Cadence Design Systems, Inc. (CDNS) Earnings Call Transcript & Summary

Okay. I think we'll get started. Welcome, everybody, to the Cadence fireside chat. Very excited to have a long-standing participant to the London Nasdaq Conference with Cadence Design Systems. Today, we are very excited to have John Wall, who is the CFO, John served as the Senior Vice President and CFO since 2017. He previously was the controller and in other positions at Cadence for more than 20 years. So lots of in-depth knowledge about the company. And we also have Richard Gu, who's Head of Investor Relations. And I think, Richard, you need to read something, right?

Yes. Just quickly we remind everybody of the safe harbor, right? Today's discussion will contain forward-looking statements and will make use of certain non-GAAP financial measures. Please see our most recent 10-K, 10-Q and website for a discussion of risk factors in our use of non-GAAP financial measures. With that, Mark, let me turn it back to you.

Great. Great. John, Richard, welcome. Thanks for coming today. .

Thanks for having us.

So John, you guys are -- you've come to the conference every year, I think, and I think a lot of investors know you, but there's a few new faces here. So for those people just to level set, why don't you just kind of bring everybody up to speed, what should investors know about cadence? What do you guys do? How do you invest?

Sure, Mark -- and thanks for having us. We love coming to this conference. Yes, I guess, Cadence, it's a computational software company. As the company was created by engineers for engineers back in the '80s. And it's the tools that other engineers use to create their electronic systems. Like I say, largely computational software, we sell a little bit of IP, a little bit of hardware. But effectively, Cadence technology is used to create all the things that you can do without. But -- so all the stuff that -- all of our electronic products. It's -- you'd be hard-pressed to go into Currys [indiscernible] and find some electronic products that haven't been used or haven't been designed used in Cadence technology. The -- ultimately, it's 85% to 90% of the revenue is recurring in nature. A lot of concentration in the top 100 customers, top 40 customers at Cadence we generate 55% to 60% of our revenue from those top 40 names. Although we don't have any heavy concentration in any one name. We have no one that's anywhere near a 10% customer. But -- and then top 100 customers, there's a significant coverage there. We have kind of 3 models that we go to market with. There's a direct -- the direct sales approach, which is basically top 500 customers really. But outside of that, we have a customer channel partner, it's a bar program. And then we just recently launched our on-cloud e-commerce model. That's the -- where you can log online to cadence and purchase technology directly from us, and we already have 4,000 customers on that. We're trying to build out the long tail, particularly on the system side that -- and the PCB side, packaging and board side that there's a lot of customers that we can gain access to through the e-commerce model.

Great. All right. And I think your customer base has expanded from just semiconductor companies over time to semiconductor and systems companies. What has been the impact on your business model as you've seen that expansion?

Yes. It's -- so our strategy is we call it the intelligent system design strategy that -- our strength is in chip design and chip verification. That's the company's core, but we were always destined to be the largest core EDA player. And then the natural next step for anyone that's using our tools to chip design, chip verification is packaging and board. And then after packaging and board, they'll use system analysis tools. So they're all near adjacencies for us. And that ultimately, where the market is going, our customers really want 1 single platform that you can do everything from chip design right through to system analysis. And ideally, concurrently, so you can track everything at once. And we feel we're best positioned to provide that to our customers. So it's seen us go probably double revenue in the last 6 years, but -- 6, 7 years. And it's tremendously profitable. The systems is mainly simulation, and that's a core expertise at Cadence. We have a lot of simulation expertise from core EDA and from chasing Moore's Law for over 30 years. And what we found is that if you go back to 2016, all of our revenue in 2016, every dollar revenue cost us $0.74. But since then, every dollar of revenue growth has cost us about $0.46, $0.47, so it scales. EDA was always a great business. Just where do you grow? And growing into systems analysis space has been a godsend for us because it provides us a path to growth and as part of the intelligent system design strategy, we think -- we can take the TAM from $10 billion where it was a couple of years ago to probably something closer to $30 billion by 2025.

Got you. And what is the relative split between the semiconductor design and the system design versus -- and maybe you can tell us from when you started 0.

