Latest / Deep Values: Value Investing Deep Dives / 【Earning Deep Dive】NVIDIA's AI Monopoly vs. The Valuation Trap
Transcript
- 0:00Welcome back to the Deep Dive. It is Wednesday, February 25th,
- 0:042026. And I just have to say, looking
- 0:07at the markets right now, the energy out there feels, it feels
- 0:12electric. Absolutely electric or or maybe
- 0:14manic is the better word for. It yeah, Manic definitely fits.
- 0:17It really feels like we're looking at a landscape that is
- 0:19just completely different than it was even, you know, two years
- 0:23ago. It certainly does.
- 0:24I mean, the dust has finally settled from that initial AI
- 0:28explosion we saw back in 2023 and 2020. 4 Right that the hype
- 0:31cycle. Exactly the hype cycle and now
- 0:33we are actually living inside the structure that explosion
- 0:37built. It's, it's less about the, the
- 0:40wow factor of a chatbot, right, a poem for you.
- 0:43And it's much more about the the industrial grade reality of who
- 0:46actually owns the infrastructure of the world right now.
- 0:49Yeah. And today, for this deep dive,
- 0:51we are tackling the absolute juggernaut of that
- 0:54infrastructure. That the elephant in the room.
- 0:56Or maybe the T Rex in the server?
- 0:57The T Rex in the server room. I like that.
- 0:59We are of course talking about NVIDIA ticker symbol NVDA.
- 1:04It's the topic everybody has been requesting and honestly,
- 1:07it's the one we couldn't avoid even if we wanted to.
- 1:09You literally cannot talk about the global economy in 2026
- 1:13without talking about this specific company.
- 1:16No, you really can't. But I want to be incredibly
- 1:19clear right off the bat for everyone listening.
- 1:22We are not just talking about a, you know, a quote UN quote chip
- 1:25company today. Far from that.
- 1:27The sources we have in front of us, and we have a really massive
- 1:30stack of documents here to go through, they are calling this
- 1:34company the foundational architect of the global AI
- 1:37factory. Yeah, that is a, that is a very
- 1:40heavy title. It is.
- 1:41It sounds like something straight out of a sci-fi novel.
- 1:43The Foundational Architect. It it does sound like sci-fi,
- 1:46but when you look at the raw numbers we have from the close
- 1:49of fiscal year 2026, it's a title they have arguably earned.
- 1:54We're looking at revenue figures that honestly they sound like
- 1:57the GDP of a small to mid sized European country.
- 2:00It's just totally detached from normal reality, right?
- 2:03We aren't dealing with normal corporate metrics anymore.
- 2:06Yeah, we are dealing with nation state level economics here.
- 2:10It is genuinely mind bending. But before we get lost in the
- 2:13billions in the trillions today, and believe me, we are going to
- 2:16get very lost in them, we have a specific mission for this deep
- 2:20dive we do. We are going to track the market
- 2:23temperature of this absolute beast.
- 2:25Because despite these massive numbers, there is a a huge
- 2:29debate raging right now. Is this a bubble or is this
- 2:32literally just the beginning? That is the trillion dollar
- 2:35question. Literally.
- 2:37And Speaking of tracking market temperature, if you want to do
- 2:39that yourself right from your phone, I want to take a second
- 2:42to mention something really cool To track the total market
- 2:45temperature, you are welcome to search and download the iOS app
- 2:49Buffett Indicator. It's available right now in the
- 2:52Apple App Store. Yeah, it is a, it's a fantastic
- 2:54tool, give you that macro view, right, the total market cap
- 2:57versus GDP. It really helps you see if the
- 2:59water you're currently swimming in is, you know, boiling or
- 3:02freezing. Exactly, it is a fantastic
- 3:05reality check when things get as crazy as they are right now.
- 3:08And for those of you who really want to go down the rabbit hole,
- 3:11our contrarian friends out there who are listening and want to
- 3:13see exactly what the algorithms are seeing.
- 3:15If you want to access more advanced contents, welcome to
- 3:19subscribe on Deep values dot AI. Highly recommended.
- 3:23Yeah, that is where we will share our latest researches and
- 3:26the most advanced AI agent tools for our contrarian friends.
- 3:30So definitely check that out. And hey, don't forget to like
- 3:33this video or podcast wherever you're listening.
- 3:35Much appreciated. It really helps us out.
- 3:37Definitely. OK, now before we unleash the
- 3:40actual analysis, we have to handle the housekeeping.
- 3:42Right the the mandatory disclaimer.
- 3:45Please listen closely everyone. All analysis discussed today is
- 3:48based on publicly available sources and may be subject to
- 3:51revisions or different interpretations.
- 3:54All contents from this channel is for educational and research
- 3:57purposes only and does not constitute investment advice.
- 4:01Conduct your own due diligence before making investment
- 4:04decisions. Perfect legalities are handled,
- 4:07so let's unpack this massive stack of paper.
- 4:10Let's do it. Our mission today is to parse
- 4:12through what is frankly a very conflicting set of documents.
- 4:16On one hand, we have an institutional equity research
- 4:19report from Q4 of fiscal year 2026 and it is basically popping
- 4:24champagne bottles. Oh, it's incredibly bullish.
- 4:27It reads like a victory lap, but then on the other hand we have a
- 4:30separate report generated by the Deep Value Intelligence system.
- 4:33That's the AI powered multi agent framework and it is
- 4:36actively waving a yellow flag. In fact, it's waving a yellow
- 4:40flag and blowing a whistle at the exact same time.
- 4:42Right. And that is the central tension
- 4:44of the current market that we need to figure out for you guys
- 4:46today. Today we are looking at a
- 4:48company with record-breaking literally history making
- 4:51performance unprecedented execution.
- 4:54Nobody denies that. But the investment analysis is
- 4:56split completely wide open. One side, says Advocate, which
- 5:00basically means buy with both hands, leverage up and do not
- 5:04look back. And the other side says avoid or
- 5:07at best wait. And that's what is so incredibly
- 5:10confusing for a normal person looking at this.
- 5:12How can a company doing this well, executing this perfectly
- 5:16be in a void? Usually that avoid label is
- 5:19reserved for companies that are, you know, bleeding money or they
- 5:23have ACEO under federal indictment.
- 5:25NVIDIA is just printing money. It's because we have crossed a
- 5:28threshold. As the sources highlight, we
- 5:31have officially left the hype face.
- 5:33OK, explain that. The hype phase was the Wild West
- 5:35of 2023 and 2024. Everybody was just throwing
- 5:39money in anything with AI in the name.
- 5:41But now we are deep into what the sources call the proof of
- 5:45value phase. Proof of value, right?
- 5:47The question for 2026 is no longer is AI cool or can it pass
- 5:53the bar exam? The market doesn't care about
- 5:54that anymore. The question is, is it actually
- 5:57making money for the people who are buying the chips?
- 5:59The market is finally demanding a receipt.
- 6:02OK, that makes a lot of sense. So let's look at the actual
- 6:06industrial revolution happening right now on the ground.
- 6:08That is the exact phrase the institutional report uses, by
- 6:11the way. Quote a structural break from
- 6:14the traditional computing era and quote.
