You're looking at an economy supported by very strong earnings.
We expect the earnings season to be positive.
Can you be bearish when you have overall earnings growth that's going to be probably mid-teens?
We've seen very few signs of weakness into this reporting season.
It has been earnings, not valuations.
This is Bloomberg Surveillance with Jonathan Ferro, Lisa Abramowicz, and Annmarie Hordern.
Good morning, good morning, for our audience worldwide, Bloomberg Surveillance starts now. SpaceX climbing following better-than-expected numbers from Super Micro, providing some relief and sending equities higher. Tesla and Meta drop later this afternoon.
SpaceX giving a lovely tone to anything related to AI. The beat wasn't just a beat; it came in astronomically higher than expected — something like $4.4 billion in income compared to the expectation of $1.8 billion. Is it the same as Meta and Tesla? They were connected to the consumer in a way they could tell a different story. What are people looking for to get excitement around the AI trade, and will they get that from Tesla, where it will be about if they can get the merger done, or Meta, which is an advertising company?
Some commentary on Tesla: get excited about a major combination potentially later on next year between Tesla and SpaceX. Tesla this morning positive in the premarket by 1.5%. This will be a major effort from Elon Musk in the coming months.
Absolutely. The fact that they have the special committee set up that is looking at it, he can't say more than he is just excited. He called it tap dancing on a minefield because he is a conflicting party, but this would be the biggest ever corporate merger. For a lot of people it makes sense: you bet on Tesla and SpaceX, and it is a bet on Elon Musk. At the same time they share a ton of the same supply chains, working towards the same thing, and at the end of the day, don't they all feel like AI companies?
Do Tesla shareholders sign up for it? This could get a little dicey in the coming months.
He put it as a 50-50 chance. The irony is people are saying, including analysts and Elon Musk, that Tesla will be the sticking point. Tesla shares were up more than 2% with the prospect of this merger. One reason why Tesla investors have been excited about this name is because of the prospect of some kind of acquisition by SpaceX. You say who takes over who? SpaceX is worth $2 trillion and Tesla is worth $1.5 trillion. It would be an interesting tie-up in many ways from a financial standpoint.
Equity futures up by around 0.1%. It's a two-speed economy: tech is doing good, bottom up, earnings are fantastic. United Airlines, a bit of a struggle — earnings from them, clearly higher fuel costs are a weight on the outlook. Scott Kirby of United coming up on the program later this morning.
The interesting aspect is, on one hand, they had a huge bill — $1.8 billion of extra fuel costs, which is astronomical and not overcome by some of the revenues — by virtue of the fact they downgraded their full-year forecast. At the same time they said people are still spending, coming in, booking flights, not necessarily offering the same capacity they were as they could, but they want to retain pricing power. At what point does the consumer start pushing back? Is this a consumer read or macroeconomic read where there are certain things they can control?
We'll catch up with Aaron Kennon of Blair Harber, Brian Gardner, and Pierre Ferragu, with the SpaceX-Tesla merger firmly on the table. We begin with stocks rising, optimism returning to the AI trade. Aaron Kennon writing: 'We will remain constructive on the durability of the AI infrastructure investment cycle; demand could extend well beyond 2030, even as the case of incremental spending inevitably fluctuates.' Good morning. Do you think earnings justify where the market is? We will have another 8% or so today.
The trends have been very positive running for the last three quarters — year to date well over 25% earnings growth — and that trend is continuing for Q3. Next year is looking quite solid too. We are expecting high-teens earning growth. The forward multiple, the PE multiple, has contracted, which is interesting. Earlier this year trading closer to 20 times, and now 19 times forward. As earnings have catapulted higher, the market becomes more attractively valued, which is an interesting thematic.
We'll have two stories competing potentially, major companies. SpaceX put up fantastic results overnight, and Meta set to follow later this afternoon. And you continue to have stories circulate around the more leveraged players getting into a little bit of trouble. Where do you put more weight at the moment?
Certainly there is a huge question about leverage at the sovereign level, the question at the AI level. The neo-cloud leverage — what is CDS doing at the CoreWeaves? You have to look under the hood at the corporate level. At CoreWeave you have different counterparty risks depending on, frankly. Undoubtedly, the real question, as it relates to this large spend in AI, as it relates to the increasing debt loads within the AI build-out, is to what degree is the monetization sooner rather than later? What is happening from the perspective of productivity? If we're going to be a corporation that spends more and more dollars on agentic AI, what is the return on investment? I think those are the important questions being asked today. We heard that last night from Ken Griffin, and today in Germany, and we will continue to hear that. Those are important things for us to be debating.
Are tech companies not the full litmus test? Is it non-tech companies adopting it to prove what the ROI is?
At the broader level, corporate America has only embraced the agentic, autonomic AI component. Where is it six months from now, a year from now? I think those are the most important questions. How does it play out through the broader economy as it pertains to inflationary pressures? Yesterday, you have inflation if you don't build data centers because the price of compute is so high right now. You need data centers to reduce spot compute pricing, but by doing that you will create other bottlenecks. You are damned if you do and damned if you don't on the inflation side, in some sense.
As the AI trade came under pressure, there was a theory that it would challenge the capital markets element, that it would ultimately restrain their ability to the same degree. Are we still getting the confidence, or are you back to they will flip the market again?
