IREN Stock Is a 5X Bet on AI Cloud. Can It Deliver?

IREN stock jumped 20% the day analysts cut their targets

0:00 Iron, the AI cloud company, jumped about 20% on November 6th to around $21. And that was the same morning Wall Street cut its price targets. So, new ones are $47 from Cantor Fitzgerald and $56 from HC Wayne. So, even after the cuts, the two newest targets on the board are more than double the stock price. And the day before, iron had hit a 52- week low. But if you think analysts are just slow to admit they were wrong, you might be right. Because before the report, those same two firms had targets of $99 and $74.

0:25 And buried in the earnings call is one detail that either makes this the best setup in AI right now or a warning sign. And I'll show you exactly which way I lean. Picture a guy who buys a fleet of really expensive food trucks with borrowed money and then rents them out to restaurants that are desperate for more kitchens. That's basically iron. My name is Blake. I dig up stocks before the bull stampede so that you're not buying at the top like WJack.

0:56 And today on 9to5 Investing, we are finding out whether those new targets are real. Iron started out as a Bitcoin miner, which basically just means it built giant data centers full of computers and cheap power, and it used them to earn Bitcoin.

From Bitcoin miner to AI cloud, and the Microsoft deal

1:06 Then AI showed up, and it turned out those same buildings and power hookups were exactly what AI companies were starving for. So Iron started filling them with Nvidia's AI chips and renting that computing power out to AI companies. And that rental business is what it calls AI cloud. The biggest renter so far is Microsoft. Last November, Iron announced a deal with Microsoft worth about 9.7 billion over 5 years, and Iron says the first part of that is now up and running.

1:30 Here's the problem. Those chips cost a fortune, and Iron has to buy them up front, long before the rent comes in. That's a big part of why the stock went from about $70 at its high to about $17 at its low on November 5th. On November 5th alone, it fell about 17% on a day when the whole market sold off over fears that the midterm results would slow down data center building. Wojack bought near the top because he saw Microsoft and stopped reading.

IREN earnings: revenue, AI Cloud and the loss

1:56 And then iron reported its earnings that night. Revenue for the quarter came in at 228 million, up 66% from the quarter before and a little ahead of the roughly 221 million analysts expected. The AI cloud business brought in 200 million of that, up 182% in a single quarter. And Iron says it keeps 86 cents of gross profit on every dollar of that revenue. Its loss per share was 54, which is still a loss, but analysts expected 72. So it lost less money than they thought.

2:19 And iron says it has about $20 billion of contracted revenue lined up for the future, which it calls its backlog, up 22% from the quarter before. In food truck terms, the trucks are rented out, the rent is coming in, and there's a long waiting list. So why did the analysts cut their targets? Because the trucks are showing up late.

Why analysts cut their IREN price targets

2:39 Iron had been aiming for more than $4 billion of yearly contracted revenue by the end of 2026 and now it expects to hit that in January 2027, maybe February 2027. The company blamed a customer changing its design and suppliers running behind. From what I can tell, Cantor's cut was mostly about that delay. But running a little late is not the same as a broken business.

The $4 billion goal and the napkin math

3:00 So, how big is the gap between where Iron is and where it says it's going? Right now, the AI cloud business is bringing in about 200 million a quarter. So call it about 800 million a year if you just multiply by four. The goal is more than four billion a year of contracts. So Iron is promising to grow that rental business about five times over in just a few months. And that's really the bet you're making. The Microsoft deal alone works out to almost $2 billion a year if you spread it out evenly over 5 years.

3:22 So that gets iron about halfway there. So here's some napkin math. And this is my math, not the company's math. At around $21 a share, the whole company was worth about 8.5 billion on the stock market. And if you add the roughly 4.4 4 billion of net debt, you're paying about 12.9 billion for the whole business, debt included. So, if Iron really does reach $4 billion a year, you'd be paying a little over $3 for every dollar of yearly revenue. For a business that keeps 86 cents of gross profit on every dollar, that would look really cheap in my opinion.

3:52 If the contracts hold and the customers keep paying, but that is a big if and it's the same if that the analysts are wrestling with. The bigger question is how Iron pays for all this.