- 6:17It sounds dramatic, but it's accurate.
- 6:19But what does it actually mean, practically speaking?
- 6:22OK, think about the original industrial revolution.
- 6:24Back in the 19th century. We moved from manual human labor
- 6:28to steam power and mechanization.
- 6:31This report argues we are seeing the exact same structural shift,
- 6:35but in the realm of information. From manual to mechanized
- 6:38information. Exactly.
- 6:39For the last 30 years we lived in an era of retrieval.
- 6:43You type a query into Google and the computer searches a database
- 6:46and fetches a pre-existing document for.
- 6:48You right, it's basically a really fast digital librarian.
- 6:51Precisely. But now we are moving to an era
- 6:53of generation. The computer isn't just fetching
- 6:55a document, it is thinking, synthesizing and creating the
- 6:58answer on the fly. And to do that, to generate
- 7:01rather than retrieve, you cannot use the old computers, they are
- 7:04built for it. You need a completely new type
- 7:06of factory. And NVIDIA basically owns the
- 7:09factory floor. They don't just own the floor.
- 7:11They built the building, poured the concrete and hold the
- 7:14patents on the machines inside it.
- 7:16OK. Let's talk about the numbers
- 7:17that actually prove this because we just saw the Q4 fiscal year
- 7:212026 numbers come in and I want everyone listening to Brace
- 7:25themselves. Revenue hit $68.1 billion.
- 7:31And we need to emphasize that is just for 1/4.
- 7:34I really want to pause on that for a second.
- 7:35So it sinks in $68.1 billion in three months.
- 7:40That's, I mean, let's put some context around that.
- 7:42That is more money than Intel, AMD and Qualcomm used to make
- 7:47combined in an entire Goodyear back in the day.
- 7:50It's completely staggering that 68 billion represents a 73%
- 7:54increase year over year. 73%. And you have to remember, last
- 7:57year's numbers were already considered astronomical.
- 8:00Every everyone on Wall Street thought, OK, surely they cannot
- 8:03double again. The law of large numbers has to
- 8:06kick in eventually. Right, You can't keep crowing
- 8:08like a startup when you're the biggest company in the world.
- 8:10Exactly. And yet they nearly did it
- 8:12again. Wall Street was expecting around
- 8:1465 billion for the quarter, which was already an incredibly
- 8:18aggressive, almost euphoric target, and NVIDIA just blew
- 8:21past it by $3 billion like it was absolute nothing.
- 8:25It's the scale of the beat that gets me.
- 8:27A 73% jump on a base that large involves moving an
- 8:32incomprehensible amount of money through the global financial
- 8:35system. We aren't talking about some
- 8:37sauce startup going from 1,000,000 to 2 million in ARR.
- 8:42We are talking about moving the equivalent of a national
- 8:45economy. It really is.
- 8:46And what's critical to understand here, and the
- 8:48institutional report makes this very, very clear, is that NVIDIA
- 8:52has successfully transitioned its entire corporate identity.
- 8:54How so? Well, if you are listening to
- 8:56this and you still still think of NVIDIA as the company that
- 8:59makes the cool graphics card for your gaming PC so you can play
- 9:03Call of Duty at high frame rates, you are looking at a
- 9:06ghost. That company does not exist
- 9:08anymore. Wait, they aren't a gaming
- 9:09company at all anymore? Not really, no.
- 9:11They are a full stack ecosystem provider.
- 9:14They sell the custom silicon, the networking systems and the
- 9:18enterprise software. Their utility company now.
- 9:20OK, Let's breakdown where that 68 billion is actually coming
- 9:24from, because the revenue mix has changed completely over the
- 9:27last few years. I'm looking at the revenue
- 9:29breakdown table from the Q4 report.
- 9:31Right now, data center revenue is now wow, it is 88.27% of the
- 9:37entire business. Almost 90% of their revenue is
- 9:39just selling into data centers is the dominant engine.
- 9:43In fact, it's really the only engine that matters to their
- 9:45stock price anymore. But compare that to the gaming
- 9:48segment. I mean, I remember the green eye
- 9:50logo on all the PC game loading screens growing up.
- 9:52The way it's meant to be played, that was their whole identity in
- 9:55the 2000s and twenty 10s. It absolutely was.
- 9:58Now gaming is sitting at just 8.7% of the revenue pie.
- 10:03That's. Wild.
- 10:04It's totally eclipsed. It is shrinking now.
- 10:06To be fair, it's not necessarily shrinking in absolute dollars.
- 10:11Gamers are still out there buying high end RTX cards.
- 10:14But relatively speaking, compared to the absolute monster
- 10:18that is the data center segment, gaming is becoming a niche.
- 10:22It's almost like a side hobby for the executives at this
- 10:25point. A very lucrative side hobby, but
- 10:26a side hobby nonetheless. And looking at the other
- 10:29segments on this table, professional visualization,
- 10:31automotive, they are incredibly tiny.
- 10:35Automotive is sitting at just one point 3% of revenue.
- 10:38Yeah, that surprises a lot of people.
- 10:40It surprises me, I thought self driving cars and autonomous
- 10:43vehicle brains were going to be the next massive leg of growth
- 10:46for them. And they might be eventually,
- 10:47but right now, compared to the insatiable black hole level
- 10:51hunger of the AI data centers, the entire global auto industry
- 10:55is just a rounding error on Nvidia's balance sheet.
- 10:58A rounding error, literally. So when we talk about NVIDIA in
- 11:022026, we are effectively analyzing a monopoly utility
- 11:06provider for the global intelligence economy.
- 11:08That is the only lens that matters.
- 11:09They are the electric company, the water company, and the steel
- 11:13mill for the AI age all rolled into one single stock.
- 11:17Precisely. If you want to build an AI
- 11:19application anywhere on Earth, you essentially have to pay a
- 11:22toll to plug into their grid. OK, so who are the actual
- 11:26customers keeping the lights on at this giant factory?
- 11:29Because it's obviously not you and me buying a single card at
- 11:32Best Buy anymore. It's the hyperscalers, right?
- 11:34Yes, the BIG4, Amazon, Google, Meta and Microsoft.
- 11:39The customer concentration here is something you rarely see in a
- 11:42company this large. How concentrated are we talking?
- 11:45Well, the sources projected these specific companies
- 11:48collectively are going to spend between 650 billion and $700
- 11:52billion in AI related capital expenditures CapEx in 2026
- 11:57alone. I'm.
- 11:58Sorry, I need to stop you, $700 billion, yes, that is.
- 12:00That is an incomprehensible amount of money for corporate
- 12:03capital expenditures in a single year.
- 12:06Where is that money actually physically going?
- 12:09It's going into the physical earth, it's going into pouring
- 12:12concrete for massive new facilities, it's going into
- 12:14building dedicated power plants just to run the servers, and
- 12:18mostly it's going directly into Nvidia's pockets.
- 12:20It's basically a Cold War arms race, isn't it?
