Power Integrations, Inc. (POWI) Earnings Call Transcript & Summary

All right. I think we're on. So welcome, everyone to the -- close to the end of a very long day, a productive day for all of us. I'm Chris Caso, I cover Semiconductors for Credit Suisse. Our next presentation is Power Integrations. With us from Power Integrations is Sandeep, CEO (sic) [ CFO ]. So, thanks for coming. And you're going to have some slides to start for a little bit of introduction, and we'll get into a Q&A. So, please take it away.

Thanks, Chris. Well, I'll just start off with the forward-looking statement for a second. What I'll try to do here is give you a bit of an overview. We just had our Analyst Day in September, and I would really encourage you to go to our website because this was our first ever Analyst Day, and we really laid out the exciting opportunities that lay ahead of us. I've been with the company for 13 years, and all I can tell you is I've never been more excited about the future than I am today. One of the earlier slides that we talked about that has been how we have performed from a very long time from 2001. We have had a CAGR of nearly 2x that of the industry. And this has not come by accident. This has been as a result of focus. We are one of the unique companies that is very focused on high voltage. Many other companies do this, but they have a lot of other things to do. As a result of this, we have developed a lot of IP. In this area, we are different than the others that we provide a system solution. Many companies have a controller or a transistor, but we provide a complete solution. If you will look, we have our own process technology, we have IP in our packages, in design and the system knowledge that we have developed. In fact, half our sales force is field application engineers. And as I go through this presentation, I want to give you kind of a summary of what happened at the Analyst Day, and I'll give you a perspective of the revolutionary products that we have and the system-level solutions that we provide with InnoSwitch as one of the key examples. The 4 key takeaways from our Analyst Day was that high voltage is a unique specialty and PI is a pure-play player there. The other thing you will see in the slides ahead, where I'll talk about how our SAM that has already in the last decade doubled is going to double pretty much again in the next 5 to 6 years. We also talked about GaN as being the future of high voltage and how we are a leader. And basically, high voltage semiconductors are really crucial to the low carbon future that is a big talk about. And as I talk about it, I'll show you, we've been doing this for 20 years, though it's getting more vote today. So here, you'll see what are power converters. There are 2 types. One is the AC to DC taking the power from the grid and converting it efficiently for all electronics. And we go into different applications across, right, from a cell phone to an appliance to an industrial USB plug in the wall. The other is the DC to AC, where we have the renewables, which converts the DC into AC and provides it to the grid. We play in both aspects of this. The other unique thing about our company is the range of applications that we serve. We have 4 end applications that we are in. In our communication, it's really cell phone, in case of computers, it's PCs, tablets, monitors, in case of consumer segment, it's all white goods and industrial is a very widespread area with high power renewables as well as USB plugs in the board, home building automations, battery-operated tool. Generally, a lot of companies have hundreds of products. What we have is that we have a handful of products that actually addresses and we basically, have a silicon carbide or GaN switch that could go with our controller. And as a result, we have a very few products that serve this end market. So, there's tremendous leverage that we have in our R&D dollars. This is -- InnoSwitch is a perfect example of the revolution. InnoSwitch is a product that came out nearly 8 years ago and the unique thing it has, a transmission link between the high voltage and the low voltage side. And typically, you have to have an isolation. As you see the light green area on the screen, our chip actually sits right on the isolation barrier, being able to integrate stuff on the high voltage and low voltage side and able to regulate between the output and the input very, very efficiently. And this has become, I will show you in a chart a leading product for us. And I talked about the power of the switch earlier and that we have a proprietary control that increases the efficiency quite dramatically. Now this is a slide which is to show you what we do differently versus a competitor. And this is using a power supply of a competitor. And you have to watch, I'm going to highlight the different areas and what our competitor does and then I'll show you how dramatically we integrate this and how we have a lower component time. So, the competitor has -- what is highlighted in green here and then there are a lot of other things that go into this power supply, which are highlighted in the different colors. And what I'll show you for all 4 of them, they only produce the green one and all the others are external components that have to be bought if you use their application. In our case, what that does is the InnoSwitch just integrates all those colored areas. And what it enables you to do is come up with a solution that is highly integrated and do not have any of those external components. As a result, our component count comes down for our product from 100 what you're using for a competitor to 45. That is the level of integration we provide, which obviously leads to very high reliability, it's ease of design. And this is the chart I was talking to you about is if you look, our products actually last a long time. And here, this thing came out nearly 8 years ago. And even after 8 years, there is no product that even comes close. And you can see it even after 8 years, it hasn't peaked. If you look at our slides and the Analyst Day, and I don't want to go into, nearly 40% of our revenue come from products which are 20 years old. Our products last a very long time. And in fact, sometimes we find out hard to replace even our own products because people find them so reliable and have been so used to using them for a very long time. And this is an example of the growth that we have had, one of the fastest ramp with InnoSwitch, and there's a lot more to go. The