SiTime Corporation (SITM) Earnings Call Transcript & Summary

All right. We'll go ahead and get started. I'm Blayne Curtis, Semi and Semi Cap Analyst at Barclays. Next up, we have SiTime. Very happy to have Rajesh Vashist, CEO; as well as Art Chadwick, CFO. We were -- just had less time in reminiscing. We were on the road in March, and it was quite different times.

So I actually thought maybe a good way to start. Obviously, you had a very strong IPO story, penetration, high precision into some core end markets, which you did. And then obviously, the business really accelerated during the pandemic. I think that's kind of the biggest debate on the stock is just some of the acceleration, how much of that kind of sticks, how much of it goes to inventory if it comes down. Let's just start with the part of the acceleration, and we'll get into some of the correction. But I just want to understand, the last 2 years, in terms of what the IPO plan was, how did you see things change during the pandemic?

Yes. So it's quite obvious that we are a growth story, right? When we went public 3 years ago, we had spoken about being a 20% growth rate, give or take, maybe 25%, maybe 22%. The year after, we grew at 66%. The year after that, we grew at 88%. This year, we're going to grow at approximately 30%, 30-ish percent, according to what people have us at. If you look at where analysts have us for next year, 2023, and you look at where we started in 2019, that's approximately a 30% growth rate with all the peaks and troughs taken out. Essentially, at the end of the day, my point is that we are holding steady at 30-odd percent, which is more than what we said at IPO. We've gotten to this level of revenue quicker than we anticipated, thanks to COVID. What COVID did, which was so fundamental to help us, was we were -- we're still a relatively small company in the timing space. We're only, give or take, $300 million in an $8 billion market. At that time, we were an even smaller company. We were less than $100 million company in an $8 billion market. What COVID did was it accelerated customers to us, it brought customers to us, it brought opportunities for us. We have taken that opportunity to fuel our product fire, because products are the heartbeat of our innovation. We would come out with 2 or 3 products a year that went in 2020 to 4, in 2021 to 5 and 6, in 2022, another 5 in 6, and so on. So that's the heart of it. That's the part of we were going to get to a $4 billion SAM in '26. We're going to get to a $4 billion SAM now in 2024 as a consequence of this excursion while maintaining a solid gross margin and maintaining growth. So I'm very, very pleased with what the last 2, 3 years have done for us. I know we have our price very high, too high for some of you guys, I think. And I think where we are is a very fair value to the company.

Yes. Blayne, I can add to that. You're talking about what our strategy was at the IPO, and one of the pieces of the strategy was to improve our gross margins. If you remember, back at our IPO 3 years ago, our gross margins were in the high 40s. And we talked about how we were going to move them up at that time into the high 50s, and now we've gone into the high -- into the mid-60s. And we've done that by following our strategy to go into these higher end markets. If you remember, our revenue into Comms & Enterprise back 3 years ago was just a couple of million dollars. Last quarter, Q3, it was 1/3 of our business. And of course, that's the business that we're really focused on. Our newer products are focused very heavily on these higher end markets, give us higher ASPs, higher gross margins, and that's exactly what's transpired over those last 3 years.

Yes. This is a perfect lead, and there's a couple of follow-ups I have. But maybe since you mentioned the products, I mean you had a couple of announcements this week, including one today. So you need to -- you've kind of introduced these families, and they've kind of carved out a little bit more of your addressable market. I don't know if you want to mention anything with the ones that you announced today, if there's anything interesting there. But let me just kind of talk about the addressable market. It's a $9 billion, $10 billion timing market, and I think you've been chipping away, adding $100, $150 kind of little chunks of addressable.