- 12:23It is exactly an arms race. Yeah, Microsoft is absolutely
- 12:26terrified that Google will get to AGI Artificial General
- 12:29intelligence first. Google is terrified that Meta's
- 12:32open source models will destroy their search monopoly, so
- 12:36they're all spending every single available dollar they
- 12:39have to stockpile weapons. And NVIDIA is the primary Arms
- 12:42dealer sitting in the middle selling to everyone.
- 12:45Exactly, and they are the only arms dealer that has the good
- 12:47stuff. But here is a really interesting
- 12:49nuance I noticed from the institutional report regarding
- 12:52the units they're selling. In the old days, analysts just
- 12:55counted chips, right? Like how many individual GPU's
- 12:58did NVIDIA sell this quarter? Right, it was a pure volume game
- 13:02of discrete components. But now the unit of measurement
- 13:05has fundamentally changed. They aren't just selling loose
- 13:09chips anymore. They are selling what the report
- 13:11calls AI racks and DGX pods. It says they are projected to
- 13:16ship nearly 30,000 AI racks this year.
- 13:19Can you paint a picture for me? What exactly is an AI Rack?
- 13:22Sure. Imagine a standard server rack
- 13:24you'd see in any IT closet. It's roughly the size of a large
- 13:27kitchen refrigerator. But inside this specific rack,
- 13:32it's not just a bunch of loose independent computers stacked on
- 13:35top of each other. It is a tightly integrated
- 13:38single supercomputer. OK, we're talking about the GB
- 13:41200 NVL 72 racks. These things are absolute
- 13:45beasts. They consume 120 kilowatts of
- 13:47power per single rack. 120 kilowatts.
- 13:50Wait for one rack. That's like, that's like
- 13:52powering a whole neighborhood of houses.
- 13:54It is, and because they draw so much power, they create an
- 13:57incredible amount of physical heat.
- 13:58You can't just put a fan on them and blow cold air down the aisle
- 14:01anymore or the silicon would literally melt into slag.
- 14:04They have to be entirely liquid cooled, plumbed with coolant
- 14:07like a car engine. And because of all that
- 14:10integration, these racks cost millions of dollars each.
- 14:14Millions for one refrigerator sized box.
- 14:17So they've completely moved up the food chain.
- 14:19Instead of just selling you the engine block, they are now
- 14:22selling you the entire luxury car, the tires, and maybe the
- 14:26road it drives onto. Exactly.
- 14:28And what that does is it massively increases their quote
- 14:30UN quote share of wallet. Share of wallet.
- 14:33Yeah, they aren't just getting the margin on the silicon chip
- 14:35anymore. They are getting the margin on
- 14:37the high speed networking cables, the networking switches,
- 14:40the specialized cooling manifolds, the software
- 14:43integration. They capture the profit from the
- 14:45entire stack. That sounds like an incredibly
- 14:48secure position to be in financially, but let's talk
- 14:51about the defense for a second. Every castle needs a Moat, and
- 14:55in the traditional tech hardware world, hardware is usually a
- 14:58terrible mode. Historically, yes, it's a race
- 15:01to the bottom. Right.
- 15:02Someone in Taiwan or South Korea always figures out how to
- 15:06reverse engineer the circuit board and make it 20% cheaper.
- 15:09We've seen it with memory chips. We've seen it with solar panels,
- 15:12flat screen TV's. Why hasn't that happened here?
- 15:15Why can't AMD or Intel just build a cheaper, faster chip and
- 15:20steal the hyperscaler market? That is the trillion dollar
- 15:23question that every bear tasks. And the answer, according to the
- 15:27sources we're reviewing, is what they call the toll bridge mode.
- 15:31And the crucial thing to understand is the toll bridge
- 15:33isn't actually made of silicon, it's made of software.
- 15:36Software. Specifically, a software
- 15:38platform called CUDAC. UDAA.
- 15:41K We hear this acronym constantly.
- 15:42We're talking about NVIDIA. Let's really unpack it for the
- 15:45non engineers listening. Is this software really that
- 15:47special? Think of CU Day as the
- 15:49foundational operating system for accelerated computing and
- 15:52video launched this platform way back in 2006 / 15 years ago,
- 15:56back when everyone else in the industry was completely asleep
- 15:59at the wheel thinking GPU's were only good for rendering video
- 16:02game graphics. They saw the parallel processing
- 16:04potential way before anyone else.
- 16:06Exactly. And over those 15 plus years,
- 16:09millions of software engineers, AI researchers, and Grab
- 16:12students learned to code their neural networks using CUA.
- 16:17All the major AI frameworks, things like π Torch or
- 16:19Tensorflow, were optimized to run on CED first and foremost.
- 16:23So it's like it's like learning a language.
- 16:25It is exactly like learning a language.
- 16:27Imagine you run a massive international business.
- 16:29All of your thousands of employees speak fluent French.
- 16:33All your legal contracts are written in French.
- 16:35All your factory instruction manuals are in French.
- 16:37OK, I'm with you. Then a competitor comes along
- 16:40and says hey, I built a fantastic shiny new office
- 16:43building down the street. It's 20% cheaper to rent than
- 16:46your current building and the elevators are a bit faster.
- 16:48Sounds like a good deal. But there's a catch.
- 16:51To work in my new building, all of your employees have to
- 16:54immediately stop speaking French and learn to speak Esperanto.
- 16:58Are you going to move your company?
- 16:59No way. The the sheer cost of retraining
- 17:02everyone, translating all the millions of documents, the
- 17:04downtime, it would completely bankrupt me.
- 17:07That is the concept of switching costs, and right now in the AI
- 17:11hardware space, the switching cost away from CUD is
- 17:15prohibitively high. The deep value report we're
- 17:18looking at specifically describes this as a wide and
- 17:21besieged Moat. Wide and besieged.
- 17:24I like that specific imagery, it sounds very medieval.
- 17:26It's a great metaphor. It means the Moat, the CED guy
- 17:29software lock in is incredibly wide right now.
- 17:32It's very, very hard for a customer to swim across it to a
- 17:36competitor, but because the profits inside the castle are so
- 17:40insanely high and we will get to those profit margins in a
- 17:43minute, everybody is currently beseeching the castle.
- 17:46Because there's so much treasure inside.
- 17:47Exactly. You have Google spending
- 17:49billions building their own custom TP use.
- 17:51You have Amazon building their trainium chips, they're building
- 17:54rafts. They're digging tunnels.
- 17:55They are trying to find any possible way around that CUDA
- 17:59mode so they don't have to pay Nvidia's toll.
- 18:00Right, but I have to play devil's advocate here for a
- 18:02second. If the toll is so painfully high
- 18:05and NVIDIA is charging such an insane premium for these racks,
- 18:09won't Amazon eventually just force their engineers to learn
- 18:13Esperanto? I mean, if you're spending $700
- 18:15billion collectively, you can pay for a lot of corporate
- 18:18language lessons. That is the ultimate long term
- 18:20threat. Yes, the Moat is not infinite.
- 18:23Software modes rarely are forever.
- 18:26But for 2026 the problem is speed.
- 18:29That toll bridge is currently the only reliable way to get
- 18:32your AI products to market fast enough.