interesting part of this is another example, InnoSwitch product, as I said, you can have silicon GaN or silicon carbide, but depending on the application. But the InnoSwitch goes into communication, goes into computer application, can go into an appliance and an industrial application and in fact, in the future into EVs too. One of the things that we have explicitly -- been very explicit about that SAM expansion is an integral part of our growth story. And we have had very revolutionary products that have enabled us to increase not only our SAM, but enabled us to address new markets. As you can see on the slide, we have made LED bulbs about electronic we made drivers, which made the bulbs last for a much longer period of time than traditional bulbs. You have a lot of IoT applications, which are permanently attached to the grid sucking power. And with our EcoSmart technology, we do not consume as much of that power. We also have advanced charges that you would be seeing for your cell phone, which are much more compact, much high level. E-bikes is another example. And of course, the EVs is going to be a big change. We had no play in traditional cars. But in the future, we have a big player and a big dollar content. If you go to our Analyst Day slides, it'll show you there's tens of dollars of content that we will have in a car and hundreds of dollars that we'll have in trucks. And obviously, the decarbonization effort high voltages go, I mean is going to play a big role, and we have a lot of products that are going to help. And I'll show you how we've been doing this for a long time. But before I go there, I talked about how our SAM will double. And you can see we'll be going from about $4 billion to nearly $8 billion with automotive, getting in motor control with the DC brushless motors. And here, we already have a big share in appliances, so we know the customers. And we already do gate drivers and here, we are going to penetrate with new products into the captive markets, plus our existing products with the GaN technology is going to go to higher power levels, increasing our content. As I said, GaN is future. We have a very differentiated GaN and it is unique, allowing us to go to very higher power levels. And as you will see, if you go to our Analyst Day slides, you'll see most of our new products are actually coming out using the GaN technology. It allows us to put higher efficiency and higher power. We have dramatic growth in our GaN revenue and we are a leader in GaN. Apart from revenue growing very quickly over the last 3 years, the pipeline that we have is very diverse. We have 400 opportunities and you can see that the GaN application is not just in 1 application, it's going right from notebooks, smartphone appliances, home building to computing. It's very diversified and a very broad opportunity base. The IAA has -- this is a slide from the IAA and you can see from a renewable standpoint, to get to low carbon future efficiency and renewables are a big player that contribute to that and we have a big play in both these areas. We had the EcoSmart technology where we were contributing to this low carbon long ago. And with this, if you think about it, this is what we have contributed. If you just look at last year alone, with the products we sold, it could pretty much use the power of nearly 2 billion homes of power consumption in a year. And we are clearly an enabler of clean technology. I talked about LED lighting, the EVs, the brushless DC motors and of course, the renewables. Talking about our financial model, we have talked about how we have had substantial revenue growth, our SAM is going to double. We also have 1 thing that to share is we have a very good share of different served markets. We typically have managed to do greater than 15% to 20% share in many markets. And in case of appliances, we have a very much higher share in the 40% to 50%. And with the power levels going up, the dollar content is increasing, and we are getting more and more dollars in different applications. We have a great model of a fabless IDM, which gets the best of the world where we can leverage our partners without the greatest investments and partner with them. And I talked about the R&D leverage that we had in our model with very few products. And as the time goes by, we are getting into a richer mix with auto, industrial and appliances. So, if you look on our model on a 3- to 5-year basis looking ahead, our revenue would grow low double-digits. Our non-GAAP gross margin will be in 50% to 55% even though we've been at a higher end. In fact, for next year, we're guiding it on the higher end of this model. And our non-GAAP operating margin is 25% to 30%, though we are going to be more towards the higher end of this model. And our CapEx spend is anywhere between 5% to 7%. The other thing is the consistency of our model. This chart shows you how we have been able to not only do the double-digit growth, but you have to look at our operating expenses, which grow at 60% of our revenue growth rate. And this chart shows that you shouldn't look at it only on a single year. But if you look at it over a long period of time, the model holds really well, and it also shows the consistency of the operating margin. If you look how we have -- we have generated 15% free cash flow from 2008, '21, we had 26%. We have utilized our cash very, very effectively. We have invested internally by thinking long term, pandemic being the most recent, we let our inventory levels go up and that turned out to be a real boon because we were able to supply when the thing turned around. Additionally, we invested quite heavily in R&D. Way back in 2011 when we were flat, we increased our R&D spending. We also do selective M&As, but are very consistently done, very accretive buybacks. In fact, if you look from 2008, we have spent over $900 million in buyback at an average price of $27. Even though we don't have a payout ratio, we are a company that has instituted a dividend and consistently increased our dividend. So over a period of time, we have -- we are paying out roughly about 23% to 24% of our free cash flow. To summarize, we have a substantial market growth opportunity. We have tremendous moat in our innovation and tremendous IP protection. We have system knowledge that is very, very unique to us and we have new markets where we are actually just touching the service. And that is why I'm very, very excited about the future. With that, I think we will turn over to some Q&A.

