Investing in Semiconductors 2022 - Part 2

Published on 09/24/22 | Saurav Sen | 7,250 Words

The BuyGist:

  • This is a follow up to Semi-conducting Rationality.
  • In this analysis, we do a detailed scenario analysis (in light of elevated risks from China) for most of the major semiconductor companies. 
  • Some of these companies are in our portfolio. 
  • We conclude with a definitive "sell price" for each company based on the scenario analysis, with definitive portfolio actions.

Semi…radioactive

We’re revisiting our semiconductor exposure because the market has forced us to. And our subscribers deserve another deep dive during these turbulent times. It seems to us that there is just no good news out there for equity investors in general, and we’re not alone in feeling that way. This feeling is amplified in the semiconductor sector. 

A heady mix of uncertain geopolitics, tempered growth expectations, increased risk aversion, higher interest rates, and the ever-increasing possibility of a global recession have hit the semiconductor sector hard. Take a look at the year-to-date returns of some of the big guns of the industry: 


We did a deep dive on semiconductors a couple of months ago. But a couple of things have happened since then: 

  1. The Biden administration took an aggressive stance on the sector by imposing restrictions on American semiconductor exports to China. 
  2. Nvidia’s latest quarterly results surprised everyone – the big sequential decline in Gaming chip sales was a reality check for a company that seemed to be invincible just a few months ago.
  3. Central Banks around the world have taken a more aggressive stance – inflation seems to be more persistent than they previously expected, pushing them to consider risking economic recession to tame the beast. 

This potent cocktail of negative factors was enough for us to take another look at our semiconductor portfolio of 6 companies. The goal is do our work now and sit tight until the world is more normal, with a fair distribution of good and bad news. We just want to sleep well at night. In the sections below, we’ll dig into each company’s fundamentals one by one. We’ll also dig into Intel’s fundamentals even though it’s not in our portfolio. 

Hint: We’re still not buying Intel.

The dragon in the room

Our decisive bet on semiconductors – since mid 2018 – has been good for our portfolio. There have been periods of volatility, but we’ve been stoic enough to ride out the storm. That has served us well. We’ve made more than 30% in returns in each of our semiconductor bets, including Micron (which we don’t hold anymore). This year is another volatile period. We should summon our stoicism once more. But we get the sense that there’s more at stake this time.

The China Factor is the one that really keeps us up at night. There are 3 scenarios that will play out in the next few years. Let’s call them the “Good, Bad and the Ugly” (GBU) scenarios: 

  1. Good Scenario: Things go back to normal – China and US come to some sort of a trade and IP agreement which benefits both countries. 
  2. Bad Scenario: China pursues its own ambition of rivaling the US on semiconductors, especially Logic Chips, thereby fanning the “tech cold war” for the next decade and beyond. In this scenario US semiconductor sales to China will decrease.
  3. Ugly Scenario: China takes a page from the Putin Handbook and invades Taiwan – for some misplaced sense of pride, and to nationalize the crown jewel of Taiwan: TSMC

Obviously, we want GBU scenario #1 to play out. But we must be prepared for scenarios #2 and #3. It would be awesome if there was (ironically) an AI program that would spit out probabilities for each scenario. But all we have is subjective judgement. There is no way to accurately quantify this; we can try, and we will.

Dracarys!

Does anyone like the new Game of Thrones spinoff? Sorry, we digress…Why are we still invested in in this radioactive section anyway? Better question. 

AI or Artificial Intelligence is the primary reason we’re still invested in our 6 core semiconductor companies. These little chips are the engines of AI. They are incredibly difficult to design. They’re even harder to manufacture. Companies that can design or manufacture these chips – Logic Chips like CPUs and GPUs in particular – have huge economic moats. It’s very difficult for a new competitor to break in, let alone steal market share. As far as we can see, market share will shift back and forth between the handful of industry stalwarts that dominate the industry today. They’ve been widening their economic moats over the last 5-7 years. And they’re not resting on their laurels.

Here are a couple of facts that demonstrate just how hard it is to make Logic Chips. 

