$APLD

Explicit claim • strong conviction • basis: price target upside

And before we really get going, just a quick reminder for you listening, check out the VIP Algos link down in the description. We use their tools constantly for tracking institutional order flow and you know, sniffing out where the smart money is actually going. So yeah, the affiliate link is right down below, but all right, let's get into this because the biggest threat to the whole AI revolution right now, it it actually isn't some rogue supercomputer, >> Right. No, it's definitely not that. >> It's honestly, it's plumbing. >> Yeah, literally plumbing. >> Like we always talked about the cloud and AI like it's this invisible ethereal software, but the physical reality is that training these huge language models requires concrete, steel and just well, a terrifying amount of electricity and water. >> It completely shatters that illusion of the digital world, you know, people think software is weightless. >> Mhm. >> But when you're running, I mean, thousands of next-gen GPUs at maximum capacity, you're not just crunching data, you are creating a localized heat event. >> Like a literal oven. >> Yeah, you're basically running a small nuclear reactor in terms of thermal output. The physical infrastructure needed to stop these chips from just uh turning into molten slag, that's becoming the absolute bottleneck of the entire tech industry right now. >> Which is exactly why we're looking at Applied Digital today. Ticker symbol is APLD because they are sitting squarely at that physical bottleneck. They design and operate these highly specialized liquid-cooled data centers. But the financial data we're looking at, it's a massive contradiction. >> Oh, completely. It's wild. >> Right, because on one hand, you have year-over-year revenue growth of almost 130% and they're sitting on this contracted backlog of $16 billion. But then you look at the chart and it looks like a seismograph during an earthquake. I mean, it recently dropped almost 40% from its 52-week high. >> And that kind of violent price action really forces you to look under the hood. When a company is growing top-line revenue at 130% but the market just slashes its valuation by 40% in a few months, you're watching a real-time tug-of-war. >> Between what the future and the present? >> Exactly. Between astronomical future promises and the very gritty, very expensive reality of building these things today. >> Because they're bleeding billions right now just to pour the concrete. So, investors are trading entirely on future expectations. And as someone who looks at the numbers pretty cautiously, when a stock trades purely on expectation, you get this EKG of a heart attack chart. >> But before we get to the valuation swings, let's talk about the physical product because why can't Meta or Microsoft just, I don't know, plug their new AI chips into the data centers they already have? >> So, it all comes down to power density and thermal limits. Think about a traditional server rack, the kind that runs, you know, your email or host the basic website. That rack might pull somewhere between 10 to 15 kilowatts of power. And a standard air conditioning system pushing cold air up through the floor can handle that. It displaces the heat. But a rack of modern AI accelerators, those can pull 40, 50, sometimes over 100 kilowatts per rack. >> Wow, so pushing cold air on that is just what's I trying to cool down a campfire by blowing on it? Pretty much. It just doesn't work. The air itself physically cannot absorb and carry away the heat fast enough. At that density, air actually becomes an insulator, not a coolant. >> So, it traps the heat. >> Yeah. You literally have to move to liquid cooling. And we're not just talking about, you know, running a PVC pipe through the server room. >> Right, like a garden hose or something. >> No, no. This is precision engineering. They have to plumb dielectric fluids or treated water directly into these microscopic cold plates that sit millimeters above the actual silicon processors. So, building a facility that can handle miles of fluid piping, the pumping stations, the leak detection, and obviously the insane power draw. It's incredibly capital intensive. You can't just retrofit an old building for this. You have to build it from the ground up. >> And clearly the tech giants know they can't do this alone because looking at Applied Digital's operational metrics, the

Explicit claim • strong conviction • basis: fair value gap

It's now trading at a forward price to sales of just 4.3 which is near the lowest this stock has traded for according to this valuation metric going all the way back to late 2023. I also revised my discounted cash flow evaluation for the company today and I ended up revising this figure higher. The intrinsic value per share I calculated came to over $42 per share, which compares favorably to the current market price of just over $28.

Explicit claim • strong conviction • basis: price target upside

I think they're a really great long-term play and they're going to make that recovery at some point into $40 to $50 which is 100% gain. Well, average targets are $73 a share for APLD. So, as they're down here to 25, 26 bucks a share

Explicit claim • strong conviction • basis: explicit term

So $36 billion the $36 billion of contracted revenue translates to something on the order of $30 billion of lifetime net operating income on a $8.5 billion company. So if you want to just look at net operating income NOI, $8.5 billion market cap, $30 billion NOI.