A (0:00)

Welcome back to the Run the Numbers podcast. Nothing says value add like this guy sifting through S1 documents to to give you the what's up? Because Core Weave just filed for their ipo. I'm deep in the weeds here and I'm going to give you all the information you need before they go public. The biggest AI company to hit the public markets this year so far and it could be a good sign for other IPOs to come. We are venturing into unknown territory here because I just absolutely annihilated four freezer waffles with extra syrup right before hopping on the microphone because I think that's a liability. I don't know if they have insurance policies against this, but could get rocky. All right, core weave, the AI hyperscaler has filed their S1 to go public. So what is Core Weave? They're basically a supercharged cloud provider built specifically for AI. So instead of offering general purpose cloud computing like AWS, it rents out super powerful GPUs, the high performance chips that AI models need to run to companies like Meta, Microsoft and all the other great AI startups that are popping up. So like for all the operators and CFOs who may be listening, if you think about like the AWS or Azure or GCP costs that are always in your gross margin and your cost of goods sold, it's kind of like that, but hyper specific for companies that run the AI models. Okay, so if AWS is a giant all you can eat buffet for cloud computing, CoreWeave is a high end sushi bar that only serves AI workloads. And they say for AI to reach its full potential, it needs a purpose built AI cloud platform with infrastructure and managed cloud services that are delivered in an efficient, automated and highly performant way. Enter Core Weave, the AI Hyperscaler trademark. I love the tm. Okay, key metrics here. This is where it gets crazy. So revenue growth, 2022 they put up 16 million. 2023 they put up 229 million. That's 1346 year over year growth. You always see those really cool charts of time from 10 million in revenue to 100 million and how fast you can get there. They went from 16 to 229 million in just a year. But don't stop there. 2024, 1.9 billion. They said hold my beer. We're going to grow at 737% year over year. That is not a typo. 737%. They sniffed two bill profitability. What's happening? Still deep in the red. 2022 net loss of 31 million 194% of revenue 2023 net loss of 594 million oh. 259% of revenue 2024 net loss of 863 My goodness, that's a lot of dollars. But only 45% of revenue. So losses are still really big, but declining as a percentage of revenue. And speaking of that, their model, 96% of it comes from long term contracts, two to five year commitments which provides high revenue visibility. Remember, investors hate surprises. They said in their S1. We generate revenue by selling access to our AI infrastructure and proprietary managed software and application services through our Core Weave cloud platform. Access to our platform, including compute networking, managed software services and application software services is currently priced on a per channel GPU per hour basis. For all you pricing nerds. Storage is sold separately on a per gigabyte per month basis. Batteries are sold separately. Right. Remaining performance obligations, that's your contracted backlog. So it includes your deferred revenue and then anything that goes out more than 12 months. 15.1 billion 53% year on year growth. That's a lot of RPO as the pros call it. And RPO is cool because it gives companies a way to show off the multi year contracts that they've landed with customers and provide additional confidence on the company's ability to deliver on revenue expectations. So instead of just looking at what's in the bank right now, RPO gives you the full picture of what's in the pipeline. It's like seeing the whole iceberg and not just the tip. So 15, 15.1 billion in remaining performance obligations. 96% of the revenue comes from committed contracts. Four year weighted average contract duration. That's pretty awesome. And 15 to 25% of that is paid up front. You gotta love the payment terms. Hell yeah. Get that prepayment brother. They say we, we purpose built our Core Weave cloud platform to be the infrastructure and application platform for AI. So speaking of that, the AI compute land grab, that's the first theme I got for you. So there's this global shortage of high end GPUs and Coreweave is one of the few players with access to them. So as AI models grow larger and demand more Compute Corporation, Core Weave, they're cashing in by renting out the most valuable resource in AI right now, which is Nvidia GPUs. And it's kind of like chip arbitrage. If you strip away all that verbiage, it's really Arbitrage. So they're buying GPUs from Nvidia and they're renting them out at a markup. They're essentially playing the middleman in the AI compute economy. But as more companies try to secure their own GPUs or build custom chips like OpenAI's rumored Stargate, does this model have long term staying power? Can you keep doing chip arbitrage? So Core Weave's entire business depends on getting the newest and best GPUs from Nvidia. And if Nvidia decides to prioritize other customers or sell directly to AI companies, Core Weave is in