  1. There are only 4 major Logic Semiconductor companies that design chips for desktops and cloud servers: Intel, AMD (including Xilinx, which merged with AMD), Nvidia, and now Apple (for their desktops and iPhones only). Google designs its own TPU (Tensor Processing Unit) for running AI software on its cloud servers. Amazon and Microsoft have their own versions of AI chips. But TPUs and the others are not widespread.
  2. There are only 3 companies in the world that can manufacture Logic Semiconductors at any respectable scale: Intel, TSMC, and Samsung. Intel, so far, has manufactured only its own chips. TSMC and Samsung make up the big global duopoly that manufactures for other stalwarts like Nvidia, AMD, Apple etc.

China has ambitions to break into this mostly American domination of the Logic Semis space. This is no secret. But if China wants to be a major force in AI, they must design and manufacture Logic Semis at home. They don’t have the technology to do either – they’re at least a few years away from having a homegrown CPU that can rival Intel’s or AMD’s. But China doesn’t want to rely on the US for cutting-edge chips forever – for geopolitical reasons and things like national pride. 

Here’s the scary part (for investors): They know that TSMC – based in Taiwan – manufactures Logic Semis for most American stalwarts, including AMD, Nvidia, and Apple. China can spoil the American semiconductor party if they want. The unknown factor is, “at what cost?”

Back to our China scenarios – how do we position the portfolio for scenarios #2 (Bad) and #3 (Ugly)?

  1. For #2 – we’re not overtly concerned right now. We believe that China is years away from offering up a Logic Semiconductor that can rival those of Intel, AMD or Nvidia. We have time and, more importantly, these American stalwarts have time to prepare for this. They won’t stop innovating to wait for China to catch up. But yes, their sales to China could dry up, say, 5 years later.
  2. #3 is the nightmare scenario. If China invades Taiwan, it’s not just semiconductor stocks that will get hit hard but global equity markets in general. If this scenario were to come to pass, it’s best to stay invested in companies that have no negligible exposure to China or Taiwan. Well, if China invades Taiwan, it’s best to not be invested in anything. But we can’t operate in investing (and in life) by attaching a 100% probability to the direst scenarios. We’ll never end up leaving our homes.

In our last deep dive on semiconductors, we had collected data on each company’s China & Taiwan exposure. We don’t think this has changed much since then. 


For most companies, this chart is still relevant…and ominous if Scenario #3 came to pass. In Scenario #2, it would matter how much these companies sold to China vs. Taiwan. For at least 2 companies analyzed below, this split makes a difference in the decision about hold/sell. More on that later. 

And in that same analysis from July 20th, 2022, this was our conclusion: 

“Assuming China creates havoc in Taiwan in the next 5 years, we believe that ASML and KLA are safer bets than AMD, Nvidia and TSMC for reasons we’ve mentioned before. That’s a qualitative assessment. In short, we’ll be forced to move away from the semiconductor space if there is another bull-run in the markets in the next 12 months. However, if Intel's stock reaches $30, we’ll do another deep dive with the new facts/numbers we have at that point. Otherwise, we’ll hold on to the rest of our positions until they reach the prices mentioned above OR if we find much better sleep-well-at-night alternatives. We’re always on the lookout.”

So, a few things have happened since then:

  1. Semiconductor stocks have NOT rallied. They’ve dipped even further – for reasons mentioned earlier.
  2. ASML now looks safer than KLA after the latest Biden trade restrictions. But that’s just on the surface.
  3. Intel’s stock HAS dipped below $30/share. Still looks dodgy.
  4. There are some other names that have popped into our radar – from our Watch List.

So, there is much to chew on and analyze. We’ll structure the rest of this analysis on the Good, the Bad, and the Ugly scenarios. 

But first, let’s set the stage.

Rules of the game

We don’t want to waste your time by hedging our language and talking in vague generalities. One of our core values at The Buylyst is Actionability. To end this analysis with definitive action items, we need to draw some definitive lines in the sand for this analysis.