Yes. So the product that we announced today is all of a piece with our 2 strategies. One is better, cheaper, faster. So in a category, let's say, TCXO or [ make sort of ] a clock to deliver a better solution. We have also started to create entire new categories of products. This one today, the Elite RF product, is a new category of product which is a very, very high-performance TCXO of a kind that doesn't exist, particularly for RF, particularly for 5G. We had solutions for 5G earlier, but this one carves a new one out and probably brings a couple hundred million dollars of new SAM into it. A product we introduced some time back, Elite X, similarly, we created new markets for it. A product that we added 1.5 years ago, Cascade, in the clocking group also added new markets in which we have more than 200 customers now in the space of about 1.5 years. So the story of SiTime is that a company that imagines timing when nobody else thought about timing, that brought together clocking, oscillators, resonators, software, solutions, all into one, and then has proceeded to lead it with highly innovative solutions. That's -- in a market that keeps on growing because of ADAS, because of AI, because of IoT, because of 5G, all those are winds beneath our wings.

And then there's a lot of kind of levers to pull for growth. I think the pandemic was unique that, obviously, some of your competitors literally ran out of parts to ship to customers, bought on new designs. Your supply chain was different, you had availability. So can you talk about -- is there any way to think about how many customers walked in the front door? And then have you seen any customers rethink their balance of timing now that things are easing up?

So some metrics, when we went public, we had about $5 million. That's 2019. This time, around 2019, we had about 5 customers that were $1 million or more in revenue. Today, we have more than 30 customers that are $1 million or more in revenue. So in the last 3 years, we've added 25 customers of that scale. That's -- I would say, 2/3 of them are as a consequence of the shortages, that pandemic. The other part of it is that we are not a market share gain. We are a unique provider of solutions that today, there's 40 crystal companies and 20 clocking companies. We always tell our customers, we're going to charge more because we deliver value that's more. By leading with premium pricing, by leading with uniqueness, by leading with differentiation, customers understand that we are the solution that they prefer to go with rather than, oh, you were just there so I went with you. Of those customers that came to us, we don't think any have left. Even those that have reduced their revenue because we are too expensive in some categories for them have continued to stay on at places where we matter. So when we went public, we had about 200 different applications. Today, we have around 300 different -- 350 different applications. So I think it's been growth and differentiation in all areas.

Yes. And maybe just a little bit more on the competitive landscape because what they all did was these 40 run out and build more factory capacity, or at least a little bit more than a handful of them. Now we're obviously seeing a broader slowdown. So the answer might be that you're above that in terms of your performance, and that might be their problem. But are you seeing any indications that there's now more capacity in the market than they need?

Absolutely. There's always -- if you look at the history, the 60-year history of crystal, it's always been boom and bust because that's what typically happens. They respond retroactively to adding -- and add more capacity. All of them add capacity. They can all deliver similar kinds of products. When those products become available, they have to all bomb pricing, so they go down this spiral, and this has been continuous. At good times, their deliveries are 16 to 18 weeks. In tough times, they are as high as 25 weeks, and this just wasn't there for the last 2 years. They've been doing this over the last 3, 4 decades. So put it another way, nobody has ever said that they love their crystal oscillator supplier because it's not been very good. So when SiTime comes in, more performance, better performance, way better quality, better reliability, programmability, quality, focus on the timing market. Most people don't want to leave. Pricing just becomes a small thing.

Yes. Maybe following up on that, I was going to ask you about pricing. I mean if you look -- I think you once -- you said to me that it was kind of like maybe we're underpricing our products a bit, like for the performance we're delivering. But I mean you did see one of the bigger increases in terms of prices. You've not, I don't think, quantified it, but you do report units and ASPs, and it was up substantially. I know there's mix and there's other factors in there, but I'm just kind of also curious that this is -- can you just explain this discovery process you saw? And were you pricing it too low? And does that put a better floor in your gross margin going forward?

Yes. So if you go back to last year, we did have some lower margin business, and we went and raised prices. We raised prices, one, because it was a low margin business, and we were willing to walk away from it if the customers didn't want the price increases. And our experience was nobody left. So that kind of took care of the low margin business. The other piece of the ASP equation is mix, as you mentioned. Again, we've been selling more and more into these higher end markets, comms, enterprise, mil, aero, electric vehicles. That has, again, higher ASPs, higher gross margins, and just that change in mix over the last few years has also raised ASPs. So it's kind of the combination of those 2 factors.

I'll ask you, you saw a correction, now you've seen kind of 2, and it kind of started with consumer and spread into some other markets. Maybe a good place to start, just where are you seeing a correction? Are there some markets that you're still not seeing a correction? And kind of walk us through that progression as we're kind of getting to the bottom. Is this -- by March, that should be the floor or not?

