Eye on the Market
Prospettive per il 2025
Gli alchimisti
Deregolamentazione, deportazioni, dazi, sgravi, tagli, cripto, petrolio, libertà medica e censura: cosa può andare storto? Argomenti: IA, nucleare, DOGE Chisciotte, trappole cinesi, Europa, cripto e Top 10 2025.
Leggere o ascoltare le prospettive per il 2025
Condividi
Guarda la sintesi di Mike
Descrizione audio
Good afternoon, everybody. This is Michael Cembalest with the 2025 Eye on the Market Outlook podcast, so Happy New Year, everybody. The Outlook this year is called “The Alchemists,” for reasons that I will explain in a minute.
First, just want to mention this is the 20th anniversary of the Eye on the Market. I launched it in January of 2005. It was meant for clients that consume information on either a BlackBerry or a Palm Pilot. It didn’t have any charts or graphs or tables or footnotes or anything. It was text only. A lot of things have changed since then. It’s been an amazing journey. I’ve enjoyed working on it. I’ve learned a lot. I’d like to thank Jamie and Mary for supporting me in this effort, particularly when I write about controversial topics. And this year is no exception.
So, on to the Outlook itself. I have a section on “The Alchemists,” which is about Trump’s impact on markets; we have a section on AI adoption trends and milestones, given the Mag 7’s support to the markets; a section on nuclear power, “Wake me when we get there”; a section on DOGE called “DOGE Quixote,” which refers to the billionaire’s quest to reduce government spending; an update on China and the liquidity trap; a section called “Doctor Seuss goes to Europe”; we revisit the crypto markets, given the rally since our Maltese Falcoin piece in February 2022; and discuss the sky-high returns that donors have gotten on their crypto political donations; and then a top 10 list for 2025, following on last year's top 10 list, which was the first time we did that.
So, I'd like to thank GPT again for some fantastic cover art. This is a depiction of Trumpism as alchemism. And what’s going on here is you’ve got a guy in red hat, and he’s mixing together eight different things: deregulation, deportations, tariffs, tax cuts, government cost cutting, oil and gas, crypto, the medical freedom movement and agency purges. And then the question is posed, what could possibly go wrong? I also think we try to address what could possibly go right.
There's a lot more detail that we go through in the actual Outlook in the executive summary. Let me just give you a brief description of it here. We start with this. What are the Alchemists trying to change? If you take them at face value, what they're trying to change is this chart. And this chart shows that for the better part of 80 years, consumer spending, GDP growth and industrial production went up together like three doves flying the same direction, or geese or—I’m not a bird person.
Then, as soon as China joins the World Trade Organization in the year 2000, we get what I call the “silence of the plants”—that’s a Clarice Starling reference—where industrial production flatlines while consumer spending and GDP keep growing. There are some other ominous inflection points negative for the United States that also took place around the same time that China joined the World Trade Organization.
Rising manufacturing job losses in the U.S., a follow labor share of gross profits, rising suicide rates in the U.S. and rising non-metro, meaning rural, poverty rates in the U.S. So that’s what the Alchemists claim they’re trying to change. What kind of tailwinds are they trying to create? Well, more inbound foreign direct investment into the United States, part of which is already happening; a slowdown in the breakneck pace of regulation, which was taking place under the Biden administration, and that we've showed charts on and discussed many times; freeing up bank capital by ending the Basel III discussions, which could presumably lower the cost of credit; increasing oil and gas production to bring down energy prices, although I think when you look at the electricity prices in many parts of the country, the renewable transition is irreversible and is driving those energy prices up; and then building on a rebounding CapEx and confidence outlook; and also a recovery that we’re seeing in venture capital markets, which we spent some time on in the executive summary as well.
So, those are the tailwinds that I think it’s fair to say the Alchemists are trying to create. Now, what headwinds are the Alchemists likely to face with their set of policies? Well, as everybody has discussed, possible inflationary consequences, not just from tariffs themselves, but from the retaliation that the United States would face, with whether it was a China tariff, a universal tariff, any kind of tariff—and, by the way, of all of the ten states that import the most amount of goods relative to their GDP, eight of them are red states. Tightening the labor supply through deportations, and things like that, have potential consequences for both wage inflation and Fed policy that I’m not sure that the administration would like.
Now, to be clear, and we talk about this in multiple places, the Biden administration oversaw the largest uncontrolled, unmanaged and unsupervised migration surge on record. And even the Brookings Institution at one point argued that the borders needed to be closed for some period, just to absorb the immigration that had taken place. But that said, tightening labor supply too much will have consequences for wage inflation and bad policy.
And then we already have Trump’s inheriting large budget deficits despite almost full employment. And then from a policy perspective, the Republicans have almost zero room for defections in the House because, they’ve only got a 219 to 215 lead. The gate seat will have a special election sometime in March. They’ll lose a couple of seats in Florida. Bottom line is they really have no room for defections in the House.
And then I won't go into too much detail, we have a page on this. Suffice to say, the medical freedom movement that RFK, Jr. supports on occasion is not a confidence builder in the concept of U.S. exceptionalism.
So, how do we see what the market verdict on this mix of headwinds and tailwinds is? My favorite barometer is a 10-year Treasury. And so far, the news is not great. If you look at the last seven tightening, I’m sorry, easing cycles from the firm, the United States going back to the 1980s. Typically, what happens is the Fed starts to ease and the 10-year Treasury either stays around flat or and or comes down.
This is the first time in 34, 35 years or so that the 10-year Treasury has gone up, and stayed there after the first Fed cut. So the early verdict from the bond markets on this mix of Alchemist policies is not great, and suggests that the productivity benefits from tax cuts and deregulation may not be enough to offset the inflationary consequences of, tariffs and deportations and some of the other things that they’re planning on doing.
So ultimately, again, the 10-year Treasury is going to be the best barometer, in my opinion, of this mix of policies. And the Fed fund futures markets are sending a similar signal. A year ago, the markets were pricing in a funds rate that would fall below 4%. That's not the case anymore. The latest Fed fund futures markets are projecting the Fed won’t be able to ease more than down to 4% or so. And I’d be surprised if they even were able to get there sometime in 2025.
Equity investors are more optimistic. The Conference Board’s survey of households expecting higher stock prices in 12 months is at an all-time high, even higher than it was in in the mid to late 1990s. And there are some technicals that support that. Equity supply is very tight. In other words, this looks at secondary and primary issuance net of buybacks and net equity supply, particularly in the U.S., is actually declining.
So there’s a scarcity value to public equities, if you can believe that. So from a technical perspective, that’s supported. But U.S. equity markets are pretty expansive. We all know this, and at least on a valuation basis. Now, markets can still go up when valuations are high. But we’re in the mid to upper 90s in terms of percentiles for many different ways that you could look at the valuation of U.S. stocks. Median P/Es, forward P/Es, trailing P/Es, the enterprise value to cash flow, price to book, price to sales, you name it. Most of the valuation indicators are on the high side. And we’ve also had two 20%-plus years in a row. That’s only happened 10 times since 1871 and only during the 1990s bull market and the Roaring Twenties, around 100 years ago, did the good times keep going.
So my takeaway from this, and you can read this, in the executive summary of the Outlook, which goes obviously to a lot more interesting detail, is, I think that there’s a steep curve here for, for the Alchemists in terms of trying this policy mix of deregulation, tariffs and cost cuts and tax cuts and crypto and things like that. They’ve got a very limited margin for error given how much of an equity war we’ve already had, and how high valuations are.
So I expect U.S. equity markets to end the year 2024 are when they began. But I expect the Alchemists to break something. And I don't know what. I don’t think they do either, but they’ve got their sights set on everything from the FBI to Medicare to border policy to the Defense Department to weapons procurement.
I think they’re going to break something. And so I expect a correction of some kind, of 10–15%, from whatever value it’s at, sometime in 2025. That’s not that infrequent a thing, it happens in around one-third of all years, but I am expecting a correction to take place. And then for equities to end up modestly higher by a few percent at the end of next year compared to where we are right now.
Just to be clear, the deregulatory efforts that the incoming administration is talking about, they do have a long history of increasing business, spending, capital spending, business confidence. I started writing about deregulation over 20 years ago. And here’s a chart from the archives of the Eye on the Market that looked at the Clinton deregulation of telecommunications and electricity markets, and the resulting increase in nonresidential business investment, which is, essentially, you know, a capital stock increase.
Peter Orszag at Brookings has written about the benefits of deregulation, on multiple occasions. You know, he was the, I think, either the CBO or the OMB for a while. And so I’m a believer in the benefits of deregulation. And then after the election, there was a surge in the number of small companies responding positively by saying that they’re now planning on some additional capital spending over the next few months, compared to what they were before the election.
But, you know, tariffs and shrinking the labor supply and some of the other issues that the Alchemists are planning, I think will offset some of this. So it’s a really interesting policy mix. It’s certainly not anything that anyone has tried in 70 or 80 years. And again, you know, the executive summary of the Outlook gets into the details of this policy mix and what its implications for markets might be.
We’ve also got a long section on AI and adoption and milestones and things. I’m just going to share a few comments here. The U.S. exceptionalism that you read about in the equity markets, you can back that up with numbers. And I thought this was an amazing chart. It comes from Mike Gold and his team at Empirical Research, and it shows the free cash flow margins by decade for the 10 largest stocks.
So you might think the 10 largest stocks always make a lot of money. That’s true. But the 10 largest stocks make a lot more money now than they did in the past. This goes back all the way to 1950, and shows the free cash flow margins for the 10 largest stocks in each decade. And this crop of U.S. exceptional large companies, with free cash flow margins above 25%, are setting all-time highs here.
Now, the rest of the S&P has been treading water. And when you decompose earnings per share growth of the Mag 7 versus the S&P 500 ex Mag 7, you get kind of a mixed picture here. The S&P 500 ex Mag 7 had an earnings boost that took place in 2022, but it was mostly just a reversal of the declines that took place during the pandemic.
It’s only the Mag 7 that has been consistently generating since 2017 some pretty impressive earnings growth. So this is a very bifurcated market and, unfortunately, the implication is that very diversified investors that have been investing in, let’s say, an equal weighted S&P, have struggled with that. And it’s really important to understand this AI boom because we’re now looking at something which is only happened I think twice in the last 50 years or so, which is when one specific company, or small group of companies, capture a 15% or so share of all the capital spending that’s taking place in the economy.
I mean, that’s kind of an amazing thing to think about—one company or a gaggle of companies capturing 50% of all capital spending. In 2026, that's what’s projected to happen with Nvidia’s data center revenues that are supposed to roughly be equal to 15% of all capital spending in the economy. The only thing we can compare that to is IBM’s mainframe peak in the late 60s. And then obviously the Dot Com bubble in 2000.
And so we spend a lot of time looking at this particular issue of the hyperscaler revenue gap. Now, what does that mean? Well, I don't have a lot of concerns about Nvidia. Their gross margins and everything looks fantastic and their order book looks amazing. The CEO described the order book for their new Blackwell GPUs as “insane.” The issue is, who's making all that money? Well, it's the hyperscalers: Google, Microsoft, meta, Apple, Oracle, Tesla, etc. And they eventually have to earn a return on all that GPU expenditure, capital spending. And if you factor in their GPU purchases, the implied overall data center spending that goes with that and the gross margins that they typically would need to earn, these companies are missing about $400 billion in annual revenue.
So the markets have been very patient so far because we're in the very early stages of this AI adoption boom. But at some point, I think the next 18 months, markets are going to start asking some very tough questions if these hyperscalers don't start earning some of this $400 billion to sustain their gross margins, all this capital spending. And our growth managers are watching this closely.
Hyperscale capital spending in R&D as a share of revenues has been creeping up for the Big 4: Meta, Microsoft, Alphabet and Amazon. Starting to get into pretty, pretty eye popping territory where our capital spending on R&D approaching 35% on revenues. And as we all know, these models just keep getting more expensive and energy intensive to train, and almost any chart you look at on this has to be plotted in log scale because of the amounts of time, effort and money that have been poured into training these models.
Now, the good news is the surveys of companies that are ultimately going to pay the hyperscalers for all of these AI programs, we're starting to see the first crop of surveys of how those gen AI applications are working out. And, as you can see in this chart here, and we go into more detail in the piece, 90% either exceeded or met expectations. So that's a positive sign. Some of the other surveys we’ll walk through in the piece look at timelines of AI applications. But I think it’s a pretty notable thing that the early crop across multiple sectors of gen AI projects have exceeded or met expectations of people investing, and that’s a good sign.
That doesn’t tell you how much money they’re ultimately going to be spending on it. But it's a good sign. But I just want to remind everybody of this time capsule that I put together recently from our archives. If you go back to the year 2000 and you'll look at Corning’s order book, they built a lot of fiber. Now, ultimately, all that fiber was used, right? None of it was wasted. Ultimately, all that fiber got used. But because of the overbuilding, it took Corning around 20 years to get back to the nominal revenues and their optical communications business that they had in the year 2000. And similarly, if you look at a chart on undersea fiber optic cables deployed per year, it took until 2020 to exceed the 20 the year 2000 level after collapse.
The reason I'm mentioning this is when you have an overbuilding episode, even of all that capital stock ultimately ends up getting used, sometimes it can take a long time to absorb it, and then, obviously, there were a lot of implications for the companies that spent that money along the way.
Now, one of the issues that I’m fascinated by are all the green power commitments from a lot of these data centers and hyperscalers, in terms of the energy intensity. Now, you know, the sometimes people make too much out of this power demand coming from data centers over the next decade. I expect much bigger power demand numbers from electrification of vehicles and electrification of home heating through heat pumps, which will also be adopted by industry and commercial applications and office buildings. And there’s also a lot of different forecasts for data center demand.
But one of the most incredible—meaning, not believable—things about this, is that nuclear power is going to play a role. I just don’t see it. There’s been a boom recently in the performance of nuclear exposed and nuclear pure-play shares. I’m going to take the other side of this one.
Just as a bit of background, OECD nuclear power. You know, the OECD used to dominate nuclear power during the 50, 60s, 70s and then started sliding in the 80s. And by the time you get to the late 1990s, and from then until now, nuclear development is essentially something that belongs to and is done in developing economies and not developed ones. And you can see why if you look at the cost gap here, I’m looking at the capital cost of four plants that were recently completed in the West, because that’s all there have been over the last 20 years, Flamanville in France, one in Finland, the Georgia Vogtle plant and then Hinkley Point in the UK, all cost many multiples of the plants that are being built in Russia, China and Pakistan.
Now, there's lots of reasons for that. I’ve read whole studies on that. I personally don’t believe that we can start cranking these things out in the same way that you build solar panels or lithium ion batteries to bring their cost down. So I think people are way too excited about the possibilities of nuclear power having some kind of renaissance that contributes meaningfully to data center demand.
In the Outlook we get into the weeds on this. Vogtle 3, by the way, in Georgia, is already having operational problems, it just came online. The U.S. lacks a lot of the uranium and it and it has import reliance on certain kinds of uranium, which is creating problems for Terrapower. There are very limited opportunities for plant restarts. And way too much was made of the Palisades reopening and the announcement and as well, the one that’s taking place with Three Mile Island. You can only really restart plants that just recently shut down, meaning within the last couple of years, because the decommissioning process hasn't started yet. If you start the decommissioning process, it’s way too expensive to then reverse that and get that clamp back on line. So at maximum, I can think of maybe two or three other shuttered plants that would be eligible for restart, the next gen designs.
Everybody likes to talk about sodium cooling, gas cooling, molten salt. According to the DOE’s nuclear people, they’re at least a decade away from commercialization in the U.S. And then, small modular reactors. Very impressive pipelines. 6.5 GW. 200 billion in orders. They’re still lottery tickets. And the power commitments are not take or pay contracts. They’re all contingent on final cost, as was the case with the new scale plants that were abandoned in Utah last year. So let’s, you know, take some reality pills on this nuclear renaissance. Take a look at this section. It gets into the details.
We also have a section on DOGE Quixote. I am envisioning Vivek Ramaswamy and Elon Musk as Don Quixote and Sancho Panza from Don Quixote. The thing to remember is that the one of the underlying themes of the model is the pragmatism and humility are required for them to complete their quest. And I would question whether Vivek or Elon qualify for each of those, or either of them. Vaya con Dios, Doge Quixote. I think they’ll obviously get some things done, but they’re going to run into a number of different buzzsaws. And one of my top 10 predictions is they’re not going to make it anywhere near some of the targets that they’ve mentioned.
There are savings to be had from reversing some of Biden’s recent executive actions, but those wouldn't count towards DOGE targets on resisting current spending. That would just be reducing potential future spending, that hasn’t really even been penciled in yet. Non-defense discretionary spending has already been cut. Budget Control Act of 2011. It’s at its lowest level in decades. So therefore, they’re going to have to focus on entitlements. And the DOGE billionaires are now finding out that that’s a political third rail. Good luck with that.
The SEC, the Department of Labor, the Department of Education and the EPA combined represent less than 2% of all federal workers. So fine, start firing them. But that’s not going to get you very much. The vast majority of federal workers are, you know, the Postal Service and the Department of Defense.
There will be efforts underway to expand the CMS, that is, maybe expand, weight loss coverage for weight loss drugs. And those are going to offset some of the DOGE savings, which takes place elsewhere. Trump has talked about impounding or redirecting congressionally approved and congressionally appropriated spending. That’s going to face judicial challenges. There was an Impoundment Control Act passed in 1974, in response to something Nixon did. A president can’t, according to the Supreme Court and according to this legislation, can’t just say, well, I’m going to take the money that was appropriated for the Department of Education and spend it on something else, which is what he’s announced that he wants to do.
Most defense spending is people and operations on the order of 70 to 80% rather than weapons procurements. And then, with respect to weapons procurements, good luck. You know, the number of federal contractors has fallen from 51 in the early 90s to five. So it’s going to be very difficult to cut those procurement costs with that much concentration in the defense contract. And then, we also get into some of the things on improper and fraudulent payments that the government makes, sometimes accidentally, you can try to reduce them, but it’s going to cost a lot of money in terms of investing in systems. And people will do that. So, we get into the details there. It’s interesting if you want to understand it and good luck to all the billionaires that they've recruited to the team, along with Vivek and Elon, to try and get this done.
We have a section on China that looks at both the liquidity trap and the Thucydides trap, which is something we wrote about a couple of months ago. The bottom line is that China is experiencing some kind of liquid illiquidity trap that shares some kind of characteristics with Japan in the early 1990s. As you can see here, a Bloomberg maintains an economic activity monitor—exports, coal and oil consumption, steel production, car sales, real estate investment, consumption of medicine—the kind of things that are harder for the Chinese government to paper over and obscure.
And, you know, the momentum on this thing is fairly negative, which is why China announced that stimulus package. Now, the stimulus package, that was substantial, impressive—I believe there’s more fiscal stimulus coming. Bridgewater did an interesting analysis a couple months ago that showed that the total impulse from monetary and fiscal stimulus would offset roughly half of the economic drag that’s taking place in China right now. That’s a good effort. And the domestic A-share market kind of soared after the announcement and has retained its gains, whereas the MSCI China index, which is a broader index of Chinese exposed stocks, has since given up about half of the gains, now the stimulus was announced. But I’m not going to go through the details here. You can read the piece. I just think the days of 14 to 18 P/Es on MSCI China are a thing of the past and for geopolitical reasons, they’re going to be capped at about 12x, which doesn’t leave a lot of room for a rally from here.
So you could take a look at the piece. We go through the details of some of the tariff and espionage issues at the core of some of the conflict issues between the U.S. and China. I’ll just say that the increasing intensity of Chinese espionage spreading across almost every industrial sector in the U.S. and other Western countries, it makes the idea of a rapprochement that I’ve read about highly unlikely, and the same goes for some of the naval blockade exercises that China has been conducting, with regards to Taiwan. So you can read more about that.
We have a section on Europe called “Doctor Seuss goes to Europe.” It’s just two pages with up with some Doctor Seuss–style language and some charts. I just want to show you one of the charts here. We have charts and comparisons of labor productivity and energy prices and deindustrialization in Europe, and problems with monetary policy being applied to countries that are two different compared to the U.S. low labor mobility, very poor ROA and ROE numbers in Europe other than in health care, and then very negative earnings revisions.
