Go back

Why Anthropic’s IPO Didn’t Stop the Tech Sell Off

18m 5s

Why Anthropic’s IPO Didn’t Stop the Tech Sell Off

Thropic’s historic IPO on Wall Street marked a pivotal moment in the AI industry, becoming the largest initial public offering in history. The company’s shares surged from $83 to a peak of $115, reflecting strong investor enthusiasm and the first direct market valuation of a pure AI model company. However, the momentum was short-lived, as the stock dropped over 8% by the close of trading, underscoring broader market instability. This sell-off occurred amid growing investor concerns—ranging from high AI-related debt and delayed data center construction to political opposition and safety fears about AI’s future. While Thropic’s IPO signaled a shift in how AI value is being captured, the financial performance of AI companies like Thropic remains under scrutiny, especially with revenue growth still outpacing massive spending commitments. Investors are now actively differentiating between AI firms, moving capital from general AI stocks to select leaders like Thropic. Despite strong initial demand, the market’s volatility reveals deep uncertainties about AI’s commercial viability, sustainability, and long-term profitability. The IPO also highlights a key transition: while hardware and infrastructure previously captured most AI value, now a public AI model company like Thropic may represent a new source of valuation. Yet, with rising interest rates, political headwinds, and existential risks, the long-term success of such ventures remains unproven, placing immense pressure on management to demonstrate real, consistent revenue and operational resilience.