- 18:35If you stop everything to build your own bridge from scratch,
- 18:38you lose the AI arms race entirely and none of the hyper
- 18:42scalers can afford to lose. So speed is the ultimate lock in
- 18:45right now. Speed is the lock in fear of
- 18:47missing out is the glue holding it together.
- 18:49Wow. OK, Speaking of new people
- 18:51trying to cross that bridge, there is a totally fascinating
- 18:53new customer tier mentioned in both of the 2026 reports.
- 18:57It's not just tech companies buying these chips anymore, it's
- 19:00entire nations. The report calls it Sovereign
- 19:03AI. Yes, this is a major structural
- 19:05shift that really started gaining traction in late 2024
- 19:09and absolutely exploded into 2025.
- 19:11What's driving it? Countries have woken up and
- 19:14realized that artificial intelligence is no longer just a
- 19:16cool tech sector. It is a matter of fundamental
- 19:19national security. They look at the landscape and
- 19:23they do not want their citizens private data or their government
- 19:26secrets, or even their cultural heritage residing in a cloud
- 19:30owned by AUS corporation, sitting on physical servers in
- 19:33Virginia or Oregon. They want data sovereignty.
- 19:36Exactly. They want their own native AI
- 19:39factories located physically within their own borders and
- 19:42subject to their own sovereign laws.
- 19:44The report highlights Japan and Vietnam specifically as nations
- 19:48that are investing heavily right now in domestic AI
- 19:51infrastructure. So they're just buying directly
- 19:53from NVIDIA? Yes, they're buying thousands of
- 19:55NVIDIA racks to build state owned or state sponsored
- 19:58supercomputing clusters. That sounds incredibly bullish.
- 20:02I mean, it sounds like a whole new massive market opening up
- 20:05and a government contract is usually the Holy Grail of
- 20:08stability for a corporation, right?
- 20:11But I noticed the Deep value record, the skeptical AI report
- 20:15we have. It actively flags the sovereign
- 20:18AI trend as a major risk. Why?
- 20:21Why is selling billions to a country risky?
- 20:23Governments usually have very deep pockets and they pay their
- 20:26bills. They do have deep pockets, but
- 20:28they are also incredibly fickle and more importantly, they are
- 20:31subject to geopolitics. The Deep Value analyst flags
- 20:35this specifically as political risk.
- 20:37Explain that for me. How does that impact the stock?
- 20:40Look at the high fence small yard policy that the US
- 20:42government implemented a few years ago.
- 20:44We have already seen this play out extensively with China.
- 20:47U.S. government can literally overnight decide that a certain
- 20:50country is no longer allowed to buy these high end AI chips
- 20:53because they might theoretically be used for military
- 20:55modernization. Export controls.
- 20:57Exactly. So imagine NVIDIA spends 2 years
- 21:01negotiating a $5 billion contract with a sovereign nation
- 21:06in, say, the Middle East or Southeast Asia.
- 21:09They build the racks, they prep the shipment, and then
- 21:11Washington decides that country is now on the naughty list.
- 21:13The Commerce Department just steps in.
- 21:15They step in and that $5 billion in revenue vanishes instantly.
- 21:19Poof. Gone.
- 21:21It's not a market force at all. Not at all.
- 21:23It is a pure political force. You cannot model it on an Excel
- 21:27spreadsheet. If you are an investor banking
- 21:29on 50 billion in sovereign revenue over the next three
- 21:32years and a political decision in Washington shuts that market
- 21:36down, Nvidia's earnings will miss expectations significantly.
- 21:40That adds a massive layer of complexity that just doesn't
- 21:42exist when you're just selling to, you know, Microsoft or Meta.
- 21:45Microsoft isn't going to get sanctioned by the US government.
- 21:48Exactly. The rules of engagement are
- 21:49totally different with sovereign states.
- 21:51OK, so we have the macro landscape pretty well defined
- 21:54here. We've got huge revenue growth, a
- 21:57massive structural shift in computing and this new sovereign
- 22:01AI pipeline coming online. Now we have to talk about the
- 22:04actual price tag of the company because this is where the real
- 22:08fight breaks out between our two sources, the valuation debate.
- 22:12Ah yes, and I want to get deep into the leads here because this
- 22:15is exactly where retail listeners end up losing money if
- 22:18they aren't incredibly careful. This is the ultimate bull versus
- 22:21bear cage match of our era. Right, so in the bull corner we
- 22:25have the growth investor perspective.
- 22:27The institutional report explicitly references the
- 22:29investment styles of legends like Peter Lynch and Stanley
- 22:33Druckenmiller, and their basic argument seems completely
- 22:35counterintuitive at first glance.
- 22:38They're essentially saying, quote the stock looks wildly
- 22:40expensive, but it's actually cheap.
- 22:42It sounds like a paradox. It does.
- 22:44How does that math actually work?
- 22:46Because if you look at a five year chart of Nvidia's stock
- 22:49price, it literally looks like a vertical line pointing straight
- 22:52to the moon. It does look vertical, but the
- 22:55growth argument relies almost entirely on a specific metric,
- 23:00the PEG ratio. PEG, which stands for the price
- 23:04to earnings to growth ratio. OK, let's unpack the PEG ratio
- 23:08for the 5th graders in the room, which definitely includes me
- 23:10sometimes when the math gets this big.
- 23:12Sure, so everyone listening probably knows the standard PE
- 23:15ratio price divided by earnings. APE of 20 simply means you're
- 23:20paying $20.00 for every $1.00 of corporate earnings.
- 23:24APE of 40 looks exactly twice as expensive.
- 23:27Makes total sense. But the PE ratio is just a
- 23:29static snapshot in time. It doesn't tell you how fast the
- 23:32car is moving. If you have a legacy bank
- 23:35growing its revenue at 2% a year and it has APE of 20, that might
- 23:39actually be really expensive. Because it's barely growing.
- 23:41Exactly. But if you have a hyper gross
- 23:44tech company doubling its revenue at 100% a year and has
- 23:47APE of 40, well, is that expensive?
- 23:49That's where the PEG ratio comes in.
- 23:51You take the PE multiple and you divide it by the earnings
- 23:53growth. Rate Adjust the price for the
- 23:55speed. Exactly.
- 23:56It contextualizes the multiple And what is the general rule of
- 24:01thumb for investors using PEG? Usually anything under a 1.0 is
- 24:05considered quote UN quote cheap or undervalued right?
- 24:08Correct. A PEG under 1 suggests that the
- 24:10company's earnings growth rate is actually higher than its PE
- 24:13multiple. OK, so where is NVIDIA sitting
- 24:16right now in February 2026? Well, depending on the specific
- 24:20metric you pull from the source documents, whether you are using
- 24:23trailing 12 month earnings or Ford projected earnings,
- 24:27Nvidia's PEG ratio is sitting somewhere between 0.52 and 0.79.
- 24:33Wait, hold on 0.52. Yes, that is.
- 24:36I mean for a mega cap tech stock that is incredibly low, that
- 24:40suggests it's essentially A bargain basement cheap relative
- 24:42to its growth. It is historic, yeah.
- 24:44It mathematically suggests that despite the stock price climbing
- 24:47vertically to these all time record highs, the actual net
- 24:50income earnings of the company have grown even faster than the
- 24:53stock price did. That's crazy.
