Hawkish Fed Minutes Drag Stocks From All-Time Highs

Bloomberg Audio Studios.

Podcasts.

Radio.

News.

This is Bloomberg Businessweek Daily. Reporting from the magazine that helps global leaders stay ahead. With insight on the people, companies, and trends shaping today's complex economy. Plus, global business, finance, and tech news as it happens. The Bloomberg Businessweek Daily Podcast. With Carol Masser and Tim Stenebeck.

We've got a lot of big thinkers that are coming at us right now.

J.P.

Morgan's Jamie Dimon spoke exclusively to Bloomberg yesterday. Nassim Tlaib, he, of course, who popularized the phrase black swan, also out with some comments. Ray Dalio, we've been covering all of this. So too, though, is a big thinker our next guest, which is why we are delighted to have some time to actually dig into some of What ails us? What questions that are out there when we look at the marketplace?

Jeffrey Rosenberg, Managing Director, Portfolio Manager, Systematic Multi-Strategy Fund at the world's largest asset manager. We are talking about BlackRock. They've got some $ 15. 3 trillion in assets under management. It's a firm and an individual very well known to the Bloomberg audience. Welcome. Well, thank you.

That was a very esteemed company to be in on your intro. So we'll do our best.

No pressure, no pressure. But I think we are leaning on a lot of different voices to kind of understand the way forward. I love folks that have seen market cycles, seen crises, and understand some things. How would you describe the environment right now? Everybody talks, we're talking about American exceptionalism a little bit again this week. We see growth.

It seems like the Fed is managing things, but how would you describe it?

Well, I think we're really just talking about artificial intelligence. That is the theme. We're in the midst of an unprecedented historical, by historic standards, investment wave. There's been a lot of interesting stuff written about it recently, comparing you know, investment as a share of GDP to canals, to railroads, to electricity, to the telecom infrastructure build out. And this wave will eclipse all of those. So it's AI everywhere all the time.

I did not expect you to start there.

It's all we talk about. And it's a macro theme as well. I mean, I spend most of my time in interest rates and in macro, but the macro theme spillover is that AI is the primary driver of of above-trend growth. And it is an investment-led growth, but it is also consumption-led growth, because the wealth side of the AI, everything you talk about here on the program, today's equity market moves, you talked about it earlier, the concentration story, that is all downstream effects from AI. So almost everything traces back to the AI story. In interest rates, if we talk about the worry about the big increase in the bond market interest rates, what's happening? Well, part of that is AI related because it's the competition for capital

that the AI investment boom has completely turned around an old, old story, which we once had, Larry Summers' secular stagnation. Not enough investment demand, too much supply of savings. That led to very, very low interest rates, and today it's the opposite. Not enough global savings to fund an incredible demand for investment and investment opportunities.

Well, since you mentioned AI, and Carol, you mentioned before Ray Dalio, so he talked about the AI bubble.

What do you think of it? Is it too much? I mean, is it going to burst?

So I want to kind of steer away from the bubble language. There's a lot of headlines on it, a lot of people talking about it. It's loaded language because of our experience with bubbles. I think we have to think about this less from kind of market bubble psychology and much more from like the fundamental investment case, which is really unique here. And what's unique is that this is an expansion of the capacity, right? So, so there's AI, but what we're really doing in the investment world is the infrastructure build, right? That's why it's hitting the real economy.

So this is data centers and this is transmission and this is chips and this is energy. So it's affecting the real economy long before the impact. And we're talking about, you know, our adoption of AI, but that whole adoption of is in front of us. What's with us right now is the infrastructure build-out. And what's really critical is that as much infrastructure is being built, the supply side is still dwarfed massively by the amount of demand that we're seeing and the demand shocks that we continue to see.

When I say demand shock, what do I mean by that? I mean that every time we think we kind of understand AI, oh, it's a chatbot. Oh, it's replacing my search engine. Oh, it's we discover an entire new category of uses that have really kind of blown away our prior understanding of how much demand is this technology really going to deliver. So the latest in that is Muse, of course.

Yeah, and so by being involved in the infrastructure, you don't care what LLM wins, right? It's just a case of we're going to build the data centers because whoever is the winning models, we're just going to have the infrastructure there to support it.

For now, because that demand piece appears so far ahead of where the supply is. Now, that may change.

Is that going to last? Do you expect it to last forever?

Well, you know, trees don't grow to the sky, right? That's a famous saying. So at some point, we're going to have to make these two things match. We're going to have to get the scale of the investment to the scale of the demand. The problem is nobody really knows because no one really knows what the sensitivity of demand is to the reduction in price, that elasticity. And it may be big, it may be fixed, it may be small, but that's really the unknown here as to where we'll end up. But it's very early stages, I think, in that kind of debate.

And the narrative has changed so much. Jeff, hang on for a second. We do want to play Black Swan author Nassim Tlaib. He did, though, sound the alarm on AI. He says, tech companies driving markets higher may not ultimately emerge as the winners of the AI boom. He spoke at the Greenwich Economic Forum and spoke with our Natalia Kinijevic.

