Why You Should Short The Intel Fab Play

Summary

• Intel has risen on news that it will spend $20B on establishing chip manufacturing factories in Arizona, but this news is more bearish than bullish to me.
• Investing in Intel now, at a relative high and while earnings are expected to fall, just because of the potential fab catalyst is nonsensical once you consider the full situation.
• I reiterate my bearish position on Intel and recommend a diagonal put play.
• Looking for more investing ideas like this one? Get them exclusively at Exposing Earnings. Learn More »

Photo by Avalon_Studio/iStock via Getty Images

I previously stated that Intel (NASDAQ:INTC) has no answer to Advanced Micro Devices (AMD), a statement that ruffled some feathers.

I’ve more-or-less officially been an Intel bear for a few months now, and I admit my shorts on this stock have not worked out well thus far:

I don’t care about being right so much as I care about making money, which is why I usually use lopsided volatility options strategies instead of outright long/short positions. But I do care about being objective. As of late, many have asked me whether I am changing my stance on Intel due to its $20B manufacturing facility (FAB) production project.

As much as I would love to reverse directions and proclaim myself a born-again Intel bull, I must objectively call Intel’s announcement as another bearish catalyst. In light of this news, I am happy to double down on my short call. Today, I will explain my viewpoint.

The Fab Ahab

(Source: Public Domain)

Ahab was the seventh king of Israel. He inherited a powerful kingdom but was eventually brought down due to his own vices and not upholding his responsibilities. Many consider Ahab to be Israel’s worst king.

Intel shares many similarities with Ahab. The company easily dominated the chip kingdom with its x86 IP, but this made Intel a bit too complacent: Easy money from x86 led the company to focus more on maintaining the status quo than moving the foundry market forward. In its comfort, Intel has lost its partnership with Apple (AAPL), who has successfully developed M1, a chip much more efficient than the previously used Intel x86 chips; failed to keep up with the shrinking node trends, currently struggling to produce 10nm chips while others are working on 3nm; and has even been overtaken by AMD, which long sat in Intel’s giant shadow. Intel’s story – to me – looks like a downfall story in the making.

Intel finally sees the need to fight back, and the two fabs in Arizona, due 2024, at the cost of $20B is the latest strategy. However, it is a risky strategy. Adding fabs to the business is not just a one-time cost of the $20B but requires consistent upkeep and upgrade costs, something more like $20B per year, which, over the coming decade, would be more than Intel’s total current equity.

Intel is thus further adding operational costs to its business during a time when the company is expected to see shrinking revenue and earnings. Typically, the responsible reaction to such earnings decay is to cut down on costs. Intel would seemingly rather throw Hail Marys:

(Source: Simply Wall St)

Hail Mary

I call the fab announcement a Hail Mary not because adding fabs is inherently a bad idea but because Intel seems to have little time remaining on the clock to benefit from this action before it finds itself well behind its competitors. Intel should have built fabs years ago when it was still the market leader and still growing earnings. Intel could have become the go-to chip manufacturer in the US, but now it faces an uphill battle in light of how strong Taiwan Semiconductor Manufacturing Company (TSM) and Samsung (OTC:SSNLF) have grown over the past decade.

At roughly the same time as Intel’s announcement to build fabs in Arizona, TSMC (henceforth TSM) announced its own fab construction in Arizona. Clearly, there is a rush to build fabs in the US, and the reason, too, is clear: China. The US reliance on Taiwanese fabs combined with increasing concerns over a Chinese invasion of Taiwan means that the US would love a modern fab stateside. That is, Intel’s decision to create fabs at this time was not decided in a vacuum; instead of the decision being a business-focused decision, it is a situation-focused decision – a decision of convenience rather than one for the sake of shareholders.

In other words, either the dedication to building fabs at this time is outright stupid or there is an ulterior motive.

Intel’s Fab Focus In Question

The more I read about this decision, the more I feel that Intel is not attempting to increase its focus on its fabs and go head-to-head-to-head with TSM and Samsung. I believe two other possibilities are more likely here: Either Intel is attempting to go fabless or Intel is just looking for some extra-governmental protection.

The first possibility is that Intel wants to spin off its fabs and go fabless. After all, the company has had dismal results attempting to produce chips that can keep up with the current node shrinkage. A logical alternative to continue trying to compete on manufacturing while paying the maintenance for fabs is to go the AMD route – fabless. Intel could build some fabs, spin off the fab side of the business for a nice chunk of cash, and just focus on design, outsourcing manufacturing to the spun-off company.

