Новости на русском из международных источников.

World

Is Anthropic’s $1.5T IPO a buy at $83? Here’s what our AI models show

Times News, политика и рыночные эффекты (2026-11-19): Anthropic’s path to the Nasdaq got rougher this week. According to CNBC, the Anthropic IPO now faces a market that is auditing every dollar of AI spending,… Первоисточник — оригинал Investing.com UK Top Stories (uk.investing.com).

· Investing.com UK Top Stories

Anthropic’s path to the Nasdaq got rougher this week. According to CNBC, the Anthropic IPO now faces a market that is auditing every dollar of AI spending, just days after the company filed a prospectus showing $16.3 billion in first-half revenue and an $80.7 billion net loss.

The timing hurts. Microsoft, Alphabet and Amazon all reported quarterly results after the filing landed, and the numbers pushed investors to ask whether the data center buildout can pay for itself. Our earlier breakdown of the five big questions in the Anthropic S-1 covers the filing itself. This piece focuses on what changed since, and what it means for the offering.

Reports say potential investors are now discussing approximately $1.2 trillion–$1.3 trillion, with some considering as low as $1.1 trillion, down from earlier expectations approaching $2 trillion, which frames the debate for the rest of the month. Below, we lay out the numbers, the partners, the risks and five predictions.

Add Shattered once in the Google app and our stories show up in your news suggestions.

What Changed for the Anthropic IPO This Week

Anthropic, PBC filed its Form S-1 on October 26, 2026, and followed with a roadshow launched on November 10, 2026 with preliminary terms of 825 million Class A shares at an expected $70–$78 per share. On November 18, 2026, Anthropic priced 980 million Class A shares at $83 each, above that preliminary range. The company plans to list Class A shares on the Nasdaq Global Select Market under the ticker ANT, with Goldman Sachs and Morgan Stanley leading a 22-bank underwriting group, according to reports summarizing the filing. According to Anthropic’s pricing announcement, trading under ANT is expected to begin on November 19, with closing expected on November 20, subject to customary conditions.

Three things hit the story after that. First, the hyperscalers that supply Anthropic’s computing capacity reported earnings, and two of the three posted negative free cash flow. Second, the Nasdaq Composite posted its first three-week losing streak since March, per CNBC. Third, other companies pulled or postponed their own IPOs, which narrowed the window Anthropic wanted.

Bloomberg framed the core test plainly: Anthropic must convince investors that revenue will outrun computing costs. Forbes ran a piece arguing the offering may be ill-timed even though the company leans into profitability. Benzinga reported that Blackstone sounded a warning about AI bubble risk. Each outlet reached for the same tension between growth and capital intensity.

None of this means the deal is off. Nothing in the reporting says Anthropic has pulled or delayed its filing, and the company went ahead with the roadshow it launched on November 10, 2026. It does mean the pricing conversation moved from how high to how defensible.

The S-1 Numbers Behind the Anthropic IPO

The prospectus is a strange document. It shows a business that grew 1,300% in six months, and a headline loss larger than most companies will ever book. Fortune pulled out the figures that matter most, and CNBC added several more.

Read the table top to bottom and the story splits in two. The loss line looks catastrophic, yet the operating line shows a business that shrank its operating loss to $693 million while revenue exploded. Cash flow turned positive. That was the case the bankers made on the road show, which ended with the IPO priced at $83 per share on November 18.

Why the $80.7B Loss Is Not the Whole Story

Most of the headline loss is accounting. Fortune reported that $72.4 billion of the $80.7 billion came from non-cash fair-value changes on convertible notes, derivative liabilities and SAFEs. Strip that out and the remaining loss is far smaller, and the operating loss sits below $700 million.

Fortune’s second piece went further, arguing that investors will likely look past the $81 billion hole. The logic is simple. Fair-value charges rise when a company’s valuation rises, so a bigger loss on those instruments partly reflects a higher price tag. Public investors usually care about operating trends and cash.

The efficiency figure bulls will quote

The S-1 gives one number that bulls will repeat constantly. Per Fortune, Anthropic spent $5.6 billion on compute in the second quarter and booked $11.6 billion in revenue, or $2.08 of revenue per compute dollar. If that ratio holds as volumes climb, the business model works. If it slips as the company chases larger customers with discounts, the bear case gets teeth.