Listen up, everybody.

There's a class of naive investors who say, okay, AI is going to change the world, hence let's invest in firms that are linked to AI. And that's probably the most naive thinking. Because if you look.

Backwards, the thing that did change the world, in retrospect, we can see that the company that benefited We're not the pioneers.

All right, that, of course, was Nassim Tlaib of Universa Investments there at the Greenwich Economic Forum. Jeff, though, there is this thinking, too, and we had a couple professors on who've been analyzing kind of last tech waves and looking at companies that were once, whether it's Kodak or some others that were household names that are no more, and looking at this cycle moving so rapid that some of the companies and these professors naming companies take your mag seven, take your top big tech names saying there's no guarantee that all of these guys are going to be around in 10 years or 20 years. How do you look at kind of who might be the winners or losers?

Yeah.

So let's, let's bring it down a little bit to the investment implications and what, what it means for people.

About a minute and then we'll come back.

Portfolios today. So, so I mean, broadly, I think what, what the implication of all that is in portfolios is this incredible concentration we talked about in the early segment. You know, it's, it's historic concentration in the indices. So when, Talib is talking about kind of naive investing. If you're just owning equity market beta, if you're just owning the market portfolio, you have a very concentrated portfolio. And so challenges to diversification in this environment. And your earlier guest from McKinsey was talking about this. Investors need to diversify.

So we'll come back and we'll talk about that because I think that's the most important thing. investors can think about is, well, how in this environment? You've got to recognize the problem. The problem is the flip side of the success of AI is it's making my beta go through the roof, equity indices at historic highs, but it's making my portfolio more concentrated. So I have a bigger demand for diversification, and we have to really think differently about diversification in today's environment.

Do customers get it? Just got 20 seconds. Or do they say, well, wait a minute, but this is where the momentum is and that's where I want to be.

I think it varies. I do think people definitely get the idea that their equity portfolios are concentrated, that they've had great wins. And you get a lot of financial advisors who are looking at that and looking for how can I diversify my client's portfolio away from this trade. I still want to be in it, but I want to have other things in my portfolio as well.

All right. We're going to come back with Jeffrey Rosenberg, Managing Director, Portfolio Manager at Systematic Multi-Strategy Fund at BlackRock, joining us right here in our Bloomberg Interactive Broker Studio.

Stay with us.

More from Bloomberg Businessweek Daily coming up after this.

Want to get back to Jeffrey Rosenberg? He is Managing Director, Portfolio Manager, Systematic Multi-Strategy Fund at BlackRock. He's here in our Bloomberg Interactive Broker Studio. So we're talking about AI and how you've got to be important when it comes to diversification. What does that mean in this environment?

Yeah. So two things to understand about this environment. We've talked about the first one in the earlier segment, just how concentrated equity portfolios have become. And it's the flip side of the good news story, which is the valuations of these AI companies and technology companies are expanding proportionally and sometimes behind their cash flow and their earnings power.

But it's been an incredible concentration in the equity side of the portfolio. So you've got that story. You have another story on the macro side, which is what's happening to the bonds and the bond market, which is since COVID, we've had too much inflation. And so when we've moved into this era of too much inflation, that structure of having too much inflation undermines the diversification role of fixed income in your portfolio. So we've lost an anchor of asset class diversification in portfolio construction at the same time as we need that diversification from the equity side. So what is one to do Well, let's, sorry for the shameless plug, talk about another fund that I managed, which you didn't mention in my bio, but it's the iShare Systematic.

I left it for you.

Thank you.

iShare Systematic Alternatives Active ETF. And so alternatives is in the name. And what we're talking about here is not alternative asset classes. We're talking about alternative, excuse me one second.

Sorry. We'll try to get you some water.

Alternative strategies. And the alternative strategies create differentiated return streams. And basically what we're bringing to an ETF wrapper, which no one has ever done, are strategies that we run in our institutional hedge funds. Long, short, market-neutral strategies across micro and macro markets. What does that do? It gives you a differentiated source of return. Differentiated, meaning it's not about our markets going up or going down. It's not about our interest rates going up or going down. It's about what's happening inside the markets, about the idiosyncratic, the distinctive individual exposures that we can capture when we have a toolkit that is more than just being long only, which is what typical retail mutual funds and ETFs have

exposures to. So it's a different kind of toolkit. And when you have that power of that tool to be able to go long and short and take out market directionality.

You pick up nuances of the market.

You pick up nuances of the market. And really simply, it's uncorrelated returns. And having uncorrelated returns in a market environment where you've got a lot of correlation, stock and bond correlation, correlation between equities, that's an incredibly powerful thing. And it can be additive to your portfolio. as a source of diversification, helping you to do what? Well, not have to rely so much on is AI a bubble or not a bubble. No one's going to get that call right, and no one's going to know when to make that call. If you base your investment strategy on that, that's a very low probability of success. So we don't want to have one big low probability event. We want to have thousands of individual exposures. That's the idiosyncratic risk,

and that's how you add diversification.

