Latest / Deep Values: Value Investing Deep Dives / 2026 Q1 Intel's Atomic Gamble or Value Trap
Transcript
- 0:00So imagine you've got this machine right, and it produces
- 0:04the single most critical resource on the entire planet.
- 0:08Yeah, the kind of thing governments are practically
- 0:10fighting over. Exactly.
- 0:11Militaries absolutely rely on it to function.
- 0:14And you know, if this machine were to just suddenly stop the
- 0:18whole modern economy, they get everything from your car to
- 0:21hospital life support systems, it would all just instantly
- 0:24collapse. Right.
- 0:25It's like the ultimate undeniable strategic asset.
- 0:28It is, but here is the billion dollar paradox we're looking at
- 0:32today. The actual underlying economics
- 0:35of running this machine, they're completely, totally broken.
- 0:38It's a, it's a massive disconnect.
- 0:40I mean, you have this business that is vital to the national
- 0:43security, the United States, right?
- 0:45But on paper, it's burning cash at a rate that would just
- 0:49terrify a traditional value investor.
- 0:51Yeah, it's wild to think about. You've got National Defense on
- 0:54one side and then this plummeting stock chart on the
- 0:56other. You're essentially looking at a
- 0:57company trying to save the world, or at least its corner of
- 1:00it, while its own balance sheet is bleeding out.
- 1:03Right. And today is April 23, 2026, and
- 1:07we are diving into a stack of incredibly timely sources that
- 1:11capture this exact tension. We've got some great materials
- 1:14today. We do.
- 1:15We have a massive AI generated deep value Intelligence Report
- 1:19alongside this really detailed internal research guide and they
- 1:22are both entirely focused on Intel.
- 1:24Which is just a fascinating case study right now.
- 1:27Oh, absolutely. So our mission for this deep
- 1:29dive is to dissect Intel's brand new Q12026 earnings, decode this
- 1:35like radical new leadership strategy they've deployed, and
- 1:38really try to answer the ultimate value investing
- 1:40question. Which is, you know, is Intel a
- 1:42brilliant once in a generation turn around play?
- 1:45Or is it just a colossal value trap?
- 1:47OK, let's unpack this. Where do we even begin to
- 1:50understand the sheer scale of Intel's current dilemma?
- 1:54Well, to really understand the turn around, you have to
- 1:56understand the crash. And to understand the crash,
- 1:58you've got to look at the sheer hubris of their glory days.
- 2:01Right, because they weren't always struggling like this.
- 2:03Not at all. If we just rewind a bit, like
- 2:05from roughly 2016 to 2021, Intel was the undisputed Titan of the
- 2:10tech world. They were untouchable.
- 2:12They really were. We're talking about a company
- 2:13that consistently posted gross margins between 55 and 65%.
- 2:20Yeah, Their annual net income peaked at over $21 billion.
- 2:24They had this near monopoly on the world's computing power,
- 2:27mostly in data centers and PC's. So they were just swimming in
- 2:30cash. Exactly, they generated so much
- 2:34free cash flow they genuinely struggle to deploy at all.
- 2:37In fact, over the last 35 years, Intel spend a staggering 152
- 2:43billion dollars on stock buybacks.
- 2:46Wait, wait. Let me make sure I'm wrapping my
- 2:47head around this $152 billion A. 152 billion.
- 2:52They had that cash over the last few decades and they just burned
- 2:55it on stock buybacks just to, like, boost the share price for
- 2:58executives and shareholders. That is the reality of their
- 3:01capital allocation back then, yes.
- 3:03And that's like a farmer eating their seed corn instead of
- 3:05planting it for the next harvest.
- 3:06Right. It's a perfect analogy,
- 3:08honestly. Because looking at where they
- 3:10are today, literally begging for government subsidies, you have
- 3:14to wonder how much the current board regrets not saving.
- 3:18I don't know even a fraction of that money for the factory
- 3:22investments they desperately need right now.
- 3:24Oh. The regret has to be immense.
- 3:26But you know, how do they let that happen?
- 3:28It was the classic innovator's dilemma.