Yes. Well, I never knew when it was 0. When I started, it was 60-40, with 60% semis and 40% systems. And this was a number driven from our marketing department. And I was always confused by it because it was like "so where did you put Samsung here, it's like how much is systems? How much is semis for Samsung?" And so I was always skeptical about the number. But I'm told it's now 55, 45. I think it's inevitable that it gets to 50-50 at least with -- because we're growing faster in systems and mainly selling more to systems companies as we execute against our intelligent system design strategy. What I can tell you, though, is that the marketing people promise me faithfully that they use the same methodology today in calculating the 55-45 that they use when they were calculating 60-40 in the past. But I still don't know. They have a methodology for allocating revenue from customers like Samsung between systems and [ services ].

Got you. Got you. Okay. And I think 20 years ago, your margins were in -- it may have been single digits and they've expanded dramatically. What's -- what have been the biggest drivers of that? And should investors expect that expansion to continue?

Yes, I don't think there's any near-term ceiling on operating leverage since 2017. So 5 years running, we've delivered more than 50% incremental margin. I think it's averaged about 53%. This year, we're on track again to deliver more than 50% incremental margin. So that's 6 years running, 2 years of that. It's been a pandemic. There's a war. There's a lot going on. I mean, when I sit down with the R&D engineers and they're looking for more budget, it's -- nothing's changed. We're still trying to drive 50% plus incremental margin. Like you say, the business scales really, really well. I can't see anything that we're going to encounter in the next few years that we haven't really encountered in the [ last 6. ] We've had customers added to entity list. There's been a whole bunch of things that happened. You still churning out over 50% incremental margin. And naturally, if you continue to do that, it's going to drag up the overall average. Now it will slow down by the time you get in the high 40s and close to 50. But I would also ask you to bear in mind that we're achieving that while we're building out a multiphysics platform and investing heavily to build out our product portfolio for the system analysis space. So despite that heavy investment in R&D, we're still achieving more than 50%. So I don't see any near-term [ seeding ].

Got you. And if you look at the industry over 20 years, there's been a lot of consolidation, and there's kind of 2 high-quality companies that people think about in this space, you and Synopsys. How -- when investors ask you to describe the difference, like what do you do? What's the difference between those 2 companies?

So I can give you the brief history lesson, the one I got on trouble with. But so I've been at Cadence a long time, and Mark kindly said 20 years. I think it's -- of all the CFOs I've met everywhere else. I think I'm the slowest [ one ] to CFO ever, taking 20 years to get into the hot seat. But the history I have at Cadence, I remember in the early days, so Cadence was the analog EDA company. And when Synopsys was formed, they were like a digital version of cadence. And then it became clear about a decade in that's -- there were a lot of EDA companies to become a complete core EDA company, you had to do both analog and digital. So we were both encroaching in each other space. The analog to digital. It's a bit like orchestra to rock band. It's easier for orchestra players to play in rock band than it is for rock band players to come the other way. So we always felt it was our destiny to be the largest core EDA player because we figured that we had 80%, 85% of the market in analog. It was only a matter of time before we got to 50-50 on digital. And I think we're there, right, in terms of -- on digital, we were probably already 50-50, excluding significant customer we just announced in Q1 that we achieved a significant contract, a record contract with a [ Marquee], U.S. semiconductor company in Q1. So I think we're on track to achieve that 50-50 and deliver on that destiny of being the largest core EDA company that when I first talked to [indiscernible] about because -- when I moved to the U.S. in 2015, I went to this Cadence sales kickoff and [ Lipu ] had given ceremoniously handed a blank check to our head of our IP business group. And as a CFO, you're going to go no, no blank cheques, but I talked to them afterwards about the strategy and Synopsys, a 10-year head start and building out an IP portfolio, and we talked about whether IP was the right thing for Cadence, whether we should invest so heavily in IP. And what we talked about was basically -- I've given the analogy I told him that Cadence and Synopsis were like those identical twins. You only went school but wore the same clothes. And it's kind of acute when you're an elementary school or middle school, but we were in our 30s now, and it's a bit creepy if you wear the same clothes as you're identical twin, you need to kind of form your own identity. But -- and we felt that our identity was -- had to be based around if our destiny was to be the largest core EDA company, that, that should be the center of gravity in our strategy. And then with [ Anoro ] included in the conversation, we thought the natural thing for us to do, we had a -- for it to be the largest core EDA company, for chip design, chip verification. We're already in a packaging and board business. The natural area for us to go into with the simulation expertise that we had was system analysis. And it's pretty tight. I mean if you're a synopsis, and your destiny then is you're going to be a smaller player in core EDA, then it's perfectly natural for you to build out an IP business and software integrity. And I don't think there's a need for us to follow them and be an identical who wore the same clothes now that we're approaching 40.