But then what's the challenges of having all that exposure?

Well, the challenges are maintaining what we call asset breadth. So what makes those strategies work is that the individual exposures are uncorrelated. If you end up doing all this work, but at the end of the day, it's all the same trade, you haven't really created any diversification.

So uncorrelated can become correlated very easily.

And so the devil's in the detail of how do you design these strategies so that you don't end up with correlated strategies, that you're ensuring the lack of correlation. And so that's what the structure, long-short, really gives you a leg up. Because when you're basically buying one stock and shorting another stock, you're pretty much at a starting point of removing that beta exposure to the market, which is your number one source of correlation. It's not your only thing. You've got to look at other factors and exposures as well. But you get a long way to solving the problem by opening up the short side of the portfolio.

And you know this, Jeff. There's different factors that can change that. And that has to do with how high we go on rates. And I just want to bring Jamie Dimon into the conversation. He spoke exclusively with our Tom McKenzie out in London. And he talked specifically about rates and rates going higher.

Just listen up, everyone.

Rates are going up, a lot of demand for capital, a lot of government financing. In my view, a little bit expected. If it goes up higher than here, does it create risk? If credit spreads gap, yeah, it creates risk for all companies. It changes valuations. And then your market's open and closed, as you know. So there's no permanent state of affairs when it comes to markets.

That's right. Jamie Dimon at the conference in London, the J.P. Morgan Techstars Conference with our Tom McKenzie. That's right. Markets are fluid, right? That's what happens. But I do want to go back to, I know you started with AI, but AI and rates, all of this stuff is connected. So I'm just curious, how high do you think U.S.

yields go at this point? Do we go six? We've had some conversations this week, 6%, 7%. Do you have a thought?

I definitely do. And the way to anchor the rate conversation is around nominal GDP. Rates don't just... grow to the sky. They don't fall to the ground. They're anchored to economic fundamentals. Yes, they can decouple over short and longer periods of time, but ultimately that's the driving force. Why did rates go higher? Simply because nominal GDP went higher.

Why did nominal GDP go higher? Because of the combination of better real growth, that's the AI spillover effect, and higher than expected inflation. That's a little bit of AI on the components and the inflationary aspects, and it's a lot of the headline on the inflation. But before we leave Jamie's comments, let's just bring it back to this diversification piece and link the kind of the beta conversation.

Jamie's talking about rates go up, credit spreads go wider, you're going to have risks. But let's link that back to our earlier conversation around alternatives and diversification through long-short market strategies. Because here's the really, really important insight is that there's a relationship between the beta story, markets going up, markets going down, rates going up, rates going down, and the alpha story. And that relationship itself is incredibly additive to your diversification. And that relationship is mostly not always, but mostly there's a negative correlation between the opportunities that are happening in the beta space and the opportunities that happen in the alpha space to wit. So when you

have really difficult equity market environments, that tends to lead to a lot of differentiation in the underlying equity markets performance. It's a measure that we call dispersion. And when dispersion goes up, think of dispersion as how much is the markets valuing the difference between winners and losers. You tend to have a much better alpha environment. Now, before you stop me, let me just explain what's unique about this environment.

Because the markets are going straight up, but it's an AI-driven environment. And that is creating competition. winners and losers. So it's an alpha environment of lots of dispersion at the same time as the markets are going up, which is really quite unique.

I kind of love that you went there. We talked about this earlier, Lisa, that we had some records on the S & P and the NASDAQ, and yet there were... not, there were a lot of stocks that weren't part of it. You saw a lot of stocks either hitting lows or so, you know, so it's just interesting when you see and dig down below.

Exactly.

Of what's going on.

But it's not by accident, right? Back to our earlier thing, it's like all AI all the time. Because AI is a driver. for the creation of winners and losers. Who is going to benefit from this incredible wave? Who's going to be left behind? And that separation between winners and losers is highlighted by the stat that you mentioned. You got some winners, but you got a lot of losers here.

But for an alpha investor, when you're long short, that's an incredible investment opportunity to add alpha sources of returns to a portfolio. And that's what IALT, is the ticker, gives an opportunity for clients, really in the first time we've ever seen this in the ETF wrapper, for retail investors to get access to that kind of investment style, that investment toolkit.

Very granular.

Very granular, very institutional, very much what we do for institutions in the hedge fund space, bringing it to the retail space. But the value of that is just the necessity and the opportunity of the moment right now.

We were talking about rates earlier. The minutes came out not too long ago. I wanted to ask your opinion. Was there anything of note there for you? Because there wasn't much of a reaction.

It wasn't much of a reaction.

Do you care about minutes?

No, we do care about minutes. We get a lot of information about the minutes, but the issue is what's the surprise in the minutes, right? And so the minutes were basically kind of as expected. You know, there was a little bit on the AI demand overheating story, which may have been a little bit of a surprise, but not enough to really move the bond market needle. The minutes are helpful. They validate and they entrench a narrative around Fed and Fed expectations. Sometimes when we do the inside baseball of some and many and a few and do that counting, the inside baseball prognosticators care and get very excited about that. But it rarely moves the markets.