Last week, if you're the Fed, for example, looking at the dynamic of higher bond prices, lower yields — a little bit of a relief in some sense that maybe the question around monetization of the huge AI build-out is being questioned. Maybe the higher cost of capital due to the higher fed funds rate, higher inflation rate, will have some sort of pause or decelerating effect on the overall economy. Maybe the tool is working in a blunt sort of way. I think that's the bullish commentary from last week. It's an interesting moment, for sure.
How do you price how political data centers have gotten? The president at 1:00 in the morning today was talking about how he thinks these companies should give the dividend to the communities they go into.
We are two weeks from the midterms, and for sure even the president has been a big bull on AI build-out and on data centers. It is waving cautionary flags. Let's wait until November 4. This is a global race of sorts. If the United States wants to cede the ability to grow the data center business, grow their AI presence to other parts of the world, I think that would be a huge mistake economically. I suspect that sound minds will prevail in that sort of way.
Part of the problem is this has become localized, and you see specific AI issues being asked on ballots. Voters can say I want a moratorium, I want construction to slow down. Period. Before the big challenge on AI build-out is the idea you have a massive, temporary, multiyear but temporary, defined build-out happening, with the idea that perhaps the employment piece after the build-out is not as robust as if you were building a steel factory in Pennsylvania. If you had jobs that seemed more robust and more meaningfully lasting, that may play out in a more positive way going into elections and into the future debate on this.
Do you think the midterms are a clearing event or not? Scott at Citadel thinks it is a clearing event.
For the data center debate?
I don't think so. I think there's going to be a really interesting digestion of what happens with regulation if the Democrats take over the House and possibly the Senate. Do we have a lame president? Does the AI, sort of, growth story, sort of, hit the curb for a little bit of time? These are all big questions. I guess, maybe instead of me saying I don't think so, I guess we will have to see how this plays out. At the end of the day, if Talarico wins in Texas, is he really not going to be supportive of the economic build-out that is the data center build-out in Texas? Will he not curry favor with companies that can either build in Texas or go to another place, like Wyoming or Colorado or elsewhere? I doubt it. I think there's a lot of noise right now.
This is a direct quote from Scott at Citadel: 'The feedback we are getting from clients is happy to buy higher, but not buy now into a major political event.'
If there's a shift in Congress, it will be important what these people and their proposals are. It would only take until January to come out against data centers based on how much popular support there is.
Futures doing okay, rolling over a bit, up one-tenth of 1% on the S&P. News worldwide, Vonnie Quinn has more.
Elon Musk has ordered the slowdown of Starlink's satellite technology used to defend Ukraine. The Wall Street Journal reported Musk restricted the speed of Ukraine's access to the satellite system, which would hurt the ability of the country's drones to carry out defensive and offensive missions. Nvidia is thinking of investing millions in the Australian company Firmus after the company abandoned plans for an IPO. United Airlines forecasting weaker-than-expected earnings for the final three months of the year, soaring jet fuel prices outweighing the benefits the airline is seeing from higher fares and robust spending by wealthier travelers. The United CEO Scott Kirby joins Bloomberg Surveillance at 8:45 Eastern.
Looking forward to that conversation in a few hours' time. Selling super intelligence.
Artificial intelligence being spoken of so much now, hereinafter called super intelligence. Whoever wins super intelligence wins.
Next, Brian Gardner. From New York City this morning, good morning.
Five handle on French 10s, the spread over Germany blowing out to 160 and climbing up seven basis points on the session. This move in foreign exchange, the euro weaker, struggling to hold onto 112. 112.17 is the story in Europe. Keep that in mind as we work towards the U.S. session, risk aversion continuing to build on the continent. Stateside, equity futures just about positive, up by 0.1% on the S&P 500, the Nasdaq up. 2s, 10s, and 30s: 10s up by two basis points, 5.13. 30s, 5.44. Selling super intelligence.
Artificial intelligence being spoken of so much now, hereinafter called super intelligence. This is extreme intelligence. It is superior intelligence. And whoever wins super intelligence wins.
The latest this morning: President Donald Trump boosting his sales pitch for super intelligence, urging data centers to sell excess power to local grids. The president saying, 'Make deals, get them built.' Tyler Kendall has the latest.
This comes as candidates across the country face fallout over data centers, whether it is environmental concerns or higher electricity costs. Part of why we've seen President Trump recently push his ratepayer protection pledge, the nonbinding initiative earlier this year that seeks to protect companies on the hook for their own electricity needs. Still, the opposition is continuing, with the data suggesting at least 106 construction projects were blocked in the first half of this year due to community opposition. The White House often touts job creation and higher tax revenue, but the president going on in the post to suggest that communities should be paid out what he is calling a dividend, adding that data centers would be, quote, 'extremely popular if a financial deal is struck.' We should say the idea of sweeteners are underway in some places. In a Louisiana parish, teachers were paid a $50,000 bonus check funded in part by higher tax revenue from one Meta-related project, though not everyone is buying it. Last week developers in a Pennsylvania township proposed a check to residents; the local media is reporting they're still blocking the project.
The price is higher. Let's build on this with Brian Gardner: 'Congress lacks the capacity to regulate the technology and also fears killing the golden goose. It will be left to the executive branch to regulate AI.' Any appetite from the executive branch to also kill the golden goose?
