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

Good morning, everyone. My name is Elizabeth Elliott, and I'm an analyst on the Morgan Stanley Software Equity Research team. Please join me in welcoming Cadence's President, Anirudh Devgan; and CFO, John Wall. Anirudh joined Cadence in 2012 and has served as President since 2018, overseeing strategy, R&D, sales and marketing in many other functions. John became CFO in 2017, has held a broad range of leadership positions over his last 23 years of the company. Thank you so much for joining us, and we're looking forward to your insights. [Operator Instructions] For important disclosures, please see Morgan Stanley disclosure website at morganstanley.com/research disclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. And with that, we'll go ahead and get started.

So first, starting with John. How about we just start with a brief overview of Cadence for those who are not as familiar with the story.

Yes. No worries. Thanks, Elizabeth, and thanks for having us at the NASDAQ, again, this year. But before I begin, I should read the safe harbor statement. 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 and our use of non-GAAP financial measures. Yes. Just in terms of a brief overview of cadence, at cadence, we provide EDA and IP products and solutions that allow our customers to design complex chips and electronic systems. We're predominantly a software business with a revenue model that is highly recurring in nature. Around 85% to 90% of our revenues is recurring revenue, every quarter. We are a company that was created by engineers for engineers by engineers. We're a technical essentially vertical software company, largely in the EDA space. We provide software tools, intellectual property, hardware verification platforms and services to semi and electronic systems companies. We're executing to our -- what we call our Intelligent System Design strategy, and I'm sure Anirudh will expand on that a little bit later. Our customers operate across multiple verticals, including mobile consumer, cloud data center, mil/aero, automotive and upcoming industrial and medical. We have a very strong culture of innovation. We spend close to 40% of our revenue on R&D every year. Over the past 3 years, we have developed and introduced more than 20 new significant products. And we have a relentless focus on ensuring customer success and partner very closely with our customers at the very early stages of their R&D cycle to understand their needs, and that in turn then drives our road map. In our most recent earnings call, at the midpoint, we guided to $2.653 billion of revenue and approximately 34.5% non-GAAP op margin at the midpoint for the current fiscal year.

Awesome. Great. So thank you very much for the kind of that great overview. This year has certainly been more challenging than expected. However, Cadence's business has been really resilient with Q3 earnings, actually exceeding your outlook on all the metrics. So kind of what were some of the headwinds and also some of the tailwinds that the company incurred during COVID? And how have you guys adjusted to doing business in a pandemic environment?

Yes. Sure, Elizabeth. Great question. We continue to be impressed by our team's ability to adapt and operate effectively in this environment. So impressed in fact that the company is prepared to work from home indefinitely, if necessary. Our R&D and customer deliverables are tracking really well. Our sales and application engineering teams continue to engage effectively with our customers. And as you can see from those Q3 results in our revised 2020 outlook, Cadence adapt really quickly to the challenges of doing business during the pandemic. That's probably a good place to drill in to understand the headwinds and tailwinds for the year. On the sales and revenue side, we haven't seen any slowdown in design activity, thankfully, with many of our customers continue to invest heavily in R&D. The -- we had the benefit of around $40 million of unusual strength in our China revenue this year. We increased our revenue guide for the year by $53 million to that $2.653 billion in our most recent earnings call in October. And we're a very, very profitable company with around $0.50 of every dollar of revenue growth since 2016, flowing through to operating income. As a result, our operating margin has been continuously increasing over time. We're currently guiding to a range of around 34% to 35% non-GAAP operating margin this year, but that includes the benefit of that unusual strength in our China revenue for 2020. Excluding that China strength, I'd say profitability is up to a range of about 33% to 34% right now. The pandemic is actually beneficial to margins in the short term on a net basis. A prolonged pandemic would have us at the higher end of that range. That's because generally, our margins are better during the pandemic because we have less T&E expense. The cost of -- the additional cost of operating in a pandemic environment are more than offset by lower overall facility costs with so many of our facilities being empty and people working from home. Also, one of the key things benefiting margins in the second half of this year is slower hiring. We can't onboard people fast enough right now, and that's helping margins in the short term. It's probably not ideal in the long term. I know that's a good problem to have. But if you take an owner's perspective, ideally, you'd prefer to get people on board and fully productive faster than we're currently able to achieve in this pandemic. On the sales and revenue side, we haven't seen any slowdown in design activity, thankfully, with many of our customers continuing to invest heavily in R&D. The -- so in terms of headwinds and tailwinds, net-net, the pandemic has been good for Cadence, although I think we'd all like to see an end to it sooner rather than later.