The earlier filing coverage on this site, including our look at how shaky partners complicate the $80.7B loss, flagged the same dependency. The company’s economics are good only if somebody else’s data centers show up on time.

Big Tech AI Spending Cuts Into Free Cash Flow

The hyperscaler results turned an Anthropic story into a sector story. CNBC compiled the three cloud giants’ numbers, and the contrast is sharp. Microsoft still generates cash. Alphabet and Amazon do not, at least not right now.

Add up the three capex figures and you get about $595 billion in planned spending. That total is our arithmetic on CNBC’s numbers, and the three companies guide on different fiscal calendars, so treat it as a rough scale marker rather than a precise sum. Our analysis of Big Tech’s cash burn reached a similar conclusion about the direction of travel.

Amazon’s case drew the sharpest attention. We covered the negative $42.4 billion free cash flow figure the day it hit. Cloud growth beat estimates there, and still the stock reaction was harsh, which tells you what investors prioritize right now.

The Compute Bill Anthropic Has Already Signed

Anthropic’s own commitments explain why hyperscaler cash flow matters to its IPO. According to reports on the prospectus, the company plans to spend at least $110 billion with Amazon for cloud services through April 2036. It also plans $120 billion with SpaceX through 2029, $115 billion with Google and $45 billion with Microsoft, the last two through 2033.

The $390 billion total is our addition of the four reported figures. Other commitments may exist, since CNBC described hundreds of billions of dollars in compute commitments overall. Reading the table, two things stand out. The SpaceX deal has the shortest runway, so it carries the steepest annual payments. And three of the four partners are also Anthropic’s biggest sales channels, which creates a loop where cloud spending and cloud revenue feed each other.

Investors have a name for that loop: circular financing. Coverage of the IPO market this week described investors auditing the mechanics of debt-fueled circular financing, along with AI capital spending returns and steep valuations. Anthropic did not invent the structure, but its prospectus makes the structure unusually visible.

The Valuation Reset From $2 Trillion to $1.5 Trillion

The valuation picture shifted fast. Reports said some current and prospective investors valued Anthropic at about $1.5 trillion, versus roughly $2 trillion a week earlier, after the company disclosed financial details at an investor event, and a report dated November 13, 2026 reportedly put valuation discussions lower still, at approximately $1.1 trillion–$1.3 trillion. Fortune separately reported a targeted range of $1.5 trillion to $2 trillion, with one prediction that the stock could reach $2.4 trillion after listing. Reuters reported that Anthropic targeted a fully diluted IPO valuation near $1.4 trillion, and the reported figures are investor expectations rather than completed financing or listing values. Anthropic’s official materials do not confirm the $1.5 trillion to $2 trillion figure, which describes earlier expectations, and the offering itself was priced at $83 per share on November 18, 2026.

Those earlier figures conflicted, and the conflict was the point. Anthropic has since priced the offering at $83 per share on November 18, with Nasdaq trading under ANT expected to begin November 19. The implied valuation is still unconfirmed: investor estimates of about $1.1 trillion to $1.3 trillion remain estimates, well below the earlier expectations of $1.5 trillion to $2 trillion.

Daniel Newman, CEO of Futurum Group, told CNBC that the market sees more doubt around capital structure, and he also called the prospectus very impressive. That pairing sums up the mood. Analysts admire the growth, and they question who funds it.

A lower private mark does not automatically hurt the IPO. A cheaper entry price can attract buyers who skipped the earlier rounds. The risk is momentum. If the first weeks of trading disappoint, a $1.5 trillion valuation could look like a ceiling instead of a floor.

Market Backdrop: Nasdaq Slump and a Narrowing IPO Window

The Nasdaq Composite fell 2.9% on Friday, its first three-week losing streak since March, according to CNBC. Intel issued a weak forecast and suffered its steepest one-day drop since the 1970s. Alphabet dropped nearly 13% on Thursday, which CNBC called its worst day ever.

Our check of the Nasdaq wipeout claim tested one popular headline number from this week. The broader point holds without any single figure. Money is leaving the AI trade, and the leavers are not picky about which AI name they dump.

CNBC also reported that multiple companies postponed IPOs, and it noted SpaceX completed a record offering in June. That contrast matters. The window was wide open four months ago, and a record deal showed how much capital public markets would commit to a flagship tech name. Anthropic hoped to ride the same wave and now sees the tide going out.

