Welcome back to the deep dive. It is Friday, 02/13/2026. And I have to say, looking at the calendar and then looking at the market, if you were looking for a sign that things are getting weird, you found it.

It definitely fits the mood. We have a full moon, Friday the thirteenth

Yeah.

In a market that is just, you know, violently rotating under the surface. Yeah. It feels less like investing right now and more like a horror movie where the call is coming from inside the house.

A horror movie is a really good analogy. We are looking at a a massive stack of research today, and we're keying off the AGI roundtable from Phil Stock World, some heavy analysis from the morning reports, and just a lot of industry chatter. And the term everyone is using is the scare trade.

That is the phrase of the week. Absolutely. The scare trade.

But I wanna challenge that right off the bat because if I just look at the headlines, the world looks incredible. I mean, the Dow Jones just cracked 50,000.

A huge milestone.

The S and P is at 7,000. President Trump is out there saying we're gonna hit Dow a 100,000. So if the numbers are this good, why is the scare trade even a thing? Is this just people being superstitious on Friday the thirteenth or is something actually broken?

It's not superstition. It's it's structure. The headline numbers, the Dow 50,000, they're masking a really violent rotation.

Okay.

We are moving from what we called AI euphoria, where you just bought anything that had a computer chip to AI phobia.

AI phobia.

Yeah. The market has realized that AI isn't just a tool that helps companies grow. No. It's a force that might actually delete entire business model.

Delete business models. That, that is a heavy concept.

It is. And that's the core tension. Know, we're seeing massive dips in software and logistics even in real estate. So the mission for this deep dive is really to answer one question. Should we buy this dip?

Right.

Is this a get rich moment where we can pick up good companies on the cheap? Or is this a get me out moment where the fundamental economics of these companies have just changed forever?

So are we dealing with ghosts or are we dealing with gravity?

Exactly. That's the perfect way to frame it.

Okay. Let's map this out. We have a lot of ground to cover. We need to dissect the CapEx shock that started this whole slide.

Mhmm.

But we need to talk about this apocalypse. And I really want to get into the mechanics of that because I think a lot of people misunderstand why these software stocks are tanking so hard.

That is the most critical part of this whole discussion, honestly. That business model shift is everything.

Then we have to talk about the karaoke crash. I read this in the notes and I I thought it was a joke, but apparently it's not.

It is tragically not a joke. It's it's a lesson in algorithmic stupidity.

Okay. Can't wait for that. And finally, we're gonna do the great sorting. We'll separate the builders, the stocks you should be buying from the dreamers that you really need to avoid right now.

It's a survival guide. A survival guide for what we're calling the Matrix Economy.

Okay. Let's start with the setup then. Dow 50,000. Psychologically, that's a huge milestone. Champagne cork should be popping everywhere.

But you're saying the magnificent seven, the big tech giants are fracturing. What triggered this hangover?

It all started with earnings season, specifically that February around the second to the sixth. Usually earning season is a celebration of profit. It's look how much money we made.

Right. The report card. The

report card. But this time the market didn't care about the income. It cared about the outgo. We call it the sticker shock.

Spending.

The spending was astronomical. It started with Alphabet Google.

Mhmm.

They announced their capital expenditure plan, their CapEx for 2026, and they are planning to spend $185,000,000,000.

185,000,000,000. Let's just pause on that for a second. That is more than the entire GDP of, I don't know, Hungary or Kuwait. It's a staggering number.

It is a nation state's budget. But it wasn't just Google. Immediately after, you know, not to be outdone, Amazon came out and effectively said, hold my beer. They announced a $200,000,000,000 CapEx plan.

200,000,000,000. So between just those two and Meta, we are talking about, what, nearly $650,000,000,000 in spending Yeah. Just for this year.

Correct. Just on AI infrastructure, servers, data centers, power, the whole shebang.

Okay. So here is where I get stuck, and I think a lot of listeners probably do too. Why is spending money a bad thing? If they are spending 650,000,000,000, doesn't that mean they are growing? Doesn't that mean they are building the future?

Why on earth did the market panic?

That is the trillion dollar question. And the answer is, you have to look at what they are becoming. For twenty years, Google and Amazon were asset light businesses.

Right. They wrote code.