Yes, definitely. So just kind of digging in on something you would just mention there. But on China, you mentioned you had a particularly strong performance in China. I know it had a pretty significant impact to the latest reported quarter. With revenue from the region up about 96% year-over-year. So how should investors think about kind of lapping that tough comp? And then what does a more normalized growth trend and growth opportunity in the region look like going forward?

Yes, another great question. The Asia Pacific is a strong growth region for us. And we've done very well in China this year. I mean we always tell people not to focus too much on any 1 quarter or even 1 year for that matter. I mean in our CFO commentary, we've provided, that's available on our website, we've provided 3 year CAGRs. I mean, typically, our customers on a 2- or 3-year renewal cycle. We think it's kind of good to look over the longer term. Even with the recurring revenue model, 85% to 90% of our revenue is ratable in nature or recurring in nature that's what upfront revenue can spike the quarters from time to time. But -- and a pandemic environment, you get some unusual things happening to. But I mean, in terms of China itself, China is investing heavily in the semi industry, and there are many exciting startups and opportunities in China. And 2020 has been an unusual year for many reasons. I mean, during Q3, we saw higher levels of hardware and IP sales activity in China. And we assume that our guidance that, that would continue into the middle of Q4. So in conjunction with the extra week we have in this fiscal year, which falls into Q4 for us, I think we'll be lapping some pretty tough comps in the second half of next year. In our most recent earnings call, when we increased the revenue outlook for 2020 to $2.653 billion. That included, I think I said earlier that included about $40 million of strength that we saw in China and that was pretty unusual strength. To put it in context, China accounted for approximately 10% of our revenue in 2018. And again, 10% of our revenue in 2019. That jumped up to 13% of our revenue at the start of 2020. And for me, I thought that was a reasonable level. Second half levels seem to be much higher than that, and it could include some demand pull forward. So we're watching that carefully. It's very, very difficult to tell right now if that's demand pull forward or if it's sustainable, but I'll have a clearer picture of that by the time we're reporting results in our next earnings call. We had some demand pull forward in China before. That was back at the end of 2018. It was much easier to identify back then as a pull forward because we'd send people home for the holidays, and we had to call some folks back in to complete orders that customers were insisting we complete before the end of the calendar year at that time. Ultimately, what's important, though, is our strategy. I mean if we continue to execute against our strategy over the next few years, any pull forward China demand from '21 into 2020 should hopefully turn out to be noise in the greater scheme of things. Yes, I think that's probably the best way, the best context to look at China in the last quarter.

Got it. Yes, thank you very much for the context. Anirudh, something that John had mentioned, you were one of the key architects of the Intelligent System Design strategy. Kind of what drove the decision to expand beyond EDA. And can you provide us an overview of what the strategy entails?

Yes, sure. Hi, everyone. So I think what drives our strategy, what drives our growth is a combination of what customers want and what we are good at. I think that's the basis of any good strategy and expansion. So the first thing is in terms of what we are really good at, I think it's really important to know what is best in the world at. And what we believe -- what I believe what we are best in the world at is what we call computational software. So most of the software we write is numerical software. It's like mathematics plus computer science. So things like matrix solvers, optimization engines, bullion algebra, if you remember your undergrad years, okay? So that's what we do. We have more than 5,000 people in R&D, and most of their background is this kind of computation on numerical software. So that's the foundation of our expertise. And in that area, we are really, really good. Now 1 big application of that is, of course, EDA and IP, and that's our core business. And this is over the last 20, 30 years. And especially in the last 5 to 10 years, Cadence has really become the premier supplier in EDA. And I can tell you more about our core business. But the computational software aspects are there in other adjacent markets, okay. So that brings into what our customer wants. So what our customer wants is -- apart from our core business, which we call design excellence, EDA and IP, the next obvious thing they want is a system-level design. Because we are designing chips and these chips are silicon goes into some systems, so the customer want us to go into the system design, okay? So if you look at system design there, it's a big market. Maybe the software part of it is about roughly $50 billion per year. And if you take that and intersect that with our expertise, which is computational software, the area that naturally falls out is system simulation or system analysis because analysis by its nature is also computational. So -- and that's about $6 billion, $7 billion market. And so that's 1 natural expansion area for us, and it's very synergistic. And that's our customer wants, okay? So that's a key part of our strategy's expansion into system simulation. Now the second area, which the customer wants is -- so you have silicon, then you have system, and then obviously have software on top of the system. And in software, the key part is data analytics and AI. So there are -- again, a lot of areas of software. But the computational part of software is, again, AI and data analytics. And that's what we call pervasive intelligence. So if you look at these 3 concentric circles, so there's design excellence, system innovation and pervasive intelligence, and what we are focused on is the computational part of it, and that's intelligent system design. And that we believe can expand our TAM from roughly $10 billion to $30 billion over the next several years. And we are very excited about it, and we have a lot of early success as we move forward.

