My guide to the IPOs of SpaceX, OpenAI and Anthropic — including the one I really want to buy
A couple of years ago, I hit it big at the ponies in an obscure northern track. I met one of the executives there. We took some swell pictures and we now exchange holiday greetings. Out of character, I received a text from him this week. He wants to buy some SpaceX no matter what. Not that long after, a guard near the New York Stock Exchange stopped me. He said he wanted to know how much SpaceX he should buy. Then my colleague in the gardening business was puzzled over the notion of what might be worth sacrificing to make room. Microsoft ? Salesforce ? Logical choices, save for the big Friday manipulation in those stocks that was so enjoyable for those of us with these positions. Let’s step back for a second, though. There’s not one, but three expected deals that will define 2026 and maybe even 2027. Although, after the rapid-fire fundraise by Anthropic — a near triple from February, not easy when your prior valuation was already north of $300 billion — things seem pretty shot-gunned. Elon Musk’s SpaceX is in pole position and is supposed to debut within two weeks. ChatGPT creator OpenAI may be next simply because it needs the money. It smells like more than a trillion-dollar valuation, given it can’t afford to have a down round and secured a $852 billion post-money valuation in March. That could be a tall order given its heavy losses and erratic leadership. Anthropic is buttoned up and profitable. If it’s the third to go public, it might be worth waiting, if only because A) it can afford to wait and B) the money from the track exec, Wall Street guard and the handyman gardener may be spent and there’s no more to come. To be sure, only SpaceX has officially released its initial public offering prospectus known as an S-1. CNBC has reported OpenAI is working on a confidential filing , while others suggest Anthropic could come public as soon as the fourth quarter. With those caveats established, let’s dig into these to see if we can make sense of what will happen here. I want to go high level, using the best info that I can muster from my time doing syndicate work on IPOs — on both the sell-side and buy-side of things — and my knowledge of the current state of play from the executives nominally in charge of the process. First, no matter what you hear, see or learn, it is most likely wrong because no one — no matter how high up — has any idea how things will go. One of the reasons is too many cooks in the kitchen. Take SpaceX. Musk has an iron hand, and he wants there to be plenty of retail participation and he wants a valuation of at least $1.8 trillion — at least for now. This is absurd. How are we defining retail? Are close associates of early investor Ron Baron’s funds considered retail? How about the loyalist Cathie Wood? How much are they getting? We all thought Musk would choose Morgan Stanley as the lead bank for the IPO, given the firm stuck its neck out for the Twitter-now-X buyout. Maybe it all worked out, and they said let’s “call it even.” Turns out, Goldman Sachs is the lead bank on the deal, which is good for us at the Club because the fees should be huge. But Goldman’s syndicate desk knows little about retail. Then again, maybe Fidelity counts as retail? Perhaps a big slug is promised to Robinhood for its users? Seems like the stock jumped inordinately Friday, but there is nothing that’s really “inordinate” in the wild west that the market has become. I have a good word to frame the deal: chaos. With chaos comes one thing: losses. So let’s play out what will happen. It’s a three-step process. First, the pricing, ostensibly where the $1.8 trillion valuation comes from. Second, the opening and first day of trading, where the total guesswork comes from. Third, the likely early inclusion into major U.S. stock indexes such as the Nasdaq 100 , thanks to rule changes by those in charge of them. This is the really fraught portion because it has never happened and nobody knows what will happen — something you could see late in Friday’s session when rebalancing caused Club name Nvidia’s stock to fall apart in the final 10 minutes of trading. Let’s put our “1999 hats” on because that’s the last time we had the level of ignorance we’ve got now. I have been saying we have to leave room in our imaginations for SpaceX to quickly become worth as much as $4 trillion when taking into account all of the market orders put in by those who got no stock on the deal. A market order tells your broker to immediately buy or sell at the best available price. In the SpaceX frenzy, they will have no idea what price they will be buying at. They will just be buying. We have a template from recent history: the much-admired IPO of Cerebras , an analogue of Nvidia, the latter being the punching bag of the process. Cerebras shares were priced at $185 , which valued the company at $56 billion on a fully diluted basis. The deal was reportedly 20 times oversubscribed, meaning 20 times the demand for the number of shares being offered. We don’t know the allocations, but the stock opened at $350, almost 90% above the deal price. Cerebras briefly supported a valuation north of $100 billion when the stock touched $386. Well, the stock is now at $237 a share, meaning everyone who bought in what we call “the after market” is now under water. Yes, the unthinkable. Everyone’s a loser, except for those who “got in” on the deal with allocations at the $185 price. Even those people could be losers if they gave the syndicate desk some “at the market” orders to help the cause and secure additional stock. Their blended average may put them underwater, too. Which brings us to the first lesson: you might actually lose money on this piece of business if you do it wrong. That’s why you want to try to get as much as you can on the deal — just call your broker and try, it will be worth the effort — and then beg off. No market orders. Let it play out, like you would trying to catch a tarpon. The stock will fly away then tire and come in, so you can buy it