#S&P senkt Meta Kredit-Ausblick auf negativ
About S&P senkt Meta Kredit-Ausblick auf negativ
S&P hat den Ausblick für Meta-Credit auf negativ herabgestuft, da die Ausgaben für die KI-Infrastruktur weiterhin steigen. Ein repräsentativer Beitrag zeigt, dass die fünf großen Tech-Giganten zusammen mehr als 3 Billionen US-Dollar für KI-Infrastruktur zugesagt haben, darunter über 800 Milliarden US-Dollar für Meta und über 1 Billion US-Dollar für Google; Alphabet hat aufgrund der KI-Ausgaben einen Marktrückgang von 475 Milliarden US-Dollar erlitten, und das Finanzierungsziel von OpenAI wurde Berichten zufolge nach unten korrigiert. Der Markt bewertet derzeit die Nachhaltigkeit der enormen KI-Investitionen und der Fremdfinanzierung neu, während Musk sagt, dass der Anleihemarkt Unternehmen, die nur GPUs anhäufen und keine Nebeneinnahmen haben, kaum Zugang gewährt.
Angesagt
Neueste
S&P senkt Meta Kredit-Ausblick auf negativ Beliebte Beiträge
S&P keeping Meta at AA- while shifting its outlook negative puts the debate on funding capacity, not a binary verdict on AI. The projection of more than $225B in 2027 capex alongside a roughly $71B free operating cash flow deficit shows how quickly investment can test credit resilience. Execution and eventual cash conversion remain uncertain.
#MetaNegativeOutlook
That's it for China.
Today, another $300,000,000,000 was wiped off China's stock market, pushing it deeper into BEAR market territory.
The CSI 300 has officially fallen for 12 straight weeks, its longest losing streak in history.
Chinese tech stocks in bear market:
MiniMax: -85%
Zhipu: -85%
Zhongji Innolight: -60%
Eoptolink: -60%
Unitree: -50%
MetaX: -50%
Moore Threads: -50%
SMIC: -49%
Cambricon: -43%
Biwin Storage: -40%
CXMT: -36%
All public bio CEOs right now are trying to figure out how to say “AI” on their earnings calls and also not have to overcommit to even top-line impact on the company.
The only one imo who hasn’t been measured in this has been $TWST which has gone vertical (potentially past a point of reason) as a result.
It’s quite likely to me this will create substantial whiplash over the next ~year before showing up meaningfully in results.
Disclosure: Compound is long and short a variety of bio companies and at some point in the future this may include $TWST.
Scorpion Capital
Sudden reality check from $ILMN Illumina. Tanking after hours and taking "AI drug discovery" meme stocks with it, like $TWST. Evasive, weak answer from management when asked about potential contribution/visibility from their AI offering in 2027. Wouldn't even commit to a mere 1-2% revenue increase:
“I think our view right now is somewhere in 2027, there should be some sort of color we should be able to begin to add...but we anticipate that somewhere in 2027, it does become more visible. And the whole point around the new products and BioInsight adding 1 to 2 points of growth for 2027 would certainly be an area."
Hype vs. reality per their earnings call transcript:
The hype - prepared comments, excerpt:
"On the analysis side, we launched the BioInsight platform, which brings our analysis application, AI models and curated biological data together in one place, reducing the need for customers to assemble and manage separate software tools. Through the platform, customers can also access SpliceAI2, our new AI model that predicts how genetic changes affect RNA splicing. In rare disease research data sets, it identified 17% more disease-relevant splice variants than other splicing models."
"I'm proud of the innovation our teams continue to deliver. Our view is that the next era of life sciences innovation will be powered not only by sequencing, but also by the software, AI and differentiated data that unlock the power of the genome. Through BioInsight, Illumina is building the base for AI biology, enabling more powerful foundational models and virtual cell models with the potential to improve our ability to understand and predict biology. As we move nearer to predicting biology, we see a path to using those insights to move health care toward a more proactive future, identifying risks sooner, preventing disease where possible and intervening earlier."
The reality, per Q&A:
Operator
Our next question is from Dan Leonard from RBC.