trouble. Next theme. Take or pay contracts. They use that term a lot. Revenue visibility. That's what we love. 96% of Core Weave's revenue comes from multi year locked in contracts, which means customers pay whether they use the computer not. This isn't a typical cloud business with unpredictable on demand usage. It's more like a gym membership for AI companies that lasts more than one year and requires really, really big prepayments. That's a sick gem. At the same time, we're in this period where AI startups are raising billions and spending freely on compute. And so if AI funding cools off, will Core Weave's backlog hold up? Or is there a risk that some of its long term contracts don't actually get paid and companies go belly up? And so they say. Committed contracts generally have a fixed price for their duration, which is measured on a dollar per contracted GPU per hour basis and are billed monthly based on the customer's reserved usage commitments. Committed contracts generally have a predetermined term and start either on a fixed date or whether we deliver the capacity specified in the contract. As of December 31, 2024, our committed contracts had a weighted average contract duration of approximately four years. Next one. Cloud giants. They're watching. Ooh, neighbors are watching. Aws, Google Cloud and Azure could easily decide to go harder into AI Compute and squeeze Core Weave out. But right now they're spread too thin across multiple industries, while core weave is 100% focused on AI. Remember, they wake up every single morning and they say we do this but for AI, so can they stay ahead? Aws, Azure and Google Cloud can all buy Nvidia chips too soon. I mean like money be green, right? And they have even more of it. Remember Nadella said, look, all I know is I'm good for my 80 billion. That guy's got a lot of money, right? Core, we've argues that it's networking, storage And Kubernetes based Orchestration makes its offering more than just a GPU rental shop. Will customers see the difference or will they just chase the lower cost? And speaking of that, just the other day they announced that they're acquiring AI developer platform weights and biases for 1.7 billion, which hopefully enhances its cloud computer computing offerings. Which means that they're trying to verticalize more. They're trying to be, you know, make sure they're not labeled as a commodity. And Weights and Biases. Their tools are widely used in AI development and corporate Core Weave plans to integrate them with its infrastructure to streamline machining learning workflows. Okay, so the private funding boom leading up to this ipo. I mean, Core Weave, they've raised billions from firms like Magnetar, which is a sick name. Blackstone and Coatu, with Nvidia itself investing 100 million, which is probably like a dollar bill that they found in their couch. This IPO is as much about giving early investors an off ramp as it is about raising capital for expansion. And like, not to put my tinfoil hat on or anything here, conspiracy cj but like, what is this a bridge to? Let's get real about it. I mean, going public helps Core Weave raise capital, but it also puts pressure on them to prove profitability. Do they scale to profitability or is the real goal to give existing investors some liquidity and then position themselves for a merger with like a Microsoft or Google or an Oracle? Because they all need more AI compute capacity. Next theme. Taking my tinfoil hat off at least for a second. Right now, AI training and inference dominate demand. But can Core Weave expand into other compute heavy markets like high frequency trading, biotech simulations or video rendering? Or will it remain tied to the boom and bust cycle of AI funding? They say our Core Weave cloud platform is an integrated solution as Purpose built. I love that term. If you're listening and you're writing something for your company, just throw Purpose Built before whatever you do for running AI workloads such as model training and inference at maximum performance and efficiency, you got to remember their sector expertise is a double edged sword. It makes them the best at what they do by remaining hyper specific while also exposing them to AI specific shocks. So next one here. Microsoft's Core Weave play a call option on GPU demand. So I stole this from my good friend Fred Haven Meyer, former sell side analyst at Macquarie. He's now doing some cool AI stuff at Fleet, so. And when we were having coffee, he was like, look dude, Microsoft accounted for 62% of Core Weave's revenue. That's like a huge reliance on a single customer. And if you strip it down, Their deal with Core Weave is the ultimate low risk, high upside move. They're essentially leasing AI compute capacity instead of building it outright. Which keeps their options open as this AI arms race unfolds. So minimal commitment, maximum flexibility for Microsoft. So they outsource AI workloads to Core Weave and they secure compute without the capex burden of building its own infra up front. And they hedge against supply chain risks. So if GPU shortages persist, Microsoft retains access to high end AI compute without having to over commit on hardware purchases today. And it's easy to walk away. I mean, like I know they