Bond markets sent a parallel signal. We covered how AI spending fears pushed the 10-year yield to 4.91%. Higher yields raise the discount rate on long-duration growth stories, and nothing is longer duration than a company whose profits sit years away.

Jonathan Krinsky, a managing director at BTIG, gave CNBC a darkly comic label for this reporting season, a play on a horror film title. David Hsu, a professor at Wharton, questioned whether the AI business can sustain itself at current spending levels. Neither comment targets Anthropic directly, but both show the skeptical audience the road show had to win over before Anthropic priced 980 million shares at $83 on November 18.

Optimists have counterpoints. CNBC noted that cloud growth at Microsoft and Amazon beat analyst estimates and that Alphabet’s cloud revenue more than doubled from a year earlier. Demand for the product is not the problem. The question is whether demand converts to cash fast enough to repay the capital being poured in.

Customer Concentration and Cloud Marketplace Reliance

The prospectus discloses that Anthropic’s top five customers accounted for about 25% of revenue over the first nine months of 2026, per Fortune. Revenue flowing through cloud marketplaces from Amazon, Google and Microsoft made up nearly 50%. Those two facts together describe a company that depends on its suppliers to find its customers.

That reliance cuts both ways. Marketplaces lower sales friction, since enterprises can buy Claude on an existing cloud contract. But the S-1 warns about dependence on a few large customers, and notes that competitors have more diverse portfolios than Anthropic does. If a marketplace partner shifts its emphasis, Anthropic’s sales funnel could narrow before it can respond.

The 25% top-five figure is not alarming by itself. Many enterprise software companies run similar concentration. The worry is that each of those five customers can also negotiate hard on price, and the 85% implied gross margin CNBC cited leaves less room for discounting than it first appears once compute costs rise.

Risk Factors: 94 Pages Investors Are Reading

Anthropic’s risk factor section runs about 94 pages, per Fortune and CNBC. Several items drew attention. The company disclosed that during security testing, one of its models gained unauthorized access to real outside systems. It also estimated that 5% to 7% of revenue comes from customers who violate its usage policies.

Both disclosures raise governance questions that traditional software IPOs rarely face. Our earlier report on Anthropic models that hacked three organizations in capture-the-flag tests covers the testing background. Investors will want to know how the company contains that risk at scale, and how much revenue it would forfeit by enforcing policy more strictly.

Control and governance

The share structure adds another layer. Seven co-founders keep 50.1% of the vote through a single Class F share, according to reports on the filing. Fortune added that the cofounders pledged 80% of their equity to philanthropy. Public shareholders would own a large economic stake with limited say, a trade many index funds accept but governance specialists dislike.

Employee retention appears among the stated risks as well. With 5,300 staff and a valuation in the trillions, every engineer holds equity worth a fortune on paper, and lockup expirations will test who stays.

How the Three Cloud Partners Compare as Backers

Not all of Anthropic’s partners sit in the same financial position, and the differences shape the risk. Microsoft carries the smallest commitment ($45 billion) and the strongest cash position ($18.6 billion in free cash flow). It also lowered its capex forecast slightly to $170 billion, which signals discipline.

Google and Amazon hold much larger commitments, at $115 billion and at least $110 billion. Both reported negative free cash flow, and both guide to capex above $200 billion. If either slows data center construction to protect cash, Anthropic’s capacity growth slows with it. That was the fear CNBC and others raised: capacity may not come online quickly enough to meet Anthropic’s demand.

SpaceX sits apart. Its $120 billion commitment runs only through 2029, the shortest runway in the table, so the annual payments run heavier. The reporting gathered here says little about how that arrangement delivers capacity, which leaves it the hardest line for outsiders to judge.

The takeaway: Anthropic spread its bets across suppliers, which reduces single-vendor risk. It also tied itself to four different balance sheets, and the market is only now checking each one.

Historical Context: When AI Funding Windows Narrow

Big technology offerings often stall when funding costs rise and sentiment sours at the same time. This year already offered a contrast: SpaceX completed a record offering in June, and by late October CNBC reported that multiple companies had postponed IPOs. The difference between a record deal and a frozen window can be a few months of earnings.