- 3:30How so? Well, they thought their
- 3:31manufacturing lead was just insurmountable, but while they
- 3:34were, you know, printing cash in the PC and server markets, they
- 3:38completely missed the mobile revolution.
- 3:40Right, they passed on making chips for the early iPhone,
- 3:42didn't they? Exactly.
- 3:44And that one decision allowed a company like TSMC, the Taiwan
- 3:48Semiconductor Manufacturing Company, to step right in.
- 3:51And TSMC just took that mobile market and ran with it.
- 3:54Yeah, they used the sheer volume of billions of smartphones to
- 3:57rapidly iterate, they funded their own massive R&D with it,
- 4:01and eventually they simply surpassed Intel's manufacturing
- 4:04tech. So TSMC didn't just like, steal
- 4:06the crown overnight, Intel practically handed it to them by
- 4:09ignoring the phone in your pocket.
- 4:11That's the tragic irony of it. And you know, Fast forward to
- 4:14today, 2026, and the picture is starkly different.
- 4:18Yeah, those buybacks are definitely a thing of the past.
- 4:20Completely halted the dividend Gone.
- 4:24They suspended it entirely to preserve capital.
- 4:27Today, Intel is carrying over $44 billion in long term debt.
- 4:3144 billion. Ouch.
- 4:33And their gross margins have plummeted from the mid 60s down
- 4:36to around 35%. Plus, their free cash flow is
- 4:40deeply, deeply negative. Which means they're spending
- 4:43significantly more cash than they're bringing in you.
- 4:45Just you don't survive that kind of financial whiplash without
- 4:48completely changing who is at the wheel.
- 4:50And that brings us to the cultural surgery happening over
- 4:54there right now. Right, the new leadership.
- 4:55Yeah, what's fascinating here is what the sources refer to as the
- 4:59Lipa Tan doctrine. Right Lipa Tan.
- 5:01He became CEO in what, March 2025?
- 5:04Exactly. He's this highly respected
- 5:06industry veteran best known for orchestrating the legendary Turn
- 5:09Around at Cadence Design Systems.
- 5:10So he comes in and what does he find?
- 5:13He inherited an Intel that was caught between it's like bloated
- 5:16legacy as a traditional chip maker and the insanely expensive
- 5:20demands of trying to become a global foundry.
- 5:22And when we say foundry, we mean manufacturing chips for other
- 5:25people, right? Not just their own designs.
- 5:27Right, exactly. So Tan looked at the old vision
- 5:31with this whole IDM 2 point O thing, realized it was failing,
- 5:34and he pivoted incredibly hard. Yeah, looking at the research
- 5:37guide, it's obvious that pivot has been pretty brutal.
- 5:39Beyond brutal, his doctrine is all about ruthless, stripped
- 5:43down engineering execution. They have executed a massive
- 5:47reduction in the workforce. How massive are we?
- 5:49Talking well, in 2023, Intel had almost 125,000 employees.
- 5:54Today, they're down to an estimated 75,000.
- 5:57Oh wow, so that's what, a 40% decline in the headcount in just
- 6:00a few years? Yeah, 40%.
- 6:02So Tan fires 40% of the staff, rips out the middle of
- 6:05management. But he didn't do that just to
- 6:07save a few bucks on payroll, did he?
- 6:08Not at all. The primary goal was to flatten
- 6:11the corporate hierarchy. You see, Intel had become famous
- 6:14for its endless meetings, the bureaucracy, really slow
- 6:18decision making. The classic bloated corporate
- 6:21giant syndrome. Exactly.
- 6:23Tan's objective is to eliminate those middle management layers
- 6:27and return power to the actual engineers.
- 6:29The sources mention a shift to referent power.
- 6:33What does that actually mean in this context?
- 6:36It means managers should lead through deep technical
- 6:39expertise, right? Not just because they have a
- 6:41vice president title on an org chart.
- 6:44He's trying to resurrect the ruthless engineering excellence
- 6:47that defined Intel decades ago under Andrew Grove.
- 6:50Which makes perfect sense, because putting power back into
- 6:53the hands of the raw engineers is basically the only way
- 6:55they're going to solve the absolute physics nightmare of
- 6:58their next gamble. The physics nightmare is right.
