Please ensure Javascript is enabled for purposes of website accessibility Research in Action: For long-duration investors, semi stocks shine - Janus Henderson Investors
For investors in Singapore

Research in Action: For long-duration investors, semi stocks shine

The semiconductor industry has swung from supply shortages to an inventory glut, creating volatility for the stocks. Research Analyst, Shaon Baqui, offers perspective, saying investors may have to put up with volatility in the near term but that long-term growth drivers and industry rationalization are shaping an attractive future for semis.

Matt Peron

Matt Peron

Global Head of Solutions

Shaon Baqui

Shaon Baqui

Research Analyst

Oct 10, 2022
33 minute listen

Key takeaways:

  • Semiconductor stocks are experiencing their worst correction since the Global Financial Crisis as a result of shifting consumer demand, macroeconomic headwinds and trade disruptions.
  • But semi stocks historically have delivered sharp rebounds following corrections of this magnitude, and are underpinned by powerful secular growth drivers as the global economy digitizes.
  • The industry has also become more rational, with management teams focused on smoothing out inventory cycles, growing free cash flow, and returning capital to shareholders.

Technology industries can be significantly affected by obsolescence of existing technology, short product cycles, falling prices and profits, competition from new market entrants, and general economic conditions. A concentrated investment in a single industry could be more volatile than the performance of less concentrated investments and the market as a whole.

SOX (PHLX Semiconductor Sector IndexSM) is a Philadelphia Stock Exchange capitalization-weighted index composed of the 30 largest U.S. companies primarily involved in the design, distribution, manufacture, and sale of semiconductors.

Carolyn Bigda: From Janus Henderson Investors, this is Research in Action. A podcast series that gives investors a behind-the-scenes look at the research and analysis used to shape our understanding of markets and inform investment decisions.

CPU, GPU, DRAM and NAND. You guessed it, today we’re talking about semiconductors. Research Analyst Shaon Baqui covers this dynamic high-tech industry and joins the show to discuss what, this year, has become an increasingly difficult market to make sense of. On the one hand, with the global economy going digital semis’ long-term growth prospects appear stronger than ever. On the other hand, in the post-pandemic period, demand for chips is pulling back sharply, leading some companies to scale back investment in future growth.

Shaon Baqui: I’d say that, yes, there is some near-term volatility potentially ahead, but I think over the long run, semiconductors continue to offer very compelling, sort of multiyear opportunities relative to the broader market.

Bigda: I’m Carolyn Bigda.

Matt Peron: And I’m Matt Peron, Director of Research.

Bigda: That’s today on Research in Action.

Shaon, welcome to the podcast.

Baqui: Thanks for having me, great to be here.

Bigda: Forecasts for global chip sales in both 2022 and 2023 are being revised lower even though demand for computing power is projected to rise dramatically over the next decade. What’s going on, Shaon?

Baqui: You’re absolutely correct. I think the overall trend towards digitization and rising compute requirements, it’s only intensifying. So, if you look out over the next 10 years, the sheer amount of data that we consume is only going to increase exponentially. So, that’s really being driven by these big secular drivers. We’re talking like cloud computing, AI [artificial intelligence], 5G, factory automation, EVs [electric vehicles], autonomous vehicles. So, the list goes on, all your favorite buzzwords. But I think in the near to medium term, yes, we are seeing a few dynamics that are bubbling up under the surface.

So, I think first and foremost, the industry’s coming off a very large pull-forward to digital demand, as we shifted from work-from-home, learn-from-home, play-from-home during the pandemic. So, this turned out to be a real boom for everything from PCs to tablets, to TVs, to wi-fi routers. And in many cases, like myself, I went out and bought all four while I was stuck at home. So, now, we’re starting to digest some of those pull-forward effects, and really seeing kind of a pronounced shift in consumer spending patterns from goods to services, like travel.

And then secondly, and probably more importantly, the world’s really changed over the last six months. We’ve seen this confluence of macro[economic] factors come together. I mean, we’re talking rising inflation, slowing economic growth in Europe and lockdowns in China that have had a pretty profound negative impact on demand for semiconductors from both businesses and consumers.