Do you want me?

Yes.

So we're going to what I'd consider to be kind of a classic semi cycle, right? There is too much inventory out there. Well, let me back up just a little bit. Last year, there were a lot of shortages, and a lot of folks had challenges with their supply chain. Couldn't get enough parts, we all read about the issues there. Anecdotally, what I think has happened is in the first half of this year, a lot of those shortages eased and customers went out and bought probably more product than they needed based on their current run rates. Not just with us, but I think with other semis out there. Then in the back half of the year, you've seen something of a slowdown in the economy, right? We read all the news, the war in Ukraine, rising interest rates, shutdowns in China, the list kind of goes on. And what that has done is that has slowed some markets. We saw this slowing first in consumer in the summertime, and we saw some slowdown in industrial kind of early fall. Now we're seeing some slowdown, not a lot, but some slowdown in the data center and in the comms market. So you've got this situation where people bought more inventory than they needed for buffer stock, then demand goes down. Now they have more inventory than they really need. So their order rates come down even more as they work through that inventory. And that's what we're seeing. So the bad news is there's a little more inventory out there than our customers need. But the good news, in my mind, is it bounds the problem. What we did is we went out to identify how much inventory we think our customers have, and I'll explain that just a little bit. We fulfill all of our sales through a distribution network, and we get sell-through from our distributors every month. So we know what they're holding, we know what they're selling through. And that distributor inventory really hasn't changed materially over the last few quarters. The issue is at our customer's site or a customer's contract manufacturers, now they have more inventory than they need. We went out and did an analysis on our top 30 customers. It came to the conclusion that those customers, on average, have about 45 days more inventory than they need. So that's not 45 days of inventory, that's 45 days more than their current run rate needs. That's about half a quarter's worth of sales. So what we're seeing this quarter in Q4, we had to guide revenue down because some of our customers really didn't need more product and they're not ordering it. We understand that. And that same thing happens in Q1. We actually offered comments 2 quarters out, out to Q1. We think Q1 will be down at least 20% to 25% from Q4. So a number of our customers should be able to work through their inventory either in Q4 or Q1. For some, it goes into Q2, but it really bounds the problem. We think by the back half of next year, our revenue should continue on a growth curve. One, because the inventory overhang disappears, and two, with the number of new design wins that we're getting. We've talked about on our conference call, design wins doubled this last quarter versus a year ago, so a lot of activity out there. Quote activity is up 4x. We've introduced a number of new products over the last couple of months, including the one we announced today. We've got more products that we're announcing over the next couple of quarters. So all of that momentum, we think, will put us back on this kind of 30% CAGR growth path sometime in the back half of next year. So that's a long-winded answer.

No, no, it's good. It's good. I mean like a market like consumer, I think there was also a variable where some of the markets like wearables or something are probably at an unsustainably high level. But if you look at where your revenue is now, it's like back to prepandemic, which I guess I'm trying to wrap my head around, you said half a quarter inventory. You also probably had some customers that walked in the door in that market as well. So you should -- I mean you're clearly under-shipping demand. So you think...

Well, you've got the inventory issue, and you've also got somewhat slower demand in different markets. It's substantially lower in consumer. For us, that would be non-Apple consumer, because our Apple business is part of our consumer bucket. That went down considerably. That dropped 60% from Q2 to Q3, so huge decrease in demand in the consumer. As I mentioned, some weakness in industrial, and some softness now in data center and comms.

So when that happened over the summer, I think everybody was already pretty well aware that consumer was weak and it wasn't a huge surprise. I think when it spread into comm, it kind of threw some people for a loop because you have such growth. The question was, could you offset? Was it ramps, or was it inventory? I guess there's now pretty clear that there's some inventory in comm as well?

Yes, absolutely. Our largest customer in comms has more inventory than they need for the next couple of quarters. And our largest customer in data center has more inventory than they need also for the next couple of quarters. So we just need to work through that. The fundamental drivers that drive growth in our business, we think, is as strong as ever, as we talked about, but we got to get through this period.