This administration has bought into AI. I will use the old initials. They bought into AI. Look, they are not going back to see the voters any time soon. The midterms come up, the president will not be back on the ballot. I think smart minds know how important and promising the technology is, and I think the administration is wary of some voter pushback. But I differentiate between AI and data centers. They are not the same; they do not resonate with voters the same way.
At the same time, if you have a data center in your backyard, it's because of AI, and there are fears as well with AI in terms of how children and teenagers are using them, but also what does it mean for the future of your job prospects?
I agree, Annmarie. They are tied in on the business side. I just don't see that voters view it the same way. There is some apprehension from voters about the labor market, what it means to their careers. Therefore I think the issue resonates more with younger voters than older voters, who are maybe not as familiar with the technology, are not as worried about it — they see themselves closer to retirement than the beginning stages of their career. Younger voters and younger workers who are asking themselves what this means for their long-term job prospects, there is more apprehension there. That's why I think AI has the potential to grow as an issue. I just don't think it's one of the top-tier issues in this midterm election.
How do you think it will resonate with voters when the president says data centers, even if they are unpopular, pay a dividend and that will help in terms of trying to build them, and also pay for the electricity costs so that are now burdensome to that community?
I don't know it resonates nationwide, but there can be pockets of the country that can see benefits from it. Maybe not so much on the dividend angle, because I think that strikes a lot of people as typical Washington noise, promising things they can't deliver on. When you go into local communities — I was having a conversation recently with someone in Pennsylvania. There has been some pushback in local areas on data centers, but when it was put in terms of, well, it does help demand for coal, and that is a big issue in Pennsylvania, then the question is seen differently. I think it is a very geographic question. Nationally, I think voters are kind of tired of Washington. They dismiss the pie-in-the-sky rhetoric that we were just talking about.
It is definitely going to be a 2028 issue if you don't think it's going to be one of the driving forces of the midterms. If it becomes one of the biggest issues of 2028, how does JD Vance separate himself potentially from this White House that is very gung ho?
That is a good question — the conundrum that every sitting vice president faces: how do I run on my own but not be disloyal to the president under whom that person serves. We saw that two years ago. It was one of the things that sunk Kamala Harris, that voters were unhappy with inflation and the state of the economy, and she could not separate herself from President Biden, who was unpopular at the time. Vance will face the same problem. There is no easy answer to it. He will have to come to some kind of accommodation with the president if he is the nominee and wants to be successful, and pursue this line of attack in his campaign.
Your thoughts on this: what do you think will happen?
I honestly do not know, but I will say if she does run, I think I initially underestimated her ability to win the nomination. Not the general election, but the Democratic nomination. I think she still has a well of support within the party. She resonates with black voters in the Democratic Party, which is a key constituency group. Given how the party has rearranged its calendar for primaries, that South Carolina, where black voters make a huge part of the Democratic base, she could get off to an early start and build some momentum. I don't know if she is going to run or not. If she does — I'm not going to guarantee she will be the nominee, but I think a year or so, 18 months ago, I underestimated her ability to win the nomination.
She has to win the primary first, to your point. Who is on your shortlist?
Who is on your shortlist?
I think the Democratic field is wide open. Governor Newsom, Rahm Emanuel, AOC — it is a very long list. I think the party is going to have to figure out where it is going. Especially we see how the DSA, Democratic Socialists and Progressives, do. This is leading to a bigger discussion, probably at another time, of why I think 2028 has the potential to be one of the most consequential presidential elections in some time. We hear the mantra 'this is the most important election of our lifetime,' and they are all important in some form, but 2028 is more consequential than most. It is a generational change with Democrats and post-Trump Republicans.
The primary battle is going to start as soon as the midterms wrap up in a few weeks' time.
November 4 at 12:01 a.m. Up to 80 candidates are expected to put their name in the hat in terms of individuals who think they can run in a Democratic primary. Winning a Democratic primary will be different than winning the nomination.
The data center debate at the heart of some of these midterm races. It will set the tone for policy into 2028 in the minds of many. Does it turn out to be a clearing event or not? Futures doing better on the tech side, encouraged by better earnings by SpaceX and Super Micro. We'll spend time on that move in a moment. Plus, Pierre Ferragu, looking ahead to the potential Tesla-SpaceX merger. Isabelle Lee has your morning movers.
Super Micro rallying on strong AI demand, up double digits in the premarket, by 11.5%.
Live from New York City, you are watching Bloomberg TV.
The two-day winning streak. The Nasdaq, a little bounce, encouraged by better-than-expected earnings from SpaceX, lifting this market up 5.3% on the Nasdaq 100, with more still to come. Tesla and Meta after the closing bell. In the bond market, 2s, 10s, 30s: U.S. Treasury yields higher by a basis point at the long end, 5.12 on 10s. A week from today, a decision from the Federal Reserve, very small amount of risk into that meeting. The potential for a rate hike, some freezing cold water from the likes of Williams and Jefferson in the last couple of weeks. Let's see what Kevin Warsh can do in a week's time. This spread getting everyone's attention, the spread between France and Germany: 160 as of this morning, and a five handle on French 10s.
Nothing has come through that France has the fiscal and political will to bring down its debt levels. There is a choice: if you want high yield you can go somewhere else that is safer, and ultimately it doesn't look like the ECB is ready to step in yet. What is the cap to some of these yields?
The euro weaker on the back of some of this, and energy costs continue to squeeze the Europeans. Gas prices are high, energy prices across the board higher. Crude: 111 on Brent, 100 on WTI.