Great. And within the kind of the systems analysis space. You recently launched Clarity 3D and Celsius solvers. Can you just talk about some of the reception that you received from this product line, and how it compares to the competition?

Yes. We are very excited about our Clarity and Celsius, which is system analysis areas. The main thing here to look at is, so these do electromagnetics and thermal simulations, which is again critical for our customers. And they use a lot of what is called finite element methods to do, especially clarity, to do system simulation. And the key thing is in all these simulation methods, it comes down to solving like a big matrix. So in the end, that's what it boils down to. Because the physics is pretty well established. If you look at finite element, the physics is from 1940s. And a lot of the commercial products are from 1990s that people use. So the physics is pretty -- and we don't make any approximations in the physics. What we do is we just solve it in a much more efficient way, in a more modern distributed, pedal computing way. And the reason we can do that is when we go from the silicon level to the system level, the silicon level has a lot more number of elements. So when we solve things at the silicon level because they're at 5 nanometer or 7 nanometer, our algorithms can naturally do like billions and billions of nodes, okay? And we take those algorithms and go to the system level. If you want to simulate something like a phone, the size of the problem is in the millions versus the billions. So we have a lot of computational expertise that we can take from the silicon level to the system level. So if you look at Clarity, for example, and it is -- can be order of magnitude faster than existing method and can do much, much larger systems because of our expertise from the chip level, and the same algorithms can be applied at the system level. So therefore, the customer response has been fabulous. If simulation can run 10x faster, what usually happens is people simulate a lot more, and then they do a lot bigger problems. And therefore, we have a lot of customers. And the customers are naturally overlap because all these companies that designing chips are also designing systems. So they want to simulate the things together. But essentially, we are much faster and we can do much bigger problems while maintaining the same level of accuracy because we don't make any approximation.

Got it. And then Cadence has looked through many economic cycles. And just wondering if you could just add some color on how defensible R&D spend is at your customers during these periods of disruption? And how we should think about the budget for R&D going forward for your customers?

Yes, sure. I mean, R&D is critical to our customer success, and that contributes to the secular trends that we tend to see in the demand for our products and solutions. So far, even in the current environment, we've not seen any slowdown in design activity. I do believe this period could be an opportunity, especially for market-shaping customers to further invest in R&D and accelerate their innovation. Our business is predominantly tied to semiconductor R&D. And in addition, our broadly diversified customer base, about $3.8 billion of backlog and a highly ratable business model. Also, have to highlight the resilience of the business, particularly in these challenging times.

And then, John, can you -- you've previously talked about the opportunity to improve pricing. Kind of what is your strategy there? And kind of what inning would you say we are in today on your road map there?

Yes. I still think it's early innings. I mean, effectively, EDA is a competitive business and pricing varies from sector to sector and product to product. We try to be very disciplined and value-driven in our pricing. Typically, we believe the best way to drive value from our products is to collaborate deeply with customers and deliver innovative, clearly differentiated solutions. But we're very data-driven at Cadence, and we're very thoughtful in terms of how we price our products. We measure contracts on a -- for deal quality or contract quality against average selling prices, and we've tightened up contract structures to simplify contract discussions with our customers as well in the past few years. If a customer has been asked for a large significant increase in their annual fee to Cadence. It's very likely that the customers previously enjoying pricing that was much lower than the average for the products provided in their configuration. So I think the application of a data-driven process to pricing provides a more kind of fair pricing structure in our -- among our competitors -- or among our customers. And I think it's important. I think the competitive environment is such that it's not just us that are very disciplined and value driven, but I believe our competitors are very similar in that respect. And I think it bodes well for the industry to see that logic being applied to pricing.

Got you. And then we've actually had a question coming through the portal just on that kind of competitive dynamic. So can you guys spend a few minutes on the competitive environment and any changes you've seen, and kind of where you find yourself with some opportunities?