- 24:55The growth analyst in the report calls this the earnings catch
- 24:59up. Let's say the stock price went
- 25:01up to 100% in a year, which sounds like a bubble, but if the
- 25:04actual bottom line, net income went up 300% in that same year.
- 25:08Then the stock effectively got cheaper while it was going.
- 25:11Up exactly the valuation multiple actually compressed
- 25:15because the fundamentals at ran the hype.
- 25:17That is the bull case in an absolute nutshell.
- 25:20Don't look at the sticker price of the stock, look at the speed
- 25:23of the earnings engine. OK, that makes sense.
- 25:25What else are the Bulls pointing to?
- 25:27They also point heavily to the rule of 40.
- 25:30This is a metric that is very often used by venture capitalist
- 25:33to evaluate software companies trying to balance their growth
- 25:36rate with their financial health.
- 25:38How do you calculate it? It's very simple.
- 25:40You take the revenue growth rate percentage and you add the
- 25:43profit margin percentage. If the combined total is over
- 25:4640, you have a very healthy, thriving business.
- 25:50OK. And where does NVIDIA score on
- 25:52the rule of 40? NVIDIA score right now is
- 25:54roughly 137. 137 on a rule of 40.
- 25:58Yes, it is elite tier. It is almost completely off the
- 26:02standard charts. Usually in the real world, you
- 26:05see software companies with incredibly high growth but zero
- 26:08profit margin because they're burning cash.
- 26:11Or you see legacy companies with high profit margins but very low
- 26:14growth. Right.
- 26:15It's usually a trade off. But NVIDIA breaks the standard
- 26:18financial models. They have massive triple digit
- 26:21growth and massive double digit margins simultaneously.
- 26:24And what about the technical traders?
- 26:26Do they like the setup? The report notes that
- 26:28technically, the stock looks very healthy.
- 26:31It is currently trading well above its 200, a simple moving
- 26:34average, which is sitting as a strong support base at $173.94
- 26:40and the Relative Strength index, the RSI, is sitting right at
- 26:4358.16. Which means it's not technically
- 26:47overbought yet. Correct.
- 26:49It has technical room to run higher before it overheats.
- 26:51That's the complete bowl case. Everything is firing on all
- 26:54cylinders, the growth is unprecedented, and the valuation
- 26:58metrics haven't actually caught up to the new reality of the
- 27:01earnings power. OK, honestly that sounds
- 27:03incredibly convincing. I'm almost ready to open my
- 27:06brokerage app right now. But wait, I can see the value
- 27:09investor, the Warren Buffett style persona in the deep value
- 27:12report just aggressively shaking their head in the corner.
- 27:16Why? What are the value guys seeing
- 27:19that the growth guys are completely missing?
- 27:21The Safe analyst or the Buffett style persona in the AI report
- 27:25looks at this exact same data completely differently.
- 27:29They operate using a series of strict investment gates or
- 27:32safety tests, and they argue that NVIDIA fails the very first
- 27:35gate, the predictability test. The predictability test, is that
- 27:39just a fancy way of saying, can you predict the future?
- 27:42Because if so, nobody passes that test.
- 27:44Well, it's a bit more specific than that.
- 27:46The test is, can you predict with high confidence what the
- 27:49specific industry landscape will look like in exactly 10 years?
- 27:53OK, if you look at a classic Buffett stock like Coca-Cola,
- 27:56you know with near absolute certainty that human beings will
- 27:58still be drinking carbonated sugar water in 2036.
- 28:02The fundamental technology of putting liquid in a bottle
- 28:04hasn't changed much in a century.
- 28:06The cash flows are highly predictable.
- 28:08True, it's a very stable business model.
- 28:10But if you look at NVIDIA and the semiconductor landscape,
- 28:14yeah. Can you tell me with high
- 28:16confidence what the dominant computing architecture will be
- 28:19in 2036? I.
- 28:20Have no idea. Nobody does.
- 28:22Will we even still be using traditional GPU's?
- 28:25Will quantum computing have taken over the data center?
- 28:27Will Google and Amazon has successfully moved entirely to
- 28:30their own proprietary in house silicon?
- 28:33Will some crazy new biology based neural computing be a
- 28:37thing? The tech industry basically
- 28:38reinvents itself from the ground up every five to seven years.
- 28:41Exactly. It is what the report calls a
- 28:43hard game. Innovation is incredibly rapid
- 28:47and rapid innovation is fantastic for us as consumers.
- 28:49We get amazing new phones and apps, but it's absolutely
- 28:52terrible for the kind of long term financial predictability
- 28:55that deep value investors crave. They want a sure thing and
- 28:59NVIDIA, by the very nature of its fast moving industry, is not
- 29:02a sure thing over a 10 year horizon.
- 29:04That makes total sense. And then reading further into
- 29:06the bear case here, there is the price to book ratio.
- 29:09I saw the specific number in the deep value report and I actually
- 29:11gasped out loud 50 times. Yes, the stock trades at nearly
- 29:1750 times it's book value. Let's make sure everyone
- 29:19listening understands the true gravity of that metric book
- 29:22value is basically the literal physical value of your
- 29:26factories, your unsold inventory, your cash in the
- 29:29bank. It's your tangible stuff, right?
- 29:32If you theoretically bankrupted and liquidated the company
- 29:36today, if you sold every office building, auctioned off every
- 29:39server, sold every patent, the cash you would have leftover is
- 29:44the book value. And right now, the stock market
- 29:46is paying $50.00 for every $1.00 of that tangible stuff.
- 29:50So you are basically paying $49.00 for pure air or well
- 29:55sentiment and future hopes. You are paying entirely for
- 29:57future earnings that have not actually happened yet.
- 29:59The expert analysis in the report has a great phrase for
- 30:02this. They call it a total absence of
- 30:04a physical floor. A total absence of a physical
- 30:07floor. Think about it.
- 30:08If the market sentiment suddenly shifts, if investors wake up
- 30:11tomorrow and decide they don't believe in the AI narrative
- 30:15anymore, there is no underlying physical asset base to catch the
- 30:19falling stock price. The stock could theoretically
- 30:22drop 80% in a panic, and at 10 times book value, value
- 30:26investors would still consider it expensive.
- 30:29That is legitimately terrifying if you're holding a large
- 30:32position. But wait, the bulls would
- 30:34counter that, wouldn't they? They'd argue that for a modern
- 30:36software and chip design company, physical book value is
- 30:40a useless metric because all their real value is in their
- 30:42intellectual property and their engineering talent, right?
- 30:45That is the standard counter argument, yes, but the bears
- 30:48point out that NVIDIA is increasingly becoming a massive
- 30:51hardware company again. They have physical inventory,
- 30:53they have complex physical supply chains for these massive
- 30:57racks, and paying A50X multiple for a hardware heavy business is
- 31:01completely historically unprecedented.
- 31:04OK, I see the risk there. And then the report brings up
- 31:06something called the quality trap.
- 31:09This is a concept I had honestly never heard before reading the
- 31:12source. Explain this to me.
- 31:14How on earth can quality be a trap for an investor?