Transcription

2886 Words, 16094 Characters

English
Yesterday was a big day on Wall Street, and Thropic made its debut on the stock market. In Thropic, the company behind the chatbot clawed, began trading on Wall Street. The expected range at $83 a share, raising more than $81 billion from the offering piece of money. Five, five minutes away from the closing bell and Thropic right now at $96.56. A record setting IPO and ended its first day of trading worth more than $1.7 trillion. Our colleague Peter Rudigier covers finance and investing. This is really the first time investors got a crack at owning a piece of a pure AI model company, and they really have been jobbing at the bit to get that for really years. So how big of a deal, like historically speaking, is this moment, I mean, it's the largest IPO ever? On opening day, and Thropic's share price reached a high of $115 by the time the market closed, and Thropic was one of the most valuable companies in the world. This is the first time investors have really gotten to say in what they think the value should be on the companies that are kind of the backbone of the AI revolution, and their verdict is we want more of that. But despite its strong start, and Thropic's stock price dropped, the stock closed down today more than 8%, and at one point nearly fell back to its pre IPO price of $83 a share. And Thropic's IPO took place amid a broader decline in the tech sector that's been going on for weeks. Today, the tech having NASDAQ fell more than 5%, meaning that investors hoping that the largest IPO ever would end the tech sell-off were sorely disappointed. And Thropic is entering the world as a public company at a time when the questions around AI are maybe as unsettled as they've ever been. Welcome to the Journal, our show about money, business, and power. I'm Ryan Koonison. It's Friday, November 20th. Coming up on the show, how the biggest IPO in history couldn't stop a massive tech wipe out. For Thropic, the road to an IPO has been rocky. At the start of the year, it was in a neck and neck race with open AI to see who would go public first. The most high-profile rivalry in tech is heating up in 2026 as both open AI and anthropic race ahead on what are poised to be historic IPOs. What ultimately gave Anthropica leg up over open AI? Anthropica was always a much more focused organization. They were building for companies. They were catering to coders that wanted to use AI tools to write software. And they were very focused on that. Open AI had a lot of different bets going on. They were a little more scattered and that meant they weren't growing as quickly in the meantime. Anthropica was able to pull ahead. Why does it matter who IPOs first? Some people want to be the first company in a new technology to go public because you can set the terms. You can define it on your terms. There's also the advantage of you kind of have a blank sheet of paper to tell your story. And also I imagine there's a question of how much money is out there that's willing to be invested in this. And if we go first we might soak up more of that potentially. Yes, there is definitely the element of a competition for capital. Well, investors waited for Anthropica or open AI to go public. They threw money at pretty much any other AI stock that could get their hands on. I think last year some of the best performing stocks in the SB 500 were companies that touched the AI supply chain from companies that owned power plants supplying power to data centers to the companies making the chips that go into the data centers to the companies that make the equipment that make the chips that go into the centers. Everything that touched AI just got bit up in the markets. Investing in AI was essentially a no-brainer. Yeah, you just had to be in the trade. You didn't have to think too hard about which stocks to buy if you owned any of them. You probably did pretty well. But as the year wore on, investors started having reasons to worry. The first issue was just this sheer amount of money that AI companies are spending. We are spending so much money building data centers in infrastructure AI that the question of can it generate an appropriate return investment just gets louder by the day. It's not a new concern. This has been the question for years. Will people pay for AI? Will companies pay for AI? Will they get enough value out of it? Across the AI industry, a lot of the money for all this build out is coming from debt. Big tech companies just don't earn enough from their existing businesses to build the data center capacity without showing up. They've had to issue a lot of debt just to raise the money to complete some of those data center builds. I think the figure last year was something like $400 billion in AI related debt issued this year. It's forecast to be 700. Next year it's forecast to be a trillion. So we're borrowing a lot of money to build the infrastructure needed for AI. While spending on AI data centers and chips is going through the roof, revenue at companies like Anthropic hasn't been keeping up. The question gets louder and louder of, are you going to actually be able to charge enough to your customers to make good on that? I think that's certainly the case with Anthropic. They are growing very quickly. They're most recent quarter. They earned about $17 billion in revenue, but they also disclosed in their perspectives that they have $900 billion in spending commitments over the next few years. It's a pretty big discrepancy. $17 billion in revenue in a quarter, but $900 billion in committed expenses. Right. Where is that money going to come from? They are locked into a trajectory that they need very fast growth just to keep up with the spending commitments that they've made. So that is one big answer question. Then in September, the Federal Reserve raised interest rates with the goal of fighting inflation. This was bad news for the AI industry since it's so reliant on taking out loans. Higher rates meant those loans are now more expensive. And with rates going up, lenders are becoming chooseier. They can't fund everything. They're going to start charging a higher price. They're going to be chooseier about which projects to back and which ones not to back. And that just means it's going to get more expensive to borrow. Another reason investors were worried about AI stocks is that the industry can't seem to build data centers fast enough. There was this larger problem of data centers were taking longer to get built than we originally thought. There were construction delays at various sites. We wrote about one in New Mexico that Oracle was leasing that had trouble getting the right permits. There were supply shortages. So already the direction to travel is this is going to take longer than we thought. Add on top of it the fact that there's this grass roots opposition to data centers. People don't like how they sound. They don't like what they're doing to the water. And then there was the blue tsunami. Like how did that hit? Yeah. You had a lot of candidates just say, I want a day center moratorium. No new builds that haven't been approved. You have people in elected to Congress with a similar philosophy. I mean, Texas went blue for the first time in decades. Texas is the state that is building the most data center capacity in the country. And the governor there campaigned on saying we're going to slow this down. Yeah, that was an explicit campaign message. I would issue a moratorium on any new data centers until the legislature can pass laws to protect Texans. Everything from how would it impact the AI companies like Anthropic if there's a slowdown in data center construction? So what you hear from these companies is they are constrained already by the amount of compute capacity that they have. They are projecting to have way more compute in the future. And they're basing all of their strategic planning off of that, right? New models that they're training, new models are developing. It requires a given