then. If not, just wait. I think because of a flood of at-the-market orders and the knowledge of the company because of Musk, you can expect at least double the double you got on Cerebras, which means that you could anticipate a $4 trillion price point that first day after the chaotic opening. You do not want to be a part of that scrum other than as a seller if you got stock. If you are, unfortunately, hung because of your lack of faith in my analysis or your overexuberance, you might have a chance to buy in a surprising way. The keepers of various indexes are going to machine gun SpaceX into their devices. In order to do that, you can expect a huge drawdown from all other stocks, especially megacap names like Nvidia, Apple and Microsoft . That’s where the money is. They will get clubbed, and I would say that’s a great time to buy, but we don’t know how close the other two deals will come. They could create similar selling pressure on these other names. The newly public stocks will pop on admission and then drop after joining the index. If you don’t get any stock on the SpaceX deal, perhaps that would be the time to think about buying. Will it be worth buying at all? Again, on the deal, yes. After the deal, not likely in the short run. There will be multiple tranches of insider stock peeling off as part of the novel way the deal is structured. You can expect a pummeling each time because nobody knows what is going to happen, including Musk and the bank syndicate desks, as you must remember the whole thing is very much like 1999. Will it be a good stock? That actually depends on the fundamentals . The company is losing so much money that I think you may only want it for bragging rights until it is through the process and is seasoned. It will not be an earnings story. The next one on the chute will most likely be OpenAI. Here’s a company that does not share the marquee of Musk. It has the personality of CEO Sam Altman, who is hard to figure out, and I am being extremely diplomatic. This one will be an actual fundraise because OpenAI needs the money. Again, if you can get in on the deal at the offer price, take it. If you can’t, don’t bother. You could have a Cerebras situation on your hands. This will be a traditional piece of business from a money-losing company, which means I expect it to be a money-losing IPO. I predict that your best price might be after the inclusion of the indexes. I would love to be more optimistic, but when a company needs the money, the institutions who get in on the deal will be sellers. They have to consider the company’s prospects, which means they would rather not own it. Finally, there is my favorite, Anthropic, a business-to-business company that is perhaps the fastest-growing company of all time, at least at this scale. It’s annual revenue run rate has crossed $47 billion, up from $10 billion in revenue last year. And it’s on pace to turn an operating profit this quarter, according to The Wall Street Journal . You have to play by a different set of rules here. You will most likely not get any stock on the deal. The first price will not be the highest price. If you want to try getting some stock, I strongly urge the use of a limit order , which gives your broker a specific price to buy or sell a stock. Pick a limit by dollar. Maybe say you will not pay a per-share price that is above the price of an Nvidia share. You calculate that price and you put in your limit order at that level. You do not use a market order, even if the opening will most likely not be the highest price of the day. You may have to be like a successful Cerebras holder flipping it at $380. At the least, because Anthropic is such a terrific company, I don’t expect the kind of losses that the Cerebras market buyers are now experiencing. Because I like Anthropic so much, I might have to break discipline on one of my buys, just to be sure I get some that first day. But if I wanted 200 shares, I would not buy more than 50 because I do fear some Cerebrus action. You can buy 50 at any price, provided it is worth less than Nvidia, which is another extremely fast growing company with a very low price-to-earnings multiple on 2027 consensus estimates. In the end, these three will be governed by their fundamentals. Go back over SpaceX — a money-losing Musk wunderkind — and expect that there will be a gulf, if not a gulch, where you will have to ride through, but it might actually be worth it to buy some if it breaks below its opening price and then where the deal came. All possibilities. You can console yourself that you are buying Musk’s dreams. It could work, unlikely, but it could be like owning a share of the Green Bay Packers until it nears some sort of profitability. OpenAI is the hardest to game because I hate paying up for money-losing companies, except ones run by Musk. We all know what he’s done for Tesla believers. OpenAI may need to offer a huge amount of stock so it can last long enough to get it to where it is losing less money. That’s the real guesswork. The computing power needs, the competition from Anthropic, the potential for a more business-to-business stream of revenue, the fear that all of the big institutions that own it will want to cash out, makes this one plain fraught. If you want to make money in it, then think of where Cerebras is now. I know, it sounds dire. Look, there’s an obvious twofold nature here — I want to make you money and not lose you money. Finally, if Anthropic is the third to go public, remember the market will be severely depleted of cash. The deal will, per se, be cheaper than if it were to come, say, first or even second. The company is a money maker. The big mutual funds will want the stock on the deal, at the opening price and even after that. I bet the low will be the price the stock opens. Yes, it will be that good. So good that I will want it for the Charitable Trust. Enough said. (Jim Cramer’s Charitable Trust is long MSFT, CRM, NVDA, and AAPL. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust’s portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.