Alternatively, Intel could be attempting to cozy up to the US government with its new fabs. The Department of Defense certainly wants some fabs stateside as China’s shadow grows over Taiwan. Intel might be playing the Tesla (TSLA) game by looking for taxpayers to help subsidize its business, gaining a novel edge over AMD.

Conclusion

In any case, I don’t see the news fabs to be transformative for Intel in the positive sense. If Intel is really going all-in on fabs, we will need many years before we see a significant payoff. Investing in INTC now, during a time of nose-bleed tech valuations and during earnings contractions is not a smart investment decision. You would be best waiting to see benefits from earnings reports in the coming years before going in long.

If Intel is looking to go fabless and spin-off its fabs, it’s the same situation: many years before perceivable benefits to the bottom line. And if Intel is looking for government subsidies by being the main US chip manufacturer, this is essentially a white flag; Intel cannot compete in the game of technological advancements and needs a handicap. This is not Tesla being subsidized as it forays into a fledgling industry but an old, stubborn company attempting to stay afloat via means outside its own merits. You would be better off investing in AMD or TSMC if you want exposure in US chip design or manufacturing.

Overall, I reiterate my bearish outlook on Intel relative to the rest of the chip industry. The upward momentum in Intel and its current price are not sustainable by the news: The fabs and other aspects I’ve mentioned in previous articles are perceived as bullish while they are in fact bearish. Nor is price sustainable by Intel’s business elementals. As I’ve described in previous articles, Intel’s tech is underperforming relative to its competitors, and the company is losing business due to this (Apple’s ditching Intel exemplifies this decade’s Intel).

I’m going short again, and I probably will continue doing so even if Intel continues to rise on what I consider to be other factors (TINA, the tech bubble, macro issues – all of which I have discussed in previous articles). But this time my options strategy will allow for an upward movement without potential losses, as I admit Intel’s fundamentals and its stock price have become detached.

My Trade

Here is my trade recommendation:

• Buy 3x Apr16 $66.50 puts at $127 each (as of Apr6)
• Sell 1x Apr9 $70 put at $445 each (as of Apr6)
• Total cost: $64 credit

This is a volatile options strategy that allows profits in three directions: large profits on a downward movement and smaller profits should INTC continue upward or sideways. Delta here is nearly 50, meaning a $50 profit per every $1 movement down in the underlying, but you can increase delta to over 100 by buying back the short puts should you see INTC beginning to drop. You can also increase your profits by selling an extra 2 puts at a lower strike price (say $62), but this would cap your maximum gains in the event of a selloff.

This is not just a weekly play, as I intend to roll it every week. That is, I will close out the trade every Friday and open the equivalent for the next week (sell 1x ITM weekly put and 3x ATM 2-week-out puts). It should be rather clear how to do this but let me know in the comments section below should you need assistance.

In any case, with no risk on the upside, I’ll be much less embarrassed should I be wrong this time.

Happy trading.

Analyst’s Disclosure: I/we have no positions in any stocks mentioned, but may initiate a short position in INTC over the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Comments

[Polyhedron]

@aryareaisparsi @st@stockroach sorry for my incomplete thought: Intel is and will be behind in x86 design until 2023, mostly due to its monolithic die approach to server design and its long design cycle. Pat has indicated the return of tic-tok product development approach, which means two staggerd teams in parallel, each design takes twice as long as the product introduction rate. This was practiced to great success in 1990s - 2000s (while Pat was at Intel). Its weakness, though is a LONG time-to-market, and vulnerable to changing requirements. Times have changed, but seems Pat has not.

[It's 6:09]

This author has been short Intel for nearly 35% upside and now recommends a diagonal put play. The hits just keep coming. Clearly little acumen regarding the business.

[Always Bullish]

Comparing an ancient king in Israel to intel makes as much sense as squiggly lines from technical analysts indicating its downfall. Intel if bought a short time ago when it went on sale is a winner. As investors not CEOs we shouldn’t try to think we can run the company better than gessinger who spent 30 years there. He is competent and capable and INTC if bought at the right price is a safely sleep at night winner.

[dawgydaddy]

$20B a year to maintain a $20B fab, in capital costs? That’s nonsensical. Try $1B.

[Just Some Guy]

@dawgydaddy, try $0B, because in operation it turns a profit, it doesn't need external funds at all. Yes, that was a goofy statement in the article.