And then third, there’s still this pesky issue of the semiconductor shortage, right? As anyone who’s tried to buy a car in the last year can attest, this is still a huge problem with lengthy wait times. So, while chip supply, yes, has broadly improved, there’s still pockets of areas like microcontrollers, logic, analog, and power where lead times still exceed 52 weeks or longer.

Bigda: Okay, so, it sounds like there’s a little bit of a pandemic hangover, there’s macroeconomic issues, and supply chain issues that are still going on. So, lots of different crosscurrents. Maybe let’s talk a little bit about where we’re seeing that oversupply and which semi suppliers are most impacted at the moment.

Baqui: It’s really interesting, we’ve seen this bifurcation of end demand between sort of these enterprise/business-to-business end markets, like cloud [computing], automotive, industrial and telecom, which has, frankly, been very resilient year to date; and then there’s the consumer end markets where things have deteriorated and [are] falling a little bit faster.

So, I guess peeling back the onion a little bit, we’ve probably seen the fastest and frankly, the most pronounced impact in PCs and smartphones, which collectively, that’s roughly half of the overall semiconductor market. So, starting right around June [2022], we saw demand for PCs and smartphones fall off a cliff, where you had this perfect storm of, number one, rising fuel prices, lockdowns in China and the fallout in Europe from the Russia/Ukraine conflict. And on top of that, as I mentioned earlier, you’re coming off of one to two years of above-trend growth in PCs. I think stepping back, pre-pandemic, you’re looking at around 260 to 270 million PCs shipped per year, on average. As the world pivoted to remote work and e-learning, that number spiked at 300 million in 2020, and 350 million in 2021. So, naturally some digestion is taking place now as well.

You know, on the smartphone side, we’re seeing similar impacts. So, a lot of folks went out and spent stimulus checks on upgrading to 5G smartphones over the last couple of years. So, that’s leaving us with a little bit of a hold this year as well.

So, kind of a long-winded way of saying that, you know, almost overnight, we went from a world where the PC supply chain and China smartphone makers were stockpiling components in anticipation of another year of robust growth to, frankly, being left with a glut of both finished goods and component inventories. So, basically, we went from a supply shortage to supply glut in just a matter of months. And to be honest, it’s probably going to take until early to middle of 2023 before the supply chain can work down all that excess inventory of things like CPUs, memory, storage, et cetera.

So, that’s what’s going on, on the negative side. So, enough doom and gloom. You’re probably going to ask me next, so what about the rest of the semi industry? What about the other half of the market? And I think over here, things have been remarkably resilient. The secular digital tailwinds I talked about earlier have really overcome those cyclical macro headwinds. I talked about cloud, auto, industrial and telecom. So, let me flesh through each of those individually because they each have their own attributes.

Starting on cloud, what we’ve seen is this continued shift from on premise data centers to cloud, as well as record investments in AI. So, that’s driving kind of massive growth in cloud capex [capital expenditure] from the big, U.S. four hyper-scalers. Very strong demand for things like servers, storage, networking that consume a ton of semis.

Bigda: And that’s even with inflation pressures, prices going up – demand for that is still robust.

Baqui: If you think about, if you’re facing some uncertainties in business, that’s only going to accelerate your pivot to the cloud versus investing in your own IT infrastructure in-house. So, definitely, we see that continuing.

Secondly, automotive, right? So, production hasn’t been great this year. So, the industry continues to grapple with the chip shortage. But underneath the surface, we’ve actually seen a couple of interesting dynamics with respect to semiconductor content per vehicle. First, this sort of scarcity of chips, that’s driven car makers to pivot their production to the highest end. So, these are models that consume a ton of semis. And then secondly, rising fuel prices have really accelerated this shift to EVs, where, on average, these EVs consume roughly 2x to 3x the content of an internal combustion engine vehicle. You layer on some pretty attractive EV subsidies through the Inflation Reduction Act; that’s a trend we don’t see slowing down anytime soon, right?

Industrial: Again, this area is a little more fragmented. What we’re seeing is digitization in areas like factory automation and robotics, which are frankly deflationary when the rest of the world is dealing with huge labor shortages and wage inflation.

And then fourth and finally, is telecom. We’ve seen global telecom providers really in a rush to build out these 5G networks and deploy all that spectrum they’ve put in place in the last couple of years. So, a nice tailwind for semis to process, move and store all that 5G data that’s coming down the pipe.