Even though we have loads of oil moving through, the Wall Street Journal is talking about how experts say flows have passed prewar levels because of the logistics — that more flows are able to move through because they are being escorted by the U.S. military. At the same time there is a massive geopolitical risk premium when it comes to the Strait of Hormuz, the Red Sea. The real fireworks are under the hood: diesel, gasoline, and the airlines — what's going on with jet fuel.
'We are ready for diplomacy, but we will defend ourselves.'
You are potentially going to have the oil market look at that as optimistic in terms of getting a diplomatic deal done. We have seen the president talk about that he is still happy to engage with the Iranians and getting a deal done. The problem is we talk a lot about this program, but they have lost sight of what they are trying to agree: free flow through the Strait of Hormuz? Iran's oil is blocked in terms of oil not going over land. The nuclear aspect — it feels like we are far away from the nuclear aspect of the debate.
Three hours from the cash open, single names and morning movers. Isabelle Lee has more.
The company reported revenue that beat estimates, surging 161% from a year prior. Driving that was the growth in the AI business and satellite internet business. The stock has been quite volatile since its blockbuster IPO in June, down by nearly 12%. Wall Street remains bullish on the stock. On the conference call yesterday, Elon Musk mentioned the prospect of merging Tesla and SpaceX.
Next, we are looking at Super Micro Computer; shares are up in premarket trade, higher by nearly 12% after the server maker reported preliminary sales that beat estimates. The sales are expected to come in between $8.5 billion and $9.5 billion. Demand shows no sign of abating. Let's take a look at Neo Cloud. The green on my screen — yesterday it had really choppy trading, a lot of crosscurrents in the space. This is a South Korean company that had a dismal IPO. It finished today lower by 32%. It had a better second day, finishing higher by 30%. The CEO dismissed any concerns about short-term volatility, and more broadly the neo-cloud sector, by the overall good earnings we are seeing so far.
Risk appetite certainly there at the moment for those stocks. Anthropic potentially spending $750 billion on additional computing power, according to the Wall Street Journal. A draft of the IPO filing revealed $60 billion in revenue for the first half of this year. OpenAI disclosing its own agents weakened monitoring systems that flag rogue activity. It is taking steps to strengthen monitoring. A bubble isn't always a bad thing.
There are almost certainly bubbles within this moment in time. Any radical transformation of an economy is generally intertwined with a bubble, but what is left in the wake of these bubbles is often game-changing for the world. The companies that are on the other side of the bubble are the market leaders for decades to come.
The CEO Ken Griffin agreeing AI is probably a bubble, while highlighting technological advancements that follow market distortions. In many ways stating the obvious. I have tremendous respect for Ken Griffin, but stating the obvious, and a lot of people agree with that statement.
Tech bulls have been saying we are overspending in places, and you need overbuild for any industrial push, and that's what you are seeing. How big is this bubble? Is it a poppable bubble? How much different companies pay for compute — does that mean that it's a bubble, or there will be winners and losers, which is exactly what the market is trying to distinguish right now.
A winner this morning is SpaceX, for sure. Elon Musk weighs a Tesla tie-up that could become the biggest corporate merger ever.
I think there's a lot of merit to combining companies as we are pretty much in the super intelligence singularity. I think it would be a good idea, but it's really up to the special committees of both companies.
Maintaining a buy on the stock with a 225 price target. Pierre joins us. It's a shame fantastic results are buried by the headline story. Let's start with the headline story before the fantastic results. What is the potential, and who is acquiring who?
From what you see in the comments of Elon, it is extremely tangible. Our committees are talking about it. They have to come up with a proposal. There would be a proposal in terms of governance; any decision at SpaceX is down to Elon. He has extremely tight control on the company. On the Tesla side it is about getting 50% of all shareholders, not the ones that vote — you have to get them to vote. Elon has about a 20% share. Retail investors have close to 60% share, and then you have a load of institutional investors with low single-digit shares to their name. That is something that can be... It will be interesting to see how they are going to go after these proxies and get the vote aligned.
Voting control of the combined company?
That is a good question, and I am probably not enough of an equity lawyer to give you a very definitive answer on that, but he owns 20% of Tesla. He owns north of 80% of the voting rights in SpaceX. It looks like the combined entity would be an entity in which he still has, like, very tremendous control. This might be something that has to go into the conversation. Maybe Tesla shareholders are going to be very demanding and picky on the topic. I think talking to investors in Tesla, simply the fact that SpaceX is a very new business, new company, and you can have a lot of investors who have a very deep understanding of Tesla who don't necessarily understand as well SpaceX. That is an aspect of it that will be educating investors on the merger.
It could be difficult to close, you articulated that. It could risk becoming an unnecessary distraction. Are the synergies worth pursuing when SpaceX continues to perform so well?
The synergies between the two are very significant, yes. They are jointly developing the Terrafab, key to the future of Tesla and SpaceX. It's always difficult to approach such an ambitious project as partners. It is much easier to drag through that kind of project as a senior entity. Then, if you look at Tesla — sorry, SpaceX being connectivity and compute, and Tesla being robots on four wheels and two legs. Obviously robots need connectivity, robots need a super intelligence, artificial intelligence, you name it. Bringing them together in terms of product development is actually very beneficial. Last but not least, the strength of Tesla and the balance sheet of Tesla, and the strength of the balance sheet of SpaceX combined, makes it an even larger market share and larger balance sheet. In a world in which winners are going to be the ones that can deploy and fully control, especially financially, compute, being as large as possible is very valuable. A long list of control synergies, yes.