Daniel Louis Leonard
A bunch. Jacob and Ankur, you talked a bunch about BioInsight. When would BioInsight become more visible in the P&L? Would that be as soon as next year? Or is that a further-out event?
Jacob Thaysen
Yes. So I think, first of all, we are -- we actually established BioInsight approximately a year ago, which was the combination of our software part, which is actually doing extremely well also. And I think for those of you who visited our booth and at our symposiums at ASHG, we really presented how important software is for the ecosystem. So very excited about that. But we combine that with the data because there's a lot of synergies between the two. You have to generate data, but you also have to develop models on top of that.
So we are -- so as we mentioned in 2024 also is that the data and, of course, the multi-omics business would start to generate 1% to 2% in the latter part. So here in '27 of our strategic period when we look to 2024, and that is playing out very well. But I think you will -- depending on how that AI opportunity plays out, which we actually looks very, very promising, this could be substantially bigger. But maybe, Ankur, you have comments on that also.
Ankur Dhingra
Yes. I think our view right now is somewhere in 2027, there should be some sort of color we should be able to begin to add the Billion Cell Atlas and then the second gen, we've talked about it expanding to 5 billion cells certainly would be one of the first drivers within that. Jacob mentioned, we actually have a very broad spectrum of customers in discussion about the Cell Atlas and several other new bio AI-based applications that they are working on. But we anticipate that somewhere in 2027, it does become more visible. And the whole point around the new products and BioInsight adding 1 to 2 points of growth for 2027 would certainly be an area.
Source: CapitalIQ earnings call transcript
That changes the scale of the opportunity for companies building these models.
The AI investment debate will continue, and valuations still need to make financial sense. But dismissing the entire sector as a bubble ignores the growing demand from institutions that can not afford to fall behind.
The next major AI market might not be won by whoever builds the most popular chatbot.
Anthropic
taishiro
A vulnerability that could take security researchers weeks to discover might be identified much faster with advanced AI assistance.
Of course, these models are not a replacement for security teams, and finding a vulnerability is only one part of defending a system. Implementation, human oversight and responsible access still matter.
But the direction is clear.
Governments are beginning to see frontier AI as part of their defensive infrastructure, not just another technology to experiment with
META’S AI SPENDING RAISES QUESTIONS ABOUT CASH FLOW
S&P revised Meta’s outlook from stable to negative while maintaining its AA- rating.
The concern centers on projected 2027 capital expenditure above $225 billion and an estimated $71 billion free operating cash flow deficit.
This doesn’t automatically mean Meta’s AI strategy is failing. It does highlight the financial pressure that comes with scaling infrastructure at this pace.
#MetaNegativeOutlook
S&P just put a number on $META's AI bill, and it's kinda ugly
After the close Friday, S&P cut Meta's credit outlook to NEGATIVE
It kept the AA- rating, but it now sees Meta burning cash for 3 straight years
It models free cash flow at -$22B in 2026, -$71B in 2027 and -$39B in 2028. It used to expect Meta back in the positive by 2028
The reason is capex. S&P now sees over $225B in 2027 and $252B in 2028, way above its old forecasts of $163B and $178B
It also sees leverage topping 1.0x as early as 2027, the level that puts the AA- rating at risk
And S&P says Meta's ability to make money from AI is more limited than its peers. Meta says it "respectfully disagrees"
Yesterday I said the fear right now is the BILL. S&P just made it official
Chinese models run 112 times cheaper than Anthropic for every million tokens.
Chamath broke it down on CNBC: Anthropic's "barrel of intelligence" runs $56, OpenAI $26, Meta $1.50, xAI and Google $1, and Chinese models $0.50.
That gap amounts to the steepest commodity curve in the recorded history of technology.
Oil needed 40 years to compress that way. Semiconductors needed 20. AI inference is getting there in months.
The firms priced at $26 and $56 aren't careless. They're gambling that trust, safety and enterprise ties keep the markup in place long enough for their costs to fall.
The timeline is the one thing they can't bet on. The $0.50 model is a working product.
Interested in AI?
Follow @FeliciaRios_ and never fall behind.
I track ChatGPT, Claude, and how every new agent is changing how we work and think, then hand you the pure signals.
