Power Integrations Inc Q3 FY2023 Earnings Call

Thank you for standing by. We do apologize for the delay in starting today. My name is Christina, and I will be your conference operator. At this time, I would like to welcome everyone to the Power Integrations Q3 Earnings Conference Call. All lines have been placed on mute to prevent any background noise, and after the speakers' remarks, there will be a question-and-answer session. I would now like to turn the floor over to Joe Shiffler, Director of Investor Relations. Joe, you may begin your conference.

Thank you. Good afternoon, everyone. Thanks for joining us. With me on the call today are Balu Balakrishnan, Chairman and CEO of Power Integrations; and Sandeep Nayyar, our Chief Financial Officer. During the call, we will refer to financial measures not calculated according to GAAP. Non-GAAP measures for the third quarter exclude stock-based compensation expenses, amortization of acquisition-related intangible assets, and the tax effects of these items. A reconciliation of non-GAAP measures to our GAAP results is included in today's press release. Our discussion today, including the Q&A session, will include forward-looking statements denoted by words like will, would, believe, should, expect, outlook, vision, view, forecast, anticipate, prospects, and similar expressions that look toward future events or performance. Such statements are subject to risks and uncertainties that may cause actual results to differ materially from those projected or implied. Such risks and uncertainties are discussed in today's press release and in our most recent Form 10-K filed with the SEC on February 7, 2023. Finally, this call is the property of Power Integrations, and any recording or rebroadcast is expressly prohibited without the written consent of Power Integrations. Now I'll turn it over to Balu.