- 31:17Usually we are desperate for high quality companies.
- 31:20It's a fascinating economic concept.
- 31:22We mentioned earlier that NVIDIA has roughly 56% net margins,
- 31:27yeah, Meaning for every single dollar they take in from a
- 31:30customer, they keep $0.56 as pure bottom line profit.
- 31:34Which sounds absolutely incredible.
- 31:36It is incredible. It is a sign of an unbelievably
- 31:38high quality business operation. So where is the trap?
- 31:41The value investor argues that profit margins that high act as
- 31:45a giant flashing neon sign for global competition.
- 31:48It's literally like dumping buckets of blood into shark
- 31:51infested waters. OK, I see where this is going.
- 31:54Capitalism fundamentally hates a vacuum, and it hates excess
- 31:58profit. When every other tech CEO on
- 32:01earth looks at Nvidia's earnings and sees them making $0.56 on
- 32:04the dollar, they say I want a piece of that action.
- 32:08It heavily incentivizes massive companies like Amazon, Google,
- 32:12Intel, and hundreds of well funded venture startups to spend
- 32:16billions and billions of dollars to attack those specific
- 32:19margins. So the trap for the retail
- 32:21investor is simply extrapolating the present into the future.
- 32:25Exactly. The trap is assuming those 56%
- 32:28margins are a permanent law of physics.
- 32:31If free market competition actually works the way it's
- 32:33supposed to, and over a long enough timeline it almost always
- 32:37does, those astronomical margins will eventually be forced to
- 32:40revert to the historical mean. Because competitors will offer
- 32:43cheaper alternatives to steal market share.
- 32:46Right. Let's say over the next three
- 32:47years, competition forces NVIDIA to lower prices and their
- 32:51margins drop from 56% down to say 30 percent, 30%.
- 32:54Still a great business. But if margins drop to 30%,
- 32:59their total net earnings effectively drop in half.
- 33:02And if the earnings drop in half, the stock price crashes,
- 33:05even if their overall top line revenue continues to grow.
- 33:09Wow, that is a very sober, very dark thought.
- 33:12The very thing that makes them a great company right now, those
- 33:15massive monopoly margins, is the exact thing that mathematically
- 33:19guarantees an army of well funded enemies is coming to
- 33:22destroy them. It is the ultimate paradox of
- 33:24corporate success. So we have this really intense
- 33:27evaluation debate. The growth team says it's cheap
- 33:30relative to its incredible speed.
- 33:32The Value team says it's wildly expensive relative to economic
- 33:36reality and the laws of capitalism.
- 33:38So let's look inside the company's own wallet to see who
- 33:40might actually be right. Section four of our dive The
- 33:44Financial Fortress. This is the one area where both
- 33:46the bulls and the bears have to sit down and agree the balance
- 33:50sheet of NVIDIA in 2026 is Immaculate.
- 33:53It is truly a thing of financial beauty.
- 33:54Let's talk about cash versus debt, because usually when you
- 33:57see a company growing at 70% a year, they're absolutely
- 34:00drowning corporate debt to fund that expansion, right?
- 34:02Not NVIDIA. They are currently sitting on a
- 34:05cash pile of $54.1 billion. $54 billion in cash just sitting in
- 34:13the bank. Yes.
- 34:15And that cash position is significantly higher than their
- 34:18total debt load. In fact, their debt to equity
- 34:21ratio right now is 0.08. 0.08 that's basically non existent.
- 34:26It's entirely negligible. They basically have 0 meaningful
- 34:29debt compared to the sheer size of their operations.
- 34:32And what about the Altman Z score?
- 34:34I absolutely love this specific metric for deep dives.
- 34:38For those who don't know, it's a formula that predicts the
- 34:40probability of a company going bankrupt within two years.
- 34:43Right, created by Edward Altman and the score for NVIDIA right
- 34:46now is 78.4. 78.4 OK context please.
- 34:50For context, any score over a 3.0 is generally considered to
- 34:54be in the safe zone. Any score below 1.8 is the
- 34:57distress zone, meaning bankruptcy is a real risk.
- 35:00So 3 is safe and NVIDIA is at 78.
- 35:03Yes, the expert analysis in the report.
- 35:05Literally calls this score nearly invincible in terms of
- 35:08traditional bankruptcy risk. That is hilarious.
- 35:11Nearly invincible. And their interest coverage
- 35:15ratio backs that up. It's sitting at 401 times.
- 35:19Meaning what exactly? Meaning they could theoretically
- 35:22pay the annual interest expense on their current debt 401 times
- 35:26over just using their operating income from a single year.
- 35:29They are an impenetrable financial fortress right now,
- 35:32but. And here is where I get a little
- 35:34skeptical again, putting on my value hat.
- 35:37They are taking all that fortress wealth and they are
- 35:39pouring it into stock buybacks. They are historically, you know,
- 35:42NVIDIA took every single spare dollar and plotted aggressively
- 35:46back into research and development.
- 35:48That's how they built the CUA mode in the 1st place.
- 35:51But in fiscal year 2026, they completely shifted strategy.
- 35:55They returned a record $41.1 billion directly to shareholders
- 36:00through stock buybacks. 41 billion in buybacks.
- 36:02And the report notes they still have a massive $58.5 billion
- 36:06authorization remaining in the chamber.
- 36:08OK, play the skeptic with me here for a second.
- 36:11Why on earth are they aggressively buying back their
- 36:13own stock when it is trading at absolute all time historic
- 36:17highs? Usually a smart management team
- 36:20buys back stock when the market crashes and they think their
- 36:23company is severely undervalued. If they are spending 40 billion
- 36:27buying at the absolute peak, are they just lighting shareholder
- 36:29money on fire or do they know something we don't?
- 36:32That is precisely the bear argument that executing massive
- 36:36buybacks at 50 times book value is the definition of long term
- 36:40capital destruction. They argue that cash should be
- 36:43saved for a rainy day or used for strategic acquisitions.
- 36:47What's the bull counter argument?
- 36:48The bull argument is much simpler.
- 36:50They argue that NVIDIA is currently generating so much
- 36:53free cash flow every single quarter that the management
- 36:56literally does not know what else to do with it.
- 36:58They cannot physically spend it all on R&D, even if they tried
- 37:01to hire every engineer on the planet.
- 37:03So they just give it back to the investors who own the stock.
- 37:05Right. But there is a much more
- 37:07strategic angle to this that the report highlights that remaining
- 37:11$58.5 billion authorization is absolutely critical for the
- 37:16listener to understand. You have to think of that
- 37:18authorization as a floor. A floor like a safety net under
- 37:22the stock price. Exactly like a safety net.
- 37:25Let's say the macro market suddenly panics, maybe inflation
- 37:28spikes, or there's geopolitical news and the NASDAQ queue drops.
- 37:33If Nvidia's stock price starts to rapidly drop during that
- 37:35panic, say a 10 or 15% correction, the company has $58
- 37:40billion of firepower ready to deploy.
- 37:43Oh, I see. They can step into the open
- 37:44market and buy billions of dollars of their own stock on
- 37:47the way down. That creates massive artificial
- 37:50demand. It supports the price.