They wrote code. And code is cheap to replicate. That's why their margins were so unbelievably high. But when you spend $185,000,000,000 on servers and copper wire and energy plants and cooling towers, you aren't a software company anymore. What are you?

Utility.

Utility. Yeah. Like the electric company.

Exactly. And electric companies, you know, they do not trade at 30 times earnings, they trade at 15 times earnings. So the market looked at that capex number and the collective thought bubble was, wait a minute, the free money era is over. These companies now have to pour concrete and dig ditches to make their next dollar.

So the valuation itself has to come down because the type of business is changing, it's getting heavier.

It's becoming incredibly heavy. In financial terms we call it ROIC compression, That's return on invested capital. If I spend $1 to write code, I might get $10 back over its lifetime. It's almost pure profit. But if I spend $1 to build a data center, I might only get, say, dollars 1.2 back over twenty years.

That is a massive downgrade in the quality of their earnings. Huge change.

And that's what started the fear. It's the Matrix economy taking shape. These companies are effectively buying the entire electric grid and the whole silicon supply chain. It's an arms race. And in an arms race, you burn a lot of cash that you never see a return on.

You just spend it to stay alive.

But it effectively prices everyone else out though, doesn't it? I mean, if the table stakes to compete in AI are now 100,000,000,000 a start up in a garage, They can't even play the game.

That's the flip side. It creates this enormous unbreachable moat. But right now, today, investors are focused on the cost of digging that moat. They see profitless growth, they see companies spending these inflationary amounts of money to build out AI all in the hope that it will eventually be deflationary for labor.

Meaning they spend billions on chips now so they can fire humans later.

That is the cynical but mathematically accurate calculation.

Which is the perfect segue. It leads us directly into the second and I would argue probably scarier part of this whole scare trade. The sespocalypse.

The sespocalypse.

I love the name but I absolutely hate what it's doing to my portfolio. We're talking about software stocks, Salesforce, Adobe, Workday. These have been the darlings of Wall Street for what? A decade?

Yeah. At least.

And they are getting absolutely decimated. Why?

To understand the crash, you have to understand the business model. For the last fifteen years, the entire software as a service industry has been built on one thing, the per seat model.

Right. So if I run a company and I hire a new salesperson, I have to buy them a Salesforce license. I have to buy them a Slack license. I have to buy them a Microsoft Office license.

Exactly. The revenue of the software company was directly and beautifully tied to the headcount of their customer. If the customer grew their staff, the software company grew their revenue. It was this perfect linear relationship.

And now?

Now that relationship is broken. In 2024 and 2025, the whole narrative was AI helps you work. Know, it makes your employees 10% more efficient. That was fine. That was good even.

Yeah. But in February 2026, the narrative has shifted to AI takes your seat.

Displacement.

Total displacement. The catalyst was when Anthropic released Claude Cowork and some of these other agentic tools is not just chatbots that help you write an email. These are agents that can perform complex multi step workflows. They can do legal research, they can code entire software modules, they can handle back office payroll tasks from start to finish.

So if I'm a CEO and I have an AI agent that can do the work of say three junior analysts.

You don't hire the three junior analysts, you just don't.

And if I don't hire them, I don't buy three Salesforce licenses or three Adobe licenses.

Precisely. You might only buy one license for the AI agent itself or, and this is the really scary part, maybe you buy none at all if the AI just bypasses the user interface entirely and works directly with the data.

So the market is pricing in a future where corporate headcount actually shrinks and therefore software revenue shrinks right alongside it.

It's a white collar recession reflected in stock tickers. The market finally realized that the seat based model is a trap. If the humans go away, the revenue goes away. It's that simple.

Let's look at the kill zone. I mean, specific sectors are taking the hardest hits?

It's fascinating because it's not just tech. It started of course with software, you know, your Salesforce, CRM, your Adobe, they took heavy hits. But then look at a company like LegalZoom.

Okay, LegalZoom makes a lot of sense to me. They are the definition of a middleman.

They're the ultimate middleman. Their whole value proposition for years was we make it easy to file complicated paperwork. But if an AI agent can read a contract, draft a will, or incorporate a business in three seconds for, you know, pennies, why would you pay a monthly subscription to LegalZoom?