Thank you, Joe, and good afternoon. Third quarter revenues were up 2% from the prior quarter but came in at the low end of our guidance range at $125 million. And for the fourth quarter, we expect a sequential decrease to $90 million at the midpoint of the range. Our results and outlook largely reflect the broad-based weakness cited by many of our peers this earnings season. In the industrial category, we are seeing strength in renewable energy, thanks to recent design wins, but the broader industrial market has weakened as several of our peers have noted. The appliance market, which now accounts for about a quarter of our revenues, has been affected by the slowdown in home sales and the residual effects of the pandemic when many appliance purchases were pulled forward. We also had an unexpected cancellation late in the quarter, affecting both third quarter revenues and fourth quarter backlog in the computer and communications categories. We believe this reflects efforts on the part of an OEM to reduce charger and component inventories. Notwithstanding the short-term outlook, we continue to see strong design activity and design wins that position us well for the eventual upturn in demand, and we feel as good as ever about our long-term growth prospects. We are on track to double our addressable market by 2027, driven by electric vehicles, motor drives, renewable energy, expanding dollar content, and a host of upcoming products featuring our proprietary GaN technology. We are full speed ahead on GaN development, more confident than ever that GaN will not only overtake silicon as a technology of choice for most high-voltage applications but will also be a more cost-effective and greener alternative to silicon carbide over the long term. Most GaN suppliers rely on the technology of a single foundry, leaving little room for differentiation and affording them no control over the technology roadmap. Our GaN is proprietary, which means that we not only control the roadmap but that we can tailor our GaN switch for optimal system performance in each application. Our GaN technology is also unique with characteristics that make it better suited for higher voltages than any other technology in the market today. In March, we introduced a 900-volt version of our GaN InnoSwitch products. And last week, we took the next step on the roadmap with our latest InnoSwitch featuring a 1,250-volt GaN switch. Higher voltage InnoSwitch ICs are ideal for many industrial applications and for geographies with unstable main voltages, as well as power supplies in 400-volt electric vehicles. These higher voltage GaN technologies will also enable us to expand our SAM with new products that address higher power applications up to 10 kilowatts, such as EV onboard chargers and DC to DC converters and a range of industrial applications. Many of these applications are served today by silicon carbide because there is no viable alternative that delivers the necessary level of efficiency. However, we have products in our pipeline that will offer a system-level GaN solution at a much higher level of performance than silicon carbide. GaN is fundamentally more cost-effective than silicon carbide because it uses lower-cost raw materials and requires a tiny fraction of the energy needed to produce silicon carbide, which is processed at extremely high temperatures. That also makes GaN a more sustainable technology than silicon carbide, which quantifies a portion of the efficiency benefits in its own manufacturing process. Our GaN roadmap does not end at 1,250 volts. We expect to introduce even higher voltage GaN in the near future, and our vision includes the potential to drive GaN beyond 10 kilowatts, making it a viable replacement for silicon carbide in a wider range of applications. While pushing GaN beyond 10 kilowatts is a longer-term proposition, our current game products are making significant gains in power supplies. We won more than a dozen smartphone and notebook designs in Q3, including a 100-watt inbox designed for a top notebook OEM which uses our hyper PFS 5 power factor chip in tandem with InnoSwitch, both incorporating GaN switches. We won an even greater number of GaN designs in non-mobile applications, including a multimillion-dollar design for a new platform at a top European appliance OEM as well as designs in home automation, lighting, audio, and industrial controls. GaN also features prominently in our roadmap for motor drive products and will enable us to more than double the addressable market for our BridgeSwitch products. Meanwhile, current BridgeSwitch products, which incorporate our proprietary silicon threaded technology, continue to win designs despite strong headwinds in the appliance market. In addition to exceptional efficiency in active mode, BridgeSwitch offers very low standby consumption, which is attracting strong interest from appliance customers in light of upcoming changes in the European eco-design standards. As we mentioned last quarter, the allowable standby consumption for a wide range of electronic products will be reduced beginning in 2025, and we expect the new standards to be especially impactful in the appliance market. The low standby performance of BridgeSwitch perfectly complements our EcoSmart technology, which has helped us win a dominant share in the appliance auxiliary power supplies. In September, we introduced our latest EcoSmart product, LinkSwitch-XT2 SR, which offers no-load consumption of less than 5 milliwatts. In Q3, we won a design at a top European customer for chipsets combining LinkSwitch-XT2 SR and BridgeSwitch in a refrigerator compressor scheduled to begin production in the middle of 2024. At a higher level, customer interest in our products has never been stronger. On last quarter's call, we said that we had added more potential revenue to our design pipeline than any quarter in our history. And we beat that record again in the third quarter. This reflects the broader range of applications we are addressing and our rising dollar content as evidenced by the fact that our average selling price has increased by almost 70% over the past six years. We are well-positioned for growth once demand returns, thanks to recent design wins, including a Q3 design win that will give us a significant role in India's 5G fixed wireless rollout, which is expected to ramp over the next several years. Finally, we demonstrated the superior efficiency and ruggedness of GaN InnoSwitch ICs last month in the Bridgestone World Solar Challenge, where we co-sponsored a team of engineering students in a race across the Australian outback in a solar-powered car. With the help of our applications team, the team implemented a DC to DC converter that achieved almost 96% efficiency at full power and a 50% improvement in light load efficiency compared to an earlier solution, greatly reducing the car's energy use. I'm happy to report that of the nearly 30 entrants in the race, the team we sponsored was one of only 12 to complete the seven-day 300-kilometer journey. To conclude, in spite of the tough demand environment, our long-term outlook is unchanged, and we are focused on what we control, sticking to the playbook we have followed in every downturn. We are keeping inventory elevated to retain foundry capacity and to be ready for the strong upturn, knowing that we will be among the first to see the turn when it comes. We are managing expenses prudently but investing in the products and technologies that will drive our long-term growth, such as GaN, automotive, motor drive, and our next-generation gate drivers. And we are buying back our stock when it's down and growing our dividend, knowing that we will continue to generate strong cash flow as revenues recover. With that, I will turn it over to Sandeep for a review of the financials.