- 37:52It acts as a structural stabilizer.
- 37:53It essentially means the company itself is standing there with a
- 37:57giant catcher's Mitt, ready to catch the falling knife so the
- 37:59retail investors don't have to bleed.
- 38:01That's a great way to visualize it.
- 38:03It reduces the downside volatility.
- 38:05OK. That is somewhat comforting if
- 38:06you're holding shares. So, summarizing, so far, they
- 38:08have infinite money. They have the CUDA software
- 38:12mode, at least for now. But let's pivot to the really
- 38:15crucial part, Section 5, the So what phase?
- 38:19The source material talks extensively about a turning
- 38:22point arriving in February 2026. This is where the rubber
- 38:26actually meets the road for the AI narrative.
- 38:28Yes, this brings us back to the macro analysis we touched on at
- 38:32the start. For the last three full years,
- 38:362023-2024 and 2025, the entire tech world has been stuck in
- 38:42build mode. Right, building the
- 38:43infrastructure. Exactly.
- 38:45Buying the physical racks, laying the fiber optic cables,
- 38:48securing the massive power contracts and training the
- 38:50foundational models. Everybody from Google to small
- 38:53startups was terrified of being left behind, so they bought
- 38:56absolutely everything NVIDIA could produce.
- 38:58It was pure FOMO spending fear of missing out.
- 39:00Massive, unprecedented FOMO spending.
- 39:04But now as we sit here in 2026, the report notes we are
- 39:08transitioning from build mode to use mode.
- 39:11And this is exactly where the systemic risk lies.
- 39:13The report calls it the CapEx Cliff.
- 39:16The CapEx Cliff, that sounds incredibly ominous.
- 39:19It is the biggest risk in the market right now.
- 39:22The source document includes a quote from an analyst that is
- 39:24quite stark. Quote.
- 39:26If they don't see a bottom line productivity explosion soon,
- 39:29their multi billion dollar orders will be the first thing
- 39:32cut. End Quote.
- 39:33When they say they, they're talking about the hyperscalers.
- 39:36Microsoft Meta, Google. Yes, think about it.
- 39:40Imagine a tense boardroom meeting at Google headquarters
- 39:43later this year. The chief financial officer
- 39:45looks across the table at the CEO and says, look, we spent $50
- 39:49billion on NVIDIA hardware over the last two years.
- 39:52Where exactly is the $50 billion in new bottom line profit to
- 39:56justify that spend? And if the CE OS answer is well,
- 39:59our internal chatbot writes slightly better marketing copy
- 40:01now and our search summaries are a bit faster.
- 40:04Then the CFO is going to slam their fist on the table and say
- 40:06cut the hardware budget immediately.
- 40:08We have enough compute. We need to monetize what we
- 40:11already bought before we buy a single rack more.
- 40:14And if Microsoft and Google both decide to cut that budget
- 40:17simultaneously? Nvidia's entire revenue growth
- 40:21story collapses overnight. Because remember what we
- 40:23discussed earlier? 40 to 50% of their entire
- 40:26revenue comes from just four or five massive companies.
- 40:29If those four companies unilaterally decide we have
- 40:32enough chips for now, the hyper growth narrative hits a solid
- 40:35brick wall. The stock price would adjust
- 40:38violently. So for the listener who is
- 40:40heavily invested or thinking about jumping in, we need a
- 40:43solid game plan. We can't just sit here and hope
- 40:46Google keeps spending forever. What exactly should we be
- 40:49watching in the coming quarters of 2026 to see if this CapEx
- 40:52Cliff is actually approaching? The Deep Value Intelligence
- 40:55Report actually provides a Fantastic Four point monitoring
- 40:58plan for exactly this scenario. Perfect.
- 41:01Let's run through that checklist right now.
- 41:03Item number one on the monitoring plan customer ROI.
- 41:06This is paramount. Do not just blindly watch
- 41:09Nvidia's quarterly earnings reports.
- 41:11You have to watch their customers earnings.
- 41:13You need to read the transcripts for Microsoft's earnings calls.
- 41:16Watch Meta's earnings. Yeah, what are we listening for?
- 41:18Are they explicitly, cleanly stating that their AI
- 41:22investments are driving measurable profitability?
- 41:25Or are they hiding those AI costs inside of vague other bets
- 41:30categories? If the hyper scalers are
- 41:32suddenly silent on actual AI profits, you need to be very,
- 41:36very worried about Nvidia's future orders.
- 41:38It's a great tip. OK, item number 2 on the
- 41:41checklist. Sovereign AI realization.
- 41:43We talked heavily about Japan and Vietnam earlier.
- 41:46The question here is, is that sovereign revenue actually real?
- 41:50What do you mean real? They announced it.
- 41:51Announcements are cheap. We need to see the actual cash
- 41:54hitting Nvidia's bank account. Are these sovereign mega
- 41:57projects actually breaking ground and taking delivery of
- 42:00the rags, or they're just vague memorandums of understanding and
- 42:04flashy political press releases? Right politicians love a press
- 42:07release. Exactly.
- 42:08Governments are world famous for grandly announcing billion
- 42:12dollar infrastructure projects that quietly get cancelled three
- 42:16years later due to budget constraints.
- 42:19Watch the realization rate. Makes sense.
- 42:21OK. Item number 3 on the monitoring
- 42:23plan margin mean reversion. This goes back to the quality
- 42:27trap we discussed. You must aggressively watch
- 42:30those 56% net margins every single quarter.
- 42:34Looking for weakness. Yes.
- 42:36If next quarter they slipped to 54% and then the quarter after
- 42:39that they slip again to 51%, that is the trap springing shot.
- 42:43It proves the Bears right it. Proves that aggressive
- 42:45competition is finally biting them.
- 42:48It means they're actively having to lower their prices to keep
- 42:51customers from fleeing to AMD or custom silicon.
- 42:55If you see sequential margin compression, that is a massive
- 42:58sell signal for the value investor crowd.
- 43:00And finally, item number 4 on the checklist Blackwell yields.
- 43:04Blackwell is Nvidia's absolute newest bleeding edge chip
- 43:07architecture, rolling out in 2026, and it is incredibly,
- 43:11almost unimaginably complex to physically manufacture.
- 43:14The physical physics of it are pushing the limits.
- 43:16Exactly. The key question is can they
- 43:19actually produce these new chips fast enough to meet the demand?
- 43:23There is a huge backlog of orders if TSMC that's their
- 43:27primary manufacturing partner in Taiwan, if TSMC has any
- 43:30production yield issues or if the advanced packaging process
- 43:34proves too difficult to scale. Then NVIDIA literally can't ship
- 43:37the product and they miss their revenue targets.
- 43:40It is a pure supply chain execution risk.
- 43:42Demand doesn't matter if you can't physically build the box.
- 43:45OK. We have thoroughly unpacked the
- 43:47macro landscape. We've debated the software mode,
- 43:49the valuation metrics, the financial fortress and the CapEx
- 43:52risks. It is finally time for Section
- 43:556, the final verdicts. The sources were very split on
- 43:58what to do with this stock. Let's summarize the positions
- 44:00and see exactly where we land. Let's do it so the fundamentals
- 44:04and growth team that Peter Lynch crowd their official verdict is
- 44:07advocate. Yes, they look at NVIDIA and see
- 44:10a classic GARP scenario growth at a reasonable price, or in
- 44:15this case, a wonderful company at a reasonable price.