Sure. But yes, I think what Anirudh spoke to earlier, in terms of their overarching secular trends and generation drivers for growth that we're excited about in the near-term that's -- the EDA and semi industry are benefiting from 5 technology waves that we see as generational trends. It's kind of machine learning and AI, 5G, hyperscale computing, autonomous driving and industrial IoT. And we're seeing opportunities across all of those 5 technology waves for ourselves and our competitors in that. And there are other strong drivers, I mean, such as domain-specific computing, system companies building, custom silicon. There's lots of new silicon start-ups. And the digital transformation of industries such as automotive and industrial, aerospace and medical. I think all those trends have silicon at their foundation, and they're driving strong design activity. I don't know, Anirudh, would you add anything?

Yes, John, sure. So absolutely. One good thing is that we are in a good market because silicon -- this renaissance or silicon, like John talked about. Now in terms of our competition, I think what we believe has happened and will continue to happen is we have become the premier EDA company. Because Cadence historically used to be good in analog or some parts of EDA. But if you look at last few years, we have become good in -- we've continue to stay strong in analog, and we have become the premier supplier in digital and verification. So overall, we believe we have the best and the most complete portfolio. So at this point in any major company, we cover analog/digital verification. So -- and if you look at our market share in EDA, it has significantly increased in the last 5 years, okay? And I think that trend is likely to continue as the premier EDA company. So that's in our core business. And in these new areas, there are some other competitors, but we are gaining a lot of share because of our expertise from EDA. So both from a market standpoint and from a competition standpoint, we are reasonably confident at this.

Great. And then Anirudh, this is probably more for you. One of the things that the whole semiconductor industry has benefited is Moore's law. However, that seems to be slowing down. Now is that a positive or a negative for Cadence?

Well, the great thing is that Moore's Law -- I used to -- I was in IBM, and I used to work with Bob Dennard, he's the guy who came up with DRAM, okay the other thing he came up with is called Bernat scaling -- or Dennard scaling, which is the classical semiconductor scaling method, okay? And Moore's Law is the marketing term for it. But the technical term is this classical scaling, which used to happen from 90 to 65 to 45, okay? And that classical scaling, I think, has stopped already about 5 years ago, okay? This is what I believe. And we used to argue about it earlier. When it would stop. And if you look at last 5 years, most of the improvement in performance is not driven by the transistors getting much, much faster, which is the classical Dennard scaling, it's driven by integration. So you put more things on the chip. Even if we go from 7 nanometer to 5 nanometer, the transistor may not get that much faster, but we put more things on the chip, okay? So instead of 1 CPU, you have 8 CPUs. And so, you put like a neural engine in the chip, you put like more display and things like that. So that is what has happened in Moore's law anyway already in the last few years, okay? And 1 trend which I think is significant going forward for the next 5 years is that integration is moving to the package level. So if you -- what we see now is this 3D-IC or interposer, which is multiple chips on a package is driving the next level of integration. So that's true that the transistor scaling on a particular technology may slow down because we are at 5, go to 3, go to 2 nanometer. There's only so many nanometers left, okay? But the number of -- there will be multiple chips, either in a 2.5D or at 3D package, which will drive this integration going forward. And the best thing there, okay -- so I believe that -- so "Moore's Law will continue because of this integration trend". And the best thing from a Cadence standpoint is that we -- apart from the things I mentioned earlier, and John mentioned, analog/digital verification. We are the premier provider of packaging technology through Allegro. So -- and there is a natural integration happening from the chip level to the package level, and that also positions Cadence well as we go forward. So long story short, I think Moore's law, anyway, last 5 years is driven by integration, and this can continue by 3D packaging.

Yes. Elizabeth, I'd just add there that I think that slowing Moore's Law is probably a very good thing for margins for a company like Cadence. I mean we record R&D expense upfront at the time we perform the R&D. So the R&D costs associated with most of the products and solutions that we're selling today has already been sunk in the past. But as a result, if customers remain on a specific process node for longer, that gives us more time to harvest what we've created in the past. So -- which, in turn, that should drive improved profitability, and I think we've seen that at Cadence over the last few years.

Great. And we've had a customer -- an investor question kind of related to that, dynamic. But you -- the question was, how should we think about the payback on R&D and as growth matures kind of what is your normalized R&D spend as a percent of revenue?