The chart I really wanted to show you was this new one, which is the creation of new public companies in the 21st century comparing the U.S. to Europe. This kind of tells you all you need to know about the outperformance of the U.S. versus Europe over the last 20 years. It's kind of an amazing chart to see. I can’t like, describe it verbally. You just kind of have to look at it.
The crypto presidency. I wrote a piece called The Maltese Falcoin a couple of years ago. It was February 2022, and I was skeptical of Bitcoin as a store value and then even more skeptical on the transactions-focused coins and the DeFi coins. and if you’re not familiar with the jargon, you can look at the Eye on the Market on this section and it gets into the details here. But, I said I would take another look at Bitcoin if it fell. And it fell, it fell hard and it fell 65% after I wrote the piece. And my mistake for not looking at it again. It has since rallied pretty sharply. And a lot of the surveys of institutional investors are fairly positive on the prospects for increased digital ownership as it relates to the transactions-focused coins and the DeFi coins. I don't believe it. I was right that about those coins for three years. Nothing much has happened in terms of actual utilization of either of those kinds of crypto coins.
There’s been this spectacular rally and in a handful of that ripple in particular, but it looks highly speculative to me. And then, specifically, I’ve got this chart here looking at the volumes of utilization of ripple like coin, Monero and Stellar. None of them have changed since two years ago. Right. And so if the value proposition of these coins is that people are going to use them more, you expect to see that happening, and you don't. So that that means that this is a very tactical and speculative rally. Same thing for Ethereum. The number of daily verified contracts in Ethereum and being executed on a blockchain are now unchanged versus what they were in early 2022. So to me, what’s really happening is that people are earning a sky-high return on their political donations.
Crypto PACs received almost $250 million in donations, around half of all corporate donations made during the election. And Coinbase gave $75 million to a super PAC called Fair Shake. Their candidates won almost every race they got involved in. Look, they're playing the game and participating in the system in ways that they’re allowed to do that, I’m not criticizing it.
I'm just saying that this, to me, is a better explanation for the pop in crypto than the presumed proposition that their acceptance and utilization has been changing. All that’s really been changing is the politics. And, you know, the GOP majority of the year of the House is pledging a pro-crypto bill in the first hundred days that Trump has announced a crypto council and a crypto czar. The new SEC head will probably roll back some of the crypto accounting disclosure rules. So this is a shift in sentiment, rather than a shift in utilization. And you could take a closer look.
And then, just to close out here and thank you all for listening. I started a top 10 list last year in honor of Byron Wiem, who passed away. He did this every year for forever. I made a bunch of predictions last year. Six of them were right. Three of them were kind of, you know, either too soon to tell, hard to say. And then one was wrong. I’ve made a new list for this year. I will tick through them very quickly and sign off.
At least three confirmed Trump cabinet members will be gone by the end of the year. Renewables will underperform traditional energy again in 2025, like they’ve done in the last three years. MSCI Japan will outperform China despite China’s higher P/E multiple. The discontinuation rates, which are approaching 50% of weight loss drugs, are going to hurt some of the GLP stocks, specifically Novo Nordisk, I think will end the year below consensus. Our price targets, despite the order book referenced earlier, I don’t think any small modular reactors in the United States will even begin construction in 2025. I think DOGE cost cuts are going to be no more than $150 billion a year, excluding reversal of executive actions compared to targets of $2 trillion. Hedge funds I expect to continue to outperform the stock bond benchmarks, and we explain what we mean by that. But you know, they’ve been doing better than expected. Housing inflation, I think because of completions, both single family and multifamily will start to finally bring housing inflation down below 4%, the way that the Fed computes it. Migration will continue. Over the last few years, 70% of all movers in the country in the U.S. are moving from either blue to red or purple states. I think that’s going to continue. And then lastly, I think there’s a chance that the Trump administration launches a targeted military action against drug cartels in Mexico. So that's my top ten list, thank you for listening.
Take a look at the Outlook. There’s a lot in there that we can't get into in a podcast like this. And it’s been a privilege. Thank you very much. Bye.
(DESCRIPTION)
Logo: J.P. Morgan, Title: Eye on the Market
An image at left shows a scientist in a lab hovering over colorful vials. He wears a red cap with two letters visible: G, A. A plume of bright magenta smoke fills the air. Title: Outlook 2025, The Alchemists.
(SPEECH)
[WHOOSHING]
(DESCRIPTION)
The speaker, Michael Cembalest, has shortly cropped brown hair and wears glasses. He appears in an office space with a window behind him.
(SPEECH)
Good afternoon, everybody. This is Michael Cembalest with the 2025 Eye on the Market Outlook podcast. A Happy New year, everybody. The outlook this year is called The Alchemists for reasons that I will explain in a minute.
(DESCRIPTION)
Images of a Blackberry and Palm Pilot appear at left. Text: Eye on the Market launched January 2005
(SPEECH)
First, I just want to mention this is the 20th anniversary of the Eye on the Market. I launched it in January of 2005. It was meant for clients that consume information on either a BlackBerry or a Palm Pilot. It didn't have any charts, or graphs, or tables, or footnotes, or anything. It was text only. A lot of things have changed since then. It's been an amazing journey. I've enjoyed working on it. I've learned a lot. I'd like to thank Jamie and Mary for supporting me in this effort, particularly when I write about controversial topics. And this year is no exception.
So on to the outlook itself.
(DESCRIPTION)
Text: 2025 Eye on the Market Outlook. The Alchemists deregulation, deportations, tariffs, tax cuts, cost cutting, oil & gas, crypto, medical freedom, and Agency purges. What could possibly go wrong? The Golden Goose: AI adoption, trends and milestones. What nuclear renaissance? Wake me when we get there. DOGE Quixote, the billionaires quest to reduce US government spending. China: the liquidity trap and the Thucydides trap. Dr. Seuss Goes to Europe. The crypto Presidency: on the crypto rally and the sky high return on political donations. Top Ten List for 2025 and a scorecard on 2024's list.
(SPEECH)
I have a section on The Alchemists, which is about Trump's impact on markets. We have a section on AI adoption, trends and milestones given the Mag 7's importance to the markets; a section on nuclear power-- wake me when we get there; a section on DOGE called DOGE Quixote, which refers to the billionaire's quest to reduce government spending; an update on China and the liquidity trap; a section called Dr. Seuss Goes to Europe. We revisit the crypto markets given the rally since our Maltese Falcon piece in February of 2022 and discuss the sky high returns that donors have gotten on their crypto political donations and then a top 10 list for 2025 following on last year's top 10 list, which was the first time we did that.
Anyway, so I'd like to thank GPT again for some fantastic cover art. This is a depiction of Trumpism as alchemism. And what's going on here is you've got a guy in a red hat. And he's mixing together eight different things-- deregulation, deportations, tariffs, tax cuts, government cost cutting, oil and gas, crypto, the medical freedom movement, and agency purges. And then the question is posed, what could possibly go wrong? I also think we try to address what could possibly go right.
There's a lot more detail that we go through in the actual outlook in the executive summary. Let me just give you a brief description of it here.
(DESCRIPTION)
Title: What are the Alchemists trying to change? A line graph appears that shows US Consumer Spending ad US Industrial Production from 1930 to 2020. They stay together until 2000, when spending exceeds production by 60 points. Text: Other US data with sharp inflection points around the year 2000: Rising manufacturing job losses, Falling labor share of gross profits, Rising suicide rates, Rising non-metro poverty rates. Source: B E A Federal Reserve, J P M A M 2024.
(SPEECH)
We start with this. What are the alchemists trying to change if you take them at face value? What they're trying to change is this chart. And this chart shows that for the better part of 80 years, consumer spending, GDP growth, and industrial production went up together like three doves flying in the same direction or geese. I'm not a bird person.
Then as soon as China joins the World Trade Organization in the year 2000, we get what's called-- what I call the silence of the plants-- that's a Clarice Starling reference-- where industrial production flatlines while consumer spending and GDP keep growing. There are some other ominous inflection points, negative for the United States, that also took place around the same time that China joined the World Trade Organization-- rising manufacturing job losses in the US, a fall of labor share of gross profits, rising suicide rates in the US, and rising non-metro, meaning rural, poverty rates in the US. So that's what the alchemists claim they're trying to change.
What kind of tailwinds are they trying to create? Well, more inbound foreign direct investment into the United States, part of which is already happening; a slowdown in the breakneck pace of regulation, which was taking place under the Biden administration that we've showed charts on and discussed many times; freeing up bank capital by ending the Basel III discussions, which could presumably lower the cost of credit; increasing oil and gas production to bring down energy prices, although I think when you look at the electricity prices in many parts of the country, the renewable transition is irreversible and is driving those energy prices up; and then building on a rebounding CapEx and confidence outlook; and also recovery that we're seeing in venture capital markets, which we spent some time on in the executive summary as well. So those are the tailwinds that I think it's fair to say the alchemists are trying to create.
Now what headwinds are the alchemists likely to face with their set of policies? Well, as everybody has discussed, possible inflationary consequences, not just from tariffs themselves, but from the retaliation that the United States would face with whether it was a China tariff, a universal tariff, any kind of tariff. And by the way, of the 10 states that import the most amount of goods relative to their GDP, eight of them are red states.
Tightening the labor supply through deportations and things like that have potential consequences for both wage inflation and Fed policy that I'm not sure that the administration would like. Now, to be clear-- and we talk about this in multiple places-- the Biden administration oversaw the largest uncontrolled, unmanaged, and unsupervised migration surge on record.
And even the Brookings Institution at one point argued that the borders needed to be closed for some period just to absorb the immigration that had taken place. But that said, tightening labor supply too much will have consequences for wage inflation and bad policy. And then we already have Trump's inheriting large budget deficits despite almost full employment.
And then from a policy perspective, the Republicans have almost zero room for defections in the House because they've only got a 219 to 215 lead. The Gates seat will have a special election sometime in March. They'll lose a couple of seats in Florida. Bottom line is they really have no room for defections in the House.
And then I won't go into too much detail. We have a page on this. Suffice to say, the Medical Freedom Movement that RFK, Jr., supports on occasion is not a confidence builder in the concept of US exceptionalism.
So how do we see what the market verdict on this mix of headwinds and tailwinds is? My favorite barometer is the 10-year Treasury. And so far, the news is not great.
(DESCRIPTION)
A line graph appears. Title: 10 year Treasury yield change after the first Fed cut. The x axis shows measurements from 0 to 75 in increments of 5. Text: Business days since first Fed rate cut. The y axis shows negative 0.8% to positive 1.0% in increments of 0.2%. Colored lines represent September 2024, the highest, then July 1995, June 1989, September 2007, September 1998, September 2001, July 2019. Source: .Bloomberg, J P M A M, December 26, 2024
(SPEECH)
If you look at the last seven tightening-- I'm sorry-- easing cycles from the Fed in the United States going back to the 1980s, typically, what happens is the Fed starts to ease and the 10-year Treasury either stays around flat or comes down.
This is the first time in 35 years or so that the 10-year Treasury has gone up and stayed there after the first Fed cut. So
(DESCRIPTION)
A line graph appears. Title: Same from Fed fund future markets. Text: Fed funds target rate: current vs prior market expectations. The x-axis shows January 2024 to December 2025. The y axis shows 3.5% to 5.5% in increments of 0.25%. A line appears labeled December 2023 that slopes down from 5.5% to 3.5% at February 2025. And a line labeled December 26, 2024 slopes down from 4.25% to 4.0% through February 2025. Source: .Bloomberg, J P M A M, December 26, 2024
(SPEECH)
the early verdict from the bond markets on this mix of alchemist policies is not great and suggests that the productivity benefits from tax cuts and deregulation may not be enough to offset the inflationary consequences of tariffs, and deportations, and some of the other things that they're planning on doing.
So ultimately, again, the 10-year Treasury is going to be the best barometer, in my opinion, of this mix of policies. And the Fed fund futures markets are sending a similar signal. A year ago, the markets were pricing in a funds rate that would fall below 4%. That's not the case anymore. The latest Fed fund futures markets are projecting the Fed won't be able to ease more than down to 4% or so. And I'd be surprised if they even were able to get there sometime in 2025.
Equity investors are more optimistic.
(DESCRIPTION)
A line graph appears. Text: Share of US households expecting higher stock prices in 12 months, Percent of respondents. The y axis measures 15% to 60% in increments of 5% and the x axis shows 1987 to 2022. The line is up and down with an increase in 2022. Source: Conference Board Bloomberg, J P M A M, November 30, 2024
(SPEECH)
The Conference Board survey of households expecting higher stock prices in 12 months is at an all-time high, even higher than it was in the mid to late 1990s. And there are some technicals that support that. Equity supply is very tight. In other words, this looks at secondary and primary issuance net of buybacks.
(DESCRIPTION)
A bar graph shows Net Equity Supply Globally from 1999 to 2023. In US billions per year, the values are negative after 2021.
(SPEECH)
And net equity supply, particularly in the US, is actually declining. So there's a scarcity value to public equities if you can believe that. So from a technical perspective, that's supported.
(DESCRIPTION)
A bar graph shows, text: Current market valuation percentile vs history: 100% = most expensive, 0% = least expensive. The x axis is divided into three segments: US equities, non-US equities, fixed income, showing market variable at year of inception. The y axis shows 0 to 100% in increment of 10%.
(SPEECH)
But US equity markets are pretty expensive. We all know this. And at least on a valuation basis now, markets can still go up when valuations are high. But we're in the mid to upper 90s in terms of percentiles for many different ways that you could look at the valuation of US stocks. Median PEs, forward PEs, trailing PEs, the enterprise value to cash flow, price to book, price to sales, you name it-- most of the valuation indicators are on the high side. And we've also had two 20%-plus years in a row. That's only happened 10 times since 1871. And only during the 1990s bull market and the roaring '20s around 100 years ago did the good times keep going.
(DESCRIPTION)
A table shows S&P 500: Two consecutive years of 20% plus total returns and what followed.
(SPEECH)
So my takeaway from this-- and you can read this in the executive summary of The Outlook, which goes obviously into a lot more interesting detail, is I think that there's a steep curve here for the Alchemists in terms of trying this policy mix of deregulation, and tariffs, and cost cuts, and tax cuts, and crypto, and things like that. They've got a very limited margin for error, given how much of an equity one we've already had and how high valuations are.
So I expect US equity markets to end the year in 2025 higher than when they began. But I expect the Alchemists to break something. And I don't know what. And I don't think they do either. But they've got their sights set on everything from the FBI, to Medicare, to border policy, to the Defense Department, to weapons procurement. I think they're going to break something. And so I expect a correction of some kind of 10% to 15% from whatever value it's at some time in 2025. That's not that infrequent a thing. It happens in around 1/3 of all years, but I am expecting a correction to take place and then for equities to end up modestly higher by a few percent at the end of next year compared to where we are right now.
(DESCRIPTION)
A line graph appears. Text: Deregulation and business capital spending. Product market regulation index: 0 = least regulated, 6 = most regulated. The x axis shows 1987 to 2001 in 2-year increments. The left side of the y axis shows 0 to 6 in increments of 1 and the right side shows 8.5 to 12.0 in increments of 0.5. Lines show Telecommunications Act of 1986 and Electricity, both sloping downward. Source: O E C D, E T C R Database, B E A
(SPEECH)
Just to be clear, the deregulatory efforts that the incoming administration is talking about, they do have a long history of increasing business spending, capital spending, business confidence. I started writing about deregulation over 20 years ago. And here's a chart from the archives of the Eye on the Market that looked at the Clinton deregulation of telecommunications and electricity markets and the resulting increase in nonresidential business investment, which is, essentially, a capital stock increase.
Peter Orszag at Brookings has written about the benefits of deregulation on multiple occasions at Brookings. And he was, I think, either the CBO or the OMB for a while. And so I'm a believer in the benefits of deregulation. And then after the election, there was a surge in the number of small companies responding positively by saying that they're now planning on some additional capital spending over the next few months compared to what they were before the election.
(DESCRIPTION)
Line graph, text: Small business capex plans, % of N F I B Survey respondents planning capex in next 3 to 6 months, seasonally adjusted. The x axis shows years 2021 to 2024 with the y axis showing percentages of 18% to 32% in increments of 2%. Source: Bloomberg J P M A M, November 30, 2024. The line dips up and down with a sharp rise from 19 to 28% at the end.
(SPEECH)
But tariffs, and shrinking the labor supply, and some of the other issues that the alchemists are planning, I think, will offset some of this. So it's a really interesting policy mix. It's certainly not anything that anyone has tried in 70 or 80 years. And again, the executive summary of The Outlook gets into the details of this policy mix and what its implications for markets might be.
(DESCRIPTION)
Bar graph, text: Free cash flow margins by decade for the ten largest stocks, 1952-2024. Percent. The x axis shows decades 1950s to 2020s and the y axis shows negative 5% to positive 30% in increments of 5%. The increase is steady each decade. Source: Empirical Research, September 2024
(SPEECH)
We've also got a long section on AI, and adoption, and milestones, and things. I'm just going to share a few comments here. The US exceptionalism that you read about in the equity markets, you can back that up with numbers.
And I thought this was an amazing chart. It comes from Mike Goldstein and his team at Empirical Research. And it shows the free cash flow margins by decade for the 10 largest stocks. So you might think the 10 largest stocks always make a lot of money. But that's true. But the 10 largest stocks make a lot more money now than they did in the past. This goes back all the way to 1950 and shows the free cash flow margins for the 10 largest stocks in each decade. And this crop of US exceptional large companies with free cash flow margins above 25% are setting all-time highs here.
(DESCRIPTION)
Line graph, text: Earnings per share growth, rolling 2-year growth. Lines represent Mag 7 and S&P 500 ex Mag 7. The x axis shows 2017 to 2024. The y axis shows negative 50% to positive 250% in increments of 50%. Mag 7 peaks in 2022 at 250% above ex Mag 7 at 50%. Source: Bloomberg J P M A M, September 30, 2024
(SPEECH)
Now the rest of the S&P has been treading water. And when you decompose earnings per share growth of the Mag 7 versus the S&P 500 ex Mag 7, you get kind of a mixed picture here. The S&P 500 ex Mag 7 had an earnings boost that took place in 2022, but it was mostly just a reversal of the declines that took place during the pandemic.
Only the Mag 7 has been consistently generating since 2017 some pretty impressive earnings growth. So this is a very bifurcated market. And unfortunately, the implication is that very diversified investors that have been investing in, let's say, an equal weighted S&P have struggled with that.
(DESCRIPTION)
Bar graph, text: AI capital spending in context, Share of market wide capital spending
(SPEECH)
And it's really important to understand this AI boom, because we're now looking at something which has only happened, I think, twice in the last 50 years or so, which is when one specific company or a small group of companies capture a 15% or so share of all the capital spending that's taking place in the economy. I mean, that's kind of an amazing thing to think about, one company or a gaggle of companies capturing 15% of all capital spending.
And in 2026, that's what's projected to happen with NVIDIA's data center revenues are supposed to roughly be equal to 50% of all capital spending in the economy. The only thing we can compare that to is IBM's mainframe peak in the late 60s and then, obviously, the dotcom bubble in 2000.
(DESCRIPTION)
Bar graph, text: A $400b gap: AI infrastructure spending vs revenues. The y axis shows US dollars in billions from 0 to 500 in increments of 100. Measurements on the x axis are: NVIDIA year 25 revenue, Implied Data Center Spending, Required AI Revenue, Projected AI Reveue
(SPEECH)
And so we spend a lot of time looking at this particular issue of the hyperscaler revenue gap. Now what does that mean? Well, the hyperscalers-- I don't have a lot of concerns about NVIDIA. Their gross margins and everything looks fantastic. And their order book looks amazing.
The CEO has described the order book for their new Blackwell GPUs as insane. The issue is who's making all that money. Well, it's the hyperscalers-- Google, Microsoft, Meta, Apple, Oracle, Tesla, et cetera. And they eventually have to earn a return on all that GPU expenditure, capital spending.
And if you factor in their GPU purchases, the implied overall data center spending that goes with that, and the gross margins that they typically would need to earn, these companies are missing about $400 billion in annual revenue. So the markets have been very patient so far because we're in the very early stages of this AI adoption boom. But at some point, I think in the next 18 months, markets are going to start asking some very tough questions if these hyperscalers don't start earning some of this $400 billion to sustain their gross margins on all this capital spending. And our growth managers are watching this closely.