We drew lines in the sand in our last analysis as well. We specified definitive selling prices for our semiconductor holdings. In fact, we had started selling some of our AMD position when the stock approached $100 a few weeks ago, but then there was another turbulent couple of weeks which brought the stock price back down – to about $70 in a matter of weeks. So, we still hold some – maybe we need to top up back to our original portfolio allocation of 3%. Anyway, we’ll get to that. But first, this was our action plan in our previous analysis: 


The goal of this analysis is to update this chart – with the new GBU sceanrios we described earlier. As a recap here they are: 

  1. Good scenario: We’re back to a normal world when there are no major wars, and the US and China are on reasonably good trade terms. 
  2. Bad scenario: US and China go their separate ways in terms of semiconductors – the engines of AI. They barely trade with each other in this specific domain. It’s a legitimate Tech Cold War.
  3. Ugly scenario: US invades Taiwan. This will be painful. 

Keeping these scenarios in mind, we will adjust our Approximate Selling Prices (ASP). If our new ASPs are lower than current prices, we will start selling. To get to our new ASPs, we need to draw some more lines in the sand. 

Here’s what we need: 

  1. Specific probabilities for each scenario: the Good, the Bad the Ugly. 
  2. Specific ASPs for each company in each scenario: the Good, the Bad, the Ugly. 

Here’s our first definitive marker. Our SUBECTIVE assessment for probabilities in each of the GBUs are: 

  1. Good: 50%
  2. Bad: 30%
  3. Ugly: 20%

One can play around with these probabilities until the cows come home. One can even get fancy and superimpose some Monte Carlo simulations. But we don’t want to get caught up in analysis paralysis. To calculate ASPs, we will need to put down some more definitive markers. This is a departure from our normal Buycasting process – where we back solve instead of forecasting – so it deserves its own section. 

Expected Values

As the title suggests, this section is a little nerdy in a somewhat nauseating finance-y way. So, if you’re not interested in these details, we don’t blame you – please feel free to skip to this section and move on to our company-specific results. This section sets up a few reasonable broadlines within which we’ll estimate our new ASPs. As a reminder, here are the old ASPs (Approximate Sell Prices) from the July 20th analysis


Now, we need to superimpose the new GBU scenarios to adjust these ASPs. Again, to do that, we need to put down some more markers. The main markers, or assumptions, about each company in each GBU scenario will be for:

  1. Revenue Growth
  2. Cost Structure/Growth

Based on those 2 assumptions, we’ll get an estimate for Free Cash Flow in each scenario. We’ll apply the following multiples on Free Cash Flow to estimate our Sell Prices: 

  1. Good Scenario: 20X Free Cash Flow
  2. Bad Scenario: 15X Free Cash Flow
  3. Ugly Scenario: 10X Free Cash Flow

Here are the Revenue Growth assumptions: 

  1. Good Scenario: 
    1. We will wholeheartedly believe each company’s Management Guidance for long-term revenue growth. Most management teams are nice enough to provide these estimates every couple of years.
    2. Based on these estimates of long-term (5-year) revenue growth, we will estimate rational stock prices for each company – that will be our Exit Price.
  2. Bad Scenario: 
    1. We assume that the China part of each company’s revenue will decline by 100% cumulatively over 5 years, while the other geographic segments will grow as per Management Guidance. 
  3. Ugly Scenario: 
    1. We assume that the China+Taiwan revenue segment will decline by 100%, while the other segments will grow as per Management Guidance. This is a horrible scenario. There are exceptions to this rule, which we’ll discuss below.
    2. For semiconductor equipment manufacturers, we need to make an exception – we believe that they will replace some of the Taiwan revenue segment. More on that when we discuss each company individually.

Here are the Cost assumptions: 