I want to stick on SpaceX's earnings given how much they blew people's expectations out of the water, with pretty much coming in in line. What can you take from there that gives you confidence about the rest of the tech sector? How much of this is transferable, and how much is SpaceX-specific in terms of the outperformance?
Great question. SpaceX has this particularity. They deploy capex, they deploy compute, and they absolutely don't care about pre-committing it. They are navigating a world that most of the players — first OpenAI, Anthropic, Google with a five-year contract — and that makes SpaceX the marginal buyer, the free float of the market, equivalent to the spot market in memory. The relationship between supply and demand. The strength of their members — they can now also sell a watt of compute for $50 — gives you a high-quality signal on the strength of demand for compute. Three months ago, they weren't talking about it. Now, contracts are being signed. Demand at the tip of the industry strengthening further. Suddenly you're like, everyone is not going to be able to deploy because everyone needs more compute. If you control the neo-cloud, it is good for the chip manufacturers, it is good for the upstream of the supply chain. SpaceX becomes this canary in the coal mine in terms of getting a sense for how much strength do we have from the very top to the very bottom.
Good to see you. This morning: revenue up triple digits from a year ago.
Shocking beats from across the board, not just in connectivity, also with respect to artificial intelligence, which gives them confidence to continue putting 86% of the total capex into the AI — or as Elon Musk would say, the SI unit. Meta coming up next.
They're taking big swings, and we didn't even get to it — they are looking to take on the established telecom players. There could be a big success in other places in years to come. Futures up by 0.5% this morning. With an update on news worldwide, let's get your Bloomberg brief.
Thank you. U.S. political prisoner American Stephen Huppert has been released after a visit by Washington special envoy Steve Witkoff to Russia. President Donald Trump announcing the 74-year-old former schoolteacher is headed back now to the United States. Huppert was sentenced two years ago by a Moscow court for allegedly serving in the Ukrainian military. U.K. inflation hitting an 11-month high: consumer prices rising 3.5% in the 12 months through September, slightly above economists' expectations. The Bank of England widely expected to raise rates in two weeks, the first hike since 2023. Netflix shares falling in the premarket: the streaming giant posting its slowest sales growth in three years while forecasting a fourth consecutive quarter of slowing growth. Pressure.
Next: weathering the neo-cloud storm.
I'm confident how we can use our capital raise efficiently. Looking at basically doubling the revenue compared to last year.
Next, Gil Luria of D.A. Davidson. You are watching Bloomberg Surveillance.
The tech trade doing better this morning; we will see if the moves hold. Positive 0.52% on the S&P, the Nasdaq 100 higher by 0.5%. Under surveillance: weathering the neo-cloud storm.
In the long term, this year, next year, we are looking at explosive growth. I'm not worried about this volatile issue in the first week. I'm confident how we can efficiently use our capital raise. We are looking at basically doubling the revenue compared to last year.
The latest: shares of South Korea's Elice rebounding after suffering one of the worst debuts on record. Gil Luria writing: 'The pressure on public neo-cloud stocks and the recent failures of the neo-cloud IPOs may be a function of the market realizing the perilous nature of that business model.' Welcome to the program. It's good to see you again, buddy. Let's get to the perilous nature of that business model.
These businesses generate a sufficient return to pay their cost of capital — the real basic thing about business. You don't need to calculate a hypothetical here. If you just say adjusted operating income at a yearly rate divided by property and equipment, these are numbers available to anybody who goes and looks. You will see that this company has never generated more than a 5% return, currently generating a 1% return, and yet they are borrowing at a 9% return. That is going up. A lot of the panic last week was about the fact that higher interest rates are going to make that go higher. The CDS rates could go into the double digits, even. That means this company is unlikely to ever generate the returns to pay their cost of capital.
There is a tension. In the short term the CEO of Elice is correct: the demand and prices of compute are so high — we heard that from SpaceX as well — that the problems on the operational side are not going to happen to CoreWeave or the neo-clouds in the short term. It is the longer term. If you don't create value it will be very hard for you to pay down that debt. These companies, many of them, are unlikely to be here five years from now.
If that is the case, can you break down how important they are as a source of demand for the chip players, and what it would mean if those companies went away?
If there was a disruption it could cause some short-term disruption for the chip companies, but really the demand goes beyond those companies. The demand is how much we are using AI and willing to pay for that. It is likely to keep going up for next year.
In a world of no neo-clouds, Microsoft, Amazon, Google would build the data centers. SpaceX and Meta would build the data centers. They wouldn't spill over to these companies. The only reason they are there is companies like Microsoft and Meta are saying there is only so much capex we are willing to spend in a year. The rest of the spillover we will put on three- to five-year contracts with these types of companies as opposed to the 15-year contracts we are taking upon ourselves. At the end of those three- to five-year contracts we will have built enough data center capacity we won't need the neo-clouds. They are a tool the hyperscalers and large internet companies are using in order to not overbuild themselves right now. They don't need them long-term. It's a matter of time before they have enough of their own capacity that they are not going to spill over into those companies.
Can you see some neo-clouds go bankrupt, become insolvent, and the tech trade do well elsewhere?