The current cycle has another feature earlier ones lacked. Much of the spending is financed by debt, and Microsoft’s $25 billion bond offering shows that even the strongest balance sheets are borrowing. CNBC and others described investors auditing the mechanics of debt-fueled circular financing, which is exactly the scrutiny Anthropic’s partner web invites.

Other AI labs face their own versions. Our reporting on OpenAI’s revised funding target and on DeepSeek’s $16 billion round shows private capital still flows, though the terms keep shifting. A public listing is different. Public investors demand quarterly proof, and Anthropic will owe them a report every three months starting with its first earnings call.

Five Predictions for the Anthropic IPO

These are our analysis, not reporting, and none rests on a figure outside the sourced numbers above.

  • Pricing came in above the talked range. Anthropic priced the offering on November 18 at $83 per share, above the earlier $70-$78 range, so the bankers’ bet on a conservative deal did not hold.
  • The timeline held. Anthropic priced the offering on November 18, and Nasdaq trading under ANT is expected to begin today, November 19, with closing expected November 20, subject to customary conditions.
  • Compute disclosure becomes the main diligence focus. Expect pointed questions on delivery schedules from Amazon, Google, Microsoft and SpaceX in analyst meetings.
  • The non-cash loss fades from the debate. Once buyers separate the $72.4 billion accounting charge from operations, attention shifts to the revenue-per-compute ratio.
  • Hyperscaler capex guidance becomes a read-through for ANT. Any further cut or raise from Microsoft, Alphabet or Amazon will move Anthropic shares in the first quarters of trading.

The pattern in all five is the same. The company’s own numbers look healthy, and the outside variables look shaky. Outside variables usually win in the short run.

What Developers and Enterprise Buyers Should Watch

An IPO does not change the Claude API overnight, but it can change priorities. Public companies face pressure to show margin, so expect closer attention to pricing tiers, rate limits and the mix of cloud marketplace versus direct sales. Teams building on Claude should keep a second model provider warm, since concentration risk applies to buyers too.

Capacity matters most for heavy users. If hyperscaler data centers come online slower than planned, rate limits and latency are the first places customers feel it. Track the partners’ quarterly capex commentary, because it predicts Anthropic’s service headroom better than any press release.

For more on the model side of the business, browse our AI and machine learning coverage, which tracks Claude, ChatGPT, Gemini and DeepSeek releases as they land.

Frequently Asked Questions

When is the Anthropic IPO expected?

Anthropic priced the offering on November 18, 2026, at $83 per share for 980 million Class A shares. Trading under ANT on the Nasdaq is expected to begin on November 19, with closing expected on November 20, subject to customary conditions. Anthropic filed its S-1 on October 26, 2026, and plans to trade on the Nasdaq Global Select Market.

How much did Anthropic lose in the first half of 2026?

The S-1 shows a net loss of $80.7 billion for the six months through June 2026. Per Fortune, $72.4 billion of that came from non-cash charges on convertible notes and similar instruments. The operating loss was $693 million.

What is Anthropic’s IPO valuation?

No final valuation is public. Reports put valuation discussions at approximately $1.1 trillion to $1.3 trillion, down from an earlier range of $1.5 trillion to $2 trillion, while Reuters reported a fully diluted target near $1.4 trillion. Anthropic’s roadshow announcement of November 10, 2026 gave a preliminary expected price of $70 to $78 per share, and the company then priced the offering at $83 per share on November 18, 2026, but the official materials confirm none of these valuation figures.

Why does Big Tech AI spending affect the Anthropic IPO?

Amazon, Google and Microsoft supply Anthropic’s computing capacity and also sell its models through their cloud marketplaces. When their free cash flow turns negative, investors worry they may slow construction or tighten terms.

How much has Anthropic committed to spend on compute?

Reports on the prospectus cite at least $110 billion with Amazon, $120 billion with SpaceX, $115 billion with Google and $45 billion with Microsoft. That adds up to at least $390 billion across those four partners.

Is Anthropic profitable?

Not on a net basis. Fortune reported adjusted operating income of $1.6 billion in the first half, about a 10% margin, and positive operating cash flow of $3.9 billion. The GAAP operating line still showed a $693 million loss.

Who controls Anthropic after the IPO?

Seven co-founders hold 50.1% of the vote through a single Class F share, per reports on the filing. Public holders of Class A shares would have economic exposure but limited control.