[MRW.Doddsville]

''You would be better off investing in AMD or TSMC if you want exposure in US chip design or manufacturing.''
Wrong on so many levels, unless you like daytrading your money over to brokers.

[millz1987]

@G.Doddsville lol what? How are those day trading stocks? Those are buy and hold stocks that you wouldn't trade, unless you either wanted to lose money, or lose out on potential large gains.

I'd argue intel is better for trading. Been in the same 40 - 65 channel for as long as I can remember :D. Buy at 40, sell at 60 rinse and repeat.

I wouldn't personally do that. But in terms of trading, it's probably more lucrative.

[danielaqwe]

I will not get much love with this one...

IMO, the "semiconductor shortage" is exaggerated.
To be sure, the latest and greatest in new chips are always difficult to get. When a circuit designer has a new project, first thing he does is to make sure he can buy the chips. You cannot tell me, that Ford or GMC have designed electronics for their cars and suddenly out of the blue they cannot get them... nonsense.
besides, very few chips are really one of a kind. If you cannot get this one, you design with another one. Worst case you have to make a SW patch or a new PCB.
But even then, electronics inside a car is not really sate of the art (EV is different). The designs are a few years old and the supply chains are planned ahead for month or even years.
Even without a "chip shortage" as a designer or manufacturer you always risk that some other player buys up all the remaining stock and you got nothing.

There are literally millions of different semiconductors and electronic components which find their way into products. But the industry does NOT produce every kind "all the time". Instead, depending on the orders, the fab is configured to make one type for a certain time, then retools and makes a different one. If you do not have stock... you may have to wait for weeks... it never happens, because no manufacturer is so stupid. There is no such thing as "just in time" when it comes to electronic components.

Car manufacturers go along with the "ship shortage myth" because this gives them an excuse to stop the assembly line for retooling for a different model as well as maintenance.

I am an electronics design engineer and have been in this situation many times before. If you are a very large manufacturer, you have planned ahead and if you are a smaller manufacturer (my former employer) you run a new set of printed circuit boards using different chips.... you never shut down the manufacturing because of a component shortage.

[Watcher of the Fray]

@danielaqwe "IMO, the "semiconductor shortage" is exaggerated."
Que?

[millz1987]

@danielaqwe If you've been following the story. I believe the thesis is that the automakers cut orders at the beginning of 2020 because of the pandemic and assumed they wouldn't sell many vehicles because of it. Their supply went to feed other end markets due to the pick up in the WFH demand.

I guess it turns out that folks don't like sitting on crowded busses in the middle of the plague and vehicle sales rebounded hard but their supply had already been sold out to other buyers and we now have a shortage. DRAM makers and other chip companies don't really care much for the auto market and are instead building inventory to feed the hyperscalers who do their buying in 2H as this is where most of their revenue comes from. Auto makers buy in tiny volumes compared to the cloud guys.

Couple this with crypto demand and you have a severe problem. You can't just build more chips when your fabs are running 100%. Companies like TSMC & Micron have to do their best at estimating demand years beforehand and if they overbuild, they will likely lose money.... See the DRAM oligopoly's wild boom & bust cycles for proof that overbuilding causes massive losses.

Car manufactures wouldn't just idle their plants just for funsies as I would imagine the fixed costs of running said fab are very high and this would only be a last resort. I'm assuming car sales are very strong this year... just anecdotal experience... I traded in my 2018 leased vehicle for a pretty hefty profit. Dealer was willing to buy it from me for $5,000 more than what my contract stated.

[danielaqwe]

@millz1987
-- I believe the thesis is that the automakers cut orders at the beginning of 2020 because of the pandemic and assumed they wouldn't sell many vehicles because of it. ---

If your refrigerator is empty, don't call it a food shortage, just because you were too lazy to go shopping.

Semiconductors are made in batches not in a continuous assembly line.
Car manufacturers do not build their own electronics. This is done by contract manufacturers. The PCBs are made by yet another company.
Some of the chips are custom designs (ASICs).
No contract manufacturer, PCB manufacturer or custom chip manufacturer will produce a single piece or even place orders for parts and materials until they have a firm order.
You may be right, F and GM dropped the ball and did not place their orders... sounds plausible because they always behave as their subcontractors should bankroll their purchases.
The "Just in time" manufacturing philosophy is a bit misleading. It only means that the parts arrive at the final assembly plant... just in time.
The orders were placed months in advance.
For long lead times like semiconductors this is even worse.
A custom chip, even after it has been designed takes still 3-6 month from order to delivery.
Purchasing managers just dropped the ball, but don't call it a chip shortage, because this would imply that the cannot deliver chips... they can, but it takes time.... like it always does.
A few years back, there was a real chip shortage, because one of the few manufacturers of chip encapsulation material went up in flames (literally the factory caught fire). We have none of this today.