amount of compute. If they don't get that compute, they have to reassess. As the year progressed, more concerns started piling up, including an international freak out over whether or not AI would end the world. The researcher at leading AI firm Anthropic says he resigned posting, "The people building AI earnestly believe that it could kill us all by the end of the decade." Dario Amade also penned a letter warning misaligned AI tech could take over the internet in six to twelve months, floated the idea of a pause on developing the technology. Around this time, open AI said it would delay its planned IPO until 2027 or later. And then the CEO of NVIDIA stepped down after revelations he had multiple affairs, including one with a suspected Chinese intelligence asset. As these worries mounted, it started taking a toll on the Outlook for Anthropics IPO. There had been whispers on Wall Street that it could be valued at $2 trillion in the offering, but as the weeks were on, that figured dropped. I think the bankers wanted to go out relatively conservatively. There was a lot of volatility in the market, so they wanted to go out smoothly. They thought, "Let's just play it safe. 1.2 trillion. We'll go out from there." But then, as the IPO got closer, everyone got excited again. And throbbing, that's why the huge IPO that comes tomorrow night is going to make a lot of money for a lot of people. What happened was as the bankers started canvassing investors and collecting orders and seeing what kind of put you down for, demand was even better than they were anticipating. They were able to raise the price range. They were able to sell more shares than they thought. And that's how you ended up at around a trillion and a half. How the IPO went is next. So this fall, there were all these concerns about the AI market, and throbbing lowered its IPO target price. But then in the days leading up to it, it raised it again, and it seemed like there was going to be more demand. So what happened? What was it that got people excited again? People were kind of satisfied with new anthropic disclosures that their growth, even though it was slowing in the third quarter, relatively early in this year, was really strong. That there wasn't this, their best days might still be out of it. That they were able to navigate some of these challenges, and they were talking about higher borrowing costs, slower construction timelines. There's still a lot of demand for their own product. So I think there was a lot of comfort in the fact that even if they are slowing from the insane pace they were in the first half of the year, it's still a really good profile. On Thursday, and throbbing shares started trading publicly around midday, and the company CEO, Dario Amade, and President Daniela Amade, his sister, rang the bell at NASDAQ. It went really well. It opens so the IPO priced at $83 a share, and it opened well over $100 a share. It's kind of an instantaneous pop. Everybody's investment is worth a good deal more than it was 12 hours before. In its IPO, anthropic had raised close to $100 billion. What is anthropic going to use all this money for? All they have, all those compute commitments that they have to meet. They got to pay rents on data centers. They have to acquire chips. They have existing contracts with companies supplying compute that they have to pony up the cash form. Right. All that's spending we talked about earlier. But there've already been some bumps along the way. Yesterday, while anthropic stock shot up, other tech stocks were going down. We turn out at your money, and today's major stock market sell off. The big three indexes all down by more than 2% the Dow. It's interesting. While anthropic traded pretty well, a lot of the other kind of AI stocks did not. A lot of semiconductor stocks are down other sectors that had been bit up a lot in the past few months exposed AI didn't react. As we talked about before, there was no hard choice to make earlier this year. If you owned any AI stock, you did well. Now, we might seeing some bifurcation there where investors are picking winners and losers among different players. We're seeing investors saying I'm going to take money out of some of my existing AI stocks, and I'm going to move it over to anthropic right now. Who captures the value in AI is the question investors now have to grapple up now that we have a public company that sells the underlying models. Before, it seemed like a lot of the values being captured by people making the hardware, people making chips, people making the data centers, people making the power. Is that going to be true in the future? Now we have a public company that can show that it can generate a lot of value just the model layer. Today, that tech sell off continued. As AI related stocks like Nvidia and SpaceX dropped. Overall, the NASDAQ has lost more than $4 trillion in market value this week alone. An anthropic is given up most of its initial gains. It closed just under $89. Only about five bucks over its original list price. Take a look at shares of anthropic. Falling this morning after a historic initial trading day yesterday. Crazy. It has come all the way back down to earth. If you were one of the first trades, you are now down more than 20 percent and with underwriters. Now that anthropic is a public company, it's going to have to continue to answer to investors and show that AI isn't just hype and that it can actually sustain a business. The public markets are different beasts, right? There's every single day the market is telling you what they think you're worth. It's very different to kind of manage a public shareholder base that every headline is going to be a referendum on your stock. It's just a lot harder to run a public company than a private company. Given everything that's happened this year and all the pressure that AI has been under that we've been talking about, do you think anthropic launched its IPO at the right time or is the enthusiasm from investors a lot more muted now? It's hard to say. I think six or ten months ago anthropic was a very different company. It certainly makes a lot more money now than it did six or ten months ago. There's a lot of concern about there being an AI bubble. Did we learn anything about that after this anthropic IPO? There's definitely been a lot of worry about an AI bubble over the past couple of weeks, just as a lot of those stocks have treated lower. I feel like a lot of the argument for that has gotten weaker in a bubble. Investors kind of lose rationality. They value a certain company at a multiple of the earning power of that company. So 20 times, 30 times what this company makes is transit their stock market value. The past couple of weeks as the market's kind of sold off, like those valuations don't really look irrational. With anthropic, it's one day. It's hard to say, you know, it does this for against a bubble. I think importantly, unlike the last time we had a bubble in tech stocks in 2000, anthropic generates real revenue, right? 18 billion dollars in the last quarter. I think investors appreciate that. Anthropic shows that it has real business. That's all for today, Friday, November 20th. Additional reporting in this episode of a Juliet Chung, Hannah Aaron Lang, and Sam Goldfarb. The journal is a co-production of Spotify and the Wall Street Journal. The show is made by Catherine Brewer, Evelyn Fahardo Alvarez, P.A. Gedkari, Sophie Codner, Matt Kwong, Trina Menino, Colin McNulty, Jessica Mendoza, Imani Moise, Laura Morris, Enrique Perez-Delerosa, Sarah Platt, Alan Rodriguez Espanosa, Heather Rogers, Katie Shepard, Pier Singy, Jiveka Verma, Tatiana Zamise, and me, Ryan Connitson. Our engineers are Griffin Tanner, Nathan Singer-Pock, and Peter Leonard. Our theme music is by So Wiley. Additional music this week by Catherine Anderson, Peter Leonard, Emma Munger, Nathan Singer-Pock, Griffin Tanner, and Budoth Sessions. Fact checking this week by Nicole Pasolco. Thanks for listening. See you Monday. [BLANK_AUDIO]