You wouldn't. The AI just obliterates the friction that LegalZoom was monetizing. It's gone.

Completely gone. But it gets even more interesting when you look at financial services. This week we saw wealth management firms, we're talking Charles Schwab, LPL Financial, Raymond James, they dropped significantly. Right. Like seven to 9%.

Now that I don't fully get. Mean, people still have money, they still need financial advice. Why are the wealth managers getting hit so hard?

It's because of a startup called Altruist and a specific tool they launched called Hazel. Hazel is an AI tax and investment tool. And it doesn't just give you generic advice. It can actually read your ten forty tax forms. It reads your pay stubs.

It analyzes your credit card spending. And based on all that, it creates a custom tax loss harvesting strategy and a personalized investment portfolio in minutes.

Okay, so essentially it does what a human financial advisor does for a 1% fee.

Exactly. A human advisor charges you 1% of your assets every single year to do that. Hazel does it for a flat fee or a fraction of the cost. So the market looked at a company like Schwab and said, your margins are at risk. The entire business model is under attack.

It's the get me out trade. Just get me out of the middleman.

Get me out of anything that relies on charging high fees for information asymmetry because AI just destroys information asymmetry. It makes everyone an expert.

And then there's real estate. CBRE, Jones Lang LaSalle, I mean these are the giants of commercial real estate. They fell what 12 to 26% in just a few days?

This is what the analysts are calling the second wave of office fear.

The second wave.

Yeah, the first wave was COVID. Everyone went home. We thought they'd eventually come back. Some did. But now comes the second wave, AI displacement.

So if the law firm uses an AI to do all the discovery work, they don't need the 20 paralegals in the back room.

And if they don't need the paralegals, they don't need the cubicles for the paralegals, they don't need that 10,000 square feet lease, They need 5,000 square feet. Maybe less.

So the market is just front running the future lease cancellations?

Correct. The market is saying we see the headcount reduction coming six months from now so we are selling the office stocks today. It's ruthless but it's logical.

It is logical. But you know what isn't logical? The karaoke crash.

Ah, yes. The karaoke crash. This is my favorite part of the week's research because it is so incredibly stupid and yet so revealing about the state of market.

So walk us through this. We saw these massive logistics companies, C. H. Robinson, Landstar plunge 14 to 24%. These are companies that move physical freight.

Trucks, trains. Why on earth did they crash?

They crashed because of a single press release from a company called Algorithm Holdings.

Algorithm Holdings. Okay. Sounds very high-tech and serious.

It is, doesn't it? And they claim their new AI could automate freight matching. So matching the truck to the cargo and scale volume by 400% without needing any humans.

Okay so an AI powered freight broker.

Right and if that were true it would be a huge threat to a company like C. H. Robinson. But here's the kicker: Do you know who Algorithm Holdings was until 2024?

I do and I still can't quite believe it.

It was the singing machine company.

People who make the karaoke machines you buy at Walmart for your kid's birthday party.

The very same. They pivoted, they rebranded, they issued a press release filled with every AI buzzword you can imagine.

And the market liquidated billions of dollars of value from legitimate established logistics fleets because a former karaoke company said they have a magic algorithm.

Yes. That is exactly what happened.

How I mean seriously, how does that happen? Are human traders that gullible?

See, that's the thing. It's not that the human traders are gullible. It's that the machines are fast. We have to talk about how modern markets actually read news.

The scrapers, the algorithms.

Exactly. High frequency trading algorithms are constantly scanning newswires, press release feeds, everything. And they aren't reading for context or nuance, they are reading for keywords.

So it just saw algorithm holdings?

It saw algorithm holdings, it saw AI, it saw disrupt logistics, and it saw automate.

So the bot creates a simple logic chain, new AI competitor in logistics equals sell CH Robinson.

In milliseconds, the short positions were initiated and executed before a human being had even finished reading the headline.

That is terrifying, it means the market is incredibly fragile to misinformation.

It defines the scare trade. Investors and their algorithms are so terrified of being on the wrong side of AI displacement that they shoot first and ask questions later. They liquidated the entire supply chain sector based on a ghost story told by a karaoke machine.