Thanks, Balu, and good afternoon. In light of the weak demand environment, we are managing expenses prudently while continuing to invest in the opportunities that will drive our long-term growth. We are also moderating our manufacturing volumes, but as always, we are cognizant of maintaining foundry capacity and having inventory to respond to an upturn when it comes. Revenues for the third quarter were $125.5 million, up 2% from the prior quarter and down 22% year-over-year. On a sequential basis, the communication category was up mid-teens, driven by design wins and channel restocking associated with the China handset market. Chinese OEMs, their charger ODMs, and distributors have run at unsustainably low levels of inventory over the past several quarters, and we view the restocking as a further sign of normalization in that market. The industrial category was also up mid-teens sequentially, driven by high power, where we have seen strong growth in utility-scale solar. Automotive, while still small, was also up from the prior quarter. Computer revenues were down more than 30% sequentially, driven by tablets and, to a lesser extent, notebooks and aftermarket chargers. Consumer revenues were down mid-single digits, driven mainly by seasonality in air conditioning and the continued softness in the overall appliance market. Revenue mix for the quarter was 32% industrial, 32% communication, 26% consumer, and 10% computer. Non-GAAP gross margin of 53.3% was modestly below our expectation due to end-market mix, but nevertheless increased by 150 basis points from the prior quarter, driven by manufacturing efficiency and the weaker yen. Distribution inventory ended at 11.6 weeks, up 1.5 weeks from the prior quarter, driven primarily by the channel restocking for China handset customers, as mentioned earlier, while distributors for other end markets remain at elevated levels due to weaker sell-through. However, we did see a reduction in channel inventories in October. Non-GAAP operating expenses for the quarter were $41.8 million, down more than $2 million sequentially and well below our forecast, as we continue to adjust the pace of hiring and manage discretionary spending carefully. Non-GAAP operating margin for the quarter was 20%, up almost 4 points from the prior quarter. Non-GAAP earnings were $0.46 per diluted share, up $0.10 from the prior quarter. Cash flow from operations for the quarter was $26.7 million. Inventory dollars on the balance sheet were essentially flat from the prior quarter and rose by four days to 230 days at the end of the quarter. Uses of cash during the quarter included $8 million for CapEx, $11 million for dividends, and $2 million for share repurchase. The buyback at $73 million remaining at quarter end has been significantly more active since the end of the quarter as dictated by our preset price volume matrix. As noted in our press release, our board has increased the quarterly dividend to $0.20 per share beginning with the fourth quarter payout in December. Turning to the outlook, we expect revenues for the fourth quarter to be $90 million, plus or minus $5 million. Non-GAAP gross margin should be similar to the Q3 level of 53.3%. Non-GAAP operating expenses should be around $42.5 million. That puts us on a course for a full year expense growth of only 3% despite this year's high inflation. Finally, I expect a non-GAAP effective tax rate for the fourth quarter to be around 7%. And now, operator, let's begin the Q&A.

And your first question comes from the line of Tore Svanberg from Stifel.

Yes. My first question is on the channel inventory. Sandeep, I think you mentioned it was 11.6%, obviously, up sequentially. But you also said you've started to see it come down this quarter. And I assume with the decline in revenues, is going to come quite a bit more. So where do you expect the channel inventory to be as you exit the year?

So I think we should get a benefit of at least 1.5 weeks or so in the coming quarter. As you know, this last quarter and the quarter before, I've been a little off. Normally, we generally have it. But I think the slowdown has kind of surprised us a little bit on the sell-through. But October was a welcome turn where the sell-through definitely exceeded the sell-in. So I think my guess is about 1.5 weeks down.

Very good. And related to that, I don't know if you want to answer or perhaps Balu here, but $90 million is obviously half of where the peak was, actually even more so. And I'm just wondering if you have a good feel for where your true consumption is. Obviously, this has been going up for a few quarters now. You had a little bit of a relief, obviously, from restocking perhaps in communications and PC. But any read you have on true consumption? And some companies have talked about sort of ripping up the band-aid and just guiding down pretty hard for Q4. I'm sort of wondering if that $90 million fits into that profile. Or do you think this could continue to be challenging into the first half of the year?

