English translation of the Dutch original Read the original (Nederlands) |
|
|
Column by Trader Guy
Little room for disappointment at Anthropic
Sunday 15 November 2026
|
|
|
I work with Claude every day and I enjoy it. That is exactly why I want to be level headed about the Anthropic IPO, because a strong company is not the same as a strong entry point.
|
|
|
The facts at a glance
Anthropic sets the price on 18 November and lists on the Nasdaq on 19 November under the ticker symbol ANT. The price range is between $70 and $78 per share, for 825 million shares. If you count all options and convertible instruments (fully diluted), the market value is between roughly $1.26 and $1.41 trillion. Based on outstanding shares alone, the top end comes to roughly $1 trillion. At the top of the range Anthropic raises roughly $64 billion through the IPO, plus $26 billion through a private placement with strategic investors, together roughly $90 billion. According to the Financial Times the banks are even considering a price above the range, which could lift the valuation to $1.5 trillion or more. At $1.5 trillion it would, according to AlphaSense, be the 3rd largest IPO ever by valuation, after SpaceX and Saudi Aramco.
A strong company, that is a given
Revenue in the 3rd quarter was $17 billion, against $1.4 billion a year earlier. There is little to take away from that. So the question is not whether Anthropic is a good company. The question is what is in the price. If I extrapolate 3rd quarter revenue to a full year, I arrive at roughly $68 billion and you pay 15 to 21 times that revenue. That is my own calculation and not a figure from the prospectus. Because revenue is still growing strongly, it is a rough yardstick.
A price to earnings ratio cannot be given yet, because Anthropic is making a loss. Working with adjusted operating profit helps a little. At an adjusted margin of 15% that is roughly $10 billion on an annual basis, and then you pay roughly 100 to 140 times that profit. That too is my own calculation, and a flattering one, because it concerns an adjusted profit before tax and interest, based on 1 quarter. Turn it around and at $1.4 trillion roughly $47 billion in profit is needed for a price to earnings ratio of 30. That is almost 5 times the current adjusted profit. So you are mainly paying for what the company has to earn in a few years. That is not wrong, but it leaves little room for disappointment.
How much room is there for disappointment?
At such a price almost everything has to keep going right, and a few things are already chafing. Growth compared with the previous quarter slowed from 144% to 47%. That is still an enormous pace, but price cuts and contractual discounts weighed on revenue by roughly $600 million. There was an operating loss of $567 million, after a profit of $974 million in the 2nd quarter, and the adjusted operating margin fell from 19% to 15%. Costs rose faster than revenue in the 3rd quarter, mainly because of more expensive short term computing power.
Commitments for computing power rise to roughly $895 billion, of which some $70 billion within 12 months. The 5 largest customers accounted for roughly 25% of revenue over the 9 months through September, although that share is falling, because in the 3rd quarter 79% of revenue came from customers outside that top 5. And among the buyers of the private part are parties such as Alphabet, Nvidia and Microsoft, companies Anthropic also does business with. That does not have to be wrong, but it is my reading that part of the money then comes from its own chain.
The net loss of $23.3 billion comes largely from an accounting revaluation and says little about day to day operations, so there is no need to be alarmed by it. But suppose growth weakens further, computing power stays more expensive or 1 large customer takes less. At a price that assumes continued hypergrowth that is no longer a small setback, because there is no buffer built in.
|
|
More structure in your investing?
Guy Systeem Trading | | Polleke Trading |
Join now until 1 January for €15. After that €60 per month. | | Join now until 1 January for €25. After that €80 per month. |
- Signals on indices, stocks and ETFs with entry zone, price target and stop
- Dynamic stock portfolio and an ETF plan with €500 per month
- Both LONG and SHORT signals, you choose which ones to follow
| | - Every morning a pre market email with entry prices, stops and price targets
- At least 2 video blogs per week from Paul on the AEX and the DAX, among other things
|
| | |
|
|
The moment does not make it easier
A company goes public when it can, not when it must. Anthropic has roughly $132 billion in liquid assets and short term investments and no conventional debt, and yet roughly $90 billion is being raised. That is smart if demand is large. According to Bloomberg the offering is oversubscribed several times and up to 20% goes to retail investors. For the seller this is a fine moment. For the buyer a full order book mainly means that the good news is already in the price.
The market is also shaky. On 5 November the S&P 500 lost 4.4% and the Nasdaq 5.5%, the worst day since April 2025, and the Nasdaq stood almost 12% below the record of October. A large part has since recovered. Whether this was the exact peak, nobody knows. What did happen is that the expected valuation of Anthropic fell within a few weeks from $2 trillion to $1 to $1.4 trillion. That tells me sentiment is sensitive, and sensitive sentiment is the enemy of a price without a buffer.
What disappointment does
SpaceX is a completely different company, but it shows what happens when a high price and a turning mood come together. It listed on 12 June at $135, closed the 1st day at $160.95 and reached a high of $225.64 on 16 June. On 15 July it fell below the issue price during trading for the 1st time, in August it dropped to around $105, and after a recovery it closed below the issue price again at the end of October. On 5 November it closed at $123.62, well over 45% below the high. At the start less than 5% of the shares were freely tradable and later more supply kept arriving through lockups. In December the full lockup expires. My estimate is that there may be room to the downside there as well, but that is an opinion and not a fact.
|
Please note, this is not a comparison with what happened back then, but a warning about what can happen with an IPO around the top, when the interest is more FOMO than reality.
My example is World Online, explicitly only on sentiment and not on value or on what happened afterwards. As companies they are worlds apart. But the IPO of 17 March 2000, a week after the top of the Nasdaq, shows how quickly full interest can flip. The issue was 21 times oversubscribed, on the grey market more than €70 was bid and the share closed the 1st day at €43.20, barely above the issue price of €43. After that the price halved. The interest then seemed like proof, but it was mostly mood. A price without a buffer forgives no bad news at all.
|
What I pass on as a coach
A share that is oversubscribed several times tells you something about demand and nothing about price. A rise on the 1st day also does not make the price cheaper, but higher. With my experience I know that the 1st day can be an expensive entry. You are not missing a train. If it is a good company, it is still a good company in 2 or 3 months, and by then you know more. By then the 1st figures as a listed company are known, the market has reacted and it is clear how many shares are really being traded. I am not saying Anthropic is a bad company, and not that the share cannot go higher. I am saying the price leaves little room for disappointment, and that looking carefully is not fear but craftsmanship.
|
|
All information has been checked as well as possible, but much of it is now generated with AI, so a few mistakes may have slipped in. This is not investment advice. The sources used are Yahoo Finance, Nasdaq, CNN, AlphaSense, NBC News, Fortune, Renaissance Capital, Sacra, Reuters, AP, Axios, Bloomberg, Financial Times and ESB.
|
|
|