(DESCRIPTION)
Line graph, text: Hyperscaler capex and R&D as a share of revenues. The x axis shows years 2017 to 2025 and y axis shows 15 to 70% in increments of 5%. Lines represent Amazon, Alphabet, Microsoft, Meta
(SPEECH)
Hyperscaler capital spending and R&D as a share of revenues has been creeping up for the big four-- Meta, Microsoft, Alphabet, Amazon-- and starting to get into some pretty eye-popping territory where capital spending in R&D is approaching 35% of revenues.
(DESCRIPTION)
A plot shows estimated training cost and compute of select AI models by thousands of dollars.
(SPEECH)
And as we all know, these models just keep getting more expensive and energy intensive to train. And almost any chart you look at on this has to be plotted in log scale because of the amounts of time, effort, and money that are being poured into training these models.
(DESCRIPTION)
Bar graph, text: Gen AI projects, return on investment vs. expectations, % of survey respondents. Source: AlphaWise, Morgan Stanley Research, 2024
(SPEECH)
Now the good news is the surveys of companies that are ultimately going to pay the hyperscalers for all of these AI programs, we're starting to see the first crop of surveys of how those Gen AI applications are working out. And as you can see in this chart here, and we go into more detail in the piece, 90% either exceeded or met expectations. So that's a positive sign.
Some of the other surveys we walk through in the piece look at timelines of Gen AI applications. But I think it's a pretty notable thing that the early crop across multiple sectors of Gen AI projects have exceeded or met expectations of people investing in them. That's a good sign. That doesn't tell you how much money they're ultimately going to be spend on it, but it's a good sign.
(DESCRIPTION)
Line graph, text: Corning optical communications revenue, US dollars billions. The x axis shows the years 2000 to 2024 in increments of 3 years. The y axis shows 1 to 5 billion dollars in increments of 1 billion. Source: Bloomberg, company filings, J P M A M, 2024. Lines represent Real and Nominal with the nominal line being higher.
(SPEECH)
But I just want to remind everybody of this time capsule that I put together recently from our archives.
If you go back to the year 2000 and you look at Corning's order book, they built a lot of fiber. Now, ultimately, all that fiber was used. None of it was wasted. Ultimately, all that fiber got used. But because of the overbuilding, it took Corning around 20 years to get back to the nominal revenues in their optical communications business that they had in the year 2000.
And similarly, if you look at a chart on undersea fiber optic cables deployed per year, it took until 2020 to exceed the year 2000 level after a collapse.
(DESCRIPTION)
Bar graph, text: New undersea cables deployed per year
(SPEECH)
The reason I'm mentioning this is when you have an overbuilding episode, even if all that capital stock ultimately ends up getting used, sometimes it can take a long time to absorb it. And then obviously, there is a lot of implications for the companies that spent that money along the way.
(DESCRIPTION)
Line graph, text: Nuclear pure play and exposed stock returns, Total return index (100 = June 2021). The x axis shows 2021 to 2024 with lines representing Nuclear Exposed and Nuclear Pure Play. Both lines increase steadily with Nuclear Exposed showing more of an increase.
(SPEECH)
Now one of the issues that I'm fascinated by are all the green power commitments from a lot of these data centers and hyperscalers in terms of the energy intensity. Now sometimes people make too much out of this power demand coming from data centers over the next decade. I expect much bigger power demand numbers from electrification of vehicles and electrification of home heating through heat pumps, which will also be adopted by industry and commercial applications in office buildings.
And there's also a lot of different forecasts for data center demand. But one of the most incredible, meaning not believable, things about this is that nuclear power is going to play a role. I just don't see it. There's been a boom recently in the performance of nuclear exposed and nuclear pure play shares. I'm going to take the other side of this one.
(DESCRIPTION)
Line graft, text: During the 1980's, nuclear development shifted from developed to developing countries. The x axis shows year of construction start from 1950 to 2010 and the y axis shows 0 to 100% in increments of 10%. The line drops below 20% for 2010 to 2020.
(SPEECH)
Just as a bit of background, OECD nuclear power-- the OECD used to dominate nuclear power during the '50s, '60s, '70s and then started sliding in the '80s. And by the time you get to the late 1990s, and from then until now, nuclear development is essentially something that belongs to and is done in developing economies and not developed ones. And you can see why.
If you look at the cost gap here,
(DESCRIPTION)
Bar graph, text: Capital cost of nuclear plants in developed countries vs emerging markets.
(SPEECH)
I'm looking at the capital cost of four plants that were recently completed in the West because that's all there have been over the last 20 years, Flamanville in France, one in Finland, the Georgia Vogtle plant, and then Hinckley Point in the UK, all cost many multiples of the plants that are being built in Russia or China and Pakistan. Now there's lots of reasons for that. I've read all the studies on that.
I personally don't believe that we can start cranking these things out in the same way that you build solar panels or lithium ion batteries to bring their costs down. So I think people are way too excited about the possibilities of nuclear power having some kind of renaissance that contributes meaningfully to data center demand. And in The Outlook, we get into the weeds on this. Vogtle 3, by the way, in Georgia already having operational problems and it just came online.
The US lacks a lot of uranium and has import reliance on certain kinds of uranium, which is creating problems for Terrapower. There are very limited opportunities for plant restarts. And way too much was made of the Palisades reopening and the announcement and as well the one that's taking place with Three Mile Island. You can only really restart plants that just recently shut down, meaning within the last couple of years, because the decommissioning process hasn't started yet. If you start the decommissioning process, it's way too expensive to then reverse that and get that plant back online.
So at maximum, I can think of maybe two or three other shuttered plants that would be eligible for restart. The next gen designs-- everybody likes to talk about them-- sodium cooling, gas cooling, molten salt. And according to the DOE's nuclear people, at least a decade away from commercialization in the US.
And then small modular reactors-- very impressive pipeline. 6.5 gigawatts, $200 billion in orders. They're still lottery tickets. And the power commitments are not take or pay contracts. They're all contingent on final cost, as was the case with the NuScale plants that were abandoned in Utah last year. So let's take some reality pills on this nuclear renaissance. Take a look at this section. It gets into the details.
(DESCRIPTION)
A painting appears of two men on horseback in a desert setting. Title: Don Quixote and Sancho Panza, Daumier, 1865 to 1867
(SPEECH)
We also have a section on DOGE Quixote. I am envisioning Vivek Ramaswamy and Elon Musk as Don Quixote and Sancho Panza from Don Quixote. The thing to remember is that one of the underlying themes of the model is the pragmatism and humility are required for them to complete their quest. And I would question whether Vivek or Elon qualify for each of those, or either of them. Anyway.
Vaya con Dios, DOGE Quixote. I think they'll obviously get some things done, but they're going to run into a number of different buzz saws. And one of my top 10 predictions is they're not going to make it anywhere near some of the targets that they've mentioned. There are savings to be had from reversing some of Biden's recent executive actions, but those wouldn't count towards DOGE targets on reducing current spending. That would just be reducing potential future spending that hasn't really even been penciled in yet.
Non-defense discretionary spending has already been cut. Budget Control Act of 2011. It's at its lowest level in decades. So therefore they're going to have to focus on entitlements. And the DOGE billionaires are now finding out that that's a political third rail. Good luck with that.
The SEC, the Department of Labor, the Department of Education, and the EPA combined represent less than 2% of all federal workers. So fine, start firing them. But that's not going to get you very much. The vast majority of federal workers are the Postal Service and the Department of Defense.
There will be efforts underway to expand-- the CMS's is maybe expand weight loss coverage for weight loss drugs. And those are going to offset some of the DOGE savings, which takes place elsewhere.
Trump has talked about impounding or redirecting congressionally approved and congressionally appropriated spending. That's going to face judicial challenges. There was an Impoundment Control Act passed in 1974 in response to something Nixon did. A president can't, according to the Supreme Court and according to this legislation, can't just say, well, I'm going to take the money that was appropriated for the Department of Education and spend that on something else, which is what he's announced that he wants to do.
Most defense spending is people and operations on the order of 70% to 80% rather than weapons procurements. And then with respect to weapons procurements, good luck. The number of federal contractors has fallen from 51 in the early '90s to five. So it's going to be very difficult to cut those procurement costs with that much concentration in the defense contracting.
And then we also get into some of the things on improper and fraudulent payments that the government makes, sometimes accidentally. You can try to reduce them, but it's going to cost a lot of the money in terms of investing in systems and people to do that. So we get into the details here. It's interesting if you want to understand it. And good luck to all the billionaires that they've recruited onto the team, along with Vivek and Elon, to try and get this done.
(DESCRIPTION)
Line graph, text: China economic activity monitor (Bloomberg). The x axis shows 2022 to 2024 and y axis shows 1% to 9%. Text: Exports, coal and electricity consumption, steel production, motor vehicle sales, real estate investment, medicine consumption, telecom equipment production, S O E output private enterprise output. Source: Bloomberg, November 30, 2024. The line shows a decrease from 2022 to 2024.
(SPEECH)
We have a section on China that looks at both the liquidities trap and the Thucydides trap, which is something we wrote about a couple of months ago. The bottom line is that China is experiencing some kind of liquidity trap, shares some kind of characteristics with Japan in the early 1990s. As you can see here, Bluebird maintains an economic activity monitor-- exports, coal, electricity consumption, steel production, car sales, real estate investment, consumption of medicine, the kind of things that are for the Chinese government to paper over and obscure.
And the momentum on this thing is fairly negative, which is why China announced that stimulus package.
(DESCRIPTION)
Bar graph. Text: Impact on growth of major policies China has announced since September 2024, Annualized impact on growth, %. Source: Bridgewater November 2024. Rough estimate of drag shows negative 2%.
(SPEECH)
Now the stimulus package was substantial, impressive. I believe there's more fiscal stimulus coming. Bridgewater did an interesting analysis a couple of months ago that show that the total impulse from monetary and fiscal stimulus would offset roughly half of the economic drag that's taking place in China right now. That's a good effort. And the domestic A-share market kind of soared after the announcement. And it has retained its gains whereas the MSCI China index, which is a broader index of Chinese-exposed stocks, has since given up about half of the gains on the stimulus announcement.
I'm not going to go through the details here. You can read the piece.
(DESCRIPTION)
Line graph, text: M S C I China valuations, Forward price to earnings ratio. x axis shows 2006 to 2021 in 5 year increments. The y axis shows 5x to 30x in increments of 5. Source: Bloomberg J P M A M, December 27, 2024. The line spikes and dips below 12x after 2021.
(SPEECH)
I just think the days of 14 to 18 PEs on MSCI China are a thing of the past. And for geopolitical reasons, they're going to be capped at about 12x, which doesn't leave a lot of room for a rally from here. So you could take a look at the piece. We go through the details of some of the tariff and espionage issues at the core of some of the conflict issues between the US and China.
(DESCRIPTION)
Annotated line graph, text: Publicly reported Chinese espionage incidents sourced from open source materials. Events span 2000 to 2021 with a dip in 2015 then a sharp increase to 2021.
(SPEECH)
I'll just say that the increasing intensity of Chinese espionage spreading across almost every industrial sector in the US and other Western countries, it makes the idea of a rapprochement that I've read about highly unlikely. And the same goes for some of the Naval blockade exercises that China has been conducting with regards to Taiwan. So you can read more about that.
(DESCRIPTION)
A graph shows creation of new public companies in the 21st century with the US leading Europe. US companies listed include Meta, Alphabet, Tesla, Visa, Mastercard, Netflix. And Europe includes Airbus, Porsche, Equinor
(SPEECH)
We have a section on Europe called Dr. Seuss goes to Europe. It's just two pages with some Dr. Seuss style language and some charts. I just want to show you one of the charts here. We have charts in comparisons of labor productivity, and energy prices, and deindustrialization in Europe, and problems with monetary policy being applied to countries that are too different compared to the US, low labor mobility, very poor ROA and ROE numbers in Europe other than in health care, and then very negative earnings revisions.
The chart I really wanted to show you was this new one, which is the creation of new public companies on the 21st century compared the US to Europe. This kind of tells you all you need to know about the outperformance of the US versus Europe over the last 20 years. It's kind of an amazing chart to see. I can't describe it verbally. You just kind of have to look at it.
(DESCRIPTION)
Line graph, text: Crypto coin prices by segment, market cap weighted. Lines represent Bitcoin, Transactions Focused Basket, Decentralized Finance Basket. The x axis shows 2022 to 2025 and y axis shows 25 to 250 in increments of 25. The lines for bitcoin and transactions focused basket reach above 225 at the end of 2024. Source: Bloomberg, J P M A M, December 27, 2024
(SPEECH)
The crypto presidency-- I wrote a piece called "The Maltese Falcoin" a couple of years ago. It was February 2022. And I was skeptical of Bitcoin as a store of value and then even more skeptical on the transactions-focused coins and the DeFi coins. And if you're not familiar with the jargon, you can look at the Eye on the Market on this section and it gets into the details here.
I said I would take another look at Bitcoin if it fell. And it fell. It fell hard. It fell 65% after I wrote the piece. And my mistake for not looking at it again. It has since rallied pretty sharply. And a lot of the surveys of institutional investors are fairly positive on the prospects for increased digital ownership.
As it relates to the transactions-focused coins and the DeFi coins, I don't believe it. I was right about those coins for three years. Nothing much has happened in terms of actual utilization of either of those kinds of crypto coins. There's been this spectacular rally in a handful of them, Ripple in particular, but it looks highly speculative to me.
And then specifically, I've got this chart here looking at the volumes of utilization of Ripple, Litecoin, Monero, and Stellar.
(DESCRIPTION)
A line graph shows transactions per day on blockchains from 2022 to 2025.
(SPEECH)
None of them have changed since two years ago. And so if the value proposition of these coins is that people are going to use them more, you'd expect to see that happening and you don't. So that means that this is a very technical and speculative rally.
Same
(DESCRIPTION)
Line graph, text: Ethereum daily verified contracts, Total verified contracts per day, 7 day average. x axis shows 2022 to 2025, and the y axis shows 100 to 1000. The line peaks at 1,000 in 2023, dropping by 2025 to about 200.
(SPEECH)
thing for Ethereum. The number of daily verified contracts in Ethereum being executed on a blockchain are now unchanged versus what they were in early 2022. So to me, what's really happening is that people are earning a sky high return on their political donations. Crypto PACs received almost $250 million in donations, around half of all corporate donations made during the election. And Coinbase gave $75 million to a Super PAC called Fair Shake. Their candidates won almost every race they got involved in.
Look, they're playing the game and participating in the system in ways that they're allowed to do. I'm not criticizing it. I'm just saying that this, to me, is a better our explanation of the pop in crypto than the presumed proposition that their acceptance and utilization has been in changing. All that's really been changing is the politics. And the GOP Majority Leader of the House is pledging a crypto bill in the first 100 days. Trump has announced a crypto council and a crypto czar. The new SEC head will probably roll back some of the crypto accounting disclosure rules. So this is a shift in sentiment rather than a shift in utilization. And you could take a closer look.
And then just to close out here-- and thank you all for listening.
(DESCRIPTION)
A table appears. Text: 2024 predictions recap, Scorecard. Rankings appear from 1 to 10, January 1, 2024 Prediction. Highlighted green, 1, Despite all the hand-wringing over risks to reserve currency status and BRIC efforts to develop a competitive reserve currency, the US dollar remains stable, within a +/- 1 std deviation band of 7%. Highlighted green, 2, The DOJ/FTC win a big antitrust case. Highlighted green, 3, President Biden withdraws between Super Tuesday and the election, citing health reasons. Highlighted green, 4, Little to no progress on self driving cars in the US, no recovery on Lidar stock basket. Highlighted yellow, 5, Broadly syndicated leveraged losses rise above credit losses for the first time. Highlighted yellow, 6, Argentine dollarization will fail if implemented. Highlighted green, 7, Russian invasion of Ukraine drags on with no ceasefire in 2024. Highlighted green, 8, US regional bank stocks will do well in 2024 with stable or rising price to book values. Highlighted red, 9, Major US cities will face electricity outages and/or natural gas outages due to outdated grid infrastructure. Highlighted yellow, 10, Researchers will complete work on an inhaled COVID vaccine that will reduce transmission.
(SPEECH)
I started a top 10 list last year in honor of Byron Wien, who had passed away. He did this every year forever.
I made a bunch of predictions last year. Six of them were right. Three of them were kind of either too soon to tell or hard to say. And then one was wrong. I've made a new list for this year. I will tick through them very quickly and then sign off.
At least three confirmed Trump cabinet members will be gone by the end of the year. Renewables will underperform traditional energy again in 2025, like they've done in the last three years. MSCI Japan will outperform China despite China's higher PE multiple.
The discontinuation rates, which are approaching 50%, of weight loss drugs are going to hurt some of the GLP stocks, specifically Novo Nordisk. I think we'll end the year below consensus analysts price targets. Despite the order book referenced earlier, I don't think any small modular reactors in the United States will even begin construction in 2025.
I think DOGE cost cuts are going to be no more than $150 billion a year, excluding reversal of executive actions compared to targets of $2 trillion. Hedge funds I expect to continue to outperform the stock bond benchmarks. And we explain what we mean by that. But they've been doing better than expected. Housing inflation, I think because of completions, both single family and multifamily, will start to finally bring housing inflation down below 4% the way that the Fed computes it.
Blue state outmigration will continue. Over the last few years, 70% of all movers in the country in the US are moving from either blue to red or purple states. I think that's going to continue. And then lastly, I think there's a chance that the Trump administration launches a targeted military action against drug cartels in Mexico.
So that's my top 10 list. Thank you for listening. Take a look at the outlook. There's a lot in there that we can't get into in a podcast like this. And it's been a privilege. Thank you very much. Bye.
(DESCRIPTION)
Logo: J.P. Morgan
Descrizione audio
Good afternoon, everybody. This is Michael Cembalest with the 2025 Eye on the Market Outlook podcast, so Happy New Year, everybody. The Outlook this year is called “The Alchemists,” for reasons that I will explain in a minute.
First, just want to mention this is the 20th anniversary of the Eye on the Market. I launched it in January of 2005. It was meant for clients that consume information on either a BlackBerry or a Palm Pilot. It didn’t have any charts or graphs or tables or footnotes or anything. It was text only. A lot of things have changed since then. It’s been an amazing journey. I’ve enjoyed working on it. I’ve learned a lot. I’d like to thank Jamie and Mary for supporting me in this effort, particularly when I write about controversial topics. And this year is no exception.
So, on to the Outlook itself. I have a section on “The Alchemists,” which is about Trump’s impact on markets; we have a section on AI adoption trends and milestones, given the Mag 7’s support to the markets; a section on nuclear power, “Wake me when we get there”; a section on DOGE called “DOGE Quixote,” which refers to the billionaire’s quest to reduce government spending; an update on China and the liquidity trap; a section called “Doctor Seuss goes to Europe”; we revisit the crypto markets, given the rally since our Maltese Falcoin piece in February 2022; and discuss the sky-high returns that donors have gotten on their crypto political donations; and then a top 10 list for 2025, following on last year's top 10 list, which was the first time we did that.
So, I'd like to thank GPT again for some fantastic cover art. This is a depiction of Trumpism as alchemism. And what’s going on here is you’ve got a guy in red hat, and he’s mixing together eight different things: deregulation, deportations, tariffs, tax cuts, government cost cutting, oil and gas, crypto, the medical freedom movement and agency purges. And then the question is posed, what could possibly go wrong? I also think we try to address what could possibly go right.
There's a lot more detail that we go through in the actual Outlook in the executive summary. Let me just give you a brief description of it here. We start with this. What are the Alchemists trying to change? If you take them at face value, what they're trying to change is this chart. And this chart shows that for the better part of 80 years, consumer spending, GDP growth and industrial production went up together like three doves flying the same direction, or geese or—I’m not a bird person.
Then, as soon as China joins the World Trade Organization in the year 2000, we get what I call the “silence of the plants”—that’s a Clarice Starling reference—where industrial production flatlines while consumer spending and GDP keep growing. There are some other ominous inflection points negative for the United States that also took place around the same time that China joined the World Trade Organization.