  1. Good Scenario: 
    1. Gross Margin – over the next 5 years – remains at the same level as of the 12 months ending June 30th, 2022. 
    2. Fixed Costs grow at 5% per year for 5 years. So, that’s roughly 30% in 5 years.
    3. Capital expenditure remains the same as the last 12 months. 
    4. Cash Interest expense same as last 12 months.
    5. Cash Tax rate is 20%.
  2. Bad Scenario: 
    1. Gross Margin – over the next 5 years – remains the same as of the 12 months ending June 30th, 2022. 
    2. Fixed Costs decrease by 20% - cumulatively – over the next 5 years.
    3. Capital expenditure remains the same as the last 12 months.
    4. Cash Interest expense same as last 12 months. 
    5. Cash Tax rate is 20%.
  3. Ugly Scenario: 
    1. Gross Margin declines by a factor of 20% cumulatively over 5 years. So, if the gross margin over the last 12 months was 50%, we assume it will be 40% in 5 years.
    2. Fixed Costs also decrease by 30% cumulatively over 5 years. So, if Fixed costs amounted to $1 billion over the last 12 months, we assume they would decrease to $0.7 billion in 5 years.
    3. Capital expenditure decreases by 20%, cumulatively, over 5 years.
    4. Cash Interest expense is at the same level as of last 12 months.
    5. Cash Tax rate is 20%.

Here they are tabulated for your convenience: 


If you’ve been following us for a while, you might be surprised at this point. We don’t normally forecast, we Buycast. Buycasting means back-solving into “revenue growth we’d need to believe” to consider buying a stock. In these scenarios, it looks like we’re forecasting but we’re not. We’re not proclaiming that the selling prices we estimate represent the true value of the company. We’re simply asking that if we incorporate these elevated China risks in the way we’ve described above, “do all our holdings pass this litmus test?” If some don’t, we should be prepared to sell, so we can sleep well at night. To us investing is more about being a logical Spock rather than a maverick Captain Kirk, especially when the world is increasingly unstable.

We’ll go down our list, holding by holding, to estimate ASPs for each scenario and a weighted average ASP. Essentially, the weighted average ASP will be our estimate of Expected Value (statistics enthusiasts, if they exist, will be happy to see this). We will sell if any Expected Values are less than current trading price of the stocks in question. And lastly, we will add Intel in this analysis even though it’s not a portfolio holding. 

Let the games begin. 

AMD: Not second fiddle anymore

AMD is the scrappy competitor (to Intel) that was never afraid to fight. And based on recent numbers, the scrappy fighter has taken the fight to Intel, and won. Check out their wildly different Revenue and EBITDA growth over the last few years. 


We won’t dwell on their competitive advantages and all that jazz in this analysis. Besides, we’ve done that recently enough. This exercise or litmus test is mostly numerical – we’re trying to find out if it’s worth holding on to AMD given the elevated risk from China. 

Here’s AMD’s regional revenue breakout as of the end of 2021. We doubt this has changed significantly. This data is important for us to estimate Sell Prices for each scenario.


The split between China and Taiwan matters. That’s because in Scenario 2 – the Bad Scenario – we assume that AMD can still sell products in Taiwan even though the China segment will decline by 100%. In the Ugly Scenario, both China AND Taiwan revenue segments will go to $0. Here are the GBU scenario assumptions for AMD: 


Based on these assumptions, we estimate cash flows 5 years later. We’re not going to show the calculations down to the cash flow items here – we’ll publish them in the next few days in a Google Sheet. But here are the results that matter: corresponding ASPs (approximate sell prices) for each scenario, the weighted average ASP, and the last closing price.


Based on our GBU scenario analysis, we will stop our AMD sales, and top up back up to 3% of our portfolio. A few weeks ago, when AMD approached $100/share, we had started selling some of the stock. But the market is so volatile these days. AMD is now back down to roughly $70/share. And we believe that, based on our analysis above, we should hold on to AMD.

The Expected Value of AMD is significantly above its current trading price. If rationality prevails (if the world returns to it), the stock should get back to $100 or more. Currently, based on our numbers, it seems that the market is already pricing in an exports blockade to China. The recent Biden proposal may have something to do with it. However, AMD’s management came out with a statement that this ruling will not meaningfully affect their business. If AMD’s management is right, the stock seems to be oversold.

We will top up AMD to 3% of the portfolio.

ASML: Key to the future?

ASML has also been one of our most successful investments – a nearly 3X multi-bagger for us in 3 years…so far. Is there any more upside? As far as competitive advantages and economic moats go, we cannot think of a more irreplaceable company in the world. They are the sole producers of something called EUV machines. EUV is a highly specialized etching process in semiconductors that requires these giant machines made by ASML. EUV is in high demand by semiconductor manufacturers because it is the technological key to extending Moore’s Law – to keep multiplying the number of transistors per square nanometer of a chip, so we can all have faster, smaller, sexier devices. 