Absolutely. Well, thank you very much, and a very good overview of the company. Maybe I'll start the Q&A with some of the near-term conditions. And certainly, in the past, Power Integrations has been one of the kind of earlier companies to see some of the changes in the semiconductor cycle. And I think it's probably your exposure to power supplies, which tend to lead other markets. Maybe you could speak to that a little bit and why you think you tend to be earlier than others. And the level of conviction you have that now we've seen some degree of correction that do we have this dialed in at the moment?

Yes. And this has been always the case that we seem to see the downturn earlier. And our own theory here is that nobody waits for a final product for power supplies. So -- you don't want a cell phone and an appliance waiting for a power supply. So, they tend to make power supplies, I believe, a couple of quarters earlier than that. And I think that's the reason we tend to see it earlier. In fact, if you look at it, we were talking about directionally what was happening way back when we announced the Q2 quarter and we said the demand seemed to have got pulled in appliances, we were seeing slowdown in China. And in fact, when we guided even this quarter, we talked about Q4 and Q1 -- Q4 of '22 and Q4 of '21, 1 of the 2 quarters being where things would bottom. Ever since we have given our guidance, if you really look at it, China, things lockdowns have become worse at present. You've seen big OEMs talk about adjustment. Generally, when in China, you have the Chinese New Year, people go away for a week or so. We are hearing people will be shutting down for longer. All that leads me to believe that more likely that Q1 will be the bottom. Now, why do I believe Q2 will be a turning point. The reason for that is, I hope by that time, the inventory in the channel, which was elevated will be depleted because for both Q4 and Q1 of next year, I'm expecting sell-through to be greater than sell-in. Additionally, in appliances, if you look at it, we talked about in Q2 that demand got pulled in. But between Q2 of this year and Q2 of next year, a year would have gone by. And hopefully, that pull-in would have got digested by now. So, with having the challenge in the inventory, with the market share gains that we are having, we believe directionally, Q2 will be an up quarter from Q1 when Q1 being the bottom. And if that bodes well with all the market share gains we are having and the share gains we're getting in PC and getting share gains in the brushless DC motor with our BridgeSwitch product, we believe bodes well for the second half of 2023.