And when you are scrubbing -- I mean the -- you have this third bucket, it's like industrial, auto. But then there's also broad-based, which I think was the weakest of it. So even though it was industrial, it was really more broad-based. How do you think about those -- the rest of that bucket in terms of whether there could be inventory there well -- as well?

Yes. So we take our sales and we report kind of 3 big buckets. So the bucket you're talking about is we call it industrial, auto and aero. So it's got industrial, broad-based, auto, aero, medical and other miscellaneous things. So industrial kind of -- low-end industrial's pretty soft. higher-end Industrial, a little weak, but not that bad. Auto is actually relatively strong for us. We've got a lot of business with the EV manufacturers. Everyone knows who our largest customer is there, and we're doing very well there. Aerospace, very, very strong for us, but it's just longer lead times. We've got a number of design wins, but in aero, it takes 1 or 2 years before those go into production. And medical has actually been reasonably strong for us.

I think Arthur answered all the hard questions. I had one for Rajesh in terms of -- you mentioned the pipeline, and you would talk prior about kind of $200 million pipelines. Maybe talk about 2 things. I mean I do want you to describe that pipeline and your visibility there? But also, did the pandemic and now the correction change any of that timing for you?

No. I think if anything, it's only outlined how strong we are. Take auto, which we've talked about, when we went public, if somebody had asked me what's your market in auto, what's your SAM in automotive, we would have said sub-$20 million. By 2024, with the products that we have, we expect it to be anywhere from $800 million to $1 billion in SAM, right? So we think that when we are done with our new product, we'd probably get something like $50 million for the right car, for the right EV, ADAS car. Because in the car, the 2 things that are driving our adoption, one is electric vehicle, which is the battery management and so on, of course, but the real part is losing weight for the car, which means replacing cabling with ethernet, which means replacing mechanical switches with softswitch switches, displays and such. So I think that's a market for us, but an even stronger market is an ADAS market where the sensing element, so the camera, the radar, the LiDAR, the communication element, the GPS, the management of all this data, the processing with the ADAS processor, the domain controllers, that becomes a good market for us. And that market is right on track to keep on growing. Really, ADAS hasn't really kicked in as much as we all know, but it's going to keep on doing that. The data center market, I think, will just keep on growing because 400-gig headed to 800, headed to 1.2 -- 1.6, excuse me. So that's a market which is high performance, size, power, all of those things that we are behind. Having said all that, the -- I will say that it's not all easy life these days being management for a semiconductor company because as I've said before, in 40 years of my experience, I've never seen so many disparate factors that we, as a management team, have to go understand. We have to go understand the impact of the Russia war. We have to understand the impact of the China lockdown and when it's going to come off and what it's going to do to demand. We think it could be a really tremendous snap, cloud positive snap in the market, but we don't know. Supply chain still is a problem. There are still some car companies that can't get what they want. COVID alternatively is on, off, but mostly [ an advance ]. And then, of course, the financial markets and the inflationary turmoil. Typically, we have 1 or 2 things to juggle with. We have 5 balls in the air now, 6 balls in the air. So it is -- they are -- this is the world we live in. But in spite of that, digesting all that, we still say that by -- as Art said, by the end of this year, we should be back to our 30% extended growth CAGR, which I think is incredible for SiTime in the market while maintaining profitability at the same time.

I wanted to dive into that data center opportunity. You've been very positive. I think you even have a slide on it in your deck. Where are you today? When you talk about data centers, a lot of things. The types of products you have there today, where do you see opportunities for your timing within the data center? And is that tied to any product launches? Or are those products to address the TAM in the market today?

Yes, they are tied to product launches, all of it. Our total SAM today in networking expanding beyond data center in 5G radio access networks in core and core and edge and then finding data center is about $400 million, $500 million today. In 2 years, '23 and '24, at the end of '24, given our pipeline, we expect it to be $1.3 billion, $1.4 billion. So a massive increase in SAM. In the data center market, we are in leaf switches, spine switches, the optical piece, the servers, the timing servers. But we're coming out with more and more clocking and oscillator products in this market. And with that, we expect to do a clean sweep of this market. That should -- that's about a $400 million sub-SAM that we would go after. So clearly, we feel good about the $100 million in a few years.

