This Before Investing in Vertiv Stock

[00:00:00] Speaker 1: This is an NVIDIA AI chip, just one of them burns over a thousand watts, that's more power than your microwave, running a full blast, 24 hours a day, 365 days a year. Now stack 72 of them into a single cabinet, a rack, and that one rack pulls about 130 kilowatts, that's roughly the power of 100 American homes squeezed into the footprint of a refrigerator. And here's the thing about power, every watt that goes in counts back out as heat, so AI has created two massive physical problems. How do you deliver that much electricity into a building without anything failing, and how do you remove that much heat before the chips literally cook themselves? There's one company that sits exactly in that layer, it doesn't make chips, it doesn't write software, it makes a power and cooling system that keeps the AI boom alive, it's called Vertiv, and the stock is up over 140% in a year. So today, what does Vertiv actually does? Why AI data centers can't function without this stuff? What does the number really say? And the one question that decides everything, is this a hidden winner, or is the stock already prized for perfection? Let's get into it. This is for educational purposes, not for financial advice. First, when you ask CheckGPT a question, you'll request travel to a real physical building, a data center, a warehouse, sometimes the size of several football fields, packed wall to wall with racks of hot, humming computers, and a data center is basically a living organism. Stay with me, because this analogy is going to carry us through the whole video. The servers, the computers doing the thinking, that's the brain. The electrical system feeding those servers, that's the bloodstream. Power comes in from the grid, gets converted, cleaned up, and distributed to every single rack. If blood stops flowing for even a second, the brain dies. And the cooling system, that's a loss. All that electricity turns into heat, and the heat has to be exhaled out of the building continuously forever. Here's why this matters so much. Downtime. The industry war for when a data center goes offline, is catastrophically expensive. For a large operator, an outage can cost millions of dollars in minutes, lost transactions, broken AI chain runs, angry customers. So data centers are built with an almost paranoid level of redundancy, backup for the backup for the backup, batteries, generators, duplicate cooling loops, monitoring systems, watching everything in real time. And that paranoia is the business opportunity. Somebody has to design, build, and service the bloodstream and the lungs. For decades, that was a sleepy and glamorous industry. Then AI showed up and broke everything. For 20 years, a typical server rack drew maybe 8 to 12 kilowatts. The industry average today, across all data centers, is still under 10. And you cool it the boring way, with air conditioning, cold air in the front, hot air out the back. Done. An AI rack draws over 10 times that, 120 to 130 kilowatts, and the roadmap gets crazier. Nvidia's future racks are headed towards 600 kilowatts, and eventually a full megawatt, one rack, one million watts. At those densities, air cooling just stops working, and it's pure physics. Air is a terrible carrier of heat. Trying to cool a 130 kilowatt rack with air is like trying to put out a bonfire with a hairdryer set to cool. You can't move enough air fast enough. And even if you could, the fans alone could eat a fortune in electricity. Water, on the other hand, liquid, carries heat thousands of times more effective than air, which is why the industry is racing towards liquid cooling. Here's what that actually means in plain English. You attach a metal plate with tiny channels, a cold plate, directly onto the chip. Coolant flows through it, absorbs the heat right at the source, and carries it away through pipes. It's essentially a car radiator, but for a GPU worth $40,000. Now here's a sentence I want you to remember, because it's the entire investment thesis. AI didn't just create demand for more data centers. It made every data center harder to build. More power per rack, more heat per rack, more complex plumbing, thicker electrical arteries, more monitoring, more service. The infrastructure inside each megawatt got more valuable. And that's exactly where Vertiv lives. Vertiv isn't a startup riding the hype. Its roots go back to a company called Libert, which has been cooling computer rooms since the mainframe era. It spent years inside Emerson Electric, got spun out, and listed on NYSE in 2020. Highquarters in Ohio, operations over 130 countries. So what do they sell? Let's go back to the body. The heart, the UPS. That stands for Uninterruptible Power Supply. It's a giant battery system that sits between the grid and the servers. When utility power flickers, even for half a second, the UPS instantly carries the load until the backup generators spin up. No UPS, no data center. Period. The other is Power Distribution. This is the gear that moves electricity safely through the building. Switch gear, which is basically the industry-grade circuit breakers. And busway, which has the big metal power rails running above the racks. The lungs. Thermal management. This is Vertiv's crown jewel. Everything from traditional precision air conditioning to full liquid cooling systems, including the CDU, the coolant distribution unit. Think of the CDU as the pump station of the cooling system. It pushes coolant to hundreds of chips and exchange the heat into the building's water loop. As AI racks get denser, CDU's go from optional to mandatory. And the doctors. Services. Vertiv have over 5,000 field engineers who install, maintain, and monitor all this equipment. Service is stickier and higher margin than hardware. Once your gear is in the building, you are probably calling Vertiv for the next decade. One more thing, because it