committed, but like after a certain number of years they can walk away. If AI compute pricing normalizes or Microsoft itself scales its own internal capacity, they can dial down their Core Weave usage with minimal downside. Or they could resell it. Yes, these guys aren't going to lose money on it. They'll find a way to resell it or use it. So from Microsoft, not to bury the lead. This is a call option on GPU demand, guys. This is a call option on GPU demand. Cheap to hold, but with huge upside if AI workloads keep exploding. But for Core Weave, it means their biggest customer can leave the moment. It's more cost effective to go in house. All right, sorry, I gotta put the tinfoil hat back on, guys. We're getting some WeWork vibes here. So Cor, we've had 15 billion in long term lease commitments tied to data centers. Suspiciously like the same amount as their rpo. Weird. So that's for power contracts and equipment rentals. And these obligations don't appear as traditional debt on the balance sheet, but they function as long term liabilities. So why does this raise red flags? So similar to WeWork's model, Core Weave is signing long term leases to scale quickly while keeping the liabilities off the books. I don't know if it's the waffles I just ate, but whenever there's off the books liabilities, like I, yeah, massive heartburn. Whenever liabilities are off the books, you're like, what's going on here? So if AI compute demand weakens, Core Weave could be locked into expensive underutilized capacity. And a shift to owning data centers would change the financial profile, which would require billions in capex instead of lease agreements. So for now, I mean, investors need to treat this 15 billion in leases as hidden leverage. If the market conditions shift though, these commitments could become A major financial burden. So like don't forget about that. Okay. And speaking of financials, I'm looking at their balance sheet and what they actually included on it. So they have 1.4 billion in cash and they do cite 7.9 in debt. Doesn't include the 15 massive facilities that can tap into for capex needs with big banks and a large convertible Note to Magnetar which we will discuss momentarily. Okay. Revenue looking at the P and l grew 737% last year while cost of revenue grew at 617%, about 100% less. Showing some early leverage in the model. We like the leverage. Whoa. We like leverage. Cost of revenue includes data centers, utilities for power and the people involved in data center operations and customer success. It's also that big old depreciation and amortization charge in here that gets added back in an EBITDA bridge. I love EBITDA bridges. Do you know your taxes pay for EBITDA bridges in your community? Technology and infrastructure expense consists of costs associated with servers, switches, networking equipment and internally developed software and the engineers required to build the product. That's important. Yeah. And the sales org or lack thereof is focused on a direct named account strategy. That's an enterprise strategy to drive demand from the world's leading AI labs and AI enterprises. They supplement this with a product led growth. Doesn't everyone just supplement it? They always throw that in. But serving individual users and developers working at AI Labs and AI Enterprises, they somehow only spent 18 million on sales and marketing last year. That's less than 1% of revenue. They grew. They grew revenue 737% and they only spent 18 million on sales and marketing. I've never seen something like this. That's crazy. Interest expenses they go to paying down their debt facilities which are used to fund capex investments. They spent 332 million on interest last year, which represents 19% of revenue. That is up from 28 million the prior year. So they basically 10x their interest expenses year over year. Wow, that's a big jump. Oh, for context, last year they spent 18 times more on interest payments than on sales and marketing. There's a nugget for you. Tell your friends that at the water cooler on Monday, net loss margin was cut from 259% of revenue to just 45% of revenue year on year. So that demonstrates that while they're investing a massive amount in build out from an absolute dollar perspective, the losses look better on a relative basis to revenue. All right, who owns this thing? Well, Core Weave's three co founders, CEO Michael Intrator, CTO Brian Venturo and Chief Strategy Officer Brandon McBee hold 83% of the company's voting power. That's right. They have 10 to 1 voting shares through a dual class share structure, ensuring they maintain control. However, this dominance hinges on Intrator's position. If he leaves, their special class B shares convert to common stock, reducing their grips. Keep an eye on if he sticks around. They did cash out some. So before the IPO, they each cashed out between 11 and 16% of their stakes. Good for them. Collectively, 500 million. A lot of money. But I mean, like, their personal exposure is much, much more worth billions because they're targeting a $35 billion valuation, which we'll get to key institutional investors. Like we said, Nvidia owns about 1%. Small but strategic. Blackstone led a 7.5 billion debt facility. Magnetar Co led multiple financing