- 44:17They firmly believe the CUA mote is insurmountable in the near
- 44:21term, the global AI demand is literally just getting started,
- 44:25and they view NVIDIA as the undisputed king of an entirely
- 44:29new technological era. To them, the massive growth rate
- 44:32completely justifies the premium valuation multiple.
- 44:36They probably think buying NVIDIA today is like buying
- 44:38Apple stock back in 2010 right after the iPhone launched.
- 44:41Exactly. The runway is years long in
- 44:43their view. But then we have the deep value
- 44:46Buffett style team. Their official verdict is avoid.
- 44:49They do say avoid emphatically. They failed the company
- 44:52completely on the 10 year predictability test.
- 44:55They think the current 50X book value price tag requires a full
- 44:57decade of absolute flawless perfection from management.
- 45:01And in the hardware tech world, flawless perfection over a
- 45:03decade is basically impossible. It's vanishingly rare they see
- 45:07absolute 0 margin of safety at these prices.
- 45:10To the value team, buying NVIDIA today is exactly like buying
- 45:14Cisco Systems at the absolute peakofthe.com bubble in the year
- 45:182000. Cisco was a truly great company
- 45:21that built the Internet, but if you bought it at the top, the
- 45:23price was so terrible it took 20 years to break even.
- 45:26A great company, but a terrible investment.
- 45:29OK, so we have advocate and avoid.
- 45:31What about the final synthesis? The Deep Values Intelligence AI
- 45:36system looked at both arguments and came to a final conclusion
- 45:40of wait. Yes, wait.
- 45:43And I think this is the most highly nuanced stance of the
- 45:45three and probably the wisest path for the average listener
- 45:48trying to navigate this volatility.
- 45:50Explain the weight thesis. Weight does not mean ignore the
- 45:53company, and if you hold it, it doesn't necessarily mean panic
- 45:56sell everything. It simply means wait for a
- 45:58significantly better pitch. I really like that classic
- 46:01baseball analogy. You are standing at the plate,
- 46:04but you don't have to swing at every single ball the pitcher
- 46:06throws. There are no called strikes in
- 46:08the game of investing. You can just stand there with
- 46:10your bat on your shoulder for years.
- 46:12Exactly. The synthesis report highly
- 46:15suggests waiting for a market price that actually offers a
- 46:18true margin of safety. Give me a specific target.
- 46:21For example, if the broader market corrects, or there's a
- 46:25sector rotation and Nvidia's forward PE multiple organically
- 46:29drops below 25 X, that would be a much better pitch to swing at
- 46:34or strategically wait for the inevitable period of
- 46:37disillusionment. The period of disillusionment.
- 46:39It sounds very poetic, but I imagine it's incredibly painful
- 46:42for the people holding the stock.
- 46:44It's always painful, but it always happens with every major
- 46:47technology shift. Eventually there will be a
- 46:49catalyst. Maybe it's a massive regulatory
- 46:51fine from the EU or sudden geopolitical scare regarding
- 46:55Taiwan, or just a bad quarter where they quote UN quote only
- 46:59grow revenue by 40% instead of 70.
- 47:01Right, a slight miss against impossible expectations.
- 47:04When that happens, the media narrative will aggressively
- 47:06shift overnight from AI is God to AI is totally overhyped,
- 47:10Retail investors will panic, and the stock will likely drop 30 or
- 47:1340% in a matter of weeks. And that that blood in the
- 47:18streets moment, that is when the true value investor finally
- 47:22strikes. Precisely.
- 47:23You don't chase the FOMO on the way up.
- 47:25You stock the prey, you build your cash position, and you wait
- 47:29patiently for the market to hand you a temporary discount.
- 47:32So the take away is discipline. This has been a truly incredible
- 47:35deep dive. Today.
- 47:37We have gone all the way from unpacking a $68 billion revenue
- 47:41quarter to debating the core philosophy of value investing.
- 47:45To wrap this all up, what is the single biggest take away for our
- 47:48listener today? The ultimate take away here is
- 47:51recognizing that it is entirely possible for a specific
- 47:54corporation to be a quote UN quote gold standard business
- 47:57operation, but simultaneously be a lead standard investment if
- 48:01the entry price is fundamentally wrong.
- 48:03That's a great way to put it. NVIDIA is a spectacular,
- 48:05generational company. It might literally be the best
- 48:07executing business in the world right now, but the stock price
- 48:10is a completely different animal than the company itself.
- 48:13As an investor, you must mentally separate the raw
- 48:16business execution from the hysterical market expectations.
- 48:19That is a crucial, crucial distinction to internalize.
- 48:23Just because the product they make is world changing doesn't
- 48:26magically mean the stock is a smart buy at any arbitrary price
- 48:30the market demands. Exactly.
- 48:32Price always matters. Price always matters, and here's
- 48:35a final provocative thought for everyone to chew on as we close
- 48:37out. The sources repeatedly mentioned
- 48:40that NVIDIA is essentially becoming the fundamental
- 48:42substrate of the new intelligence economy, like they
- 48:45are the digital silicon itself or the new electricity.
- 48:50If that premise is actually true, does traditional corporate
- 48:54valuation even apply here anymore?
- 48:56Or are we collectively trying to measure an entirely new kind of
- 48:59global utility, something the human race has literally never
- 49:03seen before, using old rulers from the 20th century?
- 49:06That is the most dangerous question in finance because
- 49:10historically, every single time someone stood up and said, hey,
- 49:13the old rules of valuation don't apply anymore, it's a new
- 49:15paradigm. They usually lost a catastrophic
- 49:17amount of money very shortly after.
- 49:19Truethe.com bubble is proof of that.
- 49:21But on the other hand, looking at the sheer terrifying
- 49:26capability of these new generative AI machines, the
- 49:31underlying technology really is fundamentally different this
- 49:34time. It is a massive tension that
- 49:36isn't going away anytime soon. And that exact tension is
- 49:39exactly why we watch the markets every single day.
- 49:42All right, a quick final reminder for everyone.
- 49:44If you want to track exactly where we are in this crazy
- 49:47cycle, if you want to see if the market feels this 1999 or 2010,
- 49:51make sure to check out the Buffett Indicator iOS app on the
- 49:54Apple App Store. It is a fantastic way to keep
- 49:57your head cool when the rest of the market gets irrationally
- 50:00hot. And for all the deep dive
- 50:01research documents and the advanced AI agent tools we
- 50:04mentioned today, definitely head over and subscribe at Deep
- 50:07Values dot AI. And hey, one last standard
- 50:10reminder for the road. Please conduct your own due
- 50:12diligence. We are just two voices in your
- 50:14headphones. At the end of the day, you are
- 50:15the only one pressing the buy or sell button on your.
- 50:18Account indeed, be careful and stay disciplined out there.
- 50:21Thanks so much for diving deep with us today.
- 50:23We will see you all next time.