Absolutely. Especially the poorly constructed ones, like CoreWeave. The demand is high, and the prices they can charge are high, but so are the costs. If we got to a point where the capital markets seized up and they cannot borrow money to replace the debt they currently have, they would go away. There would probably be a deal brokered by Nvidia for Microsoft to buy the assets, but that would leave the equity wiped out, and we would continue to build. Microsoft, Meta, better-constructed neo-clouds may continue to exist, but that wouldn't change anything in terms of the trajectory of the demand for AI, how it is delivered. This would be one poorly constructed company that goes away. Other ones that are relying on debt capital and are putting themselves at this risk could go away. It is the better-constructed ones, and obviously the biggest ones, that would continue to operate.
How much of a liability is this for the Anthropics of the world? They have leaned on them the most to compete with the Googles, who are trying to create their own rivals to the frontier models?
There is a lot of competition, but we saw with SpaceX what happens when Anthropic needs infinite compute: they go to Elon. And they get that on short-term contracts until the build-out gets to a point where they have enough compute. Let's not forget CoreWeave, Oracle — three companies. There are probably 300 neo-clouds now. There is a lot of supply coming to the market. Anthropic will get what it needs. In the meantime they are finding short-term ways, like leasing short-term capacity from SpaceX and Meta.
A final question on politics dominating conversations into the midterms. If it goes in a way that is adverse to the build-out of data centers, who are the winners from your view? A note from Torsten Slok on the winners and losers if the build-out is delayed: the owners of operating data centers are winners. How do you see things?
Amazon, Microsoft, Google have the largest installed base of data centers, and they have the savvy and wherewithal to be able to build data centers even in this kind of environment. Their data centers would become more valuable, and more customers would have to go to them, and they would be the winners.
Gil Luria. The established players, if the political winds blow in that direction in the coming weeks.
If the political winds blow in the direction of the Democrats. Clearly there has been a lot of finger-pointing going on between the White House and SpaceX, depending on what you believe right now, about who is responsible for Starlink basically going down and running at a very low operating rate as they were defending themselves from Russian missiles.
Rubio speaking a moment ago, talking to Zelenskyy, suggesting they will do something about it. Still very unclear what the chain of command was that ended up with the decision ultimately made at Starlink.
The Washington Journal has been reporting that Elon Musk made the decision on his own. There is a question if he would make the decision unilaterally without input from the White House. It creates a political risk for some of the global networks, too, because you wonder what kind of hesitation there will be on national fronts to adopt some sort of technology that does have the potential to be pulled for political reasons.
A better story for tech on the earnings side — SpaceX in the premarket doing nicely. Futures up 0.5% on the Nasdaq, the S&P up 0.2%. We will speak to George Pollack, Akhil Bansal just around the corner.
We continue to have to be invested in the AI infrastructure trade; the opportunity remains massive.
The market is struggling to predict what demand will be.
More equity issuance.
Slowing things down might not be a disaster.
This is Bloomberg Surveillance with Jonathan Ferro, Lisa Abramowicz, and Annmarie Hordern.
Live from New York City this morning, good morning, good morning. The second hour of Bloomberg Surveillance starts now. Your price action looks like this: a better tone to the tech side of the equity market, equity futures pushing higher by 0.2% on the S&P, 0.6% on the Nasdaq 100. SpaceX in the premarket flying sky high. The stock is up in the premarket by more than 8%, crushing estimates. The stock up nicely. Another stock on the move this morning, Super Micro doing well. Numbers from them providing a little bit of relief for the AI ecosystem. Super Micro Computer up by 12%, a double-digit move in the premarket. Overall, the tone firming up for Meta later.
The demand for compute is strong and supporting revenues that supersede anyone's expectations on the AI deployment front as well as on the semiconductor front. A real question is how much of a tone this is setting for the ecosystem, and who is the rest of the ecosystem? The hyperscalers doing the spend, or the semiconductors, or are you seeing it come together because companies are paying this? What is interesting is the run rate came in so far above what people expected in terms of the projections. Really, it seems like SpaceX is putting a high bar.
When Meta comes out today, if it does well, it will confirm it was feeling wobbly on the AI tech trade, but this will confirm it is here to stay. Have you seen the music downloads? 10.9 million downloads since the launch on September 8, according to data released. That is an insane amount of downloads — the consumers taking up this new tool. RBC will make a decision on if its consumers can use it to access the bank. Other big players, we know what they will say, like Bank of America. Muse and Meta today will be interesting for every sector.
Not authorized news interacting with their accounts. They will make a decision, RBC, in the next week or so. Those companies that stay open to Muse and those that put up the wall — a massive case study for the next 12 months: who capitulates to what, who has success and who doesn't?
They want to get the younger share, given that is who is downloading the Muse applications. At the same time, who is worried about security, and who proves to be right?
I think the SpaceX call set the tone for questions in the Tesla call later about a potential for those companies to become one.
We catch up with John Stoltzfus of Oppenheimer, Akhil Bansal, and Alicia Reese. We begin with stocks rising as tech earnings fueled the latest rally. John Stoltzfus is maintaining his 8,100 year-end call, writing that another better-than-expected quarter is in the offering given the corporate fundamentals and increased adoption of AI. Stocks may have further upside ahead before year-end. Good morning. Is that an everything rally or a tech-led rally?