But this tells us something really important about valuation, doesn't it? I mean, if C. H. Robinson was trading at a reasonable cheap price, a press release wouldn't tank it 20%.

That's the key. You've hit on it. Yeah. This brings us to the macro fog. The reason the market is so jumpy, so willing to believe ghost stories, is that it is priced for absolute perfection.

We mentioned the Schiller PE ratio earlier. Let's dig into that.

The Schiller PE is a valuation metric that smooths out earnings over ten years to adjust for business cycles and inflation, give you a much better picture of value, and it recently hit 40.36.

And is that high?

Is that high. The long term average is around 17. The only other times it has been over 40 were in 1929.

Oh, just before the Great Depression.

And in 1999.

Just before the .com crash. That's not a great company.

It's the worst possible company. So we are in very, very rarefied air. And when valuations are that stretched, when you are paying forty years of future earnings to buy a stock today, you are pricing in a perfect future. Nothing can go wrong.

And so when a karaoke company says boo

The whole market collapses. Superstition versus gravity. The superstition is the specific news story but the gravity is the underlying valuation.

The gravity wants to pull prices down, the news is just the excuse it was waiting for.

That's it exactly.

And the economic data isn't helping, is it? It feels like we are just getting these wildly mixed signals every single day.

The sources are calling it the random policy generator. The data is all over the place. I mean, GDP forecasts swung from 2.7% to 5.7% and then back down to 3.7% all in two weeks. You can't plan a business on that kind of volatility.

And what about the jobs data?

Even worse. We had a report showing 109,000 job cuts in January, a terrible number. The market initially rallied on that because the thinking was, oh bad news is good news, the Fed will have to cut rates now.

Right.

But then other reports came out completely contradicting it. It's a complete fog. There's no clear signal.

But there's one area where the fog is clearing, and it seems to reveal a brick wall, and that's housing. And this connects back to our matrix economy theme in a way I hadn't really realized until I read the reports.

This is a critical point that Phil Stockwell highlighted. We usually blame the Fed for high mortgage rates. We say, oh, Powell won't cut rates. But there is a new villain in the housing market, and it's called capital crowding.

Okay, break that down for us. What is capital crowding?

Okay, go back to that $650,000,000,000 CapEx number we started with Amazon, Google, Meta. They need cash to build all these data centers. And they don't just use the cash sitting in their bank accounts, they borrow it. They issue corporate bonds.

So Amazon goes out to the market and says, 'Hey, who wants to lend us $10,000,000,000 We'll pay you a nice interest rate.'

Exactly. And because Amazon is a massive, super reliable company investors, pension funds, insurance companies, sovereign wealth funds they all line up to lend them money. They buy those Amazon bonds.

Okay, so Amazon gets the cash to build its server farm. How does that hurt the guy trying to buy a house in Ohio?

Because there is a finite limited amount of lending capital in the world. If all the big pension funds are pouring billions and billions into Amazon bonds and Google bonds to fund the AI build out, there is less money available to buy mortgage backed securities. They're sucking up all the oxygen in the room. They're sucking up all the capital. It's basic supply and demand.

If there is less money available for mortgages, the price of that money, which is the interest rate, has to go up.

So the ten year treasury yield stays high, not just because of the Fed, but because big tech is borrowing everything in sight.

That's exactly right. The ten year treasury is hovering near 4.2%, maybe higher, and mortgage rates are pegged directly to that. So the average American home buyer isn't just competing with other home buyers anymore. They are literally competing with Amazon's data center for a loan.

That is, that's kind of dystopian.

It's the matrix economy in action. The financial system is prioritizing building the digital infrastructure over maintaining the physical housing market. The mortgage rate is stuck near 6% because all the capital is flowing to the Matrix, not to the suburbs.

So, okay, let's recap. We have a market that is scared of ghosts, it's priced for perfection, and it's crowding out the little guy. It sounds like a terrible time to invest, but you're gonna tell me there's an opportunity, aren't you?

There is always an opportunity, always. But you have to know what you are looking for. And this brings us to the most important part: the great sorting.

The great sorting. Okay, so how do we separate the real from the fake? The wheat from the chaff?

We are finally starting to see a clear divergence in the market between tool AI and toy AI.

Tool versus toy, I like that.