- 7:00Let's talk about what they're actually building to try and
- 7:02dethrone TSMC. The sources keep mentioning that
- 7:05Intel is betting the entire ranch on their 18A and 14 A
- 7:10manufacturing nodes. Right, the new node.
- 7:12Yeah, let's break that down because the numbers here are
- 7:15mind boggling. We really have to understand the
- 7:17scale we're talking about here. When Intel says 18 A that A
- 7:20stands for Angstroms. Wait, Angstroms, that's even
- 7:23smaller than nanometers, right? Yes, 18 A means 1.8 angstroms.
- 7:28We are talking about shrinking transistors down to the literal
- 7:31atomic level. That is insane.
- 7:33At that scale, the the physics get incredibly weird.
- 7:36You can't use normal light to print the circuits onto the
- 7:38silicon anymore. The wavelengths are simply too
- 7:41big. So it'd be like trying to paint
- 7:43a microscopic masterpiece with a giant roller brush.
- 7:46Exactly. It just doesn't work.
- 7:48So how do you actually print something at the atomic level?
- 7:51You need something called high NAEUV lithography.
- 7:54High NAEUV, OK. EUV stands for Extreme
- 7:58Ultraviolet. These are massive, incredibly
- 8:01complex scanners built almost exclusively by a Dutch company
- 8:05called ASML I've. Heard of it?
- 8:07And they have a total monopoly on this stuff, don't they?
- 8:09Pretty much. High NA essentially uses these
- 8:11larger hyper precise mirrors to focus that extreme ultraviolet
- 8:15light into a microscopic scalpel.
- 8:17Wow. It is arguably the most complex
- 8:19machine ever manufactured by human beings.
- 8:21And because of that, a single one of these high NA EUV
- 8:24machines cost roughly $350 million. 350 million dollars for
- 8:31one single machine. Just one.
- 8:34And you can't just buy one to run a global factory, can you?
- 8:37No, you need a whole fleet of them for a modern fab.
- 8:40Because of this Intel's capital expenditures, their CapEx is
- 8:44running at roughly $25 billion a year. 25 billion a year just
- 8:49burning cash. Burning cash at a historic rate
- 8:53just to put out the equipment. So how are they paying for all
- 8:55this? They're leaning on what they
- 8:57call a smart capital strategy. They're tapping into government
- 9:00subsidies through the GPS Act. They're forming equity
- 9:03partnerships. They actually sold stakes in
- 9:05some of their operations to NVIDIA and SoftBank recently.
- 9:08So they're basically scraping the couch cushions of global
- 9:11finance to fund these massive factories in Ohio and Arizona.
- 9:15That's a good. Way to put it, yeah.
- 9:16But I'm looking at these GPS Act subsidies in the sources and it
- 9:19feels like a total devil's bargain.
- 9:21The silicon handcuffs. Yes, sure, they get billions in
- 9:24free cash from the US government, but they're slapped
- 9:26with these handcuffs by taking that money.
- 9:28Intel is barred from expanding any advanced production in China
- 9:32until 2036. Yup. 10 years.
- 9:35Isn't capping your access to one of the biggest technology
- 9:38markets in the world a massive strategic handicap just to get
- 9:42some short term factory funding? Oh, it's a.
- 9:44Huge handicap. They are strictly locked into a
- 9:47reliance on US and European manufacturing bases for the next
- 9:51decade. But frankly, Intel was in no
- 9:54position to negotiate. They needed the capital to
- 9:56survive. Right.
- 9:57And here's where it gets really interesting for me.
- 9:59I'm looking at this whole foundry business model.
- 10:02Catch 22. Exactly.
- 10:03Intel wants to build these incredibly expensive fabs to
- 10:07manufacture chips for other companies, right?
- 10:10Because they need massive volume to pay off those $350 million
- 10:14machines. Yes, that's the plan.