A broader rally than anything else. On a day-to-day basis we see the market going risk on to risk off, risk on to risk off — rotation happens. If you have been riding winners, all of a sudden they become a source for traders to move into other names. I think traders have been carrying the market over the last month or so, battling the intermediate- to long-term investors to catch the babies thrown out with the bathwater.
The chain reaction has gone from stocks to bonds, bonds to stocks. Worrying about a deceleration in the economy, yields dropped, then people play in the companies that got battered by high yields. Is that what you are looking for?
We are absolutely looking for improved breadth in the market as we go forward. We did see some improvement of late. The thing about it is, the earnings season looks really good. We have right now in technology triple-digit earnings growth — skewed by one stock, of course. Something like, they made this print so much longer: information technology brought revenue growth of 86%, 295% earnings growth. That's spectacular. They changed.
I think you can change it.
I can't find my usual buttons.
I will sort it out for you in the commercial break.
How concerned are you that higher oil prices will impact consumer discretionary, given that broadening out requires consumer participation? We saw some crimping around the margins.
Crimping around the margins is to be expected. With that, right away, that is going to be a concern. We can't help but think the consumer, especially when it comes to sentiment — the Michigan sentiment, the Conference Board surveys — generally speaking, it is talk the talk, very negative, but when they walk the walk, they are still spending. Not as robustly as in the past, but the consumer is still — when I look at consumer discretionary, earnings is positive. 9.62% earnings growth. One of the hated sectors this year is doing pretty good.
What do you need to see — some efficiency from the adoption of artificial intelligence? The idea that this is part of the story of better-than-expected profits?
Lisa, when I look at that, what I hear everywhere I go, whether it is businesses or consumers or leisure users of el interneto, as I call it, people are opting for the AI button and they are liking it. I can't help but think we are all on the upgrade cycle whether we like it or not. It needs to be regulated. We need to have certified and vetted AI. It is important to have some old timers, myself, to be able to have some context that matters, so you know when it is hallucinating.
When you say regulation, do you think the industry does it itself, or will this be political?
When it comes to self-regulation, just think back at the financial crisis, when we had too much self-regulation. It needs to be a combination — the government and the industry have to work together to come up with something. So it is not a pejorative; it allows for growth, at the same time it doesn't go to the wild west.
Would you like to see the government take more of a robust stake in these companies, equity stakes?
It's funny. That's the latest thing we have seen from this administration. You know, I think as long as the government, if it takes a stake in these things, there is always the question of vested interest, you know, from the administration or what have you. That is a concern. But if they would have an exit strategy, so they would sell it back into the market, their position, that might help us reduce the national debt.
Wall Street is not going to get the chance to shape regulations in the way it might like, and I wonder how you are thinking about the midterms and how this may play out over the next two years.
I am not a political strategist. I am politically agnostic. I offend all sides.
But you have to worry about the policies.
At the end of the day, while we do worry about the politics, I think the pressure from both sides of the aisle — there is a recognition that people want AI. I think oftentimes the concern is, what happens... The NIMBY business in data centers is a problem, but we are all in that upgrade cycle whether we like it or not, including the politicians, both sides of the aisle.
Which voters are you talking to?
Industry, but poll after poll, it is massive amounts of pushback.
The polls thought Kamala Harris was going to win by a fairly wide margin, and it didn't happen, because there is a silent majority out there that is actually, in many cases, afraid to talk to an anonymous pollster on the telephone because the left has used some of these things as vehicles, and they are afraid of the left. Look at France.
The left and right don't seem to like data centers, though. There is some common ground here.
True. I hope they don't do what happened with nukes, with nuclear — that they find a middle ground here. I think because of the nature of technology being so deeply embedded in the lives of both business and the consumer, I think we have to work it out. It may be data centers that are far away from communities in isolated places. Maybe they will fire something up to Elon Musk's stuff in the sky and come down on the others, but with AI a lot of things are possible. Not everything is possible.
As a strategist, would you frame it as a two-year headwind that takes time to clear, or something we forget about in early November?
I think it is a potential two-year headwind we could run into. You never know. You really don't. As we said one time — you asked me about 2022, what did you learn? I learned anything can happen, and the power of a negative pitch book from certain corners of the market.
We will leave that there. I remember that conversation too. Good to see you, buddy. It was a tough year for a lot of people. The S&P 500 up by 0.2%. With your Bloomberg Brief, Vonnie Quinn has more.
AT&T shares are lower in the premarket. The company reporting earnings that beat estimates on the top and bottom lines, and also raised its outlook for the full year. The company facing some subscriber gains as it rolls out aggressive perks and incentives amid an escalating price war in the mobile phone market. Citadel CEO Ken Griffin says investors are in the midst of an AI bubble.
There are almost certainly bubbles within this moment in time. And any radical transformation of our economy is generally intertwined with a bubble. But what is left in the wake of the bubbles is often game-changing for the world, and the companies that are on the other side of the break are the market leaders for decades to come.
The billionaire offering optimism for a post-bubble-popping opportunity. Democrats suing President Trump and the DOJ over what they say is a coordinated effort to erode confidence in the upcoming midterms. The lawsuit seeks a court order finding the administration's demands related to the list of noncitizen voters violates the separation of powers. That is your Bloomberg brief.
Up next, leaning into super intelligence.
Artificial intelligence, being spoken of so much now, hereinafter officially called super intelligence. Whoever wins the super intelligence wins.
We'll catch up with George Pollack. From New York City this morning, good morning.