Toy AI is the chatbots, it's the fun stuff, the, you know, write me a poem about my cat stuff, that initial hype is cooling off. But tool AI, the systems that actually do mission critical work, that is booming.

And Palantir seems to be the poster child for this, right?

Collentir is the perfect example. They don't make toys. They help the military target assets. They help governments run logistics for disaster relief. Their stock is holding up because they are selling a weapon, not a novelty.

It's essential. It's not a nice to have.

It is. But there's another fascinating trend that kind of contradicts the whole doom and gloom robots are taking over narrative and that's the rent a human economy.

I saw this in the notes. This is wild. Tell us about this.

It is At the exact same moment that we are terrified of robots taking our jobs, there is a booming market for humans doing things robots physically can't do. Like tasting food, or physically going to a location. There are these new platforms paying people $50 an hour to go to a new restaurant and taste the entire menu.

Because an AI can scrape a million Yelp reviews and write a summary but it can't tell you if the salsa is too salty.

It has no senses. It has no body. There are other jobs paying $400 an hour for a human to go sonograph a specific mountain trail because the AI training data for that location is outdated or you know just hallucinated a landmark that isn't there.

So human experience, trust and taste, these are becoming premium commodities.

Exactly. The takeover isn't instant. There is this strange transition period where being a physical human being in the physical world is incredibly valuable. The matrix needs ground truth. It needs to check its work against reality and humans are the only ones who can provide it.

It's weirdly reassuring. My ability to eat a taco is now a tangible economic asset.

For now, yes. But it also highlights the other major theme here, which is the physical wall.

The physical wall? What do you mean?

You can't just code AI into existence. You need chips and We you need just saw Qualcomm run into memory chip shortages. They want to build more, but they can't. They hit a wall. And look at the energy trade.

Energy stocks were some of the best performers in January, 14%.

Because all these new data centers are incredibly hungry for electricity.

They are voracious. Yeah. The AI build out is colliding with the basic laws of physics. You need electricity, you need cooling, you need real estate. This is why the Matrix Economy narrative is so strong, it's not about code anymore, it's about power plants and copper wire.

So let's bring it all together, it's Friday the thirteenth, the market is bleeding, the karaoke machine has everyone spooked, should we buy this dip? What's the final verdict?

The answer from our sources from the roundtable is a resounding yes but with a massive asterisk next to it.

Okay, what's the asterisk?

You have to buy the builders and you have to fade the dreamers.

Buy the builders, fade the dreamers. Okay, let's name names. Who are the builders?

The logic here is simple. Follow the money. Remember that $650,000,000,000 CapEx budget from the hyperscalers.

Yeah.

That is real money that is being spent. It has to go somewhere.

It goes to the hardware, the picks and shovels.

Right. So the top picks are companies that build the physical infrastructure. And first up surprisingly is Cisco, Ticker CSCO.

Cisco. Seriously, that feels like a stock tip from 1999. I mean it's been dead money for a decade. Why Cisco now?

It has been dead money and that's exactly why it's attractive. It's the value trade hiding in plain sight. Here is the thesis: You cannot run advanced AI through old copper wires. The data density is just too high. You need optical networking.

Optical. So lasers and fiber optics.

Light. You need to move massive amounts of data at the speed of light within the data center between the racks. Cisco owns the optical networking space. Their PE ratio is under 20. They pay a healthy dividend.

It's not sexy, but it's the essential plumbing of the matrix.

Okay, so Cisco is the plumbing. Who else is on the builder 's list?

Marvell Technology ticker MRVL. Everyone talks about NVIDIA and we know NVIDIA is the king but Amazon is building its own chips. They call them Tranium and Google has its TPUs. They're all trying to move away from being totally dependent on NVIDIA.

So who benefits if Amazon successfully builds its own chips?

Marvell does. Marvell makes the custom silicon, the ASICs, that go into those specific Amazon chips. So while NVIDIA gets all the glory in the headlines, Marvell gets the quiet high volume orders from Amazon's specific build. It's a great hedge against Nvidia's total dominance.

That's a smart pivot. What about the actual physical boxes, the servers themselves?