We've only got a couple of minutes. We haven't dug as much into geopolitics as I would hope. But I do think that is still a risk that is out there in a big way. And it's often... the things we talk about that worry investors most. What's the smart conversation for you guys at BlackRock around this right now, and how do you position for investors?

I'll be doing a panel tomorrow at a Hedge Week conference, and this is the topic of our panel. So let's just separate.

Got about a minute or so.

Got it.

I'll make this quick. So let's just separate what we mean and unpack geopolitics. Because if I say geopolitics, you say- Okay, so everyone is really, and Warsh made this sort of use of geopolitics to refer to Iran and the oil impact. But geopolitics means a lot of things. It means the midterm elections. It means elections in Brazil and other places.

So it means a lot of different things. Generally, in the biggest sense of geopolitics, what we have observed for a very long time is geopolitical risk tends to not be persistent. It tends to spike and then reverse. And so your playbook as an investor is, is to not overreact to the geopolitic risk writ large. Now, Iran and oil is a little bit different story because that's had a lot more.

And some days the market does respond, some days it doesn't.

Correct.

And that's a different story because that's a trend on inflation and oil and oil disruption. And that has been in its own category of geopolitical risk. Stay with us.

More from Bloomberg Businessweek Daily coming up after this.

We've seen a bit of a choppy session when it comes to oil traders weighing a pickup in Iranian attacks on vessels in the Strait of Hormuz against resilient flows from the Middle East trying to make sense. You had Yemen's internationally recognized government backed by Saudi Arabia claiming to have regained territory from the Houthis as fighting in the region near the Bab el-Mandeb Strait intensifies. It's just ongoing.

And I'm very curious... about the US role in all of this. We've got a great voice to talk to.

We definitely do. We want to do it right now and bring in her. This is, oh, I just lost it.

That's okay. Jenny Welch. Jenny Welch. Thank you for joining us, Jenny.

So is this about the US war in Iran anymore? I mean, what are we talking about right now? Help us understand where we are. Lay it out for us.

So this is another front in the war that has opened up in recent months. And this is specifically between the Houthis in Yemen and Saudi Arabia and the internationally recognized government in Yemen. And this is a conflict that was underway for over a decade. It paused in recent years, but the outbreak of war over Iran has reignited it, including over the summer with the Houthis firing on Saudi Arabia.

Now we have Saudi Arabia fighting back. And what this creates risk for is the other key waterway in the Middle East, which is the Red Sea. And that has become especially important with the Strait of Hormuz getting choked off by Iran, that more has been flowing through the Red Sea, including out through Saudi ports there. And the reignition of this campaign really puts at risk some of those diversions.

You know, one of the things, Jenny, you are our Bloomberg Economics Chief Geoeconomics Analyst. And so you look at all of this. what's going on around the world, what it means in terms of U.S. economics, global economics. I guess what I'm trying to understand, is this the next chapter in the war? Because it does feel like I'm not hearing as much about U.S. involvement and actions, and I'm curious where we are in terms of U.S. involvement, but it does just sound like we're going to another level here, and it makes me think, again, this is longer or will go on much longer than we anticipated.

I think the way to think about it is the Middle East was already a region facing security challenges. It already had pretty significant regional rivalries and conflicts happening. And the Iran war has just shook the puzzle further and opened up new fault lines and reignited other ones. And this front is a great example of that. And I think what it also points to is the US Iran war doesn't look any closer to conclusion, but even if it were to wrap up, it is leaving behind a lot of these open fissures that may take even longer to resolve.

Jenny, I got to ask you, because the Saudi-backed coalition, right, they said that the objective of their attack against the Houthis in Yemen was to eliminate the group of threat as it goes to shipping.

Is that possible?

I mean, would it remove that Houthi threat to that strait or to Saudi Arabia?

I think certainly the counteroffensive has already achieved some tactical gains, retaking territory that the Houthis took in recent weeks. but the strategic impact of those gains is going to be relatively limited. And what it would take to really fully eliminate the Houthis and their ability to threaten both Saudi Arabia and shipping in the Red Sea would be something far more dramatic. I mean, when we think about it, since Yemen's civil war broke out over a decade ago, Saudi Arabia's entrance into that, there were years of intense fighting that didn't disrupt, or I should say, didn't dislodge the Houthis.

The U.S.

Was even engaged in very heavy bombing campaigns, including at the beginning of Trump's term. Those didn't eliminate the Houthis. And so I think it's going to be very difficult. It's a very tall order to fully eliminate them. What I think Saudi Arabia is trying to do is retake some of the initiative here, put pressure on the group, maybe try and get themselves back into a better negotiating position to restore the ceasefire that had been in place prior to this reignition of the conflict.

But you know with this whole conflict, we always talk about Iran or even Israel, about things being existential. This is about, they think about their future and what it means for their people. And so when you are fighting with that kind of mission or goal, you will do everything and keep going for as long as you need to.