[PauloCostaSilva]

People hate the turtles and love the hares. Turtles live 60+ years, hares don't.
Not every stock is worth holding, not every stock is worth buying. $INTC has been a great asset to me. That philosophy that " a stock that doesn't double or triple in value in 12 months is xit " i don't buy it !!

[MRW.Doddsville]

@PauloCostaSilva finally someone with a brain in the comment section

[Spike Capital]

If INTC stands above $67.7, sky is the limit.

Everything is panning out the way I predicted, so far :)

[Just Some Guy]

@Spike Capital, well you got it, closed above $68, but it's still within the Bollinger bands on a weekly chart.

[RUBYRUBY3]

@Spike Capital
"sky is the limit."
Nope not after all the " P/E too high " people said AMD was not a buy

Next years earnings are $4.72 so P/E will climb with price
$4.72 X 20 P/E = $94
$4.72 X 25 P/E = $118
$4.72 X 30 P/E = $141
$4.72 X 35 P/E = $165
www.macrotrends.net/...
Rare over 20 P/E almost never over 25 ( last 15 years )
Expect selling by 20 P/E ( too high),
for the same reason people bought at 9 P/E ( too low )
$100 is P/E 21.18 that is the best INTC will do before major selling

Dividend also hurt by higher prices ( all are below inflation )
$1.39 / $94 = 1.47%
$1.39 / $118 = 1.17%
$1.39 / $141 = 0.98%
$1.39 / $165 = 0.84%

[Spike Capital]

@RUBYRUBY3 The thing is I believe Intel will make more than $4.72.

[Just Some Guy]

Without comment on the option strategy, which just might work, I will note that the boost to Intel price on the fab news only lasted one day and then vanished. Indeed it may be read as bearish news for the next two years that it will take to complete the new fabs, so some negative plays may work.
OTOH the announcement yesterday of the Ice Lake server chips was reasonably strong and stand a fair chance of a short term surprise on INTC revenues and profits to the upside - at next earnings report, anyway. So take that into account as you construct volatility plays.

[freddrick]

TSMC announced their AZ fab last year, not roughly the same time as Intel. The article you have linked is for a bond sale to partly fund it.

[stockroach]

opps forgot the link

newsroom.intel.com/...

[stockroach]

2nd recent super computer win with Habana chips. Keep shorting Damon

[Spike Capital]

@stockroach He's not even shorting the stock. Instead, he's playing weeklies.

Literally makes no sense because his short thesis is long term but he's actually entering very short term option trades.

There are so many better ways to park the capital...

[MICHAELSONG]

@Spike Capital good point...shorting with long term thesis...very wrong...

[matt1685]

Intel's problem, the stock, I mean, not the business, is that it was so dominant in semiconductor manufacturing for so long that when it had a misstep people seemed to think it was the logical conclusion that it must afterwards suddenly be very bad for an equally long period of time. If Intel had instead been jockeying back and forth with TSMC for years we wouldn't be hearing so much talk about the demise of Intel.

Intel's issues have to do with the insertion of EUV and the breakdown of the methods used to extend immersion DUV. Intel's inability to correct the situation more quickly was due to the exact same issues. Their hands were tied once the initial poor decisions had been made back around 2013 or so, maybe some even earlier. Therefore, the greatest chance is that Intel has not suddenly forgot how to manufacture semiconductors. Rather, as soon as Intel has enough EUV tools and the pellicles they have been waiting for to avoid the throughput-busting methods TSMC uses to enable EUV, the situation gets reset.

Intel 10 nm is going to be a bastard node. 7 nm will get them back on track but it will come out late compared to TSMC's 5 nm, its closest equivalent. After that, things are up in the air as Intel moves to nanoribbons and TSMC stays with finFETs. Although it may turn out to be correct, it's far from a sure bet that TSMC is going to continue with its superiority. Just because TSMC navigated the EUV insertion better does not mean it has superior engineering.

Intel does not need government subsidy to be a semiconductor manufacturing leader. Any subsidy will only provide a tailwind. And the US cannot strategically allow so much American semiconductor manufacturing to take place on an island the CCP is going to attack as soon as it thinks it can win the war to take it. So the tailwind could be brisk.