Logic semiconductors like CPUs and GPUs have a reached a point where they’re pushing the laws of physics – the end of Moore’s Law. If the AI revolution is to continue, ASML’s EUV machines are the only way, at least until someone invents another method. That looks unlikely. 

The danger is that one of ASML’s main clients is TSMC – Taiwan Semiconductor Manufacturing Corporation (another Buylyst holding). That Taiwan number below is basically just one client: TSMC. 


In Scenario 2 – the Bad Scenario – we assume that ASML can still sell into Taiwan even though the China segment revenue will decline to $0. In the Ugly Scenario, the China segment revenue goes to $0, but we’ve assumed that about 30% of the Taiwan revenue segment – basically TSMC – will shift to Korea or the US. We believe that TSMC’s clients have already started the diversification process by shifting some production to Samsung (and possibly to Intel or GlobalFoundries in the future). Here are the GBU scenario assumptions for ASML:


Based on these assumptions, we estimate cash flows 5 years later. We’re not going to show the calculations down to the cash flow items here – we’ll publish them in the next few days in a Google Sheet. But here are the corresponding Expected Values or ASPs (approximate sell prices) for each scenario, the weighted average ASP, and the last closing price.


This surprised us. Before we started the GBU analysis, we figured ASML will probably hold up. It turns out that it’s priced just about fairly given these elevated China/Taiwan risks. It’s a gem of a company, a true global dominator in what it does, and an integral part of the AI revolution. But if the TSMC tap is turned off, the company will take a big hit. Some of that revenue stream will be replaced by other foundries like Samsung or Intel. We’ve assumed 30%. But even with that assumption, there isn’t much upside left in the stock at all. 

If the world returns to normal, ASML is a great investment. Maybe the probabilities we attached to the Bad and Ugly scenarios will need to be adjusted if China starts thinking more about its economic well-being rather than its imperial ego. In that case, we’ll be happy to scoop up ASML again. 

We will sell our ASML position.

KLA: Redundant?

KLA is like ASML, in that they also sell equipment to semiconductor manufacturing companies. But they don’t enjoy a monopoly like ASML although they do have dominant market share (>50%). KLA sells testing equipment to semiconductor manufacturing companies, but these machines are a bit less specialized than the complex EUV machines sold by ASML. While KLA’s customer base is more diversified than ASML’s, they do sell a lot of equipment to Memory Chip manufacturers in China. This is a risk because Memory Chips is an area where China is catching up fast – Memory Chips are easier to design and manufacture than Logic Chips. But let’s not jump to conclusions yet.

Here’s KLA’s revenue breakout by region:


The split between China and Taiwan matters. That’s because in Scenario 2 – the Bad Scenario – we assume that KLA can still sell into Taiwan even though the China revenue segment will decline to $0. In the Ugly Scenario, the China revenue segment goes to $0 but we’ve assumed that about 30% of the Taiwan revenue segment – basically TSMC – will shift to Korea or the US. We believe that TSMC clients – like AMD and Nvidia – have already started the diversification process by shifting some production to Samsung (and possibly to Intel or GlobalFoundries in the future). Here are the GBU scenario assumptions for KLA:


Based on these assumptions, we estimate cash flows 5 years later. We’re not going to show the calculations down to the cash flow items here – we’ll publish them in the next few days in a Google Sheet. But here are the corresponding Expected Values or ASPs (approximate sell prices) for each scenario, the weighted average ASP, and the last closing price.


This is another surprise. We didn’t expect KLA to pass the test. But it turns out that with our (hopefully reasonable) assumptions, there is more than enough upside in KLA. The Expected Value is a good 44% above the current trading range of $306-310.

We will hold on to our 3% allocation to KLA.

Nvidia: The AI Company

Nvidia has not been a market favorite recently, after the great run-up since the pandemic. Their revenue growth numbers have been incredible over the last few years. But their last quarterly report was a negative surprise. Since then, the stock’s been pummeled. We still believe in the long-term prospects of the company. 