- 10:15But why would giant competitors like AMD or Apple or NVIDIA
- 10:20actually use them? Well, that's the $1,000,000
- 10:22question. I mean, it's like spending $5
- 10:25billion to design the world's most advanced bank vault and
- 10:28then handing the exact, highly secretive blueprints to a guy
- 10:32who currently makes a living robbing your banks and asking
- 10:34him to build it for you, right? Why wouldn't these companies
- 10:37just keep using a neutral partner like TSMC?
- 10:39You have zeroed in on the absolute core vulnerability of
- 10:44Intel's entire strategy. It is a massive industry wide
- 10:48trust deficit. A trust deficit, yeah.
- 10:51TSMC is essentially the Switzerland of semiconductors.
- 10:54They don't design their own chips, they don't sell
- 10:57processors to consumers. They only manufacture for
- 10:59others. So they don't compete with their
- 11:01own. Customers exactly because of
- 11:03that strict neutrality, they have built a highly trusted
- 11:06ecosystem and that trust allows them to command gross margins
- 11:11well above 50%. And Intel, on the other hand,
- 11:14historically competed directly with the very companies they are
- 11:16now begging to be their foundry customers.
- 11:19Right, so even if Intel builds a better, faster factory, Apple
- 11:23and NVIDIA are still going to look at them and think you're
- 11:25just going to peek at our homework and use it for your own
- 11:27chips. That is a terrifying hurdle for
- 11:29them to clear. That is the fear, and the Q
- 11:3112026 numbers prove exactly how hard this is.
- 11:35Oh, let's look at those. So Intel's total foundry revenue
- 11:38was $5.4 billion, but almost all of that was internal.
- 11:42Meaning Intel's manufacturing arm.
- 11:44Building ships for Intel's design arm.
- 11:46Yes, their external foundry revenue, actual money from
- 11:50outside customers was a tiny $174 million.
- 11:54Wow. Out of 5.4 billion, yeah.
- 11:57So beating TSMC isn't just a physics problem or an
- 12:00engineering problem. It is a massive psychological
- 12:03and business problem. You have to convince your
- 12:05fiercest rivals that you won't steal their lunch.
- 12:08Man, OK, so the risks are established, the stakes are
- 12:11existential, and that trust deficit is very real.
- 12:14But let's look at the actual pulse of the patient.
- 12:17The Q 12026 earnings just came out.
- 12:20Is this radical, incredibly expensive gamble actually
- 12:23working? Surprisingly, yes, there is a
- 12:25pulse and it's beating stronger than many analysts expected,
- 12:29really. Yeah.
- 12:29In Q12026, Intel's total revenue hit $13.6 billion, which is
- 12:34actually up 7% year over year. 7% growth for a legacy company
- 12:38of this size right in the middle of a chaotic turn around.
- 12:41I mean, that's definitely nothing to sneeze at.
- 12:42Where's that momentum coming from The?
- 12:44Brightest spot by far is the data center and AI segment, or
- 12:47DCAI. That division surged 22% year
- 12:49over year. 22%, yeah. What's happening is that the
- 12:52broader artificial intelligence market is maturing.
- 12:56For the last few years, all the big money went into AI training,
- 12:59you know, building the massive foundation models.
- 13:02Right. And that market is completely
- 13:03dominated by Nvidia's GPU's. Exactly.
- 13:07But now workloads are shifting toward AI inference.
- 13:10Break that down for us. What is the difference between
- 13:12training and inference and why does that shift help Intel?
- 13:17OK, so training is like sending thousands of computers to school
- 13:19to learn a new language. It requires massive brute force
- 13:23power. Hence the expensive NVIDIA
- 13:25chips. Right.
- 13:26Inference, on the other hand, is using that language in the real
- 13:29world. Like when you ask a chatbot a
- 13:31question and it generates an answer.
- 13:33You don't need a massive expensive NVIDIA GPU for a lot
- 13:36of inference tasks. You just need a highly
- 13:38optimized, efficient processor, and Intel's new Zeon 6
- 13:42processors are highly tuned for exactly this.
- 13:46Right now, server demand for inference is actually
- 13:49outstripping supply. Oh wow, and it's not just the
- 13:51big server farms either, right? The sources highlight this
- 13:54massive boom in what they're calling the AIPC market.