Rising manufacturing job losses in the U.S., a follow labor share of gross profits, rising suicide rates in the U.S. and rising non-metro, meaning rural, poverty rates in the U.S. So that’s what the Alchemists claim they’re trying to change. What kind of tailwinds are they trying to create? Well, more inbound foreign direct investment into the United States, part of which is already happening; a slowdown in the breakneck pace of regulation, which was taking place under the Biden administration, and that we've showed charts on and discussed many times; freeing up bank capital by ending the Basel III discussions, which could presumably lower the cost of credit; increasing oil and gas production to bring down energy prices, although I think when you look at the electricity prices in many parts of the country, the renewable transition is irreversible and is driving those energy prices up; and then building on a rebounding CapEx and confidence outlook; and also a recovery that we’re seeing in venture capital markets, which we spent some time on in the executive summary as well.
So, those are the tailwinds that I think it’s fair to say the Alchemists are trying to create. Now, what headwinds are the Alchemists likely to face with their set of policies? Well, as everybody has discussed, possible inflationary consequences, not just from tariffs themselves, but from the retaliation that the United States would face, with whether it was a China tariff, a universal tariff, any kind of tariff—and, by the way, of all of the ten states that import the most amount of goods relative to their GDP, eight of them are red states. Tightening the labor supply through deportations, and things like that, have potential consequences for both wage inflation and Fed policy that I’m not sure that the administration would like.
Now, to be clear, and we talk about this in multiple places, the Biden administration oversaw the largest uncontrolled, unmanaged and unsupervised migration surge on record. And even the Brookings Institution at one point argued that the borders needed to be closed for some period, just to absorb the immigration that had taken place. But that said, tightening labor supply too much will have consequences for wage inflation and bad policy.
And then we already have Trump’s inheriting large budget deficits despite almost full employment. And then from a policy perspective, the Republicans have almost zero room for defections in the House because, they’ve only got a 219 to 215 lead. The gate seat will have a special election sometime in March. They’ll lose a couple of seats in Florida. Bottom line is they really have no room for defections in the House.
And then I won't go into too much detail, we have a page on this. Suffice to say, the medical freedom movement that RFK, Jr. supports on occasion is not a confidence builder in the concept of U.S. exceptionalism.
So, how do we see what the market verdict on this mix of headwinds and tailwinds is? My favorite barometer is a 10-year Treasury. And so far, the news is not great. If you look at the last seven tightening, I’m sorry, easing cycles from the firm, the United States going back to the 1980s. Typically, what happens is the Fed starts to ease and the 10-year Treasury either stays around flat or and or comes down.
This is the first time in 34, 35 years or so that the 10-year Treasury has gone up, and stayed there after the first Fed cut. So the early verdict from the bond markets on this mix of Alchemist policies is not great, and suggests that the productivity benefits from tax cuts and deregulation may not be enough to offset the inflationary consequences of, tariffs and deportations and some of the other things that they’re planning on doing.
So ultimately, again, the 10-year Treasury is going to be the best barometer, in my opinion, of this mix of policies. And the Fed fund futures markets are sending a similar signal. A year ago, the markets were pricing in a funds rate that would fall below 4%. That's not the case anymore. The latest Fed fund futures markets are projecting the Fed won’t be able to ease more than down to 4% or so. And I’d be surprised if they even were able to get there sometime in 2025.
Equity investors are more optimistic. The Conference Board’s survey of households expecting higher stock prices in 12 months is at an all-time high, even higher than it was in in the mid to late 1990s. And there are some technicals that support that. Equity supply is very tight. In other words, this looks at secondary and primary issuance net of buybacks and net equity supply, particularly in the U.S., is actually declining.
So there’s a scarcity value to public equities, if you can believe that. So from a technical perspective, that’s supported. But U.S. equity markets are pretty expansive. We all know this, and at least on a valuation basis. Now, markets can still go up when valuations are high. But we’re in the mid to upper 90s in terms of percentiles for many different ways that you could look at the valuation of U.S. stocks. Median P/Es, forward P/Es, trailing P/Es, the enterprise value to cash flow, price to book, price to sales, you name it. Most of the valuation indicators are on the high side. And we’ve also had two 20%-plus years in a row. That’s only happened 10 times since 1871 and only during the 1990s bull market and the Roaring Twenties, around 100 years ago, did the good times keep going.
So my takeaway from this, and you can read this, in the executive summary of the Outlook, which goes obviously to a lot more interesting detail, is, I think that there’s a steep curve here for, for the Alchemists in terms of trying this policy mix of deregulation, tariffs and cost cuts and tax cuts and crypto and things like that. They’ve got a very limited margin for error given how much of an equity war we’ve already had, and how high valuations are.
So I expect U.S. equity markets to end the year 2024 are when they began. But I expect the Alchemists to break something. And I don't know what. I don’t think they do either, but they’ve got their sights set on everything from the FBI to Medicare to border policy to the Defense Department to weapons procurement.
I think they’re going to break something. And so I expect a correction of some kind, of 10–15%, from whatever value it’s at, sometime in 2025. That’s not that infrequent a thing, it happens in around one-third of all years, but I am expecting a correction to take place. And then for equities to end up modestly higher by a few percent at the end of next year compared to where we are right now.
Just to be clear, the deregulatory efforts that the incoming administration is talking about, they do have a long history of increasing business, spending, capital spending, business confidence. I started writing about deregulation over 20 years ago. And here’s a chart from the archives of the Eye on the Market that looked at the Clinton deregulation of telecommunications and electricity markets, and the resulting increase in nonresidential business investment, which is, essentially, you know, a capital stock increase.
Peter Orszag at Brookings has written about the benefits of deregulation, on multiple occasions. You know, he was the, I think, either the CBO or the OMB for a while. And so I’m a believer in the benefits of deregulation. And then after the election, there was a surge in the number of small companies responding positively by saying that they’re now planning on some additional capital spending over the next few months, compared to what they were before the election.
But, you know, tariffs and shrinking the labor supply and some of the other issues that the Alchemists are planning, I think will offset some of this. So it’s a really interesting policy mix. It’s certainly not anything that anyone has tried in 70 or 80 years. And again, you know, the executive summary of the Outlook gets into the details of this policy mix and what its implications for markets might be.
We’ve also got a long section on AI and adoption and milestones and things. I’m just going to share a few comments here. The U.S. exceptionalism that you read about in the equity markets, you can back that up with numbers. And I thought this was an amazing chart. It comes from Mike Gold and his team at Empirical Research, and it shows the free cash flow margins by decade for the 10 largest stocks.
So you might think the 10 largest stocks always make a lot of money. That’s true. But the 10 largest stocks make a lot more money now than they did in the past. This goes back all the way to 1950, and shows the free cash flow margins for the 10 largest stocks in each decade. And this crop of U.S. exceptional large companies, with free cash flow margins above 25%, are setting all-time highs here.
Now, the rest of the S&P has been treading water. And when you decompose earnings per share growth of the Mag 7 versus the S&P 500 ex Mag 7, you get kind of a mixed picture here. The S&P 500 ex Mag 7 had an earnings boost that took place in 2022, but it was mostly just a reversal of the declines that took place during the pandemic.
It’s only the Mag 7 that has been consistently generating since 2017 some pretty impressive earnings growth. So this is a very bifurcated market and, unfortunately, the implication is that very diversified investors that have been investing in, let’s say, an equal weighted S&P, have struggled with that. And it’s really important to understand this AI boom because we’re now looking at something which is only happened I think twice in the last 50 years or so, which is when one specific company, or small group of companies, capture a 15% or so share of all the capital spending that’s taking place in the economy.
I mean, that’s kind of an amazing thing to think about—one company or a gaggle of companies capturing 50% of all capital spending. In 2026, that's what’s projected to happen with Nvidia’s data center revenues that are supposed to roughly be equal to 15% of all capital spending in the economy. The only thing we can compare that to is IBM’s mainframe peak in the late 60s. And then obviously the Dot Com bubble in 2000.
And so we spend a lot of time looking at this particular issue of the hyperscaler revenue gap. Now, what does that mean? Well, I don't have a lot of concerns about Nvidia. Their gross margins and everything looks fantastic and their order book looks amazing. The CEO described the order book for their new Blackwell GPUs as “insane.” The issue is, who's making all that money? Well, it's the hyperscalers: Google, Microsoft, meta, Apple, Oracle, Tesla, etc. And they eventually have to earn a return on all that GPU expenditure, capital spending. And if you factor in their GPU purchases, the implied overall data center spending that goes with that and the gross margins that they typically would need to earn, these companies are missing about $400 billion in annual revenue.
So the markets have been very patient so far because we're in the very early stages of this AI adoption boom. But at some point, I think the next 18 months, markets are going to start asking some very tough questions if these hyperscalers don't start earning some of this $400 billion to sustain their gross margins, all this capital spending. And our growth managers are watching this closely.
Hyperscale capital spending in R&D as a share of revenues has been creeping up for the Big 4: Meta, Microsoft, Alphabet and Amazon. Starting to get into pretty, pretty eye popping territory where our capital spending on R&D approaching 35% on revenues. And as we all know, these models just keep getting more expensive and energy intensive to train, and almost any chart you look at on this has to be plotted in log scale because of the amounts of time, effort and money that have been poured into training these models.
Now, the good news is the surveys of companies that are ultimately going to pay the hyperscalers for all of these AI programs, we're starting to see the first crop of surveys of how those gen AI applications are working out. And, as you can see in this chart here, and we go into more detail in the piece, 90% either exceeded or met expectations. So that's a positive sign. Some of the other surveys we’ll walk through in the piece look at timelines of AI applications. But I think it’s a pretty notable thing that the early crop across multiple sectors of gen AI projects have exceeded or met expectations of people investing, and that’s a good sign.
That doesn’t tell you how much money they’re ultimately going to be spending on it. But it's a good sign. But I just want to remind everybody of this time capsule that I put together recently from our archives. If you go back to the year 2000 and you'll look at Corning’s order book, they built a lot of fiber. Now, ultimately, all that fiber was used, right? None of it was wasted. Ultimately, all that fiber got used. But because of the overbuilding, it took Corning around 20 years to get back to the nominal revenues and their optical communications business that they had in the year 2000. And similarly, if you look at a chart on undersea fiber optic cables deployed per year, it took until 2020 to exceed the 20 the year 2000 level after collapse.
The reason I'm mentioning this is when you have an overbuilding episode, even of all that capital stock ultimately ends up getting used, sometimes it can take a long time to absorb it, and then, obviously, there were a lot of implications for the companies that spent that money along the way.
Now, one of the issues that I’m fascinated by are all the green power commitments from a lot of these data centers and hyperscalers, in terms of the energy intensity. Now, you know, the sometimes people make too much out of this power demand coming from data centers over the next decade. I expect much bigger power demand numbers from electrification of vehicles and electrification of home heating through heat pumps, which will also be adopted by industry and commercial applications and office buildings. And there’s also a lot of different forecasts for data center demand.
But one of the most incredible—meaning, not believable—things about this, is that nuclear power is going to play a role. I just don’t see it. There’s been a boom recently in the performance of nuclear exposed and nuclear pure-play shares. I’m going to take the other side of this one.
Just as a bit of background, OECD nuclear power. You know, the OECD used to dominate nuclear power during the 50, 60s, 70s and then started sliding in the 80s. And by the time you get to the late 1990s, and from then until now, nuclear development is essentially something that belongs to and is done in developing economies and not developed ones. And you can see why if you look at the cost gap here, I’m looking at the capital cost of four plants that were recently completed in the West, because that’s all there have been over the last 20 years, Flamanville in France, one in Finland, the Georgia Vogtle plant and then Hinkley Point in the UK, all cost many multiples of the plants that are being built in Russia, China and Pakistan.
Now, there's lots of reasons for that. I’ve read whole studies on that. I personally don’t believe that we can start cranking these things out in the same way that you build solar panels or lithium ion batteries to bring their cost down. So I think people are way too excited about the possibilities of nuclear power having some kind of renaissance that contributes meaningfully to data center demand.
In the Outlook we get into the weeds on this. Vogtle 3, by the way, in Georgia, is already having operational problems, it just came online. The U.S. lacks a lot of the uranium and it and it has import reliance on certain kinds of uranium, which is creating problems for Terrapower. There are very limited opportunities for plant restarts. And way too much was made of the Palisades reopening and the announcement and as well, the one that’s taking place with Three Mile Island. You can only really restart plants that just recently shut down, meaning within the last couple of years, because the decommissioning process hasn't started yet. If you start the decommissioning process, it’s way too expensive to then reverse that and get that clamp back on line. So at maximum, I can think of maybe two or three other shuttered plants that would be eligible for restart, the next gen designs.
Everybody likes to talk about sodium cooling, gas cooling, molten salt. According to the DOE’s nuclear people, they’re at least a decade away from commercialization in the U.S. And then, small modular reactors. Very impressive pipelines. 6.5 GW. 200 billion in orders. They’re still lottery tickets. And the power commitments are not take or pay contracts. They’re all contingent on final cost, as was the case with the new scale plants that were abandoned in Utah last year. So let’s, you know, take some reality pills on this nuclear renaissance. Take a look at this section. It gets into the details.
We also have a section on DOGE Quixote. I am envisioning Vivek Ramaswamy and Elon Musk as Don Quixote and Sancho Panza from Don Quixote. The thing to remember is that the one of the underlying themes of the model is the pragmatism and humility are required for them to complete their quest. And I would question whether Vivek or Elon qualify for each of those, or either of them. Vaya con Dios, Doge Quixote. I think they’ll obviously get some things done, but they’re going to run into a number of different buzzsaws. And one of my top 10 predictions is they’re not going to make it anywhere near some of the targets that they’ve mentioned.
There are savings to be had from reversing some of Biden’s recent executive actions, but those wouldn't count towards DOGE targets on resisting current spending. That would just be reducing potential future spending, that hasn’t really even been penciled in yet. Non-defense discretionary spending has already been cut. Budget Control Act of 2011. It’s at its lowest level in decades. So therefore, they’re going to have to focus on entitlements. And the DOGE billionaires are now finding out that that’s a political third rail. Good luck with that.
The SEC, the Department of Labor, the Department of Education and the EPA combined represent less than 2% of all federal workers. So fine, start firing them. But that’s not going to get you very much. The vast majority of federal workers are, you know, the Postal Service and the Department of Defense.
There will be efforts underway to expand the CMS, that is, maybe expand, weight loss coverage for weight loss drugs. And those are going to offset some of the DOGE savings, which takes place elsewhere. Trump has talked about impounding or redirecting congressionally approved and congressionally appropriated spending. That’s going to face judicial challenges. There was an Impoundment Control Act passed in 1974, in response to something Nixon did. A president can’t, according to the Supreme Court and according to this legislation, can’t just say, well, I’m going to take the money that was appropriated for the Department of Education and spend it on something else, which is what he’s announced that he wants to do.
Most defense spending is people and operations on the order of 70 to 80% rather than weapons procurements. And then, with respect to weapons procurements, good luck. You know, the number of federal contractors has fallen from 51 in the early 90s to five. So it’s going to be very difficult to cut those procurement costs with that much concentration in the defense contract. And then, we also get into some of the things on improper and fraudulent payments that the government makes, sometimes accidentally, you can try to reduce them, but it’s going to cost a lot of money in terms of investing in systems. And people will do that. So, we get into the details there. It’s interesting if you want to understand it and good luck to all the billionaires that they've recruited to the team, along with Vivek and Elon, to try and get this done.
We have a section on China that looks at both the liquidity trap and the Thucydides trap, which is something we wrote about a couple of months ago. The bottom line is that China is experiencing some kind of liquid illiquidity trap that shares some kind of characteristics with Japan in the early 1990s. As you can see here, a Bloomberg maintains an economic activity monitor—exports, coal and oil consumption, steel production, car sales, real estate investment, consumption of medicine—the kind of things that are harder for the Chinese government to paper over and obscure.
And, you know, the momentum on this thing is fairly negative, which is why China announced that stimulus package. Now, the stimulus package, that was substantial, impressive—I believe there’s more fiscal stimulus coming. Bridgewater did an interesting analysis a couple months ago that showed that the total impulse from monetary and fiscal stimulus would offset roughly half of the economic drag that’s taking place in China right now. That’s a good effort. And the domestic A-share market kind of soared after the announcement and has retained its gains, whereas the MSCI China index, which is a broader index of Chinese exposed stocks, has since given up about half of the gains, now the stimulus was announced. But I’m not going to go through the details here. You can read the piece. I just think the days of 14 to 18 P/Es on MSCI China are a thing of the past and for geopolitical reasons, they’re going to be capped at about 12x, which doesn’t leave a lot of room for a rally from here.
So you could take a look at the piece. We go through the details of some of the tariff and espionage issues at the core of some of the conflict issues between the U.S. and China. I’ll just say that the increasing intensity of Chinese espionage spreading across almost every industrial sector in the U.S. and other Western countries, it makes the idea of a rapprochement that I’ve read about highly unlikely, and the same goes for some of the naval blockade exercises that China has been conducting, with regards to Taiwan. So you can read more about that.
We have a section on Europe called “Doctor Seuss goes to Europe.” It’s just two pages with up with some Doctor Seuss–style language and some charts. I just want to show you one of the charts here. We have charts and comparisons of labor productivity and energy prices and deindustrialization in Europe, and problems with monetary policy being applied to countries that are two different compared to the U.S. low labor mobility, very poor ROA and ROE numbers in Europe other than in health care, and then very negative earnings revisions.
The chart I really wanted to show you was this new one, which is the creation of new public companies in the 21st century comparing the U.S. to Europe. This kind of tells you all you need to know about the outperformance of the U.S. versus Europe over the last 20 years. It's kind of an amazing chart to see. I can’t like, describe it verbally. You just kind of have to look at it.
The crypto presidency. I wrote a piece called The Maltese Falcoin a couple of years ago. It was February 2022, and I was skeptical of Bitcoin as a store value and then even more skeptical on the transactions-focused coins and the DeFi coins. and if you’re not familiar with the jargon, you can look at the Eye on the Market on this section and it gets into the details here. But, I said I would take another look at Bitcoin if it fell. And it fell, it fell hard and it fell 65% after I wrote the piece. And my mistake for not looking at it again. It has since rallied pretty sharply. And a lot of the surveys of institutional investors are fairly positive on the prospects for increased digital ownership as it relates to the transactions-focused coins and the DeFi coins. I don't believe it. I was right that about those coins for three years. Nothing much has happened in terms of actual utilization of either of those kinds of crypto coins.
There’s been this spectacular rally and in a handful of that ripple in particular, but it looks highly speculative to me. And then, specifically, I’ve got this chart here looking at the volumes of utilization of ripple like coin, Monero and Stellar. None of them have changed since two years ago. Right. And so if the value proposition of these coins is that people are going to use them more, you expect to see that happening, and you don't. So that that means that this is a very tactical and speculative rally. Same thing for Ethereum. The number of daily verified contracts in Ethereum and being executed on a blockchain are now unchanged versus what they were in early 2022. So to me, what’s really happening is that people are earning a sky-high return on their political donations.
Crypto PACs received almost $250 million in donations, around half of all corporate donations made during the election. And Coinbase gave $75 million to a super PAC called Fair Shake. Their candidates won almost every race they got involved in. Look, they're playing the game and participating in the system in ways that they’re allowed to do that, I’m not criticizing it.
I'm just saying that this, to me, is a better explanation for the pop in crypto than the presumed proposition that their acceptance and utilization has been changing. All that’s really been changing is the politics. And, you know, the GOP majority of the year of the House is pledging a pro-crypto bill in the first hundred days that Trump has announced a crypto council and a crypto czar. The new SEC head will probably roll back some of the crypto accounting disclosure rules. So this is a shift in sentiment, rather than a shift in utilization. And you could take a closer look.
And then, just to close out here and thank you all for listening. I started a top 10 list last year in honor of Byron Wiem, who passed away. He did this every year for forever. I made a bunch of predictions last year. Six of them were right. Three of them were kind of, you know, either too soon to tell, hard to say. And then one was wrong. I’ve made a new list for this year. I will tick through them very quickly and sign off.