Yes. Normalized R&D expand as a percent of revenue is in and around the 40% level. Generally, the model that we operate is that -- from a capital structure standpoint, is we have a largely recurring revenue model, 85% to 90% of the revenue is recurring in nature. It's predominantly software. The 10% to 15% that's not -- nonrecurring would be mainly in relation to hardware in our system verification business and some of our IP. And then the way the model works is, we throw like 40% of revenue back into R&D because you've got to continue to innovate and develop products, and it's all really, really complex problems as we kind of Chase Moore's law. But -- and then the business throws off a huge amount of cash. But -- and then we use 50% of that free cash flow to repurchase some shares and 50% of the cash flow to reinvest in the business and to fund M&A. And then one of our key focuses then is not just on revenue growth for the sake of revenue growth, but we're very focused on scalable and profitable revenue growth. We pay a lot of attention to driving effectiveness and operational efficiencies. We're very data-driven, and we look at multiple metrics to make ROI-based decisions on pricing, resource allocation and investments. And that in turn then drives higher margins.

Great. Yes. And you guys have done the 50% incremental margins for 3 years now, and now going on 4 years, which is very, very impressive. And I'm going back to -- as you guys look out the stack beyond EDA, how much of the mix is semiconductors versus how much the mix of system design today? And then how do you expect that to change over the next 3 years?

Yes. So I mean, we've been tracking that for the last number of years. And basically, around 40% of our revenue comes from systems companies. For the longest time, that's been rounding up to 40%. And then over the last few years, we're rounding down to 40%. And as you heard Andrew talking about the Intelligent System Design strategy, it's naturally taking us into the system analysis space, and we'll probably see the systems portion of the business rise faster than our like core EDA semiconductor business going forward. So I'd say the trend is for the systems part of the business continue to increase more towards kind of 50-50 over time. And thanks for noticing the 50% incremental margin. I think that's key to -- it's an important part of our story, and it's key to what we've been trying to achieve at Cadence. Like I say, our focus has been not just on revenue growth, but on profitable revenue growth. 2017 was my first year as CFO, so I tend to measure from that point. But if you look at the last 4 years, 2017 through where we're guiding for at the midpoint for 2020. More than $0.50 of every dollar of revenue growth has dropped through to operating income over that 4-year period. I mean incremental margins are a little higher than intended at the moment because of some of the pandemic-related benefits I talked to earlier, like slower hiring. But -- so I believe it or not, I wouldn't mind seeing incremental margins drop a tiny bit. But I mean, I was showing our Board at our most recent meeting that at the midpoint of our outlook for 2020. I think we're on track to increase revenue by around $840 million since 2016. At 51% or 52% incremental margin, that would mean our GAAP operating income should be up by about $430 million to $440 million, but we're on track to be up by about $450 million over the 4-year period. And that's partly because of this slower hiring phenomenon that we're seeing in the pandemic. Overall, that's an average incremental margin of like more than 53% over the last 4 years, which is a little higher than we planned. I'd probably be happier at 51% or 52%, but -- if it means that we could get new hires in and productive -- and make them productive faster. But I really realize that's like having slightly higher incremental margins than find is like a top-class problem. So I won't necessarily complain it with that.

Great. And then just kind of going back to some of the generational drivers of growth. You guys called out cloud, 5G kind of within kind of the spectrum of the 5, were there any that you are kind of most excited about as near-term drivers? And then what are some of the ones that you think make take a little bit longer to fully reach maturity?

Do you want to take that one, Anirudh?

Yes, yes. First of all, it's good to be in this position that there are these multiple waves. Now in terms of which ones are happening faster and which ones are happening slower in terms of end markets. I mean, definitely, hyperscalers is happening at a very massive rate. I mean this is not a secret. So -- and our business in that segment also -- I mean this whole cloud data center, I think, is happening very, very fast. And the other area that we always watch, but it's traditionally slower though I think it has picked up, on the other side, is always things like automotive, how fast is the electrification of cars going to happen. Now some companies are really far ahead, as you know, in that electrification and -- both from a powertrain and inside the car. But I think all 5 of them are moving along pretty well. You look at like John mentioned, industrial IoT, 5G is a big year. This is a big year. So the good thing we like about Cadence is that we have a lot of customers and no particular customer or a particular segment dictates our overall position. And there's always some ebbs and flow and what is happening this year versus next year. But since we have such a diversified customer base, and we are focused on things like right products and incremental margin and execution that these goes along. But definitely, I see in the near term, cloud and hyperscaler is very positive trend among all the others, yes.

Great. So with that, we're actually over our time. There was a lot of investor questions coming through. So sorry, we didn't get to hit on all of them. But thank you so much, Anirudh and John, and for all the investors on the line. And with that, we will wrap up. So thank you very much.

Great. Thanks, Elizabeth, and thanks for having us again.

You're welcome.

Take Care.