Peron: All right, Shaon, so thanks for that deep dive. Maybe we can back up and talk about the cycle: near-term cycle and then the super-cycle, if I’m allowed to use that word. The inventory cycle that you said is unfolding, we have to work through that. But then, you talked about the intensity of semiconductor usage in things like EVs, the content per vehicles. Those are massive changes. So, there’s a case to be made for a super-cycle – a boom, if you will – on the other side, and then layer on top of that the idea to onshore that, the recent CHIPS Act. Can you put this all into context for us, near term and long term?

Baqui: Sure. So, let’s start with the CHIPS Act because I think that’s been front and center in the media and certainly garnered the most attention. I think for the listeners that might have missed it, so, back in July, Congress passed the U.S. CHIPS Act, which is, it’s a nice step towards bringing semiconductor manufacturing back home. The majority of manufacturing is done in Asia right now. And what’s interesting is this CHIPS Act includes roughly $52 billion of subsidies for semiconductor manufacturing here in the U.S., and an additional $24 billion in tax credits for semi fabs built here. And I think the EU has put something similar in place for around 43 billion euros. But the end goal is the same: to build out this sort of local supply chain and reduce geopolitical uncertainty associated with Asia. We’ve already seen a number of big projects get announced in Arizona, Texas and Ohio, as well as in Germany. Several of these have actually broken ground, and if you go through Phoenix [in Arizona], you can start to see the shells up there. So, these guys will be ready to start taking equipment late next year.

But again, to your point, these are very strategic multiyear investments that are going to happen irrespective of demand conditions over the next six to 12 months. And I think more importantly, I think because equipment lead times are really long – in some cases, more than 12 months to get a tool delivered to you. So, we’re seeing really strong investment in advanced process nodes, think of your 5 nanometer and below. If you’re a leading-edge foundry, you can’t afford to put your manufacturing roadmap on hold just because of some short-term gyrations in demand.

Now, to your original question, yes, there is excess inventory in some parts of the market, right? Namely, in memory. And here, we’ve seen a pretty substantial glut, and the only way for these memory makers to get supply back in line with demand is to slash capex. We’ve already heard from a handful of players that are saying they’re planning on doing just that. Hopefully, that helps the industry get into a better place in a few quarters, but that will take some time to work through. And then elsewhere, I alluded to the consumer weakness, but we’re hearing of some tier-two foundries in Asia that are exposed to consumer semis, and naturally, they’re going to be pulling back on capex a little bit. So, bottom line, yes, there are commodity parts of the market where capacity investment is slowing naturally, given this demand slowdown and rising inventory. But at the same time, these big strategic investments around leading-edge process technology and reshoring of semi manufacturing, I think that’s going to continue no matter what, and that should drive a base level of capex for the industry for a number of years moving forward.

Bigda: At the same time, though, I think there is an issue with some regulations that are coming out in terms of licensing requirements for shipping products or manufacturing overseas. Can you talk a little bit about that, and does that put a lot of pressure on semi manufacturers right now?

Baqui: Stepping back, so concomitant with this focus on U.S. manufacturing, we’ve also seen, definitely, a more hawkish stance from the government in terms of restricting China’s access to things like advanced AI chips, leading-edge manufacturing equipment, and even electronic design automation software, or EDA software, that could potentially be used in military applications.

So, to date, this has been a slow drip. I mean, it started a few months ago with restrictions on shipping tools capable of producing chips at 14 nanometer and below. And then we saw increased restrictions on EDA software for next-gen gate-all-around transistors. And then most recently, probably the most publicized, was this latest round of restrictions on shipping advanced GPUs or AI chips that can be used for military purposes. I guess stepping back, the financial impact here ranges from de minimis in the case of EDA today, to less than 1% for the equipment suppliers and to roughly $400 million for the GPU market, per quarter. So, the fallout’s been relatively manageable. I think the bigger concern is really around what’s next. And that sort of got the market spooked a little bit. Frankly, it adds another layer of uncertainty at a time where, really, it’s unwelcome given we’re already on pins and needles with respect to the economy.

Bigda: So, there’s concern that there could be even more regulation coming down.