The bill: $15 billion of spending and $4.4 billion of net debt

4:01 So, let's look at the bill. Iron plans to spend about $15 billion this fiscal year, mostly on chips and buildings. And it says about 9 billion of that is already covered by cash and financing. But its cash fell by about $2.4 billion in a single quarter. And its net debt, which basically just means debt minus cash, went from about 216 million to about 4.4 billion. That is a lot of trucks bought on credit. Which brings us to the detail that I promised you.

The 90% prepayment detail

4:26 After the quarter ended, Iron signed a new 5-year AI cloud contract. And the customer is prepaying about 90% of the cost of the chips. Think about it. That's like a renter walking in and paying for 90% of the food truck before he ever gets the keys. For Iron, that's amazing because for that deal, it barely has to borrow to buy the chips. For comparison, Microsoft prepaid about 20% on its deal. And Iron says it wants about 90% of its ship spending covered by financing and prepayments combined.

4:52 But why would a customer hand over that much money upfront? On that call, Iron's coco Daniel Roberts said customers with lower credit tend to make bigger prepayments. So renter who pays for most of the truck upfront might really love that truck or maybe no bank would lend him the money. So I would lean cautious on this one and that is the question that I would like answered before getting excited. Positive Pete reads that and says great, the customer is funding the growth.

5:18 Negative Nelly reads the same sentence and asks who the customer is and what happens if he stops paying. And they both can be right. Now, to be fair, if a customer like that stopped paying, Iron would already have most of the chip money in hand, depending on how the contract is written. And that's a lot better than being stuck with a loan on hardware nobody is renting. But chips, they get older every year. And a 5-year rental only works if somebody is still paying the rent in year 4.

5:41 So, now let's talk about those targets. A price target is just an analyst guess at where a stock could trade in about a year.

IREN price targets: Cantor $47, H.C. Wainwright $56, Goldman $30

5:46 If their assumptions about growth cost and share count all work out before the report. The average target was around $70, which was almost four times the price. So, the gap was already huge. So, here's what happened on November 6th. Caner Fitzgerald cut its target from $99 to $47 and kept its overweight rating, which is basically a buy. HC Wayright cut from $74 to $56 and kept its buy rating. And Goldman Sachs, the most skeptical of the big names, kept its neutral rating with a target of $30.

6:12 So, the cuts were big and they were honest about the delay, but even the lowest of those three is above where the stock closed. Chad saw $56 and started looking at boats. Trad saw $30 and put the brochure back in the drawer. Now, as a former financial accountant and auditor, the first thing I'd want to know is whether any of these banks helped Iron raise money because a bank that earned fees selling the stock might not be the most neutral opinion.

6:38 I haven't checked every one of Iron's financing deals for this video. So, take each target as one opinion, not a promise. So, what is the bullcase in plain English?

IREN stock bull case vs bear case

6:50 Iron already has the buildings and the power, which are the hardest things to get right now, and it's renting out chips at very high margins to customers who are lining up. If it hits that $4 billion goal early next year and keeps getting customers to prepay, the business could look very different from a stock that just touched a 52- week low. And here's the bare case, because I am not going to soften it. The debt went up fast. the big goal slipped and the newest deal leans on customers who may not be able to borrow on their own.

7:13 If AI spending cools down or the anti-data center talk after the midterms turns into real rules, a company that borrowed to buy chips is exactly the kind that gets hurt first. And a stock that fell about 70% from its high can fall a lot further even after a great day. Just so you know, I own Nvidia stock and Nvidia makes the chips iron rents out. But I don't own Iron. So here's what I'd watch in the next report. However, you already own it or you're just tempted.

What I'd watch in IREN's next report

7:40 First, the $4 billion goal. If iron hits it in January, like it says, the delay really was just a delay. Second, the debt and prepayments. If more of the chip spending gets covered by customer cash instead of new loans, that's a good sign. And then the share count. If it climbs a lot, Iron is selling new stock to pay for chips and your slice gets smaller. So, if you do buy, you might consider keeping it small enough that one bad quarter doesn't wreck your year.

8:02 Would you buy a stock whose customers pay upfront, knowing why they might? Tell me in the comments. I always appreciate the spicy discussion. YouTube already picked your next video, and unlike Iron's customers, it didn't need a credit check.