At least three confirmed Trump cabinet members will be gone by the end of the year. Renewables will underperform traditional energy again in 2025, like they’ve done in the last three years. MSCI Japan will outperform China despite China’s higher P/E multiple. The discontinuation rates, which are approaching 50% of weight loss drugs, are going to hurt some of the GLP stocks, specifically Novo Nordisk, I think will end the year below consensus. Our price targets, despite the order book referenced earlier, I don’t think any small modular reactors in the United States will even begin construction in 2025. I think DOGE cost cuts are going to be no more than $150 billion a year, excluding reversal of executive actions compared to targets of $2 trillion. Hedge funds I expect to continue to outperform the stock bond benchmarks, and we explain what we mean by that. But you know, they’ve been doing better than expected. Housing inflation, I think because of completions, both single family and multifamily will start to finally bring housing inflation down below 4%, the way that the Fed computes it. Migration will continue. Over the last few years, 70% of all movers in the country in the U.S. are moving from either blue to red or purple states. I think that’s going to continue. And then lastly, I think there’s a chance that the Trump administration launches a targeted military action against drug cartels in Mexico. So that's my top ten list, thank you for listening.
Take a look at the Outlook. There’s a lot in there that we can't get into in a podcast like this. And it’s been a privilege. Thank you very much. Bye.
(DESCRIPTION)
Logo: J.P. Morgan, Title: Eye on the Market
An image at left shows a scientist in a lab hovering over colorful vials. He wears a red cap with two letters visible: G, A. A plume of bright magenta smoke fills the air. Title: Outlook 2025, The Alchemists.
(SPEECH)
[WHOOSHING]
(DESCRIPTION)
The speaker, Michael Cembalest, has shortly cropped brown hair and wears glasses. He appears in an office space with a window behind him.
(SPEECH)
Good afternoon, everybody. This is Michael Cembalest with the 2025 Eye on the Market Outlook podcast. A Happy New year, everybody. The outlook this year is called The Alchemists for reasons that I will explain in a minute.
(DESCRIPTION)
Images of a Blackberry and Palm Pilot appear at left. Text: Eye on the Market launched January 2005
(SPEECH)
First, I just want to mention this is the 20th anniversary of the Eye on the Market. I launched it in January of 2005. It was meant for clients that consume information on either a BlackBerry or a Palm Pilot. It didn't have any charts, or graphs, or tables, or footnotes, or anything. It was text only. A lot of things have changed since then. It's been an amazing journey. I've enjoyed working on it. I've learned a lot. I'd like to thank Jamie and Mary for supporting me in this effort, particularly when I write about controversial topics. And this year is no exception.
So on to the outlook itself.
(DESCRIPTION)
Text: 2025 Eye on the Market Outlook. The Alchemists deregulation, deportations, tariffs, tax cuts, cost cutting, oil & gas, crypto, medical freedom, and Agency purges. What could possibly go wrong? The Golden Goose: AI adoption, trends and milestones. What nuclear renaissance? Wake me when we get there. DOGE Quixote, the billionaires quest to reduce US government spending. China: the liquidity trap and the Thucydides trap. Dr. Seuss Goes to Europe. The crypto Presidency: on the crypto rally and the sky high return on political donations. Top Ten List for 2025 and a scorecard on 2024's list.
(SPEECH)
I have a section on The Alchemists, which is about Trump's impact on markets. We have a section on AI adoption, trends and milestones given the Mag 7's importance to the markets; a section on nuclear power-- wake me when we get there; a section on DOGE called DOGE Quixote, which refers to the billionaire's quest to reduce government spending; an update on China and the liquidity trap; a section called Dr. Seuss Goes to Europe. We revisit the crypto markets given the rally since our Maltese Falcon piece in February of 2022 and discuss the sky high returns that donors have gotten on their crypto political donations and then a top 10 list for 2025 following on last year's top 10 list, which was the first time we did that.
Anyway, so I'd like to thank GPT again for some fantastic cover art. This is a depiction of Trumpism as alchemism. And what's going on here is you've got a guy in a red hat. And he's mixing together eight different things-- deregulation, deportations, tariffs, tax cuts, government cost cutting, oil and gas, crypto, the medical freedom movement, and agency purges. And then the question is posed, what could possibly go wrong? I also think we try to address what could possibly go right.
There's a lot more detail that we go through in the actual outlook in the executive summary. Let me just give you a brief description of it here.
(DESCRIPTION)
Title: What are the Alchemists trying to change? A line graph appears that shows US Consumer Spending ad US Industrial Production from 1930 to 2020. They stay together until 2000, when spending exceeds production by 60 points. Text: Other US data with sharp inflection points around the year 2000: Rising manufacturing job losses, Falling labor share of gross profits, Rising suicide rates, Rising non-metro poverty rates. Source: B E A Federal Reserve, J P M A M 2024.
(SPEECH)
We start with this. What are the alchemists trying to change if you take them at face value? What they're trying to change is this chart. And this chart shows that for the better part of 80 years, consumer spending, GDP growth, and industrial production went up together like three doves flying in the same direction or geese. I'm not a bird person.
Then as soon as China joins the World Trade Organization in the year 2000, we get what's called-- what I call the silence of the plants-- that's a Clarice Starling reference-- where industrial production flatlines while consumer spending and GDP keep growing. There are some other ominous inflection points, negative for the United States, that also took place around the same time that China joined the World Trade Organization-- rising manufacturing job losses in the US, a fall of labor share of gross profits, rising suicide rates in the US, and rising non-metro, meaning rural, poverty rates in the US. So that's what the alchemists claim they're trying to change.
What kind of tailwinds are they trying to create? Well, more inbound foreign direct investment into the United States, part of which is already happening; a slowdown in the breakneck pace of regulation, which was taking place under the Biden administration that we've showed charts on and discussed many times; freeing up bank capital by ending the Basel III discussions, which could presumably lower the cost of credit; increasing oil and gas production to bring down energy prices, although I think when you look at the electricity prices in many parts of the country, the renewable transition is irreversible and is driving those energy prices up; and then building on a rebounding CapEx and confidence outlook; and also recovery that we're seeing in venture capital markets, which we spent some time on in the executive summary as well. So those are the tailwinds that I think it's fair to say the alchemists are trying to create.
Now what headwinds are the alchemists likely to face with their set of policies? Well, as everybody has discussed, possible inflationary consequences, not just from tariffs themselves, but from the retaliation that the United States would face with whether it was a China tariff, a universal tariff, any kind of tariff. And by the way, of the 10 states that import the most amount of goods relative to their GDP, eight of them are red states.
Tightening the labor supply through deportations and things like that have potential consequences for both wage inflation and Fed policy that I'm not sure that the administration would like. Now, to be clear-- and we talk about this in multiple places-- the Biden administration oversaw the largest uncontrolled, unmanaged, and unsupervised migration surge on record.
And even the Brookings Institution at one point argued that the borders needed to be closed for some period just to absorb the immigration that had taken place. But that said, tightening labor supply too much will have consequences for wage inflation and bad policy. And then we already have Trump's inheriting large budget deficits despite almost full employment.
And then from a policy perspective, the Republicans have almost zero room for defections in the House because they've only got a 219 to 215 lead. The Gates seat will have a special election sometime in March. They'll lose a couple of seats in Florida. Bottom line is they really have no room for defections in the House.
And then I won't go into too much detail. We have a page on this. Suffice to say, the Medical Freedom Movement that RFK, Jr., supports on occasion is not a confidence builder in the concept of US exceptionalism.
So how do we see what the market verdict on this mix of headwinds and tailwinds is? My favorite barometer is the 10-year Treasury. And so far, the news is not great.
(DESCRIPTION)
A line graph appears. Title: 10 year Treasury yield change after the first Fed cut. The x axis shows measurements from 0 to 75 in increments of 5. Text: Business days since first Fed rate cut. The y axis shows negative 0.8% to positive 1.0% in increments of 0.2%. Colored lines represent September 2024, the highest, then July 1995, June 1989, September 2007, September 1998, September 2001, July 2019. Source: .Bloomberg, J P M A M, December 26, 2024
(SPEECH)
If you look at the last seven tightening-- I'm sorry-- easing cycles from the Fed in the United States going back to the 1980s, typically, what happens is the Fed starts to ease and the 10-year Treasury either stays around flat or comes down.
This is the first time in 35 years or so that the 10-year Treasury has gone up and stayed there after the first Fed cut. So
(DESCRIPTION)
A line graph appears. Title: Same from Fed fund future markets. Text: Fed funds target rate: current vs prior market expectations. The x-axis shows January 2024 to December 2025. The y axis shows 3.5% to 5.5% in increments of 0.25%. A line appears labeled December 2023 that slopes down from 5.5% to 3.5% at February 2025. And a line labeled December 26, 2024 slopes down from 4.25% to 4.0% through February 2025. Source: .Bloomberg, J P M A M, December 26, 2024
(SPEECH)
the early verdict from the bond markets on this mix of alchemist policies is not great and suggests that the productivity benefits from tax cuts and deregulation may not be enough to offset the inflationary consequences of tariffs, and deportations, and some of the other things that they're planning on doing.
So ultimately, again, the 10-year Treasury is going to be the best barometer, in my opinion, of this mix of policies. And the Fed fund futures markets are sending a similar signal. A year ago, the markets were pricing in a funds rate that would fall below 4%. That's not the case anymore. The latest Fed fund futures markets are projecting the Fed won't be able to ease more than down to 4% or so. And I'd be surprised if they even were able to get there sometime in 2025.
Equity investors are more optimistic.
(DESCRIPTION)
A line graph appears. Text: Share of US households expecting higher stock prices in 12 months, Percent of respondents. The y axis measures 15% to 60% in increments of 5% and the x axis shows 1987 to 2022. The line is up and down with an increase in 2022. Source: Conference Board Bloomberg, J P M A M, November 30, 2024
(SPEECH)
The Conference Board survey of households expecting higher stock prices in 12 months is at an all-time high, even higher than it was in the mid to late 1990s. And there are some technicals that support that. Equity supply is very tight. In other words, this looks at secondary and primary issuance net of buybacks.
(DESCRIPTION)
A bar graph shows Net Equity Supply Globally from 1999 to 2023. In US billions per year, the values are negative after 2021.
(SPEECH)
And net equity supply, particularly in the US, is actually declining. So there's a scarcity value to public equities if you can believe that. So from a technical perspective, that's supported.
(DESCRIPTION)
A bar graph shows, text: Current market valuation percentile vs history: 100% = most expensive, 0% = least expensive. The x axis is divided into three segments: US equities, non-US equities, fixed income, showing market variable at year of inception. The y axis shows 0 to 100% in increment of 10%.
(SPEECH)
But US equity markets are pretty expensive. We all know this. And at least on a valuation basis now, markets can still go up when valuations are high. But we're in the mid to upper 90s in terms of percentiles for many different ways that you could look at the valuation of US stocks. Median PEs, forward PEs, trailing PEs, the enterprise value to cash flow, price to book, price to sales, you name it-- most of the valuation indicators are on the high side. And we've also had two 20%-plus years in a row. That's only happened 10 times since 1871. And only during the 1990s bull market and the roaring '20s around 100 years ago did the good times keep going.
(DESCRIPTION)
A table shows S&P 500: Two consecutive years of 20% plus total returns and what followed.
(SPEECH)
So my takeaway from this-- and you can read this in the executive summary of The Outlook, which goes obviously into a lot more interesting detail, is I think that there's a steep curve here for the Alchemists in terms of trying this policy mix of deregulation, and tariffs, and cost cuts, and tax cuts, and crypto, and things like that. They've got a very limited margin for error, given how much of an equity one we've already had and how high valuations are.
So I expect US equity markets to end the year in 2025 higher than when they began. But I expect the Alchemists to break something. And I don't know what. And I don't think they do either. But they've got their sights set on everything from the FBI, to Medicare, to border policy, to the Defense Department, to weapons procurement. I think they're going to break something. And so I expect a correction of some kind of 10% to 15% from whatever value it's at some time in 2025. That's not that infrequent a thing. It happens in around 1/3 of all years, but I am expecting a correction to take place and then for equities to end up modestly higher by a few percent at the end of next year compared to where we are right now.
(DESCRIPTION)
A line graph appears. Text: Deregulation and business capital spending. Product market regulation index: 0 = least regulated, 6 = most regulated. The x axis shows 1987 to 2001 in 2-year increments. The left side of the y axis shows 0 to 6 in increments of 1 and the right side shows 8.5 to 12.0 in increments of 0.5. Lines show Telecommunications Act of 1986 and Electricity, both sloping downward. Source: O E C D, E T C R Database, B E A
(SPEECH)
Just to be clear, the deregulatory efforts that the incoming administration is talking about, they do have a long history of increasing business spending, capital spending, business confidence. I started writing about deregulation over 20 years ago. And here's a chart from the archives of the Eye on the Market that looked at the Clinton deregulation of telecommunications and electricity markets and the resulting increase in nonresidential business investment, which is, essentially, a capital stock increase.
Peter Orszag at Brookings has written about the benefits of deregulation on multiple occasions at Brookings. And he was, I think, either the CBO or the OMB for a while. And so I'm a believer in the benefits of deregulation. And then after the election, there was a surge in the number of small companies responding positively by saying that they're now planning on some additional capital spending over the next few months compared to what they were before the election.
(DESCRIPTION)
Line graph, text: Small business capex plans, % of N F I B Survey respondents planning capex in next 3 to 6 months, seasonally adjusted. The x axis shows years 2021 to 2024 with the y axis showing percentages of 18% to 32% in increments of 2%. Source: Bloomberg J P M A M, November 30, 2024. The line dips up and down with a sharp rise from 19 to 28% at the end.
(SPEECH)
But tariffs, and shrinking the labor supply, and some of the other issues that the alchemists are planning, I think, will offset some of this. So it's a really interesting policy mix. It's certainly not anything that anyone has tried in 70 or 80 years. And again, the executive summary of The Outlook gets into the details of this policy mix and what its implications for markets might be.
(DESCRIPTION)
Bar graph, text: Free cash flow margins by decade for the ten largest stocks, 1952-2024. Percent. The x axis shows decades 1950s to 2020s and the y axis shows negative 5% to positive 30% in increments of 5%. The increase is steady each decade. Source: Empirical Research, September 2024
(SPEECH)
We've also got a long section on AI, and adoption, and milestones, and things. I'm just going to share a few comments here. The US exceptionalism that you read about in the equity markets, you can back that up with numbers.
And I thought this was an amazing chart. It comes from Mike Goldstein and his team at Empirical Research. And it shows the free cash flow margins by decade for the 10 largest stocks. So you might think the 10 largest stocks always make a lot of money. But that's true. But the 10 largest stocks make a lot more money now than they did in the past. This goes back all the way to 1950 and shows the free cash flow margins for the 10 largest stocks in each decade. And this crop of US exceptional large companies with free cash flow margins above 25% are setting all-time highs here.
(DESCRIPTION)
Line graph, text: Earnings per share growth, rolling 2-year growth. Lines represent Mag 7 and S&P 500 ex Mag 7. The x axis shows 2017 to 2024. The y axis shows negative 50% to positive 250% in increments of 50%. Mag 7 peaks in 2022 at 250% above ex Mag 7 at 50%. Source: Bloomberg J P M A M, September 30, 2024
(SPEECH)
Now the rest of the S&P has been treading water. And when you decompose earnings per share growth of the Mag 7 versus the S&P 500 ex Mag 7, you get kind of a mixed picture here. The S&P 500 ex Mag 7 had an earnings boost that took place in 2022, but it was mostly just a reversal of the declines that took place during the pandemic.
Only the Mag 7 has been consistently generating since 2017 some pretty impressive earnings growth. So this is a very bifurcated market. And unfortunately, the implication is that very diversified investors that have been investing in, let's say, an equal weighted S&P have struggled with that.
(DESCRIPTION)
Bar graph, text: AI capital spending in context, Share of market wide capital spending
(SPEECH)
And it's really important to understand this AI boom, because we're now looking at something which has only happened, I think, twice in the last 50 years or so, which is when one specific company or a small group of companies capture a 15% or so share of all the capital spending that's taking place in the economy. I mean, that's kind of an amazing thing to think about, one company or a gaggle of companies capturing 15% of all capital spending.
And in 2026, that's what's projected to happen with NVIDIA's data center revenues are supposed to roughly be equal to 50% of all capital spending in the economy. The only thing we can compare that to is IBM's mainframe peak in the late 60s and then, obviously, the dotcom bubble in 2000.
(DESCRIPTION)
Bar graph, text: A $400b gap: AI infrastructure spending vs revenues. The y axis shows US dollars in billions from 0 to 500 in increments of 100. Measurements on the x axis are: NVIDIA year 25 revenue, Implied Data Center Spending, Required AI Revenue, Projected AI Reveue
(SPEECH)
And so we spend a lot of time looking at this particular issue of the hyperscaler revenue gap. Now what does that mean? Well, the hyperscalers-- I don't have a lot of concerns about NVIDIA. Their gross margins and everything looks fantastic. And their order book looks amazing.
The CEO has described the order book for their new Blackwell GPUs as insane. The issue is who's making all that money. Well, it's the hyperscalers-- Google, Microsoft, Meta, Apple, Oracle, Tesla, et cetera. And they eventually have to earn a return on all that GPU expenditure, capital spending.
And if you factor in their GPU purchases, the implied overall data center spending that goes with that, and the gross margins that they typically would need to earn, these companies are missing about $400 billion in annual revenue. So the markets have been very patient so far because we're in the very early stages of this AI adoption boom. But at some point, I think in the next 18 months, markets are going to start asking some very tough questions if these hyperscalers don't start earning some of this $400 billion to sustain their gross margins on all this capital spending. And our growth managers are watching this closely.
(DESCRIPTION)
Line graph, text: Hyperscaler capex and R&D as a share of revenues. The x axis shows years 2017 to 2025 and y axis shows 15 to 70% in increments of 5%. Lines represent Amazon, Alphabet, Microsoft, Meta
(SPEECH)
Hyperscaler capital spending and R&D as a share of revenues has been creeping up for the big four-- Meta, Microsoft, Alphabet, Amazon-- and starting to get into some pretty eye-popping territory where capital spending in R&D is approaching 35% of revenues.
(DESCRIPTION)
A plot shows estimated training cost and compute of select AI models by thousands of dollars.
(SPEECH)
And as we all know, these models just keep getting more expensive and energy intensive to train. And almost any chart you look at on this has to be plotted in log scale because of the amounts of time, effort, and money that are being poured into training these models.
(DESCRIPTION)
Bar graph, text: Gen AI projects, return on investment vs. expectations, % of survey respondents. Source: AlphaWise, Morgan Stanley Research, 2024
(SPEECH)
Now the good news is the surveys of companies that are ultimately going to pay the hyperscalers for all of these AI programs, we're starting to see the first crop of surveys of how those Gen AI applications are working out. And as you can see in this chart here, and we go into more detail in the piece, 90% either exceeded or met expectations. So that's a positive sign.
Some of the other surveys we walk through in the piece look at timelines of Gen AI applications. But I think it's a pretty notable thing that the early crop across multiple sectors of Gen AI projects have exceeded or met expectations of people investing in them. That's a good sign. That doesn't tell you how much money they're ultimately going to be spend on it, but it's a good sign.
(DESCRIPTION)
Line graph, text: Corning optical communications revenue, US dollars billions. The x axis shows the years 2000 to 2024 in increments of 3 years. The y axis shows 1 to 5 billion dollars in increments of 1 billion. Source: Bloomberg, company filings, J P M A M, 2024. Lines represent Real and Nominal with the nominal line being higher.
(SPEECH)
But I just want to remind everybody of this time capsule that I put together recently from our archives.
If you go back to the year 2000 and you look at Corning's order book, they built a lot of fiber. Now, ultimately, all that fiber was used. None of it was wasted. Ultimately, all that fiber got used. But because of the overbuilding, it took Corning around 20 years to get back to the nominal revenues in their optical communications business that they had in the year 2000.
And similarly, if you look at a chart on undersea fiber optic cables deployed per year, it took until 2020 to exceed the year 2000 level after a collapse.
(DESCRIPTION)
Bar graph, text: New undersea cables deployed per year
(SPEECH)
The reason I'm mentioning this is when you have an overbuilding episode, even if all that capital stock ultimately ends up getting used, sometimes it can take a long time to absorb it. And then obviously, there is a lot of implications for the companies that spent that money along the way.