Up next, after the close you'll hear from Tesla, Meta, IBM. Thursday we'll hear from Microsoft, Amazon, and Google. Where will we be a week from now, a week into it Thursday? Mark, capex the same, more capex after these companies report?
If they prove the earnings we saw from SpaceX, that will be rewarded in markets like it is being rewarded in SpaceX.
Revenue up triple digits, crushing estimates on net income. That stock is higher by 7.6%. That is lifting equities more broadly. Equity futures on the S&P up by 0.2%. The bounce on the Nasdaq has been a struggle. It is not a struggle this morning. Up by 0.5% on the Nasdaq 100. On the politics, leaning into super intelligence.
Artificial intelligence, being spoken of so much now, hereinafter officially called super intelligence. This is extreme intelligence. It's superior intelligence. And whoever wins super intelligence wins.
The latest: President Donald Trump putting a new focus on so-called super intelligence, calling on data centers to sell excess power back to local grids, the president urging companies to "make deals" heading into the midterms.
The president calling on companies to pay what he is calling a "dividend" to communities where data centers are being built. He went on in that post on Truth Social to suggest data centers could be more popular with the American electorate if there were financial incentives for local residents. We should say tech companies have been increasingly offering financial packages to cities and towns for local upgrades like infrastructure or community resources. We saw OpenAI earlier this year, for example, give $100 million for a rec center in Michigan to be built. But direct cash payments would mark a new phase in the AI buildout as the pushback to construction remains stark. A new poll from Ipsos within the last hour finds that 66% of Americans oppose data center construction, and that number jumps to 72% when respondents were asked specifically about having a data center built in their community. Few issues in the midterms have bridged Americans' partisan divide like this, as the current Congress grapples with what to do next. John, we have seen a flurry of bills introduced that would stop tech companies from benefiting from tax breaks, and another that would impose taxes on any local incentives to build data centers, though the expectation in Washington is that nothing will be passed anytime soon.
George Pollack of Cigna writing the following: not expecting Washington to regulate AI anytime soon. We will need a disaster of some sort to drive the need for regulation. George joins us for more. Welcome, it is good to see you. I want to start with the president's position. How much daylight is there between his position on the issue and the Republican candidates on the ground heading into the midterms?
I think there's quite a bit of daylight between the president and Republican candidates. They are trying to be as positive about the president's message while trying to avoid the voter pushback against data centers. That is why states like Ohio and Michigan, Republican candidates push back and try to frame it as Democrats being pro-data center. In New York, Lawler has said that Connolly is in fact pro-data center development. There is quite a bit of daylight. That doesn't mean there will be daylight in legislation, as most Republicans are too concerned to be crosswise with the president and the administration.
Is it going to come down to a localized issue, given that it is on a number of ballots? The president's Truth Social post said if companies want to make it popular, give money back to the communities, have a dividend. Do you think there will be a lot of force from the executive branch on these companies to do that?
The companies are already doing that. In Louisiana, where Meta is building a significant base of operations, they have given teachers $50,000 bonuses, and counties in Virginia are considering rolling back property taxes. In places like Pennsylvania, data centers are offering $10K to residents to allow them to build data centers in their communities, but the money is not moving votes, because the voter concern about data centers goes beyond money — to the effect on their property values, to the perception of the impact on energy costs, to the good old American NIMBYism that isn't going anywhere anytime soon.
Who is striking the tone on the campaign trail?
The right tone is the negative tone for AI; the tone for the voters is against it. Former Senator Sherrod Brown in Ohio has been the one leaning into the anti-data center messaging, and it has seen him rewarded with polling. The senators trying to push back, saying this is what I'm doing and why it is good and why it is a productive thing, and make that national security argument. That is correct, but to the voters it does not matter. I think in terms of the midterms, the right argument — unfortunately electorally — is the wrong argument for the long-term future of AI in the United States.
We are 13 days from the election. Do you still have Texas going blue?
I still have Texas going blue. While I avoid looking at early voting data because anyone can tell a story with that data, I do find that even with Republicans increasing their spending, James Talarico has been able to set out a pretty compelling message that Ken Paxton has not been able to really push against. Even if you look at Talarico's final ad, the framing was: I have never been indicted, I'm not under investigation, I have never been impeached. Even in a time when corruption is at the top of voters' minds, in a place like Texas, it will prove beneficial for Talarico.
We have been talking about data centers, and a lot of people have pinpointed this as the main agenda item for Democrats where they could get bipartisan support. I wonder about corporate taxes. We heard from Ken Griffin last night that he thinks it was a mistake to cut corporate taxes to the degree that the U.S. administration did last year because of the fiscal deficit, and that it created a problem. Do you think there could be some sort of increase in corporate taxes that comes down the pike sooner rather than later, and it could be a bipartisan issue?
Yes. You have to look at it: if we get a Senate with more, and as chair of Senate Banking, that will make markets quite nervous. Senator Warren and the president surprisingly worked together to pass the housing legislation that included a pretty populist restriction on investors purchasing single-family homes. My understanding is there is legitimate concern they could work on some kind of deal to increase the corporate tax rate from 21 to 25 to 28, and that could address concerns in the bond market, with the deficit, and politically. We are in a populist age. Who is going to step up and say we have to defend corporations from higher taxes when most Americans are more focused, on all political stripes, on keeping their own taxes lower?