- 13:57Think about the laptop you're using to listen to this right
- 14:00now. Intel is betting it's survival
- 14:03that the next time you up Jade, it has an Intel Core inside
- 14:07specifically designed for AI. Yeah, their client computing
- 14:10group is seeing a real renaissance.
- 14:13AI related businesses now make up about 60% of Intel's total
- 14:17revenue. That's a huge pivot.
- 14:19It is. They recently launched the Core
- 14:21Ultra Series 3, and it's already powering over 200 different
- 14:26laptop and desktop designs from major manufacturers.
- 14:29Which is crucial because it proves Intel is successfully
- 14:32fighting off the ARM based competitors.
- 14:34Doesn't. It it absolutely does.
- 14:36For a while it looked like everyone was going to abandon
- 14:39Intel's traditional by 86 architecture and move to mobile
- 14:42style ARM M chips for better battery life.
- 14:45Like Apple did with their M series chips.
- 14:47Exactly, but Intel is proving they can still deliver the power
- 14:50efficiency and performance needed to run AI directly on
- 14:54your device. So we have revenue growing 7%.
- 14:57We have a massive workforce reduction making them leaner.
- 15:00They are successfully launching new AI chips for servers and
- 15:03laptops. Right.
- 15:03Sounds great, doesn't it? It does, but then we look at the
- 15:06Deep Value Intelligence Report, this incredibly robust multi
- 15:11agent AI framework that analyzes the stock for value investors.
- 15:15And despite all this good news, it slaps Intel with a strict red
- 15:19letter. Avoid rating.
- 15:21Yeah, and avoid for value. Now, it gives it a watch for
- 15:24growth investors, but for value, it's a hard pass.
- 15:28So what does this all mean? If you're a value investor
- 15:30looking at that $67 share price right now, you might be tempted
- 15:34to jump in. How can the business be showing
- 15:36signs of life? But the value verdict is so
- 15:39negative. Well, if we connect this to the
- 15:41bigger picture, it all comes down to the fundamental
- 15:43difference between an exciting turn around story and a true
- 15:46value investment. OK, unpack that.
- 15:48Warren Buffett's style of value investing demands 3 non
- 15:51negotiable things, a highly durable economic Moat,
- 15:54predictable earnings and a high return on invested capital, or
- 15:58ROIC and. Intel fails those tests right
- 16:00now. Spectacularly.
- 16:01Let's look at the Moat first. Intel's Moat used to be
- 16:04manufacturing supremacy, right? That is gone.
- 16:06TSMC has it. As for predictable earnings,
- 16:09Intel's profits have been a roller coaster for five years.
- 16:12So that's two strikes. But the most damning metric is
- 16:15the return on invested capital. It is essentially 0 right now.
- 16:19Zero. Yeah, when you were pouring $25
- 16:22billion a year into new factories and your ROIC is 0.
- 16:27You are destroying economic value in the short term just
- 16:30hoping for a payoff in the long term.
- 16:32Yeah, the sources mentioned the intrinsic value model in the DVI
- 16:36report. The market is currently pricing
- 16:38the company at around $324 billion with shares at that $67
- 16:43mark, right? But the report says the
- 16:45intrinsic value based on owner earnings is drastically lower.
- 16:49How does that math work? Owner earnings are essentially
- 16:51the reality check of investing. It's the actual cash you get to
- 16:54put in your pocket at the end of the day after paying for all the
- 16:56necessary upkeep of your business.
- 16:58OK. Because Intel's CapEx is so
- 17:00astronomically high right now, remember the $350 million
- 17:03machines? Hard to forget.
- 17:05Just to build these new fabs to stay relevant, their owner
- 17:09earnings are actually negative. So it's like buying a rental
- 17:11property where the monthly rent doesn't even cover the cost of
- 17:15replacing the roof every single month.
- 17:17You might own a house, but it's making you poorer.
- 17:20Exactly. You cannot mathematically
- 17:22calculate a positive intrinsic value for a business that has
- 17:26negative owner earnings that. Makes a lot of sense.
- 17:29The report concludes that it's $67 a share there.