And so I do wonder, is that a big factor when it comes also to the Houthis? And I'm just trying to figure that out. Same thing for Iran. Certainly for the Houthis.

I mean, they look at this as an existential conflict for them. And I think they are emboldened by what they've been able to achieve, not just in Yemen, but in holding traffic in the Red Sea, this major international waterway at risk since late 2023. I think for Riyadh, they're a little bit more pragmatic, right? They did go to war against the Houthis when the civil war broke out in Yemen. in 2014. They fought that campaign for a while, but then they eventually signed a ceasefire with the group. They want to focus on their economic development.

They want to focus on rebuilding in the wake of this conflict with Iran. And I don't think they're necessarily looking for a major war. I think what they feel is they don't really have a choice here.

Jenny, how much of an impact are the attacks having on oil flows? Because we've been hearing they're close to pre-war levels.

So how much of an impact is it having?

So at the moment, Saudi Arabia looks like they've been able to restore some of the flows that were impacted by prior Houthi attacks, including on the east-west pipeline. But I think it's important to note that that infrastructure still remains very much at risk. Saudi Arabia has got a big swath of territory to protect. Many of these key infrastructure points are very large. The pipeline, for example, is a very large area to have to try and cover with air defenses that are increasingly scarce after the conflict with Iran. So this is going to be, I think, a challenge and something where we're going to see ebbs and flows in terms of how this conflict proceeds, as well as the wider Iran conflict. That being said, yes,

there has been sort of a quiet, relatively unnoticed restoration of flows through Hormuz and everything but Iranian oil. And that's probably happening for a variety of reasons.

There's shuttling.

There's maybe countries making deals with Iran. And then there's countries like Saudi Arabia using other pathways to get oil out. That being said, none of it should be taken for granted. Iran still has the ability to disrupt those flows as they've been demonstrating this week. And I think what we might be is in more of a temporary lull than a kind of permanent de-escalation of this conflict.

Jenny, that said, though, what about barrels of diesel? Because it seems like they're barely moving any barrels of diesel.

Yeah, and it's important to note, oil has been getting out. A lot of other things that normally get through the strait have not been. And probably that is in part because of where Iran is targeting. Part of it is probably also what the U.S. and other partners in the region are prioritizing for getting out. Shuttling is a pretty resource-intensive effort. It's very expensive to do. Insurance costs are very high. And so we aren't seeing a full restoration of everything that moves through the strait. I think we're seeing a prioritization of oil in particular.

Hey, one thing you, you know, you talk about ebbs and flows, whether it's the oil or the conflict itself. What about the ebbs and flows of kind of where the U.S. is on this? We did see the Trump administration, you know, Jenny, being reluctant to get more involved, specifically in Saudi Arabia's offensive against the Houthis and has tried to restrain Riyadh from sparking a major new conflict in Yemen.

And yet here we are. So I'm just curious, where is the U.S. on all of this right now?

I think Washington is focused on Hormuz. It's focused on Iran. It's focused on its blockade there. That is a major allocation of military assets already. To try and stretch them to cover the Yemeni front would be very difficult. That being said, Washington is still providing some support to Saudi Arabia, including intelligence and likely refueling assets. So it's not doing nothing, but it's certainly not what Riyadh would hope for. And I think Washington is concerned about having a two-front war, even though the fronts are quite close to each other, when it is already not accomplishing its goals in the Iran front.

Listen, since the markets are so, it seems, on a lot of days tied to the price of oil, global oil. We see oil goes up. We're worried about inflation. We're worried about more Fed rate increases. And you do often see the equity market goes down. 20 seconds. Should we, as investors, continue to think that we're going to continue to see gyrations in the global energy markets?

Just quickly.

Yes, I think we are likely to see further gyrations, if only because this conflict is far from over.

Stay with us.

More from Bloomberg Businessweek Daily coming up after this.

A measure of SpaceX's credit risk surging to a fresh high. The company's bonds tumbling in the secondary market. This is following reports, Lisa. that the company is in talks with banks and investors to raise $ 40 billion to buy chips from NVIDIA. Oh, let's do it. Let's bring in Ed Ludlow right now.

He's on Bloomberg Tech. It airs 11 a.m. Wall Street time right here on Bloomberg TV. Ed, thanks for joining us. So $ 40 billion, that's a hefty number, isn't it?

Yeah, it is.

I mean, $ 10 billion in the form of loans and $ 30 billion in the form of investment grade debt, which is important, right? Because SpaceX is newly public, but it already has that kind of IG status. And why is $ 40 billion a lot? Because the first thing that SpaceX did after its IPO was was it went out and did $ 25 billion of debt. And this $ 40 billion would be on top of that. And so I heard you guys mention credit default swaps.

Did I hear that right?

Oh, yes.

Now we're getting somewhere exciting.