And follow-on to that, you were one of the few companies that spoke about some signs of weakness in the industrial space. And we had a number of companies up on the stage yesterday that spoke otherwise. And -- but industrials -- the definition of industrial is very broad. So, maybe you could clarify how you define industrial.

Yes. So, even though our industrial does not include our appliance business, which goes in consumer, which has appliance, what we call industrial has 3 buckets. It has got the traditional industrials controls, UPSs. We have what we call the high power business, which is the renewables, solar, the DC trends, big industrial motors that are used in fracking. And then what we call our new segment, battery-operated tools, utility meters, e-bikes, home building automation, which have a little bit of a consumer touch in it. What we saw is in third quarter, the channel inventory build in the industrial segment. We are seeing a mixed bag in the industrial segment with the high power business having double-digit growth this year and continuing to be strong next year. But other applications like industrial control and the consumer time homebuilding are having what I would call the weakness. And that's why you probably are seeing a little difference in us compared to the others.

Right. So, it's still generally pointing to the consumer in terms of where the weakness is most acute.

Correct.

That makes sense. And then just finally, with regard to inventory levels, to what degree of visibility do you have that kind of gives you that confidence of Q1 is the bottom by that point, the inventory is normalized?

So if you look at it, we are very far back from right from the retail space, but we do have a little bit of idea of what's happening at the key big OEMs and what's in our channel. The way we are looking at the build and what we are actually selling in, in Q4 and now thinking what we'll sell in, in Q1, the sell-through will far exceed our sell-in for both the quarters, which will normalize the channel inventory. And we had earlier talked, especially in cell phone, we felt that the OEMs would have depleted their inventory and will start pulling from the disty. And that's why it gives us the directional guidance that we start seeing things to bottom out in Q1 when the channel inventory is evened out because now the OEMs have started pulling from them. And that leads us to believe that even after a year has elapsed, then the demand has for the appliances, which got pulled in, would have got neutralized why we may have an uptick in the -- directionally in the Q2 quarter and Q1 being the lowest.

Understood. That's very good color. So, thank you. Moving on to some of the product stuff. And I wanted to focus on GaN to start because I think that's one of the more exciting places to be in power right now. And maybe you could discuss your breakup of exposure to high-power GaN versus low-power GaN. I think the 2 markets are quite different.

We are really focused on the high voltage side of GaN pretty much. And we have our own proprietary GaN technology, which -- that's what differentiates us. Most of the other people have discrete solutions. We have an integrated solution. Many of them use TSMC GaN. We have our own proprietary GaN that we install. Now initially, we are seeing success in PCs and in cell phone chargers, but we are also seeing applications in appliances, and also in industrial applications, USB plugs in the wall. We have products in development that will further allow us to get in the future in data center as well as in EVs. So, I think we are in the early innings, but as I showed you from the chart that I presented, basically, the applications for GaN are very, very diverse.

Right. And how do you differentiate yourself? There's a number of other folks that are participating in the GaN space, not so much in I think what some of your competitors are doing are more focused on lower voltage GaN. Is that your main point of differentiation? Or what else gives you competitive advantage?

The reason is we have -- we believe the differentiator GaN is going to give us -- which will get parity with silicon over the next couple of years and the efficiency levels that are required at higher power level will what is going to differentiate us.

In parity, you mean a cost?

Yes, especially with silicon, we're going to get parity in the next couple of years. But it's going to enable us to get higher efficiency with GaN.

Right. And you also mentioned EV as well. And EV, I think most people perceive EV to be a silicon carbide solution. You mentioned you have some participation in silicon carbide as well. Where does GaN fit into EV?

Well, GaN is going to play like in the onboard charger in the future. It's not at the inverter and right now, we have an InnoSwitch silicon carbide solution. There is a lot of power supplies that go from the 800-volt or the 400-volt battery to different parts of the car. And that's where with those multiple power supplies, we have a content play in each of that.