Got you. All right. That's very helpful. So the industry is consolidated. You have your kind of capability, your competitor has, it's -- what have you been seeing more recently in the semiconductor industry. There's been lots of supply constraints. How has this impacted you at all?

Yes. I guess, I mean the -- I guess we were looking from the perspective of a few years ago, we had a large customer added to an entity list in China. And back then, we started an exercise, we formed a global strategic sourcing group, and we wanted to understand our exposure to individual suppliers given the increase in geopolitical tensions that we particularly wanted to make sure that we could identify either a safe second source supplier locally or as close to home as possible that -- or if we didn't have that, that we would bulk up on inventory so that we were less exposed. And that's the approach we took. And I think a lot of our customers are taking a similar approach, which over the -- what people tell me, people that have been around for the longest time in the industry, they tell me that the way those supply chain problems get solved is very, very slowly for a long time and then suddenly because everybody solves it by bulking up an inventory and over ordering, double ordering, triple ordering. And then you reach a point where they have enough where they stop ordering and everyone has an inventory problem. But I don't know where we're at now, but I think we're closer to the -- I think we're closer to the point where people have enough inventory now, and there will be some shock at some point where that the ordering will slow down for inventory. But with our customers, mainly, I mean, it's software. We're just providing licenses to software. There's no -- it doesn't impact the design side of the business because we sell to the designers really.

And there a lot of times what you hear at this part of the cycle where things are tight and companies can't get certain products that there's a -- they try to redesign around them. And does that impact your business? Does that mean that -- does that translate to more business for you?

Typically, we're mission-critical technology, and there isn't -- there are many alternatives that -- but you don't have supply constraints with us either. I mean on the biggest constraints we might have would be on hardware, we need certain components for hardware, and it takes time to build the hardware. We have got ahead of that by having a lot of inventory. At the end of the year, we had like $116 million in inventory, and we burned through $55 million of that in Q1, but we managed to get access to more inventory. I mean, we closed Q1 with $110 million in inventory. So it's probably a couple of quarters that we have in inventory. But I don't think we're a source of any supply constraints for our customers.

Right. And I guess I was thinking about from the design side, does it cause higher license sales for you because companies are trying to design around the shortages that they have coming up with a different solution and then redesigning a product in order to fit products that they have.

Actually, I'm not really sure. Would you...

Yes. I think overall, if you look at -- our business is quite sticky, right? The -- it tends to be like when customers try to design more like customize chips and domain specific chips, which is more powerful, they have better control. Our tools actually were coming handy in that front. So we haven't seen a whole lot of what you just described. But I think the business has weathered all kinds of ebbs and flows in the past quite a while, and we don't think it's -- we think it's going to continue this time around, too.

I think what we are seeing, Mark, is that like if you look at like major macro trends, I mean there was that globalization trend that happened. If there's any reversal of that towards deglobalization or because of supply chain issues, people wanting to do things closer to home, well, that's going to cost more in engineering talent. I mean there's inflation when you want to hire -- like if you have a group of American engineers and you want to increase that by 2% or 3%, there is inflation in that because they're also much more expensive now. But if you wanted to deglobalize and replace some engineers that were previously in India, with engineering in the U.S., it's significant inflation. And you could bifurcate our customers spend on R&D into their spend on employees, engineers and their spend on the tools that go into the tool bag that they give those engineers. And if I go back like a decade you probably had -- 92% of the spend was probably on people and 8% on the tools and the tool bag. It's probably got to a point now where it's still only like 88-12. There's probably only 12% of R&D budgets being spent on the tools. But I do think we see the types of tools that we're launching now, there's a lot of AI and machine learning. Cerebrus allows 1 digital designer instead of operating 1 Innovus license can operate 10 Innovus licenses from a cockpit, a Cerebrus cockpit. That's tremendous saving for a company. And if you're going to deglobalize, you probably need to control your -- in terms of -- we feel that spend with us is deflationary in nature. But the more you spend with us, the more you save and that we do think that our customers will be able to control their R&D growth by leveraging more on EDA tools and less on people. So although that trend, it's taken a long time to get from 8% to maybe 12%. I think that gets to 50% over time. It could take decades, but 50-50 makes sense in that -- just the growth in engineering headcount, the growth in population can be -- can keep pace with the demands of complexity and the growth in what we all require in our technology.