So Tore, the downturn has surprised us as it did surprise other people. We really thought when the bookings came out strong in March through May, we were on a rebound. And I think our customers thought that as well. And obviously, the demand didn't show up. And so now they are still stuck with inventories, especially in the appliances and industrial markets. And so that's the reason why this has become a real reset in revenue for us. So in terms of what is the true demand, if we go back to normal demand that is before COVID demand, our run rate should be in the somewhere in the $150 million per quarter. We have done a number of modeling that's what it would say. Why it is taking so long, nobody really knows other than the fact that the economy across the world is not doing so well. The interest rates are very high. So we know it's going to come back. The question is when. And that's what we are trying and struggling to figure out. But we think this is an appropriate thing to do to reset the number. We certainly don't want to grow our inventory any further. And we are optimistic that it will come down a week we're going to have in Q4, and we expect things to start recovering from Q1 onwards.

Great. Just one last question on GaN. Obviously, a bit more longer term. So you talked about the 1,250-volt GaN switch introduced last week. And you did say that obviously, that opens up the door for new applications, obviously, more markets and so on and so forth. When is the earliest something like that could be in production at 1,250 GaN Switch?

The 1,250 GaN switch could be in production sometime next year because it usually takes six months to nine months to go into production. It will be mostly in the industrial market. But when we are able to build products that go to much higher power levels, then we'll get into things like onboard chargers in cars and DC to DC converters in cars and also into data centers. That is something that is two or three years away; we are in the process of developing those products. As you know, everything we do is at a system level. And so it takes some time. However, fundamentally, this technology is very attractive to replace silicon carbide in the 1,200-volt applications. As you may know, there is no silicon solution for that other than IGBTs. IGBTs are not very efficient. But the regular MOSFETs don't go to 1,200 volts. So the only solution you have today is silicon carbide, but I believe GaN will replace silicon carbide and provide much higher performance.

Balu, could you discuss the significant push out from an OEM that affected the third and fourth quarters? Can you provide insight into the extent of that impact as we consider the fourth quarter, and is there a possibility for recovery or rebound if demand increases?

Yes. Let me address that question. I'd be cautious as we want to avoid mentioning any specific customer. In terms of the overall impact, it's around the mid-teens million dollars. Some of this was in Q3, but a bigger portion will be in Q4. However, the reduction in Q4 is also affected by other factors. This is just one of three factors. The additional factors include a slowdown in industrial and consumer demand. The third factor is that we had unexpectedly strong bookings for shipments to distributors in the Chinese cellphone market in Q3, which will not be replicated in Q4 due to unusually low inventory levels. When customers returned with orders, distributors were caught off guard and had to scramble to secure products from us to restore inventory to normal levels. This restocking of the cellphone business is one reason for the overall increase in inventory at this time.

And then maybe just another one here on the GaN, you've released it seems like quite a bit or many products over the last quarter. And it seems like the pace of those products is really picking up and congrats on the new 1250-volt GaN. And we understand the applications for that. But I guess if you think about your customers and what you're hearing. Can you talk maybe a little bit of feedback of what the early thoughts have been with customers?

Yes. I think most of them are very surprised you can actually do GaN at 1,250 volts. Our competitors have a challenge even doing 650 or 700 volts, they have struggled for a while. And our technology is uniquely suited for higher voltages. And so we were not only able to introduce a 900-volt product earlier this year, now we are able to do 1,250. If you're wondering what is the magic about 1,250 volts, that is derated by 80% will be 1,000 volts, which is what is needed in many applications, industrial applications, and that's why we rated it at 1,250. Now the technology is so flexible that we can even go to higher voltages and we plan to do so and offer even higher voltage products. And that's really the first time in the industry that GaN is able to go to these types of voltages. So what is the magic about that is that, the 1,200 volts is quite an important development because when you go to 1,200 volts, you are in the silicon carbide world. And to the power level we can go to, which currently we think we can go to 10 kilowatts, we can replace silicon carbide in many applications. In the longer term, we believe with some breakthroughs in GaN we could go to even higher power levels and potentially become a major competition to silicon carbide but at a very competitive price point but more importantly, much higher performance than silicon carbide. And that's what is exciting about that. That 1,200 volts is a magic voltage where you are really getting into the silicon carbide world.