Right. Right. So, maybe we could say that the voltage range that is sort of the sweet spot for the GaN -- for GaN, what is that sweet spot voltage range?

Well, we just -- typically, right now at the 650, 700, but it has a potential to go all the way up, 2,100.

Okay. Maybe speak a little bit about the handset business as well because that's an area where you've been very strong in the past. Again, GaN seems to be a play there as well. So, what's your level -- what's your participation and what's your outlook for where that handset charger business is going?

Well, what's happening is that all of us want to get our phones charged 70% in less than 20 minutes and going to higher power levels is the only way to do it. And the GaN enables you to get a very compact charger. What you traditionally used to see in a queue of a 5 watt, a little bigger than that can now give you a 60-watt charger. What you used to get in -- along with your PC was a brick, now with GaN, we can have a very compact, highly efficient charger giving you 60 watts. And that's the big play for us. So, it gives us higher power level, which gives us higher dollar content. And typically, anything above 30 watts, most people will use a GaN because it will be much more cost-effective than silicon in that space over a period of time.

Right. Just checking if there's any questions from the audience as I -- well, let me continue. I'll move on to the industrial space and motor efficiency is just a critical importance. We hear that from so many different areas. Where are the growth drivers for you in industrial and motors in particular?

Well, that's why the BridgeSwitch product, the brushless DC motors, there are a lot of different applications in appliances, in fans. And everywhere, people are moving to brush because it's much more efficient. We have very high efficiency, 98% efficiency. And as a result of that, the move is there. We have some very unique fixed features in our product, which have diagnostic capability to tell when it will fail and things like that. We are in the first stage of a product and we'll be going to much higher power levels. And in fact, in the future, will come out with even a GaN product in the motor control space. This is something that is adding about $400 million to $500 million of SAM already. But if you look what we presented on the Analyst Day, we'll add close to $1 billion of SAM. So, it is one of the big growth drivers. Right now, we have single-digit millions of revenue, but that is going to double in the next few years, every year.

Right. Randy, you had a question, please?

I was curious when you gave the introduction. I just wasn't as familiar with it. I think you mentioned you have the fabless IDM model. So, could you go through that? Like do you -- how much in-house versus...

We have no fabs in-house, actually, we are fabless, but the reason is it's very unique. We don't go to the merchant fabs. These are captive fabs whom we have had relationships for 20, 25 years. Very unlike other companies, we had chosen to be away from Asia and had our fabs in Japan, Europe or U.S., whom we partnered with. And it was primarily for IP reasons. And these guys are captive, but we are real partners where we use a sizable portion of the fab. We are not the leading edge. So, we are using their fully depreciated fabs. And these are not partnerships that are new. These are foundry partnerships of 20 years that we have developed. And that's how we maintain the relationship that even in a downturn, we try to utilize the fab to keep their economics correct and they are able to provide us a very cost-effective way without we having to invest a ton of money and capital.

So these are ones like…

No, this is X-FAB, Epson, Toshiba, LAPIS.

Okay. Thank you.

Sure. And maybe just a middle up, just following on how that materializes in margins. Obviously, not CapEx because you're a fabless model. And how do you compete with some of the larger players, Analog Devices, Texas Instruments that have some of their own facilities? How do you compete on cost?

It's the level of integration on our own process technology. We install our process technology, which enables us to get the cost effective. And you just saw in the example, everybody has a discrete solution. The way we price our product is value pricing. Capturing all the value of the components that we have eliminated. I showed you an example where we had 100 components of a competitor, outs was half that. But we don't price our product at half, we price our product at the value of the 100 component. That's how we are able to do. Plus, we are not using leading edge, right? So, these are all fully depreciated fabs. But yet we utilize their fabs very effectively on a very consistent basis.

Right. That's good. It looks like, unfortunately, we're out of time. But Sandeep, thank you very much for attending, and thank you, everyone.

Thank you very much for having me.

Thank you.