- 17:32There's absolutely no margin of safety.
- 17:35The market is currently pricing Intel's stock as if Lip Bhutan's
- 17:39turn around has already succeeded.
- 17:41They're jumping the gun. Completely, investors are
- 17:44pricing in the 18 A node working perfectly on day one.
- 17:47They're pricing in the foundry business suddenly winning huge
- 17:51external clients despite the trust deficit, and they are
- 17:54pricing in a return to massive profit margins.
- 17:56And if any one of those things fails or is even slightly
- 17:59delayed. The stock has massive downside
- 18:01risk. So people are buying the hype of
- 18:04the national importance, not the reality of the balance sheet.
- 18:07They're confusing a strategic national asset with a good
- 18:11economic business. That is the crux of the issue
- 18:13right there. Think about the railroad
- 18:14industry in the early 20th century.
- 18:16Railroads were absolutely vital strategic national assets.
- 18:20The country couldn't survive without them, but for a long
- 18:23time their massive capital costs made them terrible investments
- 18:27for everyday shareholders. So just because the US
- 18:31government desperately needs Intel to survive doesn't mean
- 18:34Intel shareholders are guaranteed to make money.
- 18:36Precise. That is such a vital point.
- 18:39The government cares about the microchips being made on US
- 18:41soil. They don't necessarily care if
- 18:43your individual stock portfolio goes up.
- 18:47The government might provide a floor against bankruptcy by
- 18:49offering subsidies, but they don't provide a ceiling for your
- 18:52profit margins. Exactly, Intel still has to
- 18:55actually earn those margins by out executing TSMC and NVIDIA in
- 18:59the open market, which is arguably the hardest task in
- 19:02modern technology right now. So for the discipline value
- 19:05investor, the strategy right now is just what sitting on your
- 19:08hands. Pretty much you watch the node
- 19:10progression, you watch to see if they actually sign a massive
- 19:12external foundry customer like an Apple or an AMD, and you wait
- 19:16to see if those gross margins actually climb back up toward
- 19:1950%. Right, until the underlying
- 19:21economics actually prove themselves in the real world.
- 19:23It's a massive, fascinating gamble, but it's not an
- 19:26investment. That's the final verdict.
- 19:29It's an incredible industrial drama, but it requires far too
- 19:32many optimistic assumptions to qualify as a safe value.
- 19:35Played today. Yeah.
- 19:36And that brings up one final overarching question to really
- 19:39think about as we watch this play out.
- 19:41What's that? What happens if this grand,
- 19:44deeply expensive plan doesn't work?
- 19:46Oh. Man.
- 19:48What if the 18 A process node falls behind schedule?
- 19:51Or what if the foundry business permanently fails to attract
- 19:54those big external customers because the trust deficit is
- 19:57simply too large to bridge? Will the US government simply
- 20:00let its own? Only National Semiconductor
- 20:02champion fail? Or could we eventually see a
- 20:06scenario where Intel is forced to split entirely?
- 20:08Splitting the company. Yeah, separating its highly
- 20:11profitable asset light chip design business from its heavy
- 20:15capital burning manufacturing fabs forever.
- 20:18That is a massive con and it really highlights the razor's
- 20:21edge Intel is walking right now. We will definitely be keeping a
- 20:24close eye on their progress over the next few quarters to access
- 20:28more advanced researches, applications and our latest
- 20:30public published personal intelligent multi agent Deep
- 20:33value intelligence Welcome to explore and join Deep values dot
- 20:37AI community Track total market temperature.
- 20:40Welcome to search and download iOS app Buffett indicator from
- 20:44Apple App Store. Disclaimer All analysis is based
- 20:47on publicly available sources and may be subject to revisions
- 20:50or differing interpretations. All contents from this channel
- 20:53is for educational and research purposes only and does not
- 20:55constitute investment advice. Conduct your own due diligence
- 20:59before making investment decisions.
- 21:01If you like to access wider topics, including life and
- 21:03invest in philosophy, please subscribe to our sister channel
- 21:06Deep Values. Don't forget to like this video.
- 21:08Much appreciated.