(DESCRIPTION)
Line graph, text: Nuclear pure play and exposed stock returns, Total return index (100 = June 2021). The x axis shows 2021 to 2024 with lines representing Nuclear Exposed and Nuclear Pure Play. Both lines increase steadily with Nuclear Exposed showing more of an increase.
(SPEECH)
Now one of the issues that I'm fascinated by are all the green power commitments from a lot of these data centers and hyperscalers in terms of the energy intensity. Now sometimes people make too much out of this power demand coming from data centers over the next decade. I expect much bigger power demand numbers from electrification of vehicles and electrification of home heating through heat pumps, which will also be adopted by industry and commercial applications in office buildings.
And there's also a lot of different forecasts for data center demand. But one of the most incredible, meaning not believable, things about this is that nuclear power is going to play a role. I just don't see it. There's been a boom recently in the performance of nuclear exposed and nuclear pure play shares. I'm going to take the other side of this one.
(DESCRIPTION)
Line graft, text: During the 1980's, nuclear development shifted from developed to developing countries. The x axis shows year of construction start from 1950 to 2010 and the y axis shows 0 to 100% in increments of 10%. The line drops below 20% for 2010 to 2020.
(SPEECH)
Just as a bit of background, OECD nuclear power-- the OECD used to dominate nuclear power during the '50s, '60s, '70s and then started sliding in the '80s. And by the time you get to the late 1990s, and from then until now, nuclear development is essentially something that belongs to and is done in developing economies and not developed ones. And you can see why.
If you look at the cost gap here,
(DESCRIPTION)
Bar graph, text: Capital cost of nuclear plants in developed countries vs emerging markets.
(SPEECH)
I'm looking at the capital cost of four plants that were recently completed in the West because that's all there have been over the last 20 years, Flamanville in France, one in Finland, the Georgia Vogtle plant, and then Hinckley Point in the UK, all cost many multiples of the plants that are being built in Russia or China and Pakistan. Now there's lots of reasons for that. I've read all the studies on that.
I personally don't believe that we can start cranking these things out in the same way that you build solar panels or lithium ion batteries to bring their costs down. So I think people are way too excited about the possibilities of nuclear power having some kind of renaissance that contributes meaningfully to data center demand. And in The Outlook, we get into the weeds on this. Vogtle 3, by the way, in Georgia already having operational problems and it just came online.
The US lacks a lot of uranium and has import reliance on certain kinds of uranium, which is creating problems for Terrapower. There are very limited opportunities for plant restarts. And way too much was made of the Palisades reopening and the announcement and as well the one that's taking place with Three Mile Island. You can only really restart plants that just recently shut down, meaning within the last couple of years, because the decommissioning process hasn't started yet. If you start the decommissioning process, it's way too expensive to then reverse that and get that plant back online.
So at maximum, I can think of maybe two or three other shuttered plants that would be eligible for restart. The next gen designs-- everybody likes to talk about them-- sodium cooling, gas cooling, molten salt. And according to the DOE's nuclear people, at least a decade away from commercialization in the US.
And then small modular reactors-- very impressive pipeline. 6.5 gigawatts, $200 billion in orders. They're still lottery tickets. And the power commitments are not take or pay contracts. They're all contingent on final cost, as was the case with the NuScale plants that were abandoned in Utah last year. So let's take some reality pills on this nuclear renaissance. Take a look at this section. It gets into the details.
(DESCRIPTION)
A painting appears of two men on horseback in a desert setting. Title: Don Quixote and Sancho Panza, Daumier, 1865 to 1867
(SPEECH)
We also have a section on DOGE Quixote. I am envisioning Vivek Ramaswamy and Elon Musk as Don Quixote and Sancho Panza from Don Quixote. The thing to remember is that one of the underlying themes of the model is the pragmatism and humility are required for them to complete their quest. And I would question whether Vivek or Elon qualify for each of those, or either of them. Anyway.
Vaya con Dios, DOGE Quixote. I think they'll obviously get some things done, but they're going to run into a number of different buzz saws. And one of my top 10 predictions is they're not going to make it anywhere near some of the targets that they've mentioned. There are savings to be had from reversing some of Biden's recent executive actions, but those wouldn't count towards DOGE targets on reducing current spending. That would just be reducing potential future spending that hasn't really even been penciled in yet.
Non-defense discretionary spending has already been cut. Budget Control Act of 2011. It's at its lowest level in decades. So therefore they're going to have to focus on entitlements. And the DOGE billionaires are now finding out that that's a political third rail. Good luck with that.
The SEC, the Department of Labor, the Department of Education, and the EPA combined represent less than 2% of all federal workers. So fine, start firing them. But that's not going to get you very much. The vast majority of federal workers are the Postal Service and the Department of Defense.
There will be efforts underway to expand-- the CMS's is maybe expand weight loss coverage for weight loss drugs. And those are going to offset some of the DOGE savings, which takes place elsewhere.
Trump has talked about impounding or redirecting congressionally approved and congressionally appropriated spending. That's going to face judicial challenges. There was an Impoundment Control Act passed in 1974 in response to something Nixon did. A president can't, according to the Supreme Court and according to this legislation, can't just say, well, I'm going to take the money that was appropriated for the Department of Education and spend that on something else, which is what he's announced that he wants to do.
Most defense spending is people and operations on the order of 70% to 80% rather than weapons procurements. And then with respect to weapons procurements, good luck. The number of federal contractors has fallen from 51 in the early '90s to five. So it's going to be very difficult to cut those procurement costs with that much concentration in the defense contracting.
And then we also get into some of the things on improper and fraudulent payments that the government makes, sometimes accidentally. You can try to reduce them, but it's going to cost a lot of the money in terms of investing in systems and people to do that. So we get into the details here. It's interesting if you want to understand it. And good luck to all the billionaires that they've recruited onto the team, along with Vivek and Elon, to try and get this done.
(DESCRIPTION)
Line graph, text: China economic activity monitor (Bloomberg). The x axis shows 2022 to 2024 and y axis shows 1% to 9%. Text: Exports, coal and electricity consumption, steel production, motor vehicle sales, real estate investment, medicine consumption, telecom equipment production, S O E output private enterprise output. Source: Bloomberg, November 30, 2024. The line shows a decrease from 2022 to 2024.
(SPEECH)
We have a section on China that looks at both the liquidities trap and the Thucydides trap, which is something we wrote about a couple of months ago. The bottom line is that China is experiencing some kind of liquidity trap, shares some kind of characteristics with Japan in the early 1990s. As you can see here, Bluebird maintains an economic activity monitor-- exports, coal, electricity consumption, steel production, car sales, real estate investment, consumption of medicine, the kind of things that are for the Chinese government to paper over and obscure.
And the momentum on this thing is fairly negative, which is why China announced that stimulus package.
(DESCRIPTION)
Bar graph. Text: Impact on growth of major policies China has announced since September 2024, Annualized impact on growth, %. Source: Bridgewater November 2024. Rough estimate of drag shows negative 2%.
(SPEECH)
Now the stimulus package was substantial, impressive. I believe there's more fiscal stimulus coming. Bridgewater did an interesting analysis a couple of months ago that show that the total impulse from monetary and fiscal stimulus would offset roughly half of the economic drag that's taking place in China right now. That's a good effort. And the domestic A-share market kind of soared after the announcement. And it has retained its gains whereas the MSCI China index, which is a broader index of Chinese-exposed stocks, has since given up about half of the gains on the stimulus announcement.
I'm not going to go through the details here. You can read the piece.
(DESCRIPTION)
Line graph, text: M S C I China valuations, Forward price to earnings ratio. x axis shows 2006 to 2021 in 5 year increments. The y axis shows 5x to 30x in increments of 5. Source: Bloomberg J P M A M, December 27, 2024. The line spikes and dips below 12x after 2021.
(SPEECH)
I just think the days of 14 to 18 PEs on MSCI China are a thing of the past. And for geopolitical reasons, they're going to be capped at about 12x, which doesn't leave a lot of room for a rally from here. So you could take a look at the piece. We go through the details of some of the tariff and espionage issues at the core of some of the conflict issues between the US and China.
(DESCRIPTION)
Annotated line graph, text: Publicly reported Chinese espionage incidents sourced from open source materials. Events span 2000 to 2021 with a dip in 2015 then a sharp increase to 2021.
(SPEECH)
I'll just say that the increasing intensity of Chinese espionage spreading across almost every industrial sector in the US and other Western countries, it makes the idea of a rapprochement that I've read about highly unlikely. And the same goes for some of the Naval blockade exercises that China has been conducting with regards to Taiwan. So you can read more about that.
(DESCRIPTION)
A graph shows creation of new public companies in the 21st century with the US leading Europe. US companies listed include Meta, Alphabet, Tesla, Visa, Mastercard, Netflix. And Europe includes Airbus, Porsche, Equinor
(SPEECH)
We have a section on Europe called Dr. Seuss goes to Europe. It's just two pages with some Dr. Seuss style language and some charts. I just want to show you one of the charts here. We have charts in comparisons of labor productivity, and energy prices, and deindustrialization in Europe, and problems with monetary policy being applied to countries that are too different compared to the US, low labor mobility, very poor ROA and ROE numbers in Europe other than in health care, and then very negative earnings revisions.
The chart I really wanted to show you was this new one, which is the creation of new public companies on the 21st century compared the US to Europe. This kind of tells you all you need to know about the outperformance of the US versus Europe over the last 20 years. It's kind of an amazing chart to see. I can't describe it verbally. You just kind of have to look at it.
(DESCRIPTION)
Line graph, text: Crypto coin prices by segment, market cap weighted. Lines represent Bitcoin, Transactions Focused Basket, Decentralized Finance Basket. The x axis shows 2022 to 2025 and y axis shows 25 to 250 in increments of 25. The lines for bitcoin and transactions focused basket reach above 225 at the end of 2024. Source: Bloomberg, J P M A M, December 27, 2024
(SPEECH)
The crypto presidency-- I wrote a piece called "The Maltese Falcoin" a couple of years ago. It was February 2022. And I was skeptical of Bitcoin as a store of value and then even more skeptical on the transactions-focused coins and the DeFi coins. And if you're not familiar with the jargon, you can look at the Eye on the Market on this section and it gets into the details here.
I said I would take another look at Bitcoin if it fell. And it fell. It fell hard. It fell 65% after I wrote the piece. And my mistake for not looking at it again. It has since rallied pretty sharply. And a lot of the surveys of institutional investors are fairly positive on the prospects for increased digital ownership.
As it relates to the transactions-focused coins and the DeFi coins, I don't believe it. I was right about those coins for three years. Nothing much has happened in terms of actual utilization of either of those kinds of crypto coins. There's been this spectacular rally in a handful of them, Ripple in particular, but it looks highly speculative to me.
And then specifically, I've got this chart here looking at the volumes of utilization of Ripple, Litecoin, Monero, and Stellar.
(DESCRIPTION)
A line graph shows transactions per day on blockchains from 2022 to 2025.
(SPEECH)
None of them have changed since two years ago. And so if the value proposition of these coins is that people are going to use them more, you'd expect to see that happening and you don't. So that means that this is a very technical and speculative rally.
Same
(DESCRIPTION)
Line graph, text: Ethereum daily verified contracts, Total verified contracts per day, 7 day average. x axis shows 2022 to 2025, and the y axis shows 100 to 1000. The line peaks at 1,000 in 2023, dropping by 2025 to about 200.
(SPEECH)
thing for Ethereum. The number of daily verified contracts in Ethereum being executed on a blockchain are now unchanged versus what they were in early 2022. So to me, what's really happening is that people are earning a sky high return on their political donations. Crypto PACs received almost $250 million in donations, around half of all corporate donations made during the election. And Coinbase gave $75 million to a Super PAC called Fair Shake. Their candidates won almost every race they got involved in.
Look, they're playing the game and participating in the system in ways that they're allowed to do. I'm not criticizing it. I'm just saying that this, to me, is a better our explanation of the pop in crypto than the presumed proposition that their acceptance and utilization has been in changing. All that's really been changing is the politics. And the GOP Majority Leader of the House is pledging a crypto bill in the first 100 days. Trump has announced a crypto council and a crypto czar. The new SEC head will probably roll back some of the crypto accounting disclosure rules. So this is a shift in sentiment rather than a shift in utilization. And you could take a closer look.
And then just to close out here-- and thank you all for listening.
(DESCRIPTION)
A table appears. Text: 2024 predictions recap, Scorecard. Rankings appear from 1 to 10, January 1, 2024 Prediction. Highlighted green, 1, Despite all the hand-wringing over risks to reserve currency status and BRIC efforts to develop a competitive reserve currency, the US dollar remains stable, within a +/- 1 std deviation band of 7%. Highlighted green, 2, The DOJ/FTC win a big antitrust case. Highlighted green, 3, President Biden withdraws between Super Tuesday and the election, citing health reasons. Highlighted green, 4, Little to no progress on self driving cars in the US, no recovery on Lidar stock basket. Highlighted yellow, 5, Broadly syndicated leveraged losses rise above credit losses for the first time. Highlighted yellow, 6, Argentine dollarization will fail if implemented. Highlighted green, 7, Russian invasion of Ukraine drags on with no ceasefire in 2024. Highlighted green, 8, US regional bank stocks will do well in 2024 with stable or rising price to book values. Highlighted red, 9, Major US cities will face electricity outages and/or natural gas outages due to outdated grid infrastructure. Highlighted yellow, 10, Researchers will complete work on an inhaled COVID vaccine that will reduce transmission.
(SPEECH)
I started a top 10 list last year in honor of Byron Wien, who had passed away. He did this every year forever.
I made a bunch of predictions last year. Six of them were right. Three of them were kind of either too soon to tell or hard to say. And then one was wrong. I've made a new list for this year. I will tick through them very quickly and then sign off.
At least three confirmed Trump cabinet members will be gone by the end of the year. Renewables will underperform traditional energy again in 2025, like they've done in the last three years. MSCI Japan will outperform China despite China's higher PE multiple.
The discontinuation rates, which are approaching 50%, of weight loss drugs are going to hurt some of the GLP stocks, specifically Novo Nordisk. I think we'll end the year below consensus analysts price targets. Despite the order book referenced earlier, I don't think any small modular reactors in the United States will even begin construction in 2025.
I think DOGE cost cuts are going to be no more than $150 billion a year, excluding reversal of executive actions compared to targets of $2 trillion. Hedge funds I expect to continue to outperform the stock bond benchmarks. And we explain what we mean by that. But they've been doing better than expected. Housing inflation, I think because of completions, both single family and multifamily, will start to finally bring housing inflation down below 4% the way that the Fed computes it.
Blue state outmigration will continue. Over the last few years, 70% of all movers in the country in the US are moving from either blue to red or purple states. I think that's going to continue. And then lastly, I think there's a chance that the Trump administration launches a targeted military action against drug cartels in Mexico.
So that's my top 10 list. Thank you for listening. Take a look at the outlook. There's a lot in there that we can't get into in a podcast like this. And it's been a privilege. Thank you very much. Bye.
(DESCRIPTION)
Logo: J.P. Morgan
A PROPOSITO DI “EYE ON THE MARKET”
Michael Cembalest è Direttore della strategia “Market and Investment” presso la società di gestione patrimoniale J.P. Morgan. Dal 2005, Michael è autore di “Eye on the Market, podcast che affronta un’ampia serie di tematiche relative ai mercati, agli investimenti, all’economia, alla politica, al settore energetico, alle finanze municipali ecc.
Descrizione audio
Good afternoon, everybody. This is Michael Cembalest with the 2025 Eye on the Market Outlook podcast, so Happy New Year, everybody. The Outlook this year is called “The Alchemists,” for reasons that I will explain in a minute.
First, just want to mention this is the 20th anniversary of the Eye on the Market. I launched it in January of 2005. It was meant for clients that consume information on either a BlackBerry or a Palm Pilot. It didn’t have any charts or graphs or tables or footnotes or anything. It was text only. A lot of things have changed since then. It’s been an amazing journey. I’ve enjoyed working on it. I’ve learned a lot. I’d like to thank Jamie and Mary for supporting me in this effort, particularly when I write about controversial topics. And this year is no exception.
So, on to the Outlook itself. I have a section on “The Alchemists,” which is about Trump’s impact on markets; we have a section on AI adoption trends and milestones, given the Mag 7’s support to the markets; a section on nuclear power, “Wake me when we get there”; a section on DOGE called “DOGE Quixote,” which refers to the billionaire’s quest to reduce government spending; an update on China and the liquidity trap; a section called “Doctor Seuss goes to Europe”; we revisit the crypto markets, given the rally since our Maltese Falcoin piece in February 2022; and discuss the sky-high returns that donors have gotten on their crypto political donations; and then a top 10 list for 2025, following on last year's top 10 list, which was the first time we did that.
So, I'd like to thank GPT again for some fantastic cover art. This is a depiction of Trumpism as alchemism. And what’s going on here is you’ve got a guy in red hat, and he’s mixing together eight different things: deregulation, deportations, tariffs, tax cuts, government cost cutting, oil and gas, crypto, the medical freedom movement and agency purges. And then the question is posed, what could possibly go wrong? I also think we try to address what could possibly go right.
There's a lot more detail that we go through in the actual Outlook in the executive summary. Let me just give you a brief description of it here. We start with this. What are the Alchemists trying to change? If you take them at face value, what they're trying to change is this chart. And this chart shows that for the better part of 80 years, consumer spending, GDP growth and industrial production went up together like three doves flying the same direction, or geese or—I’m not a bird person.
Then, as soon as China joins the World Trade Organization in the year 2000, we get what I call the “silence of the plants”—that’s a Clarice Starling reference—where industrial production flatlines while consumer spending and GDP keep growing. There are some other ominous inflection points negative for the United States that also took place around the same time that China joined the World Trade Organization.
Rising manufacturing job losses in the U.S., a follow labor share of gross profits, rising suicide rates in the U.S. and rising non-metro, meaning rural, poverty rates in the U.S. So that’s what the Alchemists claim they’re trying to change. What kind of tailwinds are they trying to create? Well, more inbound foreign direct investment into the United States, part of which is already happening; a slowdown in the breakneck pace of regulation, which was taking place under the Biden administration, and that we've showed charts on and discussed many times; freeing up bank capital by ending the Basel III discussions, which could presumably lower the cost of credit; increasing oil and gas production to bring down energy prices, although I think when you look at the electricity prices in many parts of the country, the renewable transition is irreversible and is driving those energy prices up; and then building on a rebounding CapEx and confidence outlook; and also a recovery that we’re seeing in venture capital markets, which we spent some time on in the executive summary as well.
So, those are the tailwinds that I think it’s fair to say the Alchemists are trying to create. Now, what headwinds are the Alchemists likely to face with their set of policies? Well, as everybody has discussed, possible inflationary consequences, not just from tariffs themselves, but from the retaliation that the United States would face, with whether it was a China tariff, a universal tariff, any kind of tariff—and, by the way, of all of the ten states that import the most amount of goods relative to their GDP, eight of them are red states. Tightening the labor supply through deportations, and things like that, have potential consequences for both wage inflation and Fed policy that I’m not sure that the administration would like.
Now, to be clear, and we talk about this in multiple places, the Biden administration oversaw the largest uncontrolled, unmanaged and unsupervised migration surge on record. And even the Brookings Institution at one point argued that the borders needed to be closed for some period, just to absorb the immigration that had taken place. But that said, tightening labor supply too much will have consequences for wage inflation and bad policy.
And then we already have Trump’s inheriting large budget deficits despite almost full employment. And then from a policy perspective, the Republicans have almost zero room for defections in the House because, they’ve only got a 219 to 215 lead. The gate seat will have a special election sometime in March. They’ll lose a couple of seats in Florida. Bottom line is they really have no room for defections in the House.
And then I won't go into too much detail, we have a page on this. Suffice to say, the medical freedom movement that RFK, Jr. supports on occasion is not a confidence builder in the concept of U.S. exceptionalism.
So, how do we see what the market verdict on this mix of headwinds and tailwinds is? My favorite barometer is a 10-year Treasury. And so far, the news is not great. If you look at the last seven tightening, I’m sorry, easing cycles from the firm, the United States going back to the 1980s. Typically, what happens is the Fed starts to ease and the 10-year Treasury either stays around flat or and or comes down.