Sure. Is that the same process that you've used on your lower voltage, lower power? Or was there an architectural change to reach that 1,250 volt?

The fundamental device remains unchanged, but there are numerous innovations that enable us to achieve higher voltages. We are confident that we can push it to even higher levels, and we have made several breakthroughs in that area.

Okay. Great. And just one last one real quick. What do you think the turns business looks like for the $90 million midpoint fare, Sandeep?

Returns for this?

Low 20s.

And your next question comes from the line of Christopher Rolland from Susquehanna.

Just given the kind of crazy inventory dynamics here, I was wondering if you can kind of maybe talk about the snapback when you think it might come broad thoughts on March versus seasonality or even if you could, kind of the slope for next year?

That's a great question. You've been asking about that within ourselves and our customers. And unfortunately, there is no clear answer from our customers. But from everything we know, we are optimistic that Q1 will be higher than Q4, but we don't know by how much. And we are hoping that from that point onwards, it will continue to grow. At some point, it has to come back. This is way below the trend line. I am surprised it's taking so long to clear the inventory, this is obviously because the demand is low or well below where everybody was expecting it to be. The demand has to come back at some point. And I think when it comes back, we'll be in fantastic shape because we are continuing to win a lot of designs in lots of areas, whether it's electric vehicles or GaN-based solutions for industrial applications and so on and appliances with BridgeSwitch with appliances. We have a lot of designs. Some of those designs have been delayed simply because of the inventory. They want to clear out the old products before they introduce new products. So many of these designs will go into production in Q4, and we are optimistic that maybe the second half of Q4 will be the time when it will come out strong. However, if it happens, we are always the first ones to come out of it. I think we'll come out very strongly just like we have done in the past downturns.

Great. And perhaps another one for you. When I look at the kind of GaN market overall, I think the sweet spot is for right now, at least, is the high-volume consumer market. I think you guys were maybe doing $30 million or $40 million there at one point. What's kind of stalled that here other than the macro? And when do you think we can get a meaningful kind of inflection there? And I think that was a part of maybe a gross margin story as well. If you want to talk a little bit more about gross margin, it was a little lighter this quarter and kind of how we should think about it a little further out?

Let me discuss GaN, after which Sandeep will cover gross margin. We are expanding GaN beyond mobile phones, which was our initial focus due to quicker design integration. We're experiencing significant interest and design wins in appliances, industrial sectors, and computers. GaN is set to replace silicon at around 30 watts. Most of our new products utilize GaN because it is superior technology and will strongly compete with silicon within the next year. We're seeing interest in markets I previously thought would not adopt GaN, due to its benefits. These include efficiency, the elimination of the need for heat sinks, very low audio noise, and the ability to manage much higher transient voltages without damage to the transistor even if the breakdown voltage is briefly exceeded. This advantage is particularly relevant for products sold in regions like India with unstable power grids. Therefore, the number of design wins outside mobile products has surpassed those within mobile. As we mentioned in the previous conference call, Sandeep, could you now address gross margin?

Yes. While we are not entirely sure about the revenue outlook for next year, we have conducted various models and estimate a non-GAAP figure around 53.5% for next year. Long term, as I mentioned during Analyst Day, the mix is expected to improve. It’s important to note that the yen is currently at a favorable position, trading in the 140s compared to the normal levels of 120. As the product mix improves and if the yen stabilizes, we anticipate being on the higher end of the model. For 2024, I believe we will also be around 53.5%, which aligns with the upper range of our margin expectations.