Baqui: Absolutely, yes.

Bigda: Okay, so, let’s talk about the market and the market impact because there has been, as we talked about, a lot of weakness in semiconductor stocks in recent months. And so how is the market digesting all of these different facets right now. And how do we think about this in the long-term context of investing in the semiconductor opportunity that seems to still be there with all this new technology?

Baqui: Yes, so again, we’ve already seen certain parts of the semiconductor industry where demand has deteriorated significantly. And it will take multiple quarters to recover, right? But I think if we do enter into a deeper recession, there’s certainly no end market that’s going to be immune. Could semiconductor units and revenues take a breather next year and be down? Absolutely. So, we’ve been through cycles before, and we’ve seen this before.

I think more importantly, if you step back and just look at this long-term trend line, it continues to be on this positive trajectory. In the past, the industry has grown, let’s call it, mid-single digits, on average. So, nicely above GDP [gross domestic product], right? And then, now, just given the sheer amount of data being generated and broader digitization coupled with more stable pricing, you’re potentially looking at an industry that could grow high-single digits over the long run, versus kind of mid-single digits in the past. Again, that’s more than 2x global GDP. So, you’re looking at better growth, better margins and higher free cash flow than the broader market, all at a lower multiple than the S&P 500® [Index]. So, yes, I’d say that, yes, there is some near-term volatility potentially ahead, but I think over the long run, semiconductors continue to offer very compelling, sort of multiyear opportunities relative to the broader market.

Bigda: You mentioned multiples, and that was going to be my next question, which is about valuation. And do the stocks now have valuations that are uniquely attractive for the industry?

Baqui: Sure, so valuation is a little more nuanced, we know the market discounts uncertainty more than bad news, and I think we’re seeing that play out in real time. We’ve heard some call this the most anticipated downturn in the history of semis. I mean, the SOX [PHLX Semiconductor Sector IndexSM] year to date is down more than 40% from the peak and, multiples have compressed around 35%. This makes it sort of the most severe correction since the Global Financial Crisis. And half the end market has not even been impacted thus far. From where we sit today, the SOX index, the semi index, trades at roughly 14x next year’s earnings, which is nicely below the historical average of 16 and down significantly from the peak of 21 last November, but still about 15% above the trough, which incidentally we did test back in June.1 So yes, valuations reflect a lot of uncertainty, like you described.

I think what’s more interesting within semis is we do see a number of high-quality franchises that are trading at or near five-year trough multiples. So, we’re trying to identify these businesses where the market has really overshot fundamentals to the downside, which creates this asymmetric risk/reward profile if you have a little bit of duration on the other side.

Long answer short: Yes, we are acutely aware of all the near- to medium-term risk you referenced earlier around supply/demand, macro, trade, and to be honest, trying to call a hard bottom is difficult. But I think the one thing the team here has learned from investing in this space for more than a decade, that anytime valuations get as extreme as they were back in June, they can snap back just as quickly on the other side of the cycle. So, we need to be mindful of that and not get too one-sided, right? And try to identify those one-offs where there is asymmetric risk/return. We’re starting to see that a little bit.

Bigda: Why has this pullback been so severe or as severe as the Global Financial Crisis? What happened? Did the stocks overshoot so much to the upside coming into this correction?

Baqui: Yes, I think to your point, part of it is market driven, right? So, market multiples were certainly elevated, so we’re starting from a higher place. And then you start talking about all the other things we talked about. Was there a certain amount of pull-forward from the pandemic? Certainly, so, I think that the market is still trying to grapple with, what is the steady state growth of this market? Number one. And number two, how long is the inventory correction going to last on the other side of the cycle, and then you layer on trade as well. So, just generally speaking, a lot of uncertainty with respect to macro, the trade and the fact that we started from a higher base. I think that we’re definitely seeing the other side of that right now.

Bigda: Matt, I’m going to let you take the next question because we always talk about innovation on this show, and I wanted to hear what you might be wondering about in terms of like where the next big breakthroughs are going to come with semiconductors, or the next innovation. What are you curious about these days?

Peron: I think there’s a really big story that we should talk about. I mentioned the word super-cycle earlier on, which semiconductor analysts are not allowed to say.

Bigda: Why is that?