This is the first time in 34, 35 years or so that the 10-year Treasury has gone up, and stayed there after the first Fed cut. So the early verdict from the bond markets on this mix of Alchemist policies is not great, and suggests that the productivity benefits from tax cuts and deregulation may not be enough to offset the inflationary consequences of, tariffs and deportations and some of the other things that they’re planning on doing.
So ultimately, again, the 10-year Treasury is going to be the best barometer, in my opinion, of this mix of policies. And the Fed fund futures markets are sending a similar signal. A year ago, the markets were pricing in a funds rate that would fall below 4%. That's not the case anymore. The latest Fed fund futures markets are projecting the Fed won’t be able to ease more than down to 4% or so. And I’d be surprised if they even were able to get there sometime in 2025.
Equity investors are more optimistic. The Conference Board’s survey of households expecting higher stock prices in 12 months is at an all-time high, even higher than it was in in the mid to late 1990s. And there are some technicals that support that. Equity supply is very tight. In other words, this looks at secondary and primary issuance net of buybacks and net equity supply, particularly in the U.S., is actually declining.
So there’s a scarcity value to public equities, if you can believe that. So from a technical perspective, that’s supported. But U.S. equity markets are pretty expansive. We all know this, and at least on a valuation basis. Now, markets can still go up when valuations are high. But we’re in the mid to upper 90s in terms of percentiles for many different ways that you could look at the valuation of U.S. stocks. Median P/Es, forward P/Es, trailing P/Es, the enterprise value to cash flow, price to book, price to sales, you name it. Most of the valuation indicators are on the high side. And we’ve also had two 20%-plus years in a row. That’s only happened 10 times since 1871 and only during the 1990s bull market and the Roaring Twenties, around 100 years ago, did the good times keep going.
So my takeaway from this, and you can read this, in the executive summary of the Outlook, which goes obviously to a lot more interesting detail, is, I think that there’s a steep curve here for, for the Alchemists in terms of trying this policy mix of deregulation, tariffs and cost cuts and tax cuts and crypto and things like that. They’ve got a very limited margin for error given how much of an equity war we’ve already had, and how high valuations are.
So I expect U.S. equity markets to end the year 2024 are when they began. But I expect the Alchemists to break something. And I don't know what. I don’t think they do either, but they’ve got their sights set on everything from the FBI to Medicare to border policy to the Defense Department to weapons procurement.
I think they’re going to break something. And so I expect a correction of some kind, of 10–15%, from whatever value it’s at, sometime in 2025. That’s not that infrequent a thing, it happens in around one-third of all years, but I am expecting a correction to take place. And then for equities to end up modestly higher by a few percent at the end of next year compared to where we are right now.
Just to be clear, the deregulatory efforts that the incoming administration is talking about, they do have a long history of increasing business, spending, capital spending, business confidence. I started writing about deregulation over 20 years ago. And here’s a chart from the archives of the Eye on the Market that looked at the Clinton deregulation of telecommunications and electricity markets, and the resulting increase in nonresidential business investment, which is, essentially, you know, a capital stock increase.
Peter Orszag at Brookings has written about the benefits of deregulation, on multiple occasions. You know, he was the, I think, either the CBO or the OMB for a while. And so I’m a believer in the benefits of deregulation. And then after the election, there was a surge in the number of small companies responding positively by saying that they’re now planning on some additional capital spending over the next few months, compared to what they were before the election.
But, you know, tariffs and shrinking the labor supply and some of the other issues that the Alchemists are planning, I think will offset some of this. So it’s a really interesting policy mix. It’s certainly not anything that anyone has tried in 70 or 80 years. And again, you know, the executive summary of the Outlook gets into the details of this policy mix and what its implications for markets might be.
We’ve also got a long section on AI and adoption and milestones and things. I’m just going to share a few comments here. The U.S. exceptionalism that you read about in the equity markets, you can back that up with numbers. And I thought this was an amazing chart. It comes from Mike Gold and his team at Empirical Research, and it shows the free cash flow margins by decade for the 10 largest stocks.
So you might think the 10 largest stocks always make a lot of money. That’s true. But the 10 largest stocks make a lot more money now than they did in the past. This goes back all the way to 1950, and shows the free cash flow margins for the 10 largest stocks in each decade. And this crop of U.S. exceptional large companies, with free cash flow margins above 25%, are setting all-time highs here.
Now, the rest of the S&P has been treading water. And when you decompose earnings per share growth of the Mag 7 versus the S&P 500 ex Mag 7, you get kind of a mixed picture here. The S&P 500 ex Mag 7 had an earnings boost that took place in 2022, but it was mostly just a reversal of the declines that took place during the pandemic.
It’s only the Mag 7 that has been consistently generating since 2017 some pretty impressive earnings growth. So this is a very bifurcated market and, unfortunately, the implication is that very diversified investors that have been investing in, let’s say, an equal weighted S&P, have struggled with that. And it’s really important to understand this AI boom because we’re now looking at something which is only happened I think twice in the last 50 years or so, which is when one specific company, or small group of companies, capture a 15% or so share of all the capital spending that’s taking place in the economy.
I mean, that’s kind of an amazing thing to think about—one company or a gaggle of companies capturing 50% of all capital spending. In 2026, that's what’s projected to happen with Nvidia’s data center revenues that are supposed to roughly be equal to 15% of all capital spending in the economy. The only thing we can compare that to is IBM’s mainframe peak in the late 60s. And then obviously the Dot Com bubble in 2000.
And so we spend a lot of time looking at this particular issue of the hyperscaler revenue gap. Now, what does that mean? Well, I don't have a lot of concerns about Nvidia. Their gross margins and everything looks fantastic and their order book looks amazing. The CEO described the order book for their new Blackwell GPUs as “insane.” The issue is, who's making all that money? Well, it's the hyperscalers: Google, Microsoft, meta, Apple, Oracle, Tesla, etc. And they eventually have to earn a return on all that GPU expenditure, capital spending. And if you factor in their GPU purchases, the implied overall data center spending that goes with that and the gross margins that they typically would need to earn, these companies are missing about $400 billion in annual revenue.
So the markets have been very patient so far because we're in the very early stages of this AI adoption boom. But at some point, I think the next 18 months, markets are going to start asking some very tough questions if these hyperscalers don't start earning some of this $400 billion to sustain their gross margins, all this capital spending. And our growth managers are watching this closely.
Hyperscale capital spending in R&D as a share of revenues has been creeping up for the Big 4: Meta, Microsoft, Alphabet and Amazon. Starting to get into pretty, pretty eye popping territory where our capital spending on R&D approaching 35% on revenues. And as we all know, these models just keep getting more expensive and energy intensive to train, and almost any chart you look at on this has to be plotted in log scale because of the amounts of time, effort and money that have been poured into training these models.
Now, the good news is the surveys of companies that are ultimately going to pay the hyperscalers for all of these AI programs, we're starting to see the first crop of surveys of how those gen AI applications are working out. And, as you can see in this chart here, and we go into more detail in the piece, 90% either exceeded or met expectations. So that's a positive sign. Some of the other surveys we’ll walk through in the piece look at timelines of AI applications. But I think it’s a pretty notable thing that the early crop across multiple sectors of gen AI projects have exceeded or met expectations of people investing, and that’s a good sign.
That doesn’t tell you how much money they’re ultimately going to be spending on it. But it's a good sign. But I just want to remind everybody of this time capsule that I put together recently from our archives. If you go back to the year 2000 and you'll look at Corning’s order book, they built a lot of fiber. Now, ultimately, all that fiber was used, right? None of it was wasted. Ultimately, all that fiber got used. But because of the overbuilding, it took Corning around 20 years to get back to the nominal revenues and their optical communications business that they had in the year 2000. And similarly, if you look at a chart on undersea fiber optic cables deployed per year, it took until 2020 to exceed the 20 the year 2000 level after collapse.
The reason I'm mentioning this is when you have an overbuilding episode, even of all that capital stock ultimately ends up getting used, sometimes it can take a long time to absorb it, and then, obviously, there were a lot of implications for the companies that spent that money along the way.
Now, one of the issues that I’m fascinated by are all the green power commitments from a lot of these data centers and hyperscalers, in terms of the energy intensity. Now, you know, the sometimes people make too much out of this power demand coming from data centers over the next decade. I expect much bigger power demand numbers from electrification of vehicles and electrification of home heating through heat pumps, which will also be adopted by industry and commercial applications and office buildings. And there’s also a lot of different forecasts for data center demand.
But one of the most incredible—meaning, not believable—things about this, is that nuclear power is going to play a role. I just don’t see it. There’s been a boom recently in the performance of nuclear exposed and nuclear pure-play shares. I’m going to take the other side of this one.
Just as a bit of background, OECD nuclear power. You know, the OECD used to dominate nuclear power during the 50, 60s, 70s and then started sliding in the 80s. And by the time you get to the late 1990s, and from then until now, nuclear development is essentially something that belongs to and is done in developing economies and not developed ones. And you can see why if you look at the cost gap here, I’m looking at the capital cost of four plants that were recently completed in the West, because that’s all there have been over the last 20 years, Flamanville in France, one in Finland, the Georgia Vogtle plant and then Hinkley Point in the UK, all cost many multiples of the plants that are being built in Russia, China and Pakistan.
Now, there's lots of reasons for that. I’ve read whole studies on that. I personally don’t believe that we can start cranking these things out in the same way that you build solar panels or lithium ion batteries to bring their cost down. So I think people are way too excited about the possibilities of nuclear power having some kind of renaissance that contributes meaningfully to data center demand.
In the Outlook we get into the weeds on this. Vogtle 3, by the way, in Georgia, is already having operational problems, it just came online. The U.S. lacks a lot of the uranium and it and it has import reliance on certain kinds of uranium, which is creating problems for Terrapower. There are very limited opportunities for plant restarts. And way too much was made of the Palisades reopening and the announcement and as well, the one that’s taking place with Three Mile Island. You can only really restart plants that just recently shut down, meaning within the last couple of years, because the decommissioning process hasn't started yet. If you start the decommissioning process, it’s way too expensive to then reverse that and get that clamp back on line. So at maximum, I can think of maybe two or three other shuttered plants that would be eligible for restart, the next gen designs.
Everybody likes to talk about sodium cooling, gas cooling, molten salt. According to the DOE’s nuclear people, they’re at least a decade away from commercialization in the U.S. And then, small modular reactors. Very impressive pipelines. 6.5 GW. 200 billion in orders. They’re still lottery tickets. And the power commitments are not take or pay contracts. They’re all contingent on final cost, as was the case with the new scale plants that were abandoned in Utah last year. So let’s, you know, take some reality pills on this nuclear renaissance. Take a look at this section. It gets into the details.
We also have a section on DOGE Quixote. I am envisioning Vivek Ramaswamy and Elon Musk as Don Quixote and Sancho Panza from Don Quixote. The thing to remember is that the one of the underlying themes of the model is the pragmatism and humility are required for them to complete their quest. And I would question whether Vivek or Elon qualify for each of those, or either of them. Vaya con Dios, Doge Quixote. I think they’ll obviously get some things done, but they’re going to run into a number of different buzzsaws. And one of my top 10 predictions is they’re not going to make it anywhere near some of the targets that they’ve mentioned.
There are savings to be had from reversing some of Biden’s recent executive actions, but those wouldn't count towards DOGE targets on resisting current spending. That would just be reducing potential future spending, that hasn’t really even been penciled in yet. Non-defense discretionary spending has already been cut. Budget Control Act of 2011. It’s at its lowest level in decades. So therefore, they’re going to have to focus on entitlements. And the DOGE billionaires are now finding out that that’s a political third rail. Good luck with that.
The SEC, the Department of Labor, the Department of Education and the EPA combined represent less than 2% of all federal workers. So fine, start firing them. But that’s not going to get you very much. The vast majority of federal workers are, you know, the Postal Service and the Department of Defense.
There will be efforts underway to expand the CMS, that is, maybe expand, weight loss coverage for weight loss drugs. And those are going to offset some of the DOGE savings, which takes place elsewhere. Trump has talked about impounding or redirecting congressionally approved and congressionally appropriated spending. That’s going to face judicial challenges. There was an Impoundment Control Act passed in 1974, in response to something Nixon did. A president can’t, according to the Supreme Court and according to this legislation, can’t just say, well, I’m going to take the money that was appropriated for the Department of Education and spend it on something else, which is what he’s announced that he wants to do.
Most defense spending is people and operations on the order of 70 to 80% rather than weapons procurements. And then, with respect to weapons procurements, good luck. You know, the number of federal contractors has fallen from 51 in the early 90s to five. So it’s going to be very difficult to cut those procurement costs with that much concentration in the defense contract. And then, we also get into some of the things on improper and fraudulent payments that the government makes, sometimes accidentally, you can try to reduce them, but it’s going to cost a lot of money in terms of investing in systems. And people will do that. So, we get into the details there. It’s interesting if you want to understand it and good luck to all the billionaires that they've recruited to the team, along with Vivek and Elon, to try and get this done.
We have a section on China that looks at both the liquidity trap and the Thucydides trap, which is something we wrote about a couple of months ago. The bottom line is that China is experiencing some kind of liquid illiquidity trap that shares some kind of characteristics with Japan in the early 1990s. As you can see here, a Bloomberg maintains an economic activity monitor—exports, coal and oil consumption, steel production, car sales, real estate investment, consumption of medicine—the kind of things that are harder for the Chinese government to paper over and obscure.
And, you know, the momentum on this thing is fairly negative, which is why China announced that stimulus package. Now, the stimulus package, that was substantial, impressive—I believe there’s more fiscal stimulus coming. Bridgewater did an interesting analysis a couple months ago that showed that the total impulse from monetary and fiscal stimulus would offset roughly half of the economic drag that’s taking place in China right now. That’s a good effort. And the domestic A-share market kind of soared after the announcement and has retained its gains, whereas the MSCI China index, which is a broader index of Chinese exposed stocks, has since given up about half of the gains, now the stimulus was announced. But I’m not going to go through the details here. You can read the piece. I just think the days of 14 to 18 P/Es on MSCI China are a thing of the past and for geopolitical reasons, they’re going to be capped at about 12x, which doesn’t leave a lot of room for a rally from here.
So you could take a look at the piece. We go through the details of some of the tariff and espionage issues at the core of some of the conflict issues between the U.S. and China. I’ll just say that the increasing intensity of Chinese espionage spreading across almost every industrial sector in the U.S. and other Western countries, it makes the idea of a rapprochement that I’ve read about highly unlikely, and the same goes for some of the naval blockade exercises that China has been conducting, with regards to Taiwan. So you can read more about that.
We have a section on Europe called “Doctor Seuss goes to Europe.” It’s just two pages with up with some Doctor Seuss–style language and some charts. I just want to show you one of the charts here. We have charts and comparisons of labor productivity and energy prices and deindustrialization in Europe, and problems with monetary policy being applied to countries that are two different compared to the U.S. low labor mobility, very poor ROA and ROE numbers in Europe other than in health care, and then very negative earnings revisions.
The chart I really wanted to show you was this new one, which is the creation of new public companies in the 21st century comparing the U.S. to Europe. This kind of tells you all you need to know about the outperformance of the U.S. versus Europe over the last 20 years. It's kind of an amazing chart to see. I can’t like, describe it verbally. You just kind of have to look at it.
The crypto presidency. I wrote a piece called The Maltese Falcoin a couple of years ago. It was February 2022, and I was skeptical of Bitcoin as a store value and then even more skeptical on the transactions-focused coins and the DeFi coins. and if you’re not familiar with the jargon, you can look at the Eye on the Market on this section and it gets into the details here. But, I said I would take another look at Bitcoin if it fell. And it fell, it fell hard and it fell 65% after I wrote the piece. And my mistake for not looking at it again. It has since rallied pretty sharply. And a lot of the surveys of institutional investors are fairly positive on the prospects for increased digital ownership as it relates to the transactions-focused coins and the DeFi coins. I don't believe it. I was right that about those coins for three years. Nothing much has happened in terms of actual utilization of either of those kinds of crypto coins.
There’s been this spectacular rally and in a handful of that ripple in particular, but it looks highly speculative to me. And then, specifically, I’ve got this chart here looking at the volumes of utilization of ripple like coin, Monero and Stellar. None of them have changed since two years ago. Right. And so if the value proposition of these coins is that people are going to use them more, you expect to see that happening, and you don't. So that that means that this is a very tactical and speculative rally. Same thing for Ethereum. The number of daily verified contracts in Ethereum and being executed on a blockchain are now unchanged versus what they were in early 2022. So to me, what’s really happening is that people are earning a sky-high return on their political donations.
Crypto PACs received almost $250 million in donations, around half of all corporate donations made during the election. And Coinbase gave $75 million to a super PAC called Fair Shake. Their candidates won almost every race they got involved in. Look, they're playing the game and participating in the system in ways that they’re allowed to do that, I’m not criticizing it.
I'm just saying that this, to me, is a better explanation for the pop in crypto than the presumed proposition that their acceptance and utilization has been changing. All that’s really been changing is the politics. And, you know, the GOP majority of the year of the House is pledging a pro-crypto bill in the first hundred days that Trump has announced a crypto council and a crypto czar. The new SEC head will probably roll back some of the crypto accounting disclosure rules. So this is a shift in sentiment, rather than a shift in utilization. And you could take a closer look.
And then, just to close out here and thank you all for listening. I started a top 10 list last year in honor of Byron Wiem, who passed away. He did this every year for forever. I made a bunch of predictions last year. Six of them were right. Three of them were kind of, you know, either too soon to tell, hard to say. And then one was wrong. I’ve made a new list for this year. I will tick through them very quickly and sign off.
At least three confirmed Trump cabinet members will be gone by the end of the year. Renewables will underperform traditional energy again in 2025, like they’ve done in the last three years. MSCI Japan will outperform China despite China’s higher P/E multiple. The discontinuation rates, which are approaching 50% of weight loss drugs, are going to hurt some of the GLP stocks, specifically Novo Nordisk, I think will end the year below consensus. Our price targets, despite the order book referenced earlier, I don’t think any small modular reactors in the United States will even begin construction in 2025. I think DOGE cost cuts are going to be no more than $150 billion a year, excluding reversal of executive actions compared to targets of $2 trillion. Hedge funds I expect to continue to outperform the stock bond benchmarks, and we explain what we mean by that. But you know, they’ve been doing better than expected. Housing inflation, I think because of completions, both single family and multifamily will start to finally bring housing inflation down below 4%, the way that the Fed computes it. Migration will continue. Over the last few years, 70% of all movers in the country in the U.S. are moving from either blue to red or purple states. I think that’s going to continue. And then lastly, I think there’s a chance that the Trump administration launches a targeted military action against drug cartels in Mexico. So that's my top ten list, thank you for listening.
Take a look at the Outlook. There’s a lot in there that we can't get into in a podcast like this. And it’s been a privilege. Thank you very much. Bye.
(DESCRIPTION)
Logo: J.P. Morgan, Title: Eye on the Market
An image at left shows a scientist in a lab hovering over colorful vials. He wears a red cap with two letters visible: G, A. A plume of bright magenta smoke fills the air. Title: Outlook 2025, The Alchemists.
(SPEECH)
[WHOOSHING]
(DESCRIPTION)
The speaker, Michael Cembalest, has shortly cropped brown hair and wears glasses. He appears in an office space with a window behind him.
(SPEECH)
Good afternoon, everybody. This is Michael Cembalest with the 2025 Eye on the Market Outlook podcast. A Happy New year, everybody. The outlook this year is called The Alchemists for reasons that I will explain in a minute.
(DESCRIPTION)
Images of a Blackberry and Palm Pilot appear at left. Text: Eye on the Market launched January 2005
(SPEECH)
First, I just want to mention this is the 20th anniversary of the Eye on the Market. I launched it in January of 2005. It was meant for clients that consume information on either a BlackBerry or a Palm Pilot. It didn't have any charts, or graphs, or tables, or footnotes, or anything. It was text only. A lot of things have changed since then. It's been an amazing journey. I've enjoyed working on it. I've learned a lot. I'd like to thank Jamie and Mary for supporting me in this effort, particularly when I write about controversial topics. And this year is no exception.
So on to the outlook itself.