If there is any company in the world that deserves the moniker “The AI company”, it is Nvidia. Their GPUs (Graphics Processing Units normally made of PCs and Gaming Consoles) were the main engines behind Machine Learning – the ability for a computer to self-program. Machine Learning and Artificial Intelligence are often interchangeable terms. Nvidia’s GPUs, perhaps inadvertently, made this revolution possible. 

The problem is that their chips are manufactured by TSMC in Taiwan. Furthermore, they sell a lot of chips to China, and, surprisingly, to Taiwan. So, one doesn’t need to do many calculations to conclude that Nvidia is risky these days. Here’s Nvidia’s revenue breakout by region:


The split between China and Taiwan matters. That’s because in Scenario 2 – the Bad Scenario – we assume that Nvidia can still sell into Taiwan even though the China segment will decline by 100% to $0. In the Ugly Scenario, both China AND Taiwan revenue segments will go to $0. Here are the GBU scenario assumptions for Nvidia:


Based on these assumptions, we estimate cash flows 5 years later. We’re not going to show the calculations down to the cash flow items here – we’ll publish them in the next few days in a Google Sheet. Those cash flows are then translated to share prices for each scenario. Here are the corresponding Expected Values or ASPs (approximate sell prices) for each scenario, the weighted average ASP, and the last closing price.


Nvidia just has too much exposure to China and Taiwan. In our Ugly scenario, it would lose 58% of its revenue. We don’t know how much of that will be replaced by non-China and non-Taiwan markets. Maybe time will tell us that we were too draconian to attach a 20% probability to our Ugly Scenario in which China invades Taiwan. But at the moment, a 20% probability seems like a reasonable assumption. 

There is still about 15-20% upside left in NVDA. So, here’s what we’ll do: if we find a replacement in the next couple of weeks, we will sell our Nvidia position. If not, we’ll wait until the stock trades close to this Expected Value estimate of $146/share.

Qualcomm: Mobile Star

Qualcomm was turning out to be one of our best calls…until Putin invaded Ukraine. The war instigated a risk-off sentiment in the market, which punished so-called Growth Stocks. Qualcomm’s stock was not immune to this sentiment despite clocking in impressive revenue growth numbers. The reason Mr. Market may have been especially nasty to QCOM was probably because of its China exposure.


But we’re here to find out if QCOM is oversold. Obviously if China actually invades Taiwan, then probably not. But we can’t function in investing (or in life) by attaching a 100% probability to that ghastly scenario. Here are the scenarios we consider, and assumptions associated with each one:


Based on these assumptions, we estimate cash flows 5 years later. We’re not going to show the calculations down to the cash flow items here – we’ll publish them in the next few days in a Google Sheet. Those cash flows are then translated to share prices for each scenario. Here are the corresponding Expected Values or ASPs (approximate sell prices) for each scenario, the weighted average ASP, and the last closing price.


Based on these numbers, we’re holding on to Qualcomm at a weighting of 3% of our portfolio. At an Expected Value of $250/share, there is more than enough upside in the stock (currently trading at about $121). In other words, there is more than enough room for us to be wrong. That’s as good as it gets in investing.

TSMC: Eye of the storm

TSMC has been a star holding in our portfolio. We had identified this opportunity in 2018 when we had decided to bet big on AI. It was still just a buzzword back then. Our rationale for continued revenue growth was essentially this: they were 1 of 3 companies that had (still has) the capability to manufacture increasingly complicated and microscopic chips for a brave, new, artificially intelligent world. And one of the three was (is) Intel, which only manufactures its own chips (although plans are in motion to open up its chip foundry to competitors). 

We’d be forgiven to say that TSMC is one of the most important companies in the world today, if not the most. They manufacture the most complex chips in the world. So, they hold the key to American dominance in technology over the next decade. Xi Jinping wants to stop this. TSMC could be his weapon – to slow down American progress in AI chips, just enough to let China catch up. That makes TSMC a risky investment. 

In our scenario analysis here, TSMC is a bit of an oddball. Essentially, in scenario 3 – the Ugly Scenario – we assume that its stock price will go to $0. That’s a tough stance, but if China invades Taiwan, we can forget about TSMC. Most of its clients are outside China and Taiwan, and the revenue tap will most likely cease to exist. 