And your next question comes from the line of Matt Ramsey from TD Cohen.

I'm trying to understand the commentary you provided about a new normal of approximately $150 million per quarter. I would like to explore how you arrived at that figure. Was it based on historical data from a top-down perspective, or was it determined from a bottom-up approach considering design wins and existing content? I'm comparing this to the peak quarter of around $150 million in 2019 before the pandemic, when circumstances changed and numbers increased. I'm attempting to grasp the elements you combined to reach this new normal, assuming that is indeed what you're predicting.

We have approached modeling in various ways. Finance, marketing, and sales have developed their own models based on several factors. Our analysis includes both bottom-up and top-down approaches. For the bottom-up side, we've analyzed the normal consumption patterns across different markets before COVID and noted that historically, the market's serviceable available market is growing. Additionally, we have new products and designs that will contribute to that growth. While the projections are not exact, it is interesting to note that all three models suggest we should reach about 150 per quarter, excluding the impact of external factors like wars and geopolitical issues. If we disregard those influences, that's the trend line for 2023 with anticipated growth beyond that. Ultimately, this is a modeling exercise that doesn't account for the unpredictable events happening globally.

Thanks, appreciate it. Just a little bit shorter term. Sandeep, could you maybe break down the guidance for December by segment just given the amount of revenue dislocation here? I just want to make sure we're all starting from the right place.

Yes. I think what the decline you should see, as we talked about, the decline will be more significant in communication and computer, but all four segments will decline.

Your next question comes from the line of Ross Seymore from Deutsche Bank.

You talked about the first quarter, you hope is going to go up, and I know it's very difficult to predict these days. But if you're going to end your channel inventory at 10 weeks, and that's still at least one, if not two weeks above your target range, why wouldn't we still have some burn there? And maybe even just seasonality to the extent that matters also being a headwind in 1Q?

Yes, we are anticipating a rebound following the cancellations we saw in Q3 and Q4. I believe this rebound will primarily come from direct sales rather than through the channel. We're also hopeful that by Q4, the channel inventory for appliances will be back to normal levels. If you examine the current state of appliances, their levels are even lower than in 2019. Although we cannot precisely predict the timing, I am optimistic that by Q1, products like air conditioning will begin to pick up, particularly in Q2, alongside the return of customers who canceled their orders. This is why Balu mentioned that we expect improvement. Historically, we see a seasonal decline of about 1% to 2%, but I believe we will perform better than that this time around. Balu referenced these factors as reasons for a more favorable outlook than in the previous quarter.

Prior question asked about the gross margin. What about the OpEx side of things? I know you're pretty tight in the fourth quarter. How do you expect that in 2024?

Yes, this has been quite challenging with inflation and various factors. Over the past four or five years, we've really tightened our budget, but we now have some exciting products coming up. We've announced our investment in GaN for automotive, and we have additional investments planned for India and other regions. It’s crucial for us to make these strategic investments. Therefore, for next year, I anticipate that our expenses will increase by 7% to 8% compared to this year. We've managed to cut our expenses by about $7 million to $8 million this year, but next year we'll see the usual raises and increased healthcare costs, reflecting the current market conditions. While we're doing our best to manage expenses, these investments are necessary. We've faced similar tight situations before, such as in 2011 and 2012, and we continued to invest, which eventually led to recovery. Typically, these downturns last about four quarters; this one may be longer, but they do turn around, and we tend to emerge stronger.

There are no further questions at this time. So I would like to turn the floor back over to Joe Shiffler.

All right. Thank you, Christina, and thanks, everyone, for listening. There will be a replay of this call available on our website, investors.power.com. Thanks again, and good afternoon.

Thank you. This does conclude today's conference call. You may now disconnect. Have a great day.