more in the way how we guide the wafer optical inspection to see defects in such a way that whenever the customers need a production worthy solution, it cannot be an E-beam inspection. It gotta be a wafer optical inspection. We use our E-beam solution to guide the wafer optical inspection to get better and better. It's a complementary in our portfolio. We sell portfolio solution. We don't sell piece of equipment, and I believe that's the concept we have.
ASML and Applied, they have different concept of e-beam. Applied, for example, is more focused on really e-beam metrology more than inspection. And of course, e-beam review, ASML does a different type of application. We see the market in a different way.
Okay.
Yeah, it's complementary. I think what I would add is that if you look at the inspection part, and sometimes the third parties will gather data and they lump the metrology applications with the inspection. If you look at the inspection part, it's been very clear for a long time, the physics restrict the ability to use e-beam in a production environment, just given that it is so slow. Like, we're talking, you know, 1,000 tim es slower than optical solutions. You can go back a number of years, we'd have investors go, "Well, Bren, we've heard competition talk about hundreds of beams, and they're gonna speed it up with multi-beam capabilities." Now there's, you know, I think people are finding that, A, that those views are very different from reality.
I think it just sort of highlights some of the challenges with the physics that are involved here. Metrology is an interesting and growing application because it allows you to kind of set those parameters around actual metrology, particularly given the density of the chips. Review is an area that's scaling, and Applied Materials has, as a competitor, a strong position in the review part of the market. That scales with the optical inspectors because you have to review those defects. We're looking at ways to how do we leverage some of that e-beam capability for us to drive, as I said earlier, the relevancy of the optical tools to be able to do fast review, as an example, where we can review and classify defects very quickly.
That allows customers to get more value out of the optical investments that they're making. When you start to break it down about where the dollars are flowing, I think you come back to it overall and you're like, "Look, 80% of the dollars tend to flow into the optical solutions," 'cause those are what ultimately drives the production investments.
Very helpful.
We're doing these other things. I think we can use it as ways to drive, like I said earlier, the relevancy more of some of the product offerings we have, more than direct kind of point product, competitors in some of these more limited markets.
We got 5 minutes left.
Yeah.
I wanna hit on two major topics. Oreste Donzella, three minutes to you. EPC
Yeah.
your new baby.
Yeah.
You know, you've got a great opportunity, both packaging and automotive.
Yes.
Can you speak to that and, you know?
Sure.
Less interested in kind of what's it gonna grow next year, but more of, you know, what's the five-year path and, you know, how are you really driving.
Well, you mentioned the right end drivers for EPC market, not only EPC, but also for KLA overall. What we represent in EPC also, we do some work with a customer in particular in automotive with OEM and Tier 1, and we represent the entire KLA portfolio, including the software solution that we have now implemented to screen the chips in the fab for reliability issues. So this is a completely new solution that's been accepted by the Automotive Electronics Council that generates the specs for all the automotive electronics in the car. Now they accepted the KLA new machine learning-based software solution called I-PAT, implemented on our inspection tool to make sure that they do proper screening of reliability failures in the fab.
On the automotive also, going back to EPC, of course, a big player in automotive is SPTS. SPTS is the specialty semiconductor division process equipment that we built and shipped out of U.K. The market is booming for SPTS, in particular for this power trend, either silicon-based or silicon carbide-based. We have a pretty bright future ahead of us because the silicon carbide transition is going to be a secular shift. It's not something that is happening right now, and then in one year from now, nobody is ordering silicon carbide process equipment anymore. It's a secular shift. Packaging requires a lot of time. This is super exciting. What we are seeing now in packaging is something that I've seen maybe 30 years ago in a semiconductor front end.
There is so much innovation in advanced packaging right now because packaging has become an enabler of the semiconductor technology roadmap overall. The chiplet integration is increasing. Chiplet, 3D heterogeneous integration and 2.5D and 2.1D, all the integration of multiple different chips on a single piece of silicon, can be organic interposer, can be a silicon interposer and so on, is creating a lot of complexity. Whenever we see complexity, of course, we see KLA playing a big role in process control, process integration. Even in packaging, we see more and more opportunities for SPTS division. In plasma dicing, how you singulate wafers in a cleaner way, that is extremely important right now. We have a new solution, plasma dicing SPTS, in, and in many other areas of packaging.
Just to let you know, last year, first time ever, KLA was in the top 10 suppliers in packaging. If you consider both the wafer level packaging, assembly and test overall. First time ever, and this is also because of EPC inception two to three years ago.
That's great. Congrats. So Bren, last question to you, and I know capital returns are near and dear to your heart. You know, co-covering you for many years, I am not surprised by your aggressiveness. I think that your business model in terms of the stability of revenues, the best gross margins in the industry, best free cash flow margin in the industry, support you know, an aggressive behavior. I think for many new investors in the last you know, year or two, they were surprised by your most recent endeavor in terms of accelerated buyback during you know, arguably a very uncertain macro period. Wanting to give you the time here to speak to-
Oh, yeah.
You know, your philosophy and you know, what gives you the confidence to be so aggressive in any type of market?
You know, it's a great question, and it's been fundamental to KLA, certainly over the last decade or so, that our perspective that every dollar in capital has to be deployed productively, and that includes the capacity of the company. If you go back ten years or so, where we decided to put a permanent tier of debt in the capital structure, we thought that this is the way you should finance the business. It makes the most sense in terms of the messages it sends. A, the dynamics you mentioned around the profitability of the company, the balance sheet, and so on.
But also to investors that invest in these kinds of businesses, to our you know our supply chain, to our employees about just that this level of debt makes sense in terms of how we finance the business and doesn't encumber our ability to invest and to execute the strategy of the company. You go through that process and make that decision, then it comes, okay, how do you best allocate the capital? We do this on a regular basis, and we run all of our simulations and everything else with our boards. I don't know how many boards are looking at Monte Carlo simulations every year about what could happen. 'Cause when you make these decisions, they're in some ways irrevocable in a lot of ways, certainly around dividends and so on.
It was pretty apparent to us, given over the last couple of years of growth in our business and the story we were telling at Investor Day, that the best thing we could do with not only the go-forward free cash flow of the company, but also the incremental capacity available to us due to the business growth, was to invest back into the businesses that we're in to create our shareholder base and drive earnings per share growth. We talked about a plan of $14 billion in revenue in 2026 and $38 in earnings. Against that backdrop, given the way our business behaves, the things you mentioned, it's an asset-light business that if you can go back 25 years, I think we had 3- free cash flow quarters in the last 25 years.
In 2009, even the worst downturn ever, and you could argue the industry, very different, the company, very different, today than it was then, 1- quarter. Given the way the cash flow behaves in the company and our views of the right way to finance, we thought it was a strong message that reinforced the story we were telling and the future that we expect from the company. Yeah, obviously, then it comes down to how do you think about that, the vehicle distribution, and we have investors that value our dividends. We think a balanced approach makes sense. We've increased the dividend for the thirteenth year in a row, and we target to grow that along with free cash flow.
It can be modeled if you're explicit about how you value it. That the incremental sort of cash flow from there on the ongoing, but also these capacity assessments periodically will flow back through share repurchase. Strategically, we look at. We like the businesses we're in. Doesn't mean we're not gonna look at, you know, opportunities as they present themselves. We always look at things, but at the end of the day, we like our position and we think that, relative to the alternatives, and the alternative being buying back our stock and returning the cash, that that's the bar that we compare everything to.
That's a great ending. I think we've run out of time, but, Oreste Donzella, Bren Higgins. Thank you very much.
[crosstalk]
Thank you all.
Thank you, sir.