Podcast Summary

Key Points:

  1. Thropic made its debut on Wall Street with a record-breaking IPO, reaching a high of $115 per share and closing at $96.56, valued at over $1.7 trillion.
  2. Despite strong initial performance, Thropic’s stock dropped over 8% on its first trading day, reflecting broader tech sector sell-offs and investor caution.
  3. The IPO marked the first time investors could directly value a pure AI model company, signaling a shift in how AI value is assessed in the market.
  4. Thropic outpaced OpenAI in the race to go public due to its focused strategy targeting developers using AI for software coding, unlike OpenAI’s broader bets.
  5. Investor enthusiasm for AI stocks waned amid concerns over high debt spending, data center construction delays, supply chain issues, and rising interest rates.
  6. Growing public and political opposition to data centers, including moratoriums in key states like Texas, has created uncertainty for AI infrastructure growth.
  7. Concerns about AI safety, including warnings from Anthropic researchers about existential risks, contributed to market skepticism ahead of the IPO.
  8. While Thropic raised nearly $100 billion, the company now faces intense pressure to demonstrate sustainable revenue and operational viability in public markets.

Summary:

Thropic’s historic IPO on Wall Street marked a pivotal moment in the AI industry, becoming the largest initial public offering in history. The company’s shares surged from $83 to a peak of $115, reflecting strong investor enthusiasm and the first direct market valuation of a pure AI model company. However, the momentum was short-lived, as the stock dropped over 8% by the close of trading, underscoring broader market instability.

This sell-off occurred amid growing investor concerns—ranging from high AI-related debt and delayed data center construction to political opposition and safety fears about AI’s future. While Thropic’s IPO signaled a shift in how AI value is being captured, the financial performance of AI companies like Thropic remains under scrutiny, especially with revenue growth still outpacing massive spending commitments. Investors are now actively differentiating between AI firms, moving capital from general AI stocks to select leaders like Thropic.

Despite strong initial demand, the market’s volatility reveals deep uncertainties about AI’s commercial viability, sustainability, and long-term profitability. The IPO also highlights a key transition: while hardware and infrastructure previously captured most AI value, now a public AI model company like Thropic may represent a new source of valuation. Yet, with rising interest rates, political headwinds, and existential risks, the long-term success of such ventures remains unproven, placing immense pressure on management to demonstrate real, consistent revenue and operational resilience.

FAQs

Thropic's IPO was significant as it was one of the largest in history and the first time investors could directly own a piece of a pure AI model company, signaling a major shift in how AI value is assessed and priced.

Thropic's stock opened strongly, reaching a high of $115, and was valued at over $1.7 trillion, but it closed down over 8% later that day, dropping back toward its initial IPO price of $83.

The stock declined due to a broader tech sector sell-off and investor concerns about AI's long-term profitability, data center construction delays, and rising borrowing costs.

Concerns included massive AI-related debt spending, slow data center construction, supply chain issues, political opposition to data centers, and fears of AI misalignment or existential risks.

Initial IPO pricing was conservative, but demand grew as investors felt confident in Thropic's growth trajectory and ability to navigate challenges like rising costs and slower infrastructure timelines.

It shows that value can now be captured directly by AI model companies, shifting from earlier models where value was mostly held by hardware and infrastructure providers.

Chat with AI

Loading...

Pro features

Go deeper with this episode

Unlock creator-grade tools that turn any transcript into show notes and subtitle files.