[Spike Capital]

@matt1685 TSM will most likely move away from FinFET because physically you just have to find another way to build transistor below 3nm.

[matt1685]

@Spike Capital Yes they will after 3nm. But TSMC plans on using finFET at 3 nm while Intel plans to use nanoribbons at their 5 nm, which will compete against TSMC's 3 nm.

[RUBYRUBY3]

@matt1685
A repeat of the 10nm mess would have them backporting
5 nm Nanoribbon design to 7nm while TSMC goes ahead again with proven FinFET on 3 nm.
Don't think the market would like a repeat of the last
3 years over the next 5
The smaller we go the harder it is going to get

[zisdead]

So to sum up. Short a hot stock in a bull market not to mention it pays a dividend.

Now you know why the author got destroyed.

[Chris Lau]

My $INTC PT was $70 for the longest time. It might cross there in days.
GLTA.
And sell $AMD.

[freddrick]

@Chris Lau ok will buy AMD now

[stringtemple]

@freddrick yep I’ll buy AMD on that note

[stok_analyzer]

@Chris Lau I do remember that prediction few months back where you had mentioned it will be around $70 by Spring 2021.
Somehow, it stuck in my head and I stuck to INTC since last fall despite being down 15% at one point, no less also because of inherent value calculator you had mentioned in the link. Today, I am up by 30%. Thank you!

[Bonjwa]

seekingalpha.com/...

You came out with that article when INTC was in the $40s. At 10x earnings and 12+% shareholder yield. I took the other side and made 40+%.

[x399]

I don't own INTC, but the narrative around the stock is crazy. We live in a world where if a company builds a chip, it will go into something. Google, AWS, and Microsoft are going to be constantly building data centers as more and more companies transitions to the cloud and the amount of data saved their continues to grow by companies already on the cloud. We haven't even really seen growth in self-driving, which will require storage of Terabyte's worth of data per day. There are also billions of people without access to internet still to this day, who slowly will gain access to the internet. The semiconductor industry is just going to constantly grow at some rate and INTC will see some share in that growth. I would buy INTC, if I didn't already own NVDA.

[bengalesq]

Agree with much of this - speculation of course but seems the logical analysis. That said, it is a cash cow that still has a huge market share. It has gone the opposite direction from what I thought made sense. Its competitors are not better plays however because they represent a completely different type of investment - despite being in the same industry - because of the multiples involved. That said I sit on the sidelines for now - a buyer around 50 and if I owned I suppose I would hold here or sell out half a stake.

[Tetragrammaton]

Moving forward from Ahab, I take the long view of Jewish history (in which I am considerably influenced by the opinions of Josephus and Gibbon) . So it seems to me that pig-headedness, determination, and go-it-alone have been attributes which have been both helpful and hurtful over the past couple of thousand years. In the more recent period of 1980-2021 both Intel and AMD could be considered to have followed different paths, but with somewhat analogous stubborn virtues and vices. For the next five years, Intel is arguably more in control of its destiny than AMD, which is why I have stock holdings in INTC, and also in TSM.

The determining factor over that period, not mentioned in the excellent article above, is going to be determined by whether state-of-the-art foundries can find sufficiently skillful labor forces to keep yields at a satisfactory level. (This supply factor will control the demand/supply equation, and will dictate pricing and thus profitability for both vendors and customers). Engineering design has tended to produce machines, manufacturing procedures, and quality control factors which are simply too demanding of skills and precision for the available human resources, in the clean-room environment where a speck of dust can ruin a whole production run. Silicon Valley has problems in this regard, Taipei not so much.

[PreCambrian]

Not only is Taiwan close to China but so is South Korea. China could easily convince North Korea to lob a missile into some of Samsung's foundries. So foundries are important to the security of the United States and $20 billion isn't even pocket change to the US government anymore. And it is possible that Intel will actually do a good job manufacturing new processors. It is difficult but others do it, it just takes the right kind of management and know how. They could also get a little homegrown expertise from some Micron employees. So altogether I think there are better places to place your bets.

[sojuofwallstreet]

@PreCambrian dude. taiwan is fulled backed by USA.

[oslo]

@sojuofwallstreet

Yet to be seen.

[jbw1937]

@oslo
Our current comrades will not go to war over Over Taiwan, just bluster. Plus, if China does pull the trigger they will go all the way and take Korea as well. Ultimate goal will probably mean Japan as well. We’ve seen this movie before with another player.