But the reason you see the five-year CDS go up like that is that this debt pile is getting big and big and big for SpaceX. And so investors looking to protect themselves against the risk of default. But I don't really think the market sees much risk of default. They just want SpaceX to have access to capital to buy lots of chips.

Well, it needs it for XAI, right? I mean, where are we with XAI? Home to Grok and home to the AI coding company Cursor.

Yeah.

Tell us where we are and what we know about what XAI is up to.

Yeah.

So like these days, it's actually probably more commonly known by SpaceX AI.

They kind of I forgot about that.

Because, you know, there's never enough time in any program to sort of get to what on earth SpaceX actually is.

But they are a.

Frontier lab that builds AI models and has tools like Grok, the chatbot, and Grokbot, the AI agent. But their data center footprint is really interesting. Like, XAI is a very big buyer and deployer of NVIDIA GPUs. And in Tennessee, they have these series of data centers, Colossus 1, Colossus 2. And they have many hundreds of thousands of NVIDIA GPUs of different generations there. And Elon Musk has been talking about this on X, the social platform, which also happens to be a part of SpaceX AI.

It's very confusing. Very confusing.

They could be adding 220,000 more of those in the month of November as well. So they need the capital just to move quickly, and that's kind of the hallmark of this company. They move quickly in building the capacity and turning the servers on, some of which they rent to others, but some of which they use themselves.

So if you build it, they will come. Are folks coming? Is this one of the popular, you know, when it comes to, there are different LLMs out there. So I'm just curious about what we know in terms of success, user, like all of that.

Yeah, so in the reporting, SpaceX AI has been kind of aggressive in trying to get into the enterprise with the underlying model, the different generations of the Grok model. What's more interesting present day is like where they are getting traction or not with the chatbot Grok and then the like AI agent Grok bot.

It's mixed, honestly. There isn't a reliable benchmark on that.

There is anecdotal evidence that it's really popular, particularly the bot side of it. Now, Cursor, as a coding tool, was very popular, and SpaceX acquired that immediately post-IPO for $ 60 billion. how they are structuring, pricing, packaging all these different tools together. That's something that we're looking at quite closely at the moment. You still need GPUs to run it all either way.

You do.

Can you talk more about this whole thing of the trend of tech companies, you know, investing hundreds of billions of dollars to either build, lease data centers? I mean, when does it get to be too much? Like what warning signs would you look for in something like this?

Well, so in our reporting about this SpaceX $ 40 billion, again, $ 10 billion of loans from the banks, $ 30 billion from pools of capital put together by investors in the form of debt.

One of the names that came up was Apollo.

And Danny Berger on Open Interest talked to one of the corporate credit leads at Apollo who said, I'm not going to comment on SpaceX, but, and went on to explain, like, the world knows what the capital expenditure picture is, not just for this year and next year, but, like, beyond that. And markets generally are pretty accepting of that.

Forget whether it's.

Debt or equity, this year has been story continuously being big tech companies raising capital in different guises. Alphabet did a ginormous equity raise. SpaceX did the biggest IPO in history and then immediately tapped the bond markets. But genuinely speaking, like, the markets are sanguine about that because everyone is at this point where they'd say they're compute constrained. If they have more data centers up and running, they would be able to offer more services and charge more and have better token throughput. And so, like, The capital around the world, Wall Street and beyond, everyone's pretty calm about that. I'm not answering your question, but the warning sign was, are they tipping into negative free cash flow? That happened to

loads of companies, Amazon included. Everyone was calm about it.

It's fine.

Yeah, what a different environment, right, with these big mega-cap tech names. Hey, one of the things I want to ask you while we have you, and it is hard to keep up with all the AI headlines. I know this was a story you talked about yesterday. Moonshot AI, Bloomberg exclusively reported it out. Final round of private fundraising, valuation of about $ 50 billion. This is a big Chinese AI name. What do we need to understand about that, especially as we think about this battle between the U.S.

and China when it comes to AI?

Yeah, so Moonshot closed a private market round that valued it at 50 billion as a precursor to an IPO soon, early next year, right? The difference with Moonshot is that its models are open models. And what a lot of people track now is the proportion of tokens out there that are generated from open versus closed models. It used to be vast majority closed, small proportion, like let's say 80-20, closed to open. Now it's reversed, 80% or 80-20 open to close. And Moonshot accounts for K3, that model that surprised all of us with its capability earlier in the year.

But what links all these stories together? If you're an AI company somewhere in the world right now, it's very easy for you to raise money somehow.

Are you an AI company?

You need money?

Here's some money.

Is that how it works? No, it's remarkable. Talking with venture people, too, or other companies that are either trying to raise money and they're not in the AI space, it's harder for them because nobody's as interested.

Well, I think that. There is capital to deploy, right?

Again, this goes back to what we're talking about in debt markets with the absolute leading edge frontier labs. The capital is available and investors are willing. The difficulty, the thing that is muddying it now is how do you assign a fair value, right? And that brings us to the story of like, Anthropix IPO is imminent and OpenAI will happen next year. And Moonshot is somewhere in the middle.