rounds and structured a convertible note deal, giving them discounted IPO shares. More on this interesting setup in a SEC Code 2 management, they led Core Weaves 1.1 billion Series C, betting heavily on AI Infra. They're a known crossover investor. And I mean, these guys can exit their existing shares or double down in the public market, so they could stick around. And digital bridge backs Core weaves real estate and data center expansion. Other investors with minority stakes. This is a. This is quite the hitters list. And it sounds like I'm reading Lord of the ring characters. Altimeter Capital, Assured Asset Management, Claridge Venture Partners, Interplay and Iron Arc Ventures hold smaller positions, likely looking for a post IPO exit. So who stands to gain the most? Who's going to make some money? The founders still hold billions in equity even after partially cashing out. Okay. Magnetar structured a sweetheart deal, and they're going to get some shares at a steep discount. Nvidia small stake like this isn't going to. Like this isn't going to make them rich. But it's strategic. And private equity players like Blackstone CO2 may dump some shares after lockup. Who knows? All right, so how high can the valuation go? Revenue growth? That's bonkers. From 16 million in 2022 to 1.9 billion in 2024. They got this massive contract backlog. Investors are going to like, drool over that. So 15.1 billion in remaining performance obligations, or RPO. And at the moment, yes, at the moment, the market loves AI infrastructure. So if investors see Core Weave as an AI infra pure play, it could command a premium. Now, AI computer is being valued like a scarce commodity in in this day and age. But if investors treat Core Weave more like an AI hardware supplier than a cloud provider, aws, the stock could pop. So the debate AWS versus Nvidia. Like what type of multiple are they going to get? So how the market values Core Weave depends on what bucket the investors will put them in. So AI hardware like Nvidia and AMD trading at 25 to 40x forward revenue. So if AI infra stays scarce, Core Weave could trade at premium. Then you get high growth SaaS, the likes of Snowflake, Databricks, they're all trading in the 15 to 20 plus forward revenue range. Maybe if you squint. But Core Weave had like they have the recurring revenue but they also have heavy capex. You did see Snowflake get their gross margin well over 60% though. But these guys have even more hardware. Cloud hyperscalers, AWS, Azure, GCP, they're trading at 30 in the 5 to 8.9x forward revenue range. Lower margins but long term stability. This wouldn't be good enterprise hardware, dell super micro 3-6x forward revenue. This would be bearish case, more like a commodity business. So let's do some IPO valuation math here. Okay, so their last private valuation was at 23 billion in November of 2024 and they've raised over 4 billion in preferred capital to date and are looking to raise another 3 to 4 at a $35 billion valuation. That's a nice step up. So if they grow revenue next year, even just 100% and I pause because I completely recognize that it's wild to say doubling is like a lower expected range of revenue to 4 billion. That would still be less than 10x. A forward revenue multiple. That's not crazy at all. Likely market multiple. I would probably peg it at like 20x current, maybe 12 to 15x forward, I don't know. Depending on hype versus fundamentals. Implied valuation of 30 to 38 billion. Very, very realistic. So if Core Weave leans into the AI scarcity hype, we could see 30 billion plus territory. If the market starts worrying about cash burn, that's when it could get dicey and then they fall to the 20 billions where the last round was. So some miscellaneous stuff of note, but very important payback period here. So Core weaves customers prepay 15 to 25% of their contract value upfront, allowing the company to fund Infra without tapping as much outside capital nugget here. The company estimates that GPUs reach full payback within 2.5 years, meaning that from year three onward they generate pure cash flow. So they'll want to keep their contracts above 2.5 years at least. Right now they're at 4, which is great. Now the shift to owning data centers could reshape the business and they hinted at it. So they currently rent the data center space and all those off balance sheet, you know well. But their S1 suggests it may begin building its own infra. This would be a massive shift in capital allocation, which would bring it closer to the AWS model. But it also require billions in upfront investment, so they'd have to raise more and dilute shareholders. IPO ratchets protect insiders at the expense of retail investors. Oh, all right. Series B and C preferred stockholders have built in downside protection if the IPO price is below a certain threshold. If that happens, they automatically receive extra shares, effectively diluting common stockholders. So this structure shields institutional investors from risk while shifting it to retail buyers like you and me. Okay. Lockup expirations could lead to a sell off. So insiders are generally locked up for 180 days post IPO. But certain conditions allow early selling. Expect significant selling pressure after six months. So