And more recently, there's concerns around inflation and higher interest rates. How does that impact your business, either from a cost side or from a customer operational side.

Well, naturally, I mean, inflation impacts all of us that we are able -- we have -- our tools are mission critical. So we have the privilege we've been able to pass a lot of that on. But -- and we've been working with our partners that -- for many, many years to improve pricing and improve our share of wallet for the value that we provide. But generally, we're very data-driven and value driven. So it immediately impacts us in our return on investment calculations in terms of what we want. We measure kind of net present values on a lot of things in terms of -- it's all net present value of future cash flows and all our investment decisions. And of course, higher interest rates impact the -- impact that. So we probably are exposed in that long retail of customers. They'll be weaker credit customers, but the big customers, the big names have no issues with paying us. But I think the first place we'll see an impact will be maybe poorer credit quality. We saw that in a couple of years ago that I had to make a $70 million reserve in bookings. So we reversed our bookings over $70 million for companies that got into trouble during the pandemic. Now what we did there was we committed to those customers that you will not grow a business for lack of tools. We will help you -- we'll provide you the tools, we'll help you through it. And that -- and to the extent that a large number of those kind of managed to work their way through it, but we created enormous goodwill with those customers. And we figured it was the right thing to do anyway just certainly joined a pandemic that -- but it benefited us not even from those customers that didn't survive that didn't survive the pandemic and ended up closing the doors on their startup, but they've ended up at other places. And we work with them again just under a different company name now.

So your top line is growing in the mid-teens. You have your multiples margins have expanded. What do you worry about?

Execution against the strategy really. I think the strategy is excellent. It's the natural place for us to go, but people can't get distracted. I mean, at work, thankfully, for the last few years, when people tell me about the stock price, they tell them stop looking at the stock price and get back to doing what you're doing, and it's the same today. You can't be distracted with that. The way we see it is that we focus on the EE, it's all your job to figure out the PE , what's the PE, but we can only control the earnings piece. So we focus on trying to drive earnings. And earnings growth the last 5 years has averaged about 23%, I think. Our latest guidance for this year is closed, it's just under 20%. And -- so I think if we continue to focus on the EE, focus on delivering that incremental margin, like you say, we're up to low teen growth now for revenue growth, certainly, when you look at the 3-year CAGR because most of our customers do 3-year baseline contracts with us. And they're ratable contracts, where we've discovered that we're better farmers than Hunter. So we're always working on trying to grow that annuity. What we found in previous times of stress with customers. You have some customers that need some help in the short term. And we're prepared to restructure some contracts to give them that help in the short term in return for a higher exit rate and larger proliferation into the account. So we can take advantage of a crisis as well and be helpful to our customers at the same time.

Got you. So you had gone through a period of consolidation, and you guys were consolidators in the industry. But you haven't done any major transformative deals recently at least. What would change in the current environment potentially with given the multiple compressions that you've seen in the industry?

Good points. So for me, I always have a preference. And again, I think the company has a preference for R&D spend. Like I said, in our DNA, the company was created by engineers for engineers. So we think organic is delicious, right? So we always want to invest in our own R&D. Outside of that, when we do anything inorganically on the M&A side, prefer -- which prefer the small tuck-in M&A where it fits and furthers our strategy where we get access to top talent or technology that we didn't have or where we can accelerate our road map. In the past few years, we've fallen down on the price there, so we can't agree a price because everything for us is based on net present value of future cash flows and some of the prices, the multiples were ridiculous. But even now, where I think it's much more same an environment. I think in terms of -- we're getting closer, but I think some of the things that we're interested in. Those people are still a little bit anchored on their value from last year. But what they thought their value was last year. That's -- so maybe it gives us opportunity over the next year or 2 if people are willing to accept more reasonable prices. But other than that, we're happy to make versus buy.

Got you. One of the biggest drivers of growth in the technology industry is machine learning, neural networks. You have a product, Cerebrus. What is -- how does that address that market?

Actually, Richard, do you want to take that one because you have done study in that?