You still have to look at the long-term economics of those businesses and say like, When and how much are you going to actually start making money, like a profit from this?

Especially when I feel like, I know we've talked about this before, but a lot of folks think that when it comes to models, companies will be agnostic. And so do you have to be the leading? Which is true too.

Right?

Yeah.

Okay, so who's going to lose and who has spent a lot of money that the stuff is going to pan out?

Yeah.

When I was at Lockheed Martin on Friday at the Skunk Works, Lockheed Martin is the biggest defense contractor in the world. They use 55 different large language models in their contained environment. And the way they describe it to me is like, it isn't clear which model is going to win in the end. Some have pros, some have cons.

So they use all of them for whichever is the best application. And the other way of looking at that is Lockheed Martin pays for access to 55 different models.

Right.

Now, for now they do.

Yeah.

Yeah, for now.

Ed, can you talk about the demand?

I know one of the stories you were talking about yesterday, too, was AMD's Lisa Su and how she was saying it's going to, you know, demand is going to continue for the next few years.

You've spoken with her before. What does that say? I mean, is demand still strong?

Yeah, I mean.

It takes a lot of time to build a data center. So when you talk about the chip makers, AMD included, they have a pretty good sense of where their sales will be going, not in the next 30 days.

But over a much longer time horizon.

And with AMD, one of the stories has been like, they could get more supply. In other words, if TSMC could allocate them more supply, then more of their gear would be out in the world. The same story is happening with NVIDIA, and you can go on to FAGO on the Bloomberg terminal, or you can listen to Jensen Huang in his official disclosures, and for quite a long period of time, they will be in this environment where the demand for their compute platforms exceeds their ability to supply them, that is typically in any industry or economic cycle a pretty good position to be in. And nothing's really changed on that front.

Hey, one of the things, I want to go back to, you said you were at Lockheed Martin. Lisa Su of AMD saying, urging the top AI companies to cooperate on ensuring that the technology causes no harm. The defense companies who are playing with so many models, We've talked so much about guardrails, and we had an off-camera interview with a guest recently who said, you know, we're kind of creating our own guardrails when we're using data to make sure.

Yeah.

So I'm assuming that a company like Lockheed Martin does the same thing.

Yeah, I mean, by the way, Lisa Su's view is all of the above. The labs should take responsibilities themselves. There should be a degree of regulation. There should be a degree of third-party evaluation and independent audit.

Everyone should work together. That's my point of view.

But Lockheed, again, the Skunk Works facility I was at in Palmdale on Friday, 97% of the work that goes on there is classified. They have 50% of their workloads on data centers they own and operate themselves. 50% is in the cloud, but they have very clear parameters, containerized environments that those models run in. So to answer your question in short, yes, like Lockheed has many controls in place and safeguards or guardrails to make sure that those models behave in the way that they are intended to.

So, Ed, where are we on this?

Since we're talking AI safety, where are we on this AI task force? I remember over the weekend, like it just came out and, you know, it kind of had a couple days to kind of sink in.

Where are we right now with it?

Well, we have some humans, you know, some names that are, I mean that seriously, but like, so Jay Clayton is leading this alongside Scott Kapoor, Emile Michael. I think what's being debated between Washington, D.C., politics, political circles, and industry is whether this is a story about the president putting a lot of expertise around this, or if this is people that are getting proximity and informing.

AI policy, right?

You know, there is concern, like, these people, are they, is that the right blend of expertise to do it from private industry? Should private industry have so much direct say in policy, et cetera? But generally, they're acting on it. When I was with Sam Altman last week, which I think we covered off on this show, right, at some point, Carol, we definitely talked about it in LA, I asked him, like, by the way, for what it's worth, what's the net result of all of this been? about the idea that the president believes you guys should keep going, you believe you should pace the frontier, like the framework the president announced last Tuesday, I think they're still working out what the blend of all of this looks like.

Stay with us.

More from Bloomberg Businessweek Daily coming up after this.

A lot going on, and one of the things that has increasingly been catching our attention takes us really to Russia and what is going on there. A story that did cross a little bit after lunchtime today, the World Health Organization pressing Russia for more information. on the death of a worker in a plague lab, saying that what's been disclosed so far makes it impossible to determine the risk that the case poses. So we are continuing to follow these cases of pneumonic plague in Siberia. And we did see Beijing making some comments. They, I think, are reaching out to Russia. The US is rushing out. So we want to get to our guest, someone who is very well known to the Bloomberg audience. We talked to

him many times throughout the COVID pandemic. I want to say delighted to have him back, but I wish we weren't talking about a plague. Joining us is Dr. Amesh Adalja. He's senior scholar and attending physician for infectious diseases, critical care and emergency medicine at John Hopkins Center for Health Security at the John Hopkins Bloomberg School of Public Health, which of course is supported by Michael R. Bloomberg, founder of Bloomberg LP and Bloomberg Philanthropies. He joins us from Baltimore.

Dr.