you always want to watch these things like was there a day one pop and then how does it do in the six months leading up to when, when, when they, they let the hounds out. Right. Early employees who sell off and how high profile investors. Okay, this one. Magnetar's 230 million cloud deposit looks more like financial engineering than anything. The waffle heartburn is it's coming up in August 2024. Core, we've received a 230 million refundable deposit from Magnatar under a pre negotiated cloud services deal. So if Magnetar's portfolio companies don't use the services, the deposit is refunded with a 1.1x to one 0.7x markup insider deal. This structured deal raises the question is this real revenue or a form of synthetic financial engineering? If it walks like a duck, talks like a duck, whatever the saying is, Magnetar special IPO discount puts retail investors at a disadvantage. These Magnetar folks, they're crushing it. They're, they're getting a sweetheart deal here. Magnetar also bought convertible notes that converted into core weave stock at the IPO price. However, they also received a penny warrant. I want a penny warrant allowing them to buy even more shares for $0.01 each. This gives them a better deal than retail investors who are buying at full price. So final thought on that. This IPO is structurally built for insiders. Between the IPO ratchets, the early liquidity options and structured deals with key investors. The IPO is designed to protect the insiders while leaving public investors exposed to more risk. The company is growing like really fast. But its financial structure does raise the question of whether this IPO is a long term public play or or just a bridge to an eventual acquisition. All right, what could this look like? I'm not an investor here, but bull case that would be if the AI compute Super cycle continues valuation 40 bill plus 20x plus forward revenue multiple. That's what it would look like. So what would have to happen? AI model sizes and compute needs keep growing, fueling GPU demand. Core Weave remains a key player. They benefit from the scarcity pricing. Their revenue then skyrockets beyond expectations. Microsoft and other hyperscalers continue outsourcing AI workloads rather than fully internalizing their compute. Gross margins expand. We love when that happens as Core Weave improves efficiency and potentially builds its own data centers to cut costs. And then the public market would reward AI infrastocks treating Core Weave more like Nvidia than a commodity cloud provider. And then that would make the price go up and Core Weave becomes a long term AI infrastructure leader. So what does it look like to me? A snowflake for AI compute narrative with some Nvidia sprinkled in commanding a premium valuation as the go to hyperscaler alternative. What's the bear case that would be if AI compute demand normalizes and the hyperscalers eat its lunch. That would put their valuation in the 10 to 15 range. 5 to 7 forward revenue. Microsoft, AWS and Google ramp up their own AI compute offerings which reduces reliance on CoreWeave GPU. Supply chain constraints ease leading to lower pricing and eroding margins. Customer concentration risk materializes. So Microsoft or major AI startup pulls back which shrinks that very valuable Core Weave backlog. Cash burn accelerates as capex commitments for GPUs and data centers outstrip revenue growth. And the stock struggles as investors realize Core Weave's margins resemble traditional cloud businesses, not AI SaaS. Ooh, robo dope. Core Weave would pivot or get acquired in this scenario at a discount by giant, you know, cloud provider looking for additional compute capacity. What does it look like? That'd be high growth capital intensive business that struggles to achieve profitability and loses its competitive moat base case. High growth, high cash burn if you're cool with that. With an uncertain endgame that would put their valuation in the 20 to 30 billion range 10 to 15x forward revenue and core wave continues growing fast, but AI infrastructure becomes more competitive. Revenue continues to go up but profitability remains elusive. Isn't it always due to constant reinvestment in GPUs and data centers? Microsoft stays a key customer but negotiates better pricing, squeezing coreweave's margins. They always win. The IPO pops but stock remains volatile as investors debate whether this is a long term business or an acquisition target and M and A rumors. Will the hyperscalers or private equity double down to buy Core Weave and absorb its valuable customer base? Why? What's this look like? A highly valuable but capital hungry AI infrastructure provider that either matures into a niche hyperscaler or gets rolled up into a larger cloud play. So that's wow. The waffles, the syrup is catching up to me. That is the S1 breakdown for Core Weave. Once again, none of this is investment advice. Just having fun here. I really appreciate everybody's support. I love doing this for you. If you could help me out and give me five stars. I want to show my mom this weekend that that number is going up. She doesn't really understand what I do for a living. I really appreciate it. Guys. Check out mostlymetrics.com for the written version of this. Peace.