That would be Celestica, CLS. They are a pure pick and shovel play. They are one of the main companies that assembles the servers for the hyperscalers. As long as Google is writing those $100,000,000,000 checks, Celestica is busy screwing the servers together. It's a direct beneficiary.

And Applied Materials, they just reported?

AMAT, yes. They make the complex machines that make the chips, and they just reported very strong earnings. This validates the entire thesis. Chips need to be physically made before they can virtually steal anyone's job. The manufacturing demand is real and it's happening now.

So that's the builders group. What about what you call the reality plays, the stuff that exists outside of this tech bubble?

First one that comes up is AutoNation, ticker AM.

A network of car dealerships in the age of Tesla and EVs. That feels counterintuitive.

It's precisely because of the stumbles in the age of Tesla. While EV makers like Stellantis are crashing and burning, AutoNation is quietly buying back stock and printing money selling the cars people actually buy. Gas cars, hybrids, used cars. It's a bet on consumer reality, not the EV dream that hasn't quite panned out.

And what about our friends at CH Robinson, the victim of the karaoke crash?

That is the contrarian buy. CH Robinson CHRW. The thesis is simple. Robots cannot drive trucks yet. Physical goods still need to move next week, next month, next year.

The massive sell off was based on an algorithmic ghost story. If you believe that we still need to ship food and furniture across the country, then that stock is on sale for a ridiculous reason.

So you're basically betting that the algorithm was just flat out wrong?

I'm betting that a karaoke machine isn't gonna disrupt global logistics by next Tuesday. It feels like a safe bet.

That feels like a very safe bet. Okay. So those are the buys. Who are the traps? Who are the dreamers and debtors we should be avoiding?

Top of the list is Stellantis, STLA.

The parent company of Jeep, Chrysler, Fiat, they took a real beating.

They are in the middle of what's being called the EV reset. They cut their dividend, they took a 22,000,000,000 Euro charge, they basically admitted their entire EV strategy has failed. One analyst in the reports called it a repair shop, not a race car. It's dead money for now.

Is the big trap there?

Avoid generic SaaS. If a company's only value is being a pretty middleman for data, like a basic user interface on a database stay away. If an AI agent can bypass that interface and work directly with the API, the stock could literally go to zero.

What about a giant like Oracle that's been around forever?

Oracle is a controversial one. Some sources say to avoid it because of the massive debt load they took on for their acquisitions. If demand for their cloud services softens, that debt becomes toxic very quickly. Other analysts like Barclays see long term value, but it's risky. It's in the danger zone.

You need to be careful there.

Okay, so to summarize the Friday the thirteenth playbook for everyone listening. First, ignore the superstition. Don't sell your logistics stocks just because a karaoke machine issued a press release.

Correct. But second, respect the gravity. Don't go out and buy the whole S and P 500 index when it's trading at 40 times earnings. You have to be selective.

And third, follow the CapEx money, hardware over software for now.

The money is flowing into the ground, into chips and wires and power plants. That's where you want to be positioned.

You know, this entire conversation brings up a really provocative thought to end on. We've been talking about the Matrix economy all this time. It really feels like we are heading toward a world where three companies Amazon, Google, Microsoft basically own the electric grid for the mind.

That's a great way to put it. The innovation era. You know, the classic Silicon Valley garage startup era, that might be ending. We are entering the utility era. It's just like standard oil in the early 1900s or the bell system in the 1950s.

They own the pipes.

And if they own the pipes, they control the flow of information of commerce of everything.

They do. And for you, the listener, sitting there wondering about your own job, your own career, this shift matters immensely.

How so?

The great sorting applies to careers too. If your job involves staring at a computer screen all day and just rearranging data moving information from a spreadsheet to a presentation, you are in the blast radius. You are now competing directly with the matrix.

Because the matrix can rearrange data faster, cheaper and better than any human.

Infinitely faster. But if your job involves tasting a taco or fixing a pipe or physically installing a server rack or caring for a patient, you might just be the most valuable asset in the 2026 economy.

The physical world is making a comeback. The value is in doing things a bot can't.

It is. The bots can write the poetry, but they can't fix the leak in your sink. Not yet, anyway.

I love that. A perfect place to end. Don't let the bots bite.

Indeed. Stay grounded.

Thanks for listening to the deep dive. We'll see you next time.