Peron: Because they’ve gotten into trouble saying that before, so I’ll say it for him [Shaon]. But I really think this story of – and Shaon can walk us through the exciting innovations happening here – but this is really a locomotive that I don’t know that everyone appreciates when they can see the amount of innovation in the sector. When I look across all of the sectors, it’s got to be one of the most innovative sectors, and we’ve had a number of really great ones on this show. So, I guess, Shaon, the question for you, maybe you can walk through some of the innovations you’re excited about that could carry this, dare I say, super-cycle through the next decade.

Baqui: Yes, so I’m not going to use the word super-cycle again, to your point, but generally speaking, we continue to see significant long-term, I’ll say secular mega trends, in areas like artificial intelligence, vehicle electrification and Internet of Things. So, those are the three themes we’re focused on. And we’re also really excited about semi cap equipment [semiconductor capital equipment], which is slightly different from the other three.

So, I guess, first on AI. I think what we’re seeing is use cases going far and beyond just identifying pictures of dogs and cats on the Internet. So, we really see AI at this tipping point where we have vertical industries really start to realize the economic benefit that AI brings, and monetize that into real cost savings. In other words, companies using AI and semiconductors to play offense, rather than in the past where compute was a back-office cost center; you’d throw a server rack in your office and forget about it. So, I think what you’re seeing now is this strong adoption of AI by vertical industries. So, whether it’s applications like genomics or drug discovery in health care, seismic exploration in oil and gas, fraud detection in banks, and pricing algorithms in retail. These monetizable use cases continue to grow for AI, right?

And I think what’s even more interesting is that the number of parameters that trade in these models is growing exponentially faster than compute. Think about it this way, the number of parameters in the model is growing roughly 3x to 4x every two months. Okay, the compute performance of these chips is only growing 10x every two years. So, that gap between these large language parameter models versus compute powers is only widening. What does that mean? That means you need more chips, more GPUs, more CPUs, and custom ASICs [application-specific integrated circuits] to train these massive transformer models. So, we’re really positive on AI. We think we’re still early innings there. We see a long runway of growth ahead as we start to unlock the benefits there.

Vehicle electrification. We know the shift to EVs is happening now. The train has sort of left the station here. And with EVs, you really see a big inflection in the semi content, on the order of 2x to 3x that of an internal combustion engine vehicle. You have new applications here like onboard charging, the main traction for your wheels and your motors, battery management systems and domain controls, right? And then you layer on these increasing government incentives. For example, the most recent Inflation Reduction Act gives you up to $10K per vehicle in incentives. So, you put all that together, we think that EV penetration can hit mid-teens percentage of vehicle production by the middle of the decade and potentially 30% by the end of the decade. So, a huge windfall for these automotive suppliers that do things like power semis and things of that nature. And then you layer on these additional content gains, with level 2-plus and level 3 autonomous vehicle function, and that’s another multiplier effect for things like processors, image sensors, lidars, radars. So, really strong growth for automotive semis for the next 10 years.

And then on industrial, right? So, whether or not you want to call it IoT, smart factory, industry 4.0; a lot of cool buzzwords here, but really what’s going on in industrial, there’s a lot of tailwinds in factory automation, robotics and edge computing, which really feed into that sweet spot for analog and embedded semiconductors. Historically, not that exciting of a space, but now when you think about all that sensor data that’s collected on a factory floor that needs to be processed and secured in real time, I think there’s a very compelling investment case there, as well. I think we’ve seen third-party data show that the number of connected endpoints is going to grow to 50 billion by 2025 versus roughly 20 billion today. And then you layer in labor shortages, wage inflation and reshoring of manufacturing. We think that just accelerates the demand for industrial semis over the long term.

So, there’s AI, there’s vehicle electrification and industrial. But I want to make sure we don’t forget about semi cap equipment because that sort of ties together those three other themes.

Bigda: And semi cap equipment, just to interrupt briefly, what is that?