Here are the GBU Scenario Analysis assumptions for TSMC:


Based on these assumptions, we estimate cash flows 5 years later. We’re not going to show the calculations down to the cash flow items here – we’ll publish them in the next few days in a Google Sheet. Those cash flows are then translated to share prices for each scenario. We should sell TSMC if we attach a high probability to the Ugly Scenario. But based on the assumptions above, here’s what we get for the ASPs. 


This number is relatively unchanged from our previous analysis. Basically, if TSMC stock reaches the $97-100 range, we’re selling. So, there’s still about 30% upside (or margin of safety), whichever way you look at it.

We’ll hang on to our allocation of TSM at 3% of our portfolio. We’ll be ready to sell our entire position if the stock reaches $97.

Intel: Contrarian bet?

Intel is an oddball in this list because it’s NOT in our portfolio. But it is one of the semiconductor OGs…an industry bellwether.

We’ve had an odd relationship with Intel. We held it for a short while before selling our position right after Apple announced its M1 chip. A year later when new CEO Pat Gelsinger announced a major departure from Intel’s modus operandi – of opening up their semiconductor fabs to other chip companies – we got curious. We took another stab at it, but we just couldn’t get ourselves to buy it. In our most recent semiconductor analysis, we proclaimed that if INTC dips below $30, we’ll consider it. Well, it’s there. So, let’s see whether INTC is worth looking into…again!

Here’s how Intel’s revenue looks when broken out by region:


Now let’s list our assumptions for our scenario analysis: 


Based on these assumptions, we estimate cash flows 5 years later. We’re not going to show the calculations down to the cash flow items here – we’ll publish them in the next few days in a Google Sheet. Those cash flows are then translated to share prices for each scenario. Here are the corresponding Expected Values or ASPs (approximate sell prices) for each scenario, the weighted average ASP, and the last closing price.


Needless to say, we’re staying away from Intel. Basically, if the world returns to normal, it may be worth a look. In this analysis, which represent elevated risks from China, it’s best to stay away from INTC. There is no margin of safety.

Action Summary

Here’s a quick run-down of the actions required for our semiconductor portfolio: 

  1. AMD: Buy back to 3% of portfolio.
  2. ASML: Sell position over next few weeks
  3. KLA: Remains at 3% of portfolio
  4. Nvidia: Remains at 3% of portfolio until further notice (see below)
  5. Qualcomm: Remains at 3% of portfolio
  6. TSMC: Remains at 3% of portfolio
  7. Intel: Not in portfolio, and we’re not buying either

We hope this analysis helped you out in positioning your own portfolio. For us, the intention was to make some adjustments so we can sleep well at night. That’s our vision of a successful investment strategy – generate double-digit returns over the long term while sleeping well at night. We wish you the same kind of success.

Before we leave, one last thing…

Worthy Replacement?

We mentioned that we’d sell Nvidia if we find a replacement. Well, we’ve been curious about Adobe since the day they announced their intent to acquire Figma for a whopping sum of $20 billion. The market punished them for the announcement. But we think this could be a good thing. In fact, when we last did our Adobe deep dive, we had cited Figma as one of the reasons we didn’t believe Adobe could keep growing as fast as it did in recent years. Figma was offering graphic designers something that Adobe ignored. Adobe is trying to remedy that now. And while the market zigs, maybe we should zag. Investing in Adobe would require another deep dive. But we wondered – what if we run Adobe through the GBU test as well? We ran the numbers, but with a twist. 

Instead of relying on Management Guidance for our revenue growth assumptions, we reverted to Buycasting. We asked ourselves, “with the cost assumptions we’ve delineated for the Good, Bad, and the Ugly scenarios, what revenue growth do we need to believe to consider investing in Adobe?” As is customary in our Buycasts, we require a return of 50% in less than 5 years. 

We’ll leave you with the result. It looks like it’s worth a deep dive. Maybe the market really did oversell ADBE. Or maybe Adobe overpaid for Figma. What do you think?



We use cookies on this site to ensure the best service possible.