Sure. So Cerebras, it is the machine learning offering we had out there which really helps engineers to really just automate the entire digital foot flow, right, from the RTL synthesis to really sort of like the sign-off process, right, end to end. And the benefit of Cerebras clearly is twofold. One is it helps really shorten the time to market, right? Because when you think about this business, the biggest enemy is time, right? And in the past, this full flow process was going through a lot of permutations and given the certain constraints and parameters. So engineers tend to do one iteration, figure the problem and then and reiterate time and time again. So with machine learning, it dramatically improves that, right, because machine can do things the human just cannot do. And also the second benefit is it improves the result, right? And the quality of the result is which matters a lot. So I think -- if you look at the digital full flow, really using service will help the customers dramatically improve the PPA, which is a huge benefit like power performance and area and also it shortens the time to market. Now as a company, we're applying the Cerebras similar kind of capabilities [ as tosses ] design on analysis side. So it's our core strategy is kind of infusing intelligence and data into the entire product portfolio, including both the chip design side as well as the system analysis side, which helps to continue to advance our strategy and stay true to our vision. So there's been consolidation in your industry, there's been consolidation in your customers as well, both in semiconductors and systems. What is the risk of the concentration of your customer base and being dependent on a smaller number of companies.

Yes. So the great question. With Cadence, I mean, top 40 customers, we generate 55% to 60% of our revenue from those names. Top 100, there's a significant portion of revenue coming from that top 100. And there -- they're high maintenance customers. I mean, they're your -- they get concierge service, white glove treatment, but one of the challenges we have is scaling out to a broad tail. But on the system analysis space, like you're competing with great companies like ANSYS that -- ANSYS, if you're listening to some of their calls, they're boasted about having 80,000 logos. But if we try to scale what we do with those top 100 or top 500 cadence to 80,000, we'll kill ourselves trying to do that. So that's why we launched our on-cloud e-commerce platform that what we wanted there was we wanted to make sure -- there were 2 things, we wanted to have like an over to-the-top model that was able to address the needs of a broader customer base without changing the experience of those VIPs and concierge treatment customers. But -- and the second thing was we call it the MC Hammer rule that we can't touch this, right? But we thought we'll kill ourselves on the G&A side if we had to review or touch any one of those individual transactions. So the IT group has set it up in a way that when we sell on cloud that it comes just seamlessly flows through our systems. But in terms of concentration, we don't have any significant concentration in any one customer, but you do have that concentration in the top 100. I'm not sure we did see a bunch of consolidation in our customer base maybe 5 years ago that's -- but even then that's -- we were paranoid about it at the start because we thought we're going to lose some business. There was one example of an American company that when they acquired the -- another customer of ours, they decided to let go of the entire analog design team in Sweden. And we thought, okay, there's about 18 months left to go on that customer's contract. I take it out of the forecast that we're not going to get that on the renewal that -- and what happened was another customer of ours picked up the engineers in Sweden had to buy licenses for them. But -- and then when the renewal came around, the customer that had acquired them, that they renewed all of those licenses again because they were doing that work somewhere else. So it's quite fluid. And what we found there was that we actually had an overlap of revenue for about 18 months, where we were getting paid for licenses that weren't be used, and then we were getting paid a second time. But -- so it's -- I think in terms of consolidation, we don't have significant consolidation, and we've tried to make it work to our advantage because you have so much revenue coming from those top 100 contracts, we have individual strategies for each one of those contracts. We call it the 4 Ps that there's 3 inputs that are: People; products; and profitability; and the output as a plan. And the whole idea is how can you -- can you give me a plan for each of those accounts to improve profitability, double profitability in 4 years? And then we work on what we can do with each of those customers. In some cases, it's really, really difficult or impossible. And then we cut our cloth accordingly in terms of investment on the account. In other cases, there's a huge opportunity, and we make sure we have the right people they have the right access to products and et cetera, at the profitable level for us.

You guys do -- I know you spend a lot of energy talking to investors. We got about 30 seconds left. So what do you think is the biggest misperception about Cadence in the investment community?

I think Cadence is a compounder of value. I mean, over time, it's very hard. It's like habits. Habits are a great compounder of value. Good habits getting up early every morning. You might not see the benefit in any particular day or any particular week. But if you do it consistently over time, you'll benefit from that considerably that it has that compounding effect. But Cadence is a great compounder. I think when we talk to a lot of our investors, if you if you want to put money somewhere and leave it there for a decade, Cadence is a great place to put it. But if you can time the market, you're probably better off with another semiconductor company, if you -- I mean, because if you really can time the cycle, you'll benefit more from them than us because we'll be a patient grower over time, I would say.

Excellent. I think that will have to be the last word. John, Richard, thank you very much for joining us.

Thanks so much, man.

Thank you. Bye-bye. Great presentation.

Thanks a million. Thank you.