Baqui: Right. So, these are essentially the picks and shovels for the semiconductor industry, right? So, these are the tools that are used to manufacture a chip, and basically, you can’t have cloud, you can’t have AI, you can’t have 5G without these tools you just can’t. And what’s great about this industry, it’s basically an oligopoly structure where five companies own, let’s call it, 75% share of the market. And each of these businesses has its own swim lane with really high barriers to entry. And the basic premise here is look, manufacturing chips is getting tougher. Chips are getting bigger. They’re getting more complex. So, this old playbook of just simply doubling the number of transistors every two years via Moore’s law is getting more and more difficult; it’s not as simple as it used to be. You need more steps to build a chip. That means more tools, right? Longer lead times, longer cycle times, excuse me, and more clean room space. So, it’s getting tougher, and you also need new innovation. We’ve seen things like extreme ultraviolet lithography, which I could spend all day talking about – I don’t want to bore you guys. 3D NAND, advanced packaging, gate all-around-transistors, I mean, you name it. All these things drive another layer of inflection on top of that base spend. And then you layer on this emerging growth factor, from utilization of semi manufacturing and all these massive fabs that are coming up in Arizona, Ohio, Texas, Europe. So, long story short, we’ve seen this big step-up in capital intensity, which historically has been around, let’s call it, 10% to 11% of semiconductor industry revenues, to about 15% today. So big numbers.

So, yes, we could see some near-term noise around capex cuts and the China restrictions you talked about earlier. But I think, overall, demand for equipment continues to be up and to the right. And to be honest, these are great businesses and super critical to the digital economy. So, a lot to like here over the long run.

Peron: And if I could give a little plug for the team, people tend to focus on semiconductors and all the GPUs and all the things that Shaon talked about, but the team has really been on this trend of no matter what happens, you need innovation in semi cap equipment. It’s the key, fulcrum of the whole industry. And our team has identified that early on, and really, it’s been a big part of our investment ecosystem, so to speak. Semi cap equipment gets lost because it’s very technical. It’s in the weeds. It’s not front and center, but it’s so important to the industry. As Shaon noted, without the innovation there, you can’t get down to these process nodes, you can’t get the gate-all-arounds. You can’t get the improved productivity that we all depend on.

Bigda: Right. So, it sounds like it’s the important foundation of the industry.

Peron: Exactly right.

Bigda: Which sounds like it’s experiencing increased demand, not just for compute power, but also volume. And so, I’m just wondering, does increasing demand, coming from all different sides, does that help potentially mitigate the cyclicality of this industry, which tends to go through ups and downs, historically?

Baqui: Yes, so, coming back to Matt’s earlier comment about the super-cycle…

Peron: You’re not allowed to use that.

Baqui: …anytime you hear someone say that the semiconductor industry is no longer cyclical, I think it’s always important to step back and think. I think, generally speaking, yes, semiconductors will always be cyclical, just given the sort of inherent supply/demand/macro component of it. But I think that, to your point, because of all the drivers you mentioned, I think the amplitudes of the cycles are going to continue to dampen going forward. Think higher highs and higher lows.

So, why is that? There’s a number of reasons here. Long story short, you’ve got a space that’s significantly consolidated. So, fewer players that are much more rational in terms of their investments. And you’ve got end markets that are significantly more diversified. So, in the past –I’m going back like 20-plus years – you’d have dozens of companies chasing demand. They’d all build up a ton of capacity at the same time. And then, when the cycle did roll over, you’d take a massive gross margin hit and EPS [earnings per share] hit from number one, factory under-absorption because you’d have all this capacity that needs to be absorbed. And then you’d also have price declines because you’d had too much supply in the market.

So, I think what you see now is this pivot to fabless, fab-light manufacturing. And that really diminishes the impact of a downturn on the other side because it’s not the semi companies taking the supply risk. And then moreover, in the past 20 years, you’d get this massive PC cycle followed by a downturn on the other side. PCs were the biggest driver of semi consumption in the past. Now, it’s below 20% of the overall semiconductor market. And at the same time, you’ve got much more diversified demand drivers.

Bigda: Right, it’s not just the smartphones and the PCs anymore. It’s now much more diversified.

Baqui: So much less sort of boom-bust now that you don’t have to worry about timing that PC cycle every couple of years. I think bottom line, yes, semis are cyclical, and they will continue to be cyclical. But at the end of the day, you’ve got a group that has higher revenue growth, better margins, and superior cash generation and capital allocation versus the broader market, but all that comes at a discount to the S&P 500.