matters. Vertiv is co-designing next-generation 800-volt power systems directly with NVIDIA, timed to NVIDIA's future chips. When a company designing the chip picks you as a power partner, you are not a bystander to the AI roadmap. You are on it. The business logic in one line. Vertiv makes money per megawatt of data center. And AI is making every megawatt more complex and more expensive. They win twice. Okay, great story. Do the numbers back it up? Short answer, yes, loudly. Latest quarter, reported April 2026, revenue $2.65 billion, up 30% year-over-year. Earnings per share up 83%. Notice something? Profits growing almost three times faster than sales. That's called operating leverage. And here's a simple version. The factories are built. The engineers are hired. The overhead is paid. So every extra order drops mostly straight to the bottom line. Operating margin hit 20.8%, up more than 4 points in a single year. And management is targeting 27% by 2030. Now my favorite number in this whole story: orders and backlog. Backlog means work customers have already ordered and paid contracts on. But Vertiv haven't delivered yet. At the end of 2025, it hit $15 billion, more than double the year before. And in the fourth quarter, the book-to-build ratio hit 2.9%. For every $1 of products shipped out the door, customer ordered almost $3 of new stuff. Imagine a restaurant where, for every meal served, three new parties joined the waitlist. That's Vertiv right now. So who is placing those orders? Mostly hyperscalers. That's the industry term for the giants building AI at massive scale. Microsoft, Amazon, Google, Meta. Combined, they have guided to roughly $700 billion of capital spending in 2026. Most of it on data centers. Vertiv got a cut of nearly every one of those buildings. Full-year guidance, $13.5 to $14 billion in revenue, roughly 30% organic growth, and over $2 billion in free cash flow, with almost no debt. So, here's the honest part. Two yellow facts. One, Vertiv recently stopped disclosing its backlog number. Companies usually hide metrics when optics get less flattering. So watch that. Two, concentration. Analysts estimate a handful of hyperscalers may account for close to half of Vertiv's revenue. When your biggest customer can be counted on one hand, one changed mind changes your year. Next, competition. Fair question. If this market is so good, why not just buy Schneider Electric or Eaton? Because of focus. Look at the landscape. The giants, Schneider, Eaton, ABB, Siemens, are bigger and more diversified. But that's the point. Data centers are one slice of their empire. For Vertiv, it's basically the whole company, around 80% of revenue. The air conditioning players, chain, carrier, drones and controls. They know cooling, but they are buying their way into liquid cooling through acquisitions. But they don't have the power side. The specialists, Invent, Modine, various liquid cooling startups, are fast but narrow. One product line, no global service army. Vertiv's pitch is that it's the only company at scale offering power plus cooling plus monitoring plus service. When you are reasoning to open a billion-dollar AI facility, buying the whole nervous system from one vendor is worth paying for. The honest counterpoint? Everyone can see this market now. Competition is flooding in. And some hyperscalers are even experimenting with designing their own cooling in-house. Vertiv is the leader in its lane. But the lane is getting crowded. So here's the real debate. And I'll tell you upfront, it's not about the business. It's about the price. The bull case. AI infrastructure is a structural decade-long build-out. 700 billion dollars a year and climbing. Every generation of chips get denser, which makes Vertiv's products more essential, not less. The cooling is still early, around 10% of Vertiv's revenue today. With most of the world's rack still on air, margins are expanding. Cash is gushing. The unvidious partnership locks them into the roadmap. And the backlog gives over a year of visibility. That's a classic picks and shovels compounder. The bear case? All of that might be true, and already in the stock. Vertiv's trees are roughly 46 times next year's earnings. Schneider eaten mid-20s to low-30s. You are paying double the going rate for industrial companies. And data center spending has always been cyclical. Booms, then digestion. If even one hyperscaler pauses because of power shortages, permitting delays, or a CFO asking where the AI revenue actually is, that 2.9 book to build unwinds fast. At customer concentration, rising competition, the risk that liquid cooling becomes a commodity, and the stock price for perfection has a long way to fall on imperfect news. So frame it this way. To own Vertiv here, you have to believe roughly 20%+ growth continues for years, margins keep climbing toward 27%, and hyperscaler spending doesn't take a single serious breather. That might happen, but that's a bet, and you should know you're making it. If you're following this stock, here's your dashboard for the next few quarters. Green lights to look for. Margins grinding toward that 27% target. Europe recovering. It's been the weak spot. And continued guidance rises. Yellow. Order growth normalizing. A 2.9 book to build can't last forever. The question is how gently it lands. Reflex. Any hyperscaler cutting hypex guidance or any earnings miss. Because at this valuation, misses get punished brutally. Here's the takeaway. Everyone stare at the chips. But the AI risk isn't just one in silicon. It's one in power and heat. Somebody has to build the bloodstreams and the lungs of every AI factory on earth. And right now, Vertiv is the purest way to own that layer. Real company, real growth, real cash. The only thing up for debate is the price tag. And now you know exactly what you'll be paying for. If you find this video helpful, please subscribe, and I'll see you in the next one.