Adalja, I did not think we would be talking about something that feels potentially could be very serious so soon after COVID. What do we need to know here?

What we need to know is that Well, and there's two different questions. So what we do know is a 28-year-old who works at the Anti-Plague Institute in Siberia died of a pneumonia of unknown etiology, meaning unknown cause. They didn't know what caused her pneumonia. And she supposedly was vaccinated against everything she worked with, that.

This wasn't plague.

But what we saw was the Russian government basically acted very proactively, very aggressively, quarantining the hospital she was treated in, asthmat suits, monitoring 200 contacts. And that is incongruous with plague. People might think of plague and think that's the worst thing it could be. That's not the worst thing it could be. Plague in general is treatable. There are protocols when people are exposed. It doesn't usually merit the type of response we saw in Russia. So that raised a lot of alarm bells that this was something novel, something new that they were worried about, that they were trying to contain. that it might have been spreading from person to person. That's what we wanted to know about. Those

are the questions that are still outstanding. And that's what Russia needs to really be transparent about.

So, Dr.

Adalja, how concerned, I mean, should we be right now? I mean, with everything that you've been saying so far, what should we know? How concerned should we be?

I've been freaked out a little bit.

Yes, I know.

She's very freaked out. So should we be freaked out? Because I feel like COVID, we did the same thing.

Not a problem.

It's not going to impact us. And then there we were.

The thing that we're waiting to understand is why this woman died. The fact that 200 of her contacts have not gotten sick and, you know, you have to take... You can't really take the Russian... messaging at face value. They are notorious for disinformation and obfuscation and opacity. But what they're saying is that the contacts of that individual, none of them have tested positive for anything, that they're all healthy. So that tells you that it's not communicating.

The WHO has said that this is a lower risk situation. But until we know what killed this woman and whether it was something that could be transmitted and why the Russians took such aggressive action, I think we're going to have a hard time risk stratifying what this actually means. And there's been some reassuring data over the last 24 hours.

I think that the Russians have given drips and drabs of information that seem to be reassuring. So 24 hours have passed since the last time I was on television talking about this, and it seems to be a better situation now than yesterday, but we still are a long way from actually having full situational awareness of what happened in this town.

Now, we keep hearing pneumonic plague specifically. I mean, what makes this form of plague different? Why is that more concerning?

So I don't think that this was plague, but pneumonic plague is plague that affects your lungs. So when plague affects your lungs, it can be spread from person to person. You can cough and cause respiratory droplets to spread from one person to the next. That's a contagious form of plague between humans. So that's why the stakes rise when you're talking about pneumonic plague versus bubonic plague, which is plague confined to the lymph nodes in your neck, in your groin, in your axilla, in your armpits.

So very, very different. China reaching out, President Trump reaching out.

Normal?

Should we expect that? That's the right thing that we want to see in terms of helping Russia?

What we want is real transparency. coordination, cooperation, anytime there's an infectious disease emergency in any country, because that's how we are. That's how we're proactive. That's how we increase our resilience. That's how we have situational awareness. That's how we mobilize when we need to mobilize. And we take a temperature down when the temperature can come down. All of that is about, you know, that's the norm that we want. And that's why you'd see the WHO and world leaders reaching out. The United States actually issued a Demarche, which is a a special type of diplomatic communication saying, we need to know these answers now. So it has kind of escalated from just friendly helping out to we

want to know now what's going on so that we can prepare if we need to prepare.

Did that happen during COVID?

During COVID, there was definitely demands for China to have increased transparency. And the WHO basically, I would say, did not get full accounting from the Chinese government. We know that China delayed in telling people that COVID was efficiently transmitting between persons. They delayed how long they knew about something odd that was going on. So no, I think that we don't often get that. Even though it's the norm, it's the aspiration. many countries will often fall short because they don't want the stigma of having an infectious disease issue in their borders. And that's the perennial problem that we face. It keeps us from being fully resilient. We always are late to the game because countries want to

hide things. Saudi Arabia did the same thing with Middle East Respiratory Syndrome in 2012. This always happens over and over again.

So what's the next step?

What would have to happen for this to move from an isolated incident to, say, an international public health concern?

If we see that there are secondary cases of this unknown pneumonia occurring in the contacts of that woman who died, if we see that there's evidence of onward transmission, that changes the game. And it doesn't seem to be that at this point, but that's certainly what Russia was trying to rule out with what they were doing with monitoring those contacts of the patient who died.

That's what I would be looking for.

Are there any other cases?

This is the Bloomberg Businessweek Daily Podcast. Available on Apple, Spotify, and anywhere else you get your podcasts. Listen live weekday afternoons from 2 to 5 p.m. Eastern on Bloomberg.com, the iHeartRadio app, TuneIn, and the Bloomberg Business app. You can also watch us live every weekday on YouTube and always on the Bloomberg Terminal.

Hawkish Fed Minutes Drag Stocks From All-Time Highs

Kapitelklicken springt direkt hin

Shownotes

Transkriptvom Podcast bereitgestellt · Zeitmarken springen hin