Bigda: I just want to clarify one thing: You said about fabless or fab-light. So, what does that mean? Because it sounded like it was important in terms of the way that industry operates.

Baqui: Absolutely. So, in the past, the majority of companies, they did manufacture their own chips; they were vertically integrated. But what we’ve seen over the last 20 years, we’ve talked about rising capital intensity. So, the follow to that is that very few companies can actually afford to stay on the Moore’s law curve and manufacture their own chips. So, they’re increasingly turning to third-party foundries to do their manufacturing. So, they’re taking less risk in-house with respect to supply, working capital and capacity, and relying on the foundries. So, what that does is, they can focus on doing what they do best: designing chips rather than manufacturing them. So, they’re not taking the supply risk on the other side. So, definitely seeing more rational investment in the industry, which has helped dampen that cyclicality.

Peron: Shaon, you did touch on it, maybe it is important to bring out, the capital allocation discipline, the cycle awareness that managements really are laser-focused on now, that is also a little bit different. They’re just, it seems to me, they’re much more adept at managing the cycle. Would you agree with that?

Baqui: Yes, absolutely. For example, I’ll use DRAM; that’s probably the most cyclical part of the semiconductor industry. There used to be 10-plus suppliers of DRAM. Today, we have three, and they all know what they’re doing, and they all understand that when times are tough and there’s oversupply, we need to pair back in capex, right? In the past, it was sort of a prisoners’ dilemma: Who’s going to pull back on capex first? One guy would pull back, and the other guy would go pedal to the metal and use that time to gain share. So, I think we’re seeing much more rational investment in the space, and that’s really helped, number one, just dampen the cyclicality. And number two is that because these guys are generating such outsized cash, we are seeing a number of these companies return 100% of it via dividends and [share] buybacks. And that’s very compelling, right?

Bigda: So, it sounds like the industry is maturing, getting a little bit more rational, and yet, it still has a long runway of growth ahead of it, which sounds like a pretty good combination to me.

Baqui: Absolutely. Yes, we could see some volatility in the next three to six months, but I think that, the group offers definitely some very compelling long-term risk/reward potential for those of us with duration and ability to withstand some near-term volatility.

Bigda: Shaon, thank you so much for joining us today. We’ve really appreciated your insights.

Baqui: Absolutely, my pleasure.

Bigda: And I think on our next episode, Matt, we have Caroline Escobar joining us. She’s a Research Analyst here who covers apparel and retail, and she’s going to be talking about all of this just before the holidays, which I think will be great timing.

Peron: Yeah, that’ll be fun. And by the way, to connect it to today’s conversation, you’ll also cover wearable watches and things like that. So, more and more semiconductors all over the place.

Bigda: Put it on the holiday shopping list, I think.

I’m Carolyn Bigda.

Peron: I’m Matt Peron.

Bigda: This has been Research in Action.

1 Source: Bloomberg. Price-to-earnings figures based on daily, two-year rolling average from 31 December 2015 through 19 September 2022. Performance figures as of 19 September 2022.


Past performance is not a guide to future performance. The value of an investment and the income from it can fall as well as rise and you may not get back the amount originally invested.
The information in this article does not qualify as an investment recommendation.
For promotional purposes.
Anything non-factual in nature is an opinion of the author(s), and opinions are meant as an illustration of broader themes, are not an indication of trading intent, and are subject to change at any time due to changes in market or economic conditions. It is not intended to indicate or imply that any illustration/example mentioned is now or was ever held in any portfolio. No forecasts can be guaranteed and there is no guarantee that the information supplied is complete or timely, nor are there any warranties with regard to the results obtained from its us.
Matt Peron

Matt Peron

Global Head of Solutions

Shaon Baqui

Shaon Baqui

Research Analyst

Oct 10, 2022
33 minute listen

Key takeaways:

  • Semiconductor stocks are experiencing their worst correction since the Global Financial Crisis as a result of shifting consumer demand, macroeconomic headwinds and trade disruptions.
  • But semi stocks historically have delivered sharp rebounds following corrections of this magnitude, and are underpinned by powerful secular growth drivers as the global economy digitizes.
  • The industry has also become more rational, with management teams focused on smoothing out inventory cycles, growing free cash flow, and returning capital to shareholders.