This post was originally published on this site.
There are two kinds of IPOs.
Most recently, all the high-profile IPOs have been cash-outs. They have been vehicles for venture capitalists to hype up an offering at a ridiculous valuation and use retail for exit liquidity. CNBC plays into this by giving wall-to-wall coverage to IPOs, despite years of them costing retail traders money, particularly the ones they have focused intensely on.
On the other side are IPOs for companies that are looking to build a long-term shareholder base. In those, the IPO isn’t priced overly aggressively and the fundraising isn’t too high.
Anthropic is showing signs that it wants to be in the latter camp. Reports today say they have already found bids for $60 billion, which is the reported size of the IPO. It wasn’t disclosed what the price on that will be, but it’s likely to creep higher as we get close to the finish line. In any case, it doesn’t look like they’re trying to maximize the valuation at this point. There was talk of $2 trillion earlier this year and now it looks to be in the $1.25 to $1.75 trillion range.
Now I’m not going to argue that isn’t a huge number and the whole thing requires some heroic assumptions but the adjustments on the valuation don’t scream that this is one to avoid.
The FT reports that Goldman Sachs, the lead bank on the deal, is telling investors Anthropic will do about $52 billion in revenue this year, up from $4.6 billion in 2025. More than 10x is always good.
From there the forecast is $136 billion in 2027, $442 billion in 2029 and more than $600 billion in 2030.
Put that in perspective. Only two companies in the S&P 500 have revenue above $600 billion: Amazon and Walmart, both a little over $700 billion. Apple is at $416 billion. Goldman is saying a company founded less than six years ago will be the third-largest in America by sales in four years.
On those numbers, the valuation math changes quickly. At $1.25 to $1.75 trillion, you’re paying 24 to 34 times this year’s revenue, which is nosebleed territory. On the 2027 number it’s 9 to 13 times. On 2030 it’s 2 to 3 times, which is cheaper than the S&P 500 trades today. I’m very skeptical but skepticism hasn’t exactly paid in the AI boom.
There is also a catch, and it’s a big one. Goldman doesn’t see Anthropic turning free cash flow positive until 2030, when it forecasts $33.6 billion. That’s a margin under 6% on $600 billion in sales, and it puts the IPO at 37 to 52 times free cash flow that’s four years away. Meanwhile, the company has committed to spend around $900 billion on computing capacity. The $60 billion raise is a rounding error against that bill.
Remember who is making these forecasts too. This is the bank selling the deal, talking to a potential buyer. Revenue has to grow 160% next year and then keep compounding at 80% a year through 2029 just to hit the script.
That’s not the kind of thing that gets me excited and I will not be even thinking about buying the opening print but — for the broader market — there are indications it will go well.
This article was written by Adam Button at investinglive.com.