You mentioned clocking and you have some organic clocks. I think there's -- you've talked in the past whether there was something you could buy in that market. M&A is usually tougher downtime, so who knows. But I mean, I guess, can you just update us as to kind of the progress of what you've introduced so far? And is this something that at this point is still an organic operation? Or...

Yes, it's definitely an organic operation for now. Because as you mentioned, there are very, very, very few, if any, off-scale clocking businesses that we could acquire that are stand-alone. And therefore, if we go up to a large analog semiconductor company and try to bring out a $50 million, $100 million business, it's harder for a variety of reasons. So we can't depend upon it even though that probably doesn't stop us from trying to do it. In the meantime, Plan A is to continue with our family of products, particularly focused on the networking, timing on the enterprise, on data center markets. Because what we build for that $1.3 billion market takes time to disperse into automotive and military, aerospace. So we have 1 product that has about 200 different customers. It starts to be fairly at scale next year, '23, and certainly in '24. In '23 and '24, we'll probably introduce a variety of new clocks for that market. Typically, ASPs for these products is around $3 to $7, higher than corporate gross margin and very, very sticky.

I want to ask, you already mention your gross margin, but I'm going to ask you another question on gross margin. So either you had -- part of it was just raising prices on those products at lower gross margins. But then like everybody else, there's a big portion that was passing through price increases that you had to deal with. What's your perspective on that part of the component? Do you expect pricing to get worse still? I mean there is still talk about companies like TSMC raising pricing. Or do you think now that the market is kind of moving some markets to oversupply, that might ease?

You're talking about [ these products ]...

Because obviously, if you get price cuts, I'm assuming you're going to have to -- you raise prices because they [ selling -- I'm assuming ] you might have to give back.

I'd love to get some price cuts. So we've talked about this. So TSMC raised prices by 20% for our CMOS wafers this year. And they do that across the board to all of their customers, including us. And we went out to our customers and raised some prices, some, but not enough to cover all of that increase. We're able to absorb the rest of that increase by other cost improvements in our manufacturing organization. Better sort yields, better test yields, better test throughput, lower package costs, things like that. So we were able to absorb that, and for the course of this year, maintain margins essentially in the mid-60s. Now I had to guide down 2 points of margin from Q3 to Q4 and then another 2 points of margin essentially from Q4 to Q1, but that's strictly related to volume, not reduced prices. We've got a fixed amount of, I call it, manufacturing overhead. It's the customer operations group, depreciation and some equipment we have at our OSATs and such. So when revenue drops, that decreases the blended margin number. And so that accounts for the decrease from Q3 to Q4 and Q4 to Q1. As revenue goes back up, that of course will reverse itself. So our decrease in margins short term have nothing to do with price pressure on prices that we sell our products to our customers for.

I'll add something to that. I really want people to think of SiTime as a high value, high differentiated product company, which means we lead with high prices. We are not afraid to talk about high prices. We're not afraid to talk about single-sourced. In the timing business, particularly in the oscillator resonator business, it's axiomatic that you're multi-sourced. So when SiTime shows up and says, you will be single source with us, we have said officially that 80% of our business is now single-sourced, and in comparison to about 2 years ago when about maybe 30% or 40% was single-sourced. And we are very clear that just like WiFi chips, just like processor chips, just like other chips in the market, our customers are always single-sourced. Why should timing be any different? So we want to make sure that they understand and get comfortable with high prices for high value, high differentiation. And single-sourced, because of high quality, high reliability, great support, that will never let them down just like any other semiconductor company like a Broadcom or a Marvell, whoever. So with that said, I think we're carving out a position for ourselves where architectural people at our customers are paying attention to timing as it becomes more and more important, and they're willing to take a chance in SiTime. That's the message that we are putting out and we'll continue to put out.

Right. Well with that, we're out of time, but appreciate it. Art and Rajesh, thank you very much.

All right. Thank you, Blayne. Thank you, everybody, for spending the time.

Thank you.