(DESCRIPTION)
Text: 2025 Eye on the Market Outlook. The Alchemists deregulation, deportations, tariffs, tax cuts, cost cutting, oil & gas, crypto, medical freedom, and Agency purges. What could possibly go wrong? The Golden Goose: AI adoption, trends and milestones. What nuclear renaissance? Wake me when we get there. DOGE Quixote, the billionaires quest to reduce US government spending. China: the liquidity trap and the Thucydides trap. Dr. Seuss Goes to Europe. The crypto Presidency: on the crypto rally and the sky high return on political donations. Top Ten List for 2025 and a scorecard on 2024's list.
(SPEECH)
I have a section on The Alchemists, which is about Trump's impact on markets. We have a section on AI adoption, trends and milestones given the Mag 7's importance to the markets; a section on nuclear power-- wake me when we get there; a section on DOGE called DOGE Quixote, which refers to the billionaire's quest to reduce government spending; an update on China and the liquidity trap; a section called Dr. Seuss Goes to Europe. We revisit the crypto markets given the rally since our Maltese Falcon piece in February of 2022 and discuss the sky high returns that donors have gotten on their crypto political donations and then a top 10 list for 2025 following on last year's top 10 list, which was the first time we did that.
Anyway, so I'd like to thank GPT again for some fantastic cover art. This is a depiction of Trumpism as alchemism. And what's going on here is you've got a guy in a red hat. And he's mixing together eight different things-- deregulation, deportations, tariffs, tax cuts, government cost cutting, oil and gas, crypto, the medical freedom movement, and agency purges. And then the question is posed, what could possibly go wrong? I also think we try to address what could possibly go right.
There's a lot more detail that we go through in the actual outlook in the executive summary. Let me just give you a brief description of it here.
(DESCRIPTION)
Title: What are the Alchemists trying to change? A line graph appears that shows US Consumer Spending ad US Industrial Production from 1930 to 2020. They stay together until 2000, when spending exceeds production by 60 points. Text: Other US data with sharp inflection points around the year 2000: Rising manufacturing job losses, Falling labor share of gross profits, Rising suicide rates, Rising non-metro poverty rates. Source: B E A Federal Reserve, J P M A M 2024.
(SPEECH)
We start with this. What are the alchemists trying to change if you take them at face value? What they're trying to change is this chart. And this chart shows that for the better part of 80 years, consumer spending, GDP growth, and industrial production went up together like three doves flying in the same direction or geese. I'm not a bird person.
Then as soon as China joins the World Trade Organization in the year 2000, we get what's called-- what I call the silence of the plants-- that's a Clarice Starling reference-- where industrial production flatlines while consumer spending and GDP keep growing. There are some other ominous inflection points, negative for the United States, that also took place around the same time that China joined the World Trade Organization-- rising manufacturing job losses in the US, a fall of labor share of gross profits, rising suicide rates in the US, and rising non-metro, meaning rural, poverty rates in the US. So that's what the alchemists claim they're trying to change.
What kind of tailwinds are they trying to create? Well, more inbound foreign direct investment into the United States, part of which is already happening; a slowdown in the breakneck pace of regulation, which was taking place under the Biden administration that we've showed charts on and discussed many times; freeing up bank capital by ending the Basel III discussions, which could presumably lower the cost of credit; increasing oil and gas production to bring down energy prices, although I think when you look at the electricity prices in many parts of the country, the renewable transition is irreversible and is driving those energy prices up; and then building on a rebounding CapEx and confidence outlook; and also recovery that we're seeing in venture capital markets, which we spent some time on in the executive summary as well. So those are the tailwinds that I think it's fair to say the alchemists are trying to create.
Now what headwinds are the alchemists likely to face with their set of policies? Well, as everybody has discussed, possible inflationary consequences, not just from tariffs themselves, but from the retaliation that the United States would face with whether it was a China tariff, a universal tariff, any kind of tariff. And by the way, of the 10 states that import the most amount of goods relative to their GDP, eight of them are red states.
Tightening the labor supply through deportations and things like that have potential consequences for both wage inflation and Fed policy that I'm not sure that the administration would like. Now, to be clear-- and we talk about this in multiple places-- the Biden administration oversaw the largest uncontrolled, unmanaged, and unsupervised migration surge on record.
And even the Brookings Institution at one point argued that the borders needed to be closed for some period just to absorb the immigration that had taken place. But that said, tightening labor supply too much will have consequences for wage inflation and bad policy. And then we already have Trump's inheriting large budget deficits despite almost full employment.
And then from a policy perspective, the Republicans have almost zero room for defections in the House because they've only got a 219 to 215 lead. The Gates seat will have a special election sometime in March. They'll lose a couple of seats in Florida. Bottom line is they really have no room for defections in the House.
And then I won't go into too much detail. We have a page on this. Suffice to say, the Medical Freedom Movement that RFK, Jr., supports on occasion is not a confidence builder in the concept of US exceptionalism.
So how do we see what the market verdict on this mix of headwinds and tailwinds is? My favorite barometer is the 10-year Treasury. And so far, the news is not great.
(DESCRIPTION)
A line graph appears. Title: 10 year Treasury yield change after the first Fed cut. The x axis shows measurements from 0 to 75 in increments of 5. Text: Business days since first Fed rate cut. The y axis shows negative 0.8% to positive 1.0% in increments of 0.2%. Colored lines represent September 2024, the highest, then July 1995, June 1989, September 2007, September 1998, September 2001, July 2019. Source: .Bloomberg, J P M A M, December 26, 2024
(SPEECH)
If you look at the last seven tightening-- I'm sorry-- easing cycles from the Fed in the United States going back to the 1980s, typically, what happens is the Fed starts to ease and the 10-year Treasury either stays around flat or comes down.
This is the first time in 35 years or so that the 10-year Treasury has gone up and stayed there after the first Fed cut. So
(DESCRIPTION)
A line graph appears. Title: Same from Fed fund future markets. Text: Fed funds target rate: current vs prior market expectations. The x-axis shows January 2024 to December 2025. The y axis shows 3.5% to 5.5% in increments of 0.25%. A line appears labeled December 2023 that slopes down from 5.5% to 3.5% at February 2025. And a line labeled December 26, 2024 slopes down from 4.25% to 4.0% through February 2025. Source: .Bloomberg, J P M A M, December 26, 2024
(SPEECH)
the early verdict from the bond markets on this mix of alchemist policies is not great and suggests that the productivity benefits from tax cuts and deregulation may not be enough to offset the inflationary consequences of tariffs, and deportations, and some of the other things that they're planning on doing.
So ultimately, again, the 10-year Treasury is going to be the best barometer, in my opinion, of this mix of policies. And the Fed fund futures markets are sending a similar signal. A year ago, the markets were pricing in a funds rate that would fall below 4%. That's not the case anymore. The latest Fed fund futures markets are projecting the Fed won't be able to ease more than down to 4% or so. And I'd be surprised if they even were able to get there sometime in 2025.
Equity investors are more optimistic.
(DESCRIPTION)
A line graph appears. Text: Share of US households expecting higher stock prices in 12 months, Percent of respondents. The y axis measures 15% to 60% in increments of 5% and the x axis shows 1987 to 2022. The line is up and down with an increase in 2022. Source: Conference Board Bloomberg, J P M A M, November 30, 2024
(SPEECH)
The Conference Board survey of households expecting higher stock prices in 12 months is at an all-time high, even higher than it was in the mid to late 1990s. And there are some technicals that support that. Equity supply is very tight. In other words, this looks at secondary and primary issuance net of buybacks.
(DESCRIPTION)
A bar graph shows Net Equity Supply Globally from 1999 to 2023. In US billions per year, the values are negative after 2021.
(SPEECH)
And net equity supply, particularly in the US, is actually declining. So there's a scarcity value to public equities if you can believe that. So from a technical perspective, that's supported.
(DESCRIPTION)
A bar graph shows, text: Current market valuation percentile vs history: 100% = most expensive, 0% = least expensive. The x axis is divided into three segments: US equities, non-US equities, fixed income, showing market variable at year of inception. The y axis shows 0 to 100% in increment of 10%.
(SPEECH)
But US equity markets are pretty expensive. We all know this. And at least on a valuation basis now, markets can still go up when valuations are high. But we're in the mid to upper 90s in terms of percentiles for many different ways that you could look at the valuation of US stocks. Median PEs, forward PEs, trailing PEs, the enterprise value to cash flow, price to book, price to sales, you name it-- most of the valuation indicators are on the high side. And we've also had two 20%-plus years in a row. That's only happened 10 times since 1871. And only during the 1990s bull market and the roaring '20s around 100 years ago did the good times keep going.
(DESCRIPTION)
A table shows S&P 500: Two consecutive years of 20% plus total returns and what followed.
(SPEECH)
So my takeaway from this-- and you can read this in the executive summary of The Outlook, which goes obviously into a lot more interesting detail, is I think that there's a steep curve here for the Alchemists in terms of trying this policy mix of deregulation, and tariffs, and cost cuts, and tax cuts, and crypto, and things like that. They've got a very limited margin for error, given how much of an equity one we've already had and how high valuations are.
So I expect US equity markets to end the year in 2025 higher than when they began. But I expect the Alchemists to break something. And I don't know what. And I don't think they do either. But they've got their sights set on everything from the FBI, to Medicare, to border policy, to the Defense Department, to weapons procurement. I think they're going to break something. And so I expect a correction of some kind of 10% to 15% from whatever value it's at some time in 2025. That's not that infrequent a thing. It happens in around 1/3 of all years, but I am expecting a correction to take place and then for equities to end up modestly higher by a few percent at the end of next year compared to where we are right now.
(DESCRIPTION)
A line graph appears. Text: Deregulation and business capital spending. Product market regulation index: 0 = least regulated, 6 = most regulated. The x axis shows 1987 to 2001 in 2-year increments. The left side of the y axis shows 0 to 6 in increments of 1 and the right side shows 8.5 to 12.0 in increments of 0.5. Lines show Telecommunications Act of 1986 and Electricity, both sloping downward. Source: O E C D, E T C R Database, B E A
(SPEECH)
Just to be clear, the deregulatory efforts that the incoming administration is talking about, they do have a long history of increasing business spending, capital spending, business confidence. I started writing about deregulation over 20 years ago. And here's a chart from the archives of the Eye on the Market that looked at the Clinton deregulation of telecommunications and electricity markets and the resulting increase in nonresidential business investment, which is, essentially, a capital stock increase.
Peter Orszag at Brookings has written about the benefits of deregulation on multiple occasions at Brookings. And he was, I think, either the CBO or the OMB for a while. And so I'm a believer in the benefits of deregulation. And then after the election, there was a surge in the number of small companies responding positively by saying that they're now planning on some additional capital spending over the next few months compared to what they were before the election.
(DESCRIPTION)
Line graph, text: Small business capex plans, % of N F I B Survey respondents planning capex in next 3 to 6 months, seasonally adjusted. The x axis shows years 2021 to 2024 with the y axis showing percentages of 18% to 32% in increments of 2%. Source: Bloomberg J P M A M, November 30, 2024. The line dips up and down with a sharp rise from 19 to 28% at the end.
(SPEECH)
But tariffs, and shrinking the labor supply, and some of the other issues that the alchemists are planning, I think, will offset some of this. So it's a really interesting policy mix. It's certainly not anything that anyone has tried in 70 or 80 years. And again, the executive summary of The Outlook gets into the details of this policy mix and what its implications for markets might be.
(DESCRIPTION)
Bar graph, text: Free cash flow margins by decade for the ten largest stocks, 1952-2024. Percent. The x axis shows decades 1950s to 2020s and the y axis shows negative 5% to positive 30% in increments of 5%. The increase is steady each decade. Source: Empirical Research, September 2024
(SPEECH)
We've also got a long section on AI, and adoption, and milestones, and things. I'm just going to share a few comments here. The US exceptionalism that you read about in the equity markets, you can back that up with numbers.
And I thought this was an amazing chart. It comes from Mike Goldstein and his team at Empirical Research. And it shows the free cash flow margins by decade for the 10 largest stocks. So you might think the 10 largest stocks always make a lot of money. But that's true. But the 10 largest stocks make a lot more money now than they did in the past. This goes back all the way to 1950 and shows the free cash flow margins for the 10 largest stocks in each decade. And this crop of US exceptional large companies with free cash flow margins above 25% are setting all-time highs here.
(DESCRIPTION)
Line graph, text: Earnings per share growth, rolling 2-year growth. Lines represent Mag 7 and S&P 500 ex Mag 7. The x axis shows 2017 to 2024. The y axis shows negative 50% to positive 250% in increments of 50%. Mag 7 peaks in 2022 at 250% above ex Mag 7 at 50%. Source: Bloomberg J P M A M, September 30, 2024
(SPEECH)
Now the rest of the S&P has been treading water. And when you decompose earnings per share growth of the Mag 7 versus the S&P 500 ex Mag 7, you get kind of a mixed picture here. The S&P 500 ex Mag 7 had an earnings boost that took place in 2022, but it was mostly just a reversal of the declines that took place during the pandemic.
Only the Mag 7 has been consistently generating since 2017 some pretty impressive earnings growth. So this is a very bifurcated market. And unfortunately, the implication is that very diversified investors that have been investing in, let's say, an equal weighted S&P have struggled with that.
(DESCRIPTION)
Bar graph, text: AI capital spending in context, Share of market wide capital spending
(SPEECH)
And it's really important to understand this AI boom, because we're now looking at something which has only happened, I think, twice in the last 50 years or so, which is when one specific company or a small group of companies capture a 15% or so share of all the capital spending that's taking place in the economy. I mean, that's kind of an amazing thing to think about, one company or a gaggle of companies capturing 15% of all capital spending.
And in 2026, that's what's projected to happen with NVIDIA's data center revenues are supposed to roughly be equal to 50% of all capital spending in the economy. The only thing we can compare that to is IBM's mainframe peak in the late 60s and then, obviously, the dotcom bubble in 2000.
(DESCRIPTION)
Bar graph, text: A $400b gap: AI infrastructure spending vs revenues. The y axis shows US dollars in billions from 0 to 500 in increments of 100. Measurements on the x axis are: NVIDIA year 25 revenue, Implied Data Center Spending, Required AI Revenue, Projected AI Reveue
(SPEECH)
And so we spend a lot of time looking at this particular issue of the hyperscaler revenue gap. Now what does that mean? Well, the hyperscalers-- I don't have a lot of concerns about NVIDIA. Their gross margins and everything looks fantastic. And their order book looks amazing.
The CEO has described the order book for their new Blackwell GPUs as insane. The issue is who's making all that money. Well, it's the hyperscalers-- Google, Microsoft, Meta, Apple, Oracle, Tesla, et cetera. And they eventually have to earn a return on all that GPU expenditure, capital spending.
And if you factor in their GPU purchases, the implied overall data center spending that goes with that, and the gross margins that they typically would need to earn, these companies are missing about $400 billion in annual revenue. So the markets have been very patient so far because we're in the very early stages of this AI adoption boom. But at some point, I think in the next 18 months, markets are going to start asking some very tough questions if these hyperscalers don't start earning some of this $400 billion to sustain their gross margins on all this capital spending. And our growth managers are watching this closely.
(DESCRIPTION)
Line graph, text: Hyperscaler capex and R&D as a share of revenues. The x axis shows years 2017 to 2025 and y axis shows 15 to 70% in increments of 5%. Lines represent Amazon, Alphabet, Microsoft, Meta
(SPEECH)
Hyperscaler capital spending and R&D as a share of revenues has been creeping up for the big four-- Meta, Microsoft, Alphabet, Amazon-- and starting to get into some pretty eye-popping territory where capital spending in R&D is approaching 35% of revenues.
(DESCRIPTION)
A plot shows estimated training cost and compute of select AI models by thousands of dollars.
(SPEECH)
And as we all know, these models just keep getting more expensive and energy intensive to train. And almost any chart you look at on this has to be plotted in log scale because of the amounts of time, effort, and money that are being poured into training these models.
(DESCRIPTION)
Bar graph, text: Gen AI projects, return on investment vs. expectations, % of survey respondents. Source: AlphaWise, Morgan Stanley Research, 2024
(SPEECH)
Now the good news is the surveys of companies that are ultimately going to pay the hyperscalers for all of these AI programs, we're starting to see the first crop of surveys of how those Gen AI applications are working out. And as you can see in this chart here, and we go into more detail in the piece, 90% either exceeded or met expectations. So that's a positive sign.
Some of the other surveys we walk through in the piece look at timelines of Gen AI applications. But I think it's a pretty notable thing that the early crop across multiple sectors of Gen AI projects have exceeded or met expectations of people investing in them. That's a good sign. That doesn't tell you how much money they're ultimately going to be spend on it, but it's a good sign.
(DESCRIPTION)
Line graph, text: Corning optical communications revenue, US dollars billions. The x axis shows the years 2000 to 2024 in increments of 3 years. The y axis shows 1 to 5 billion dollars in increments of 1 billion. Source: Bloomberg, company filings, J P M A M, 2024. Lines represent Real and Nominal with the nominal line being higher.
(SPEECH)
But I just want to remind everybody of this time capsule that I put together recently from our archives.
If you go back to the year 2000 and you look at Corning's order book, they built a lot of fiber. Now, ultimately, all that fiber was used. None of it was wasted. Ultimately, all that fiber got used. But because of the overbuilding, it took Corning around 20 years to get back to the nominal revenues in their optical communications business that they had in the year 2000.
And similarly, if you look at a chart on undersea fiber optic cables deployed per year, it took until 2020 to exceed the year 2000 level after a collapse.
(DESCRIPTION)
Bar graph, text: New undersea cables deployed per year
(SPEECH)
The reason I'm mentioning this is when you have an overbuilding episode, even if all that capital stock ultimately ends up getting used, sometimes it can take a long time to absorb it. And then obviously, there is a lot of implications for the companies that spent that money along the way.
(DESCRIPTION)
Line graph, text: Nuclear pure play and exposed stock returns, Total return index (100 = June 2021). The x axis shows 2021 to 2024 with lines representing Nuclear Exposed and Nuclear Pure Play. Both lines increase steadily with Nuclear Exposed showing more of an increase.
(SPEECH)
Now one of the issues that I'm fascinated by are all the green power commitments from a lot of these data centers and hyperscalers in terms of the energy intensity. Now sometimes people make too much out of this power demand coming from data centers over the next decade. I expect much bigger power demand numbers from electrification of vehicles and electrification of home heating through heat pumps, which will also be adopted by industry and commercial applications in office buildings.
And there's also a lot of different forecasts for data center demand. But one of the most incredible, meaning not believable, things about this is that nuclear power is going to play a role. I just don't see it. There's been a boom recently in the performance of nuclear exposed and nuclear pure play shares. I'm going to take the other side of this one.
(DESCRIPTION)
Line graft, text: During the 1980's, nuclear development shifted from developed to developing countries. The x axis shows year of construction start from 1950 to 2010 and the y axis shows 0 to 100% in increments of 10%. The line drops below 20% for 2010 to 2020.
(SPEECH)
Just as a bit of background, OECD nuclear power-- the OECD used to dominate nuclear power during the '50s, '60s, '70s and then started sliding in the '80s. And by the time you get to the late 1990s, and from then until now, nuclear development is essentially something that belongs to and is done in developing economies and not developed ones. And you can see why.
If you look at the cost gap here,
(DESCRIPTION)
Bar graph, text: Capital cost of nuclear plants in developed countries vs emerging markets.
(SPEECH)
I'm looking at the capital cost of four plants that were recently completed in the West because that's all there have been over the last 20 years, Flamanville in France, one in Finland, the Georgia Vogtle plant, and then Hinckley Point in the UK, all cost many multiples of the plants that are being built in Russia or China and Pakistan. Now there's lots of reasons for that. I've read all the studies on that.
I personally don't believe that we can start cranking these things out in the same way that you build solar panels or lithium ion batteries to bring their costs down. So I think people are way too excited about the possibilities of nuclear power having some kind of renaissance that contributes meaningfully to data center demand. And in The Outlook, we get into the weeds on this. Vogtle 3, by the way, in Georgia already having operational problems and it just came online.
The US lacks a lot of uranium and has import reliance on certain kinds of uranium, which is creating problems for Terrapower. There are very limited opportunities for plant restarts. And way too much was made of the Palisades reopening and the announcement and as well the one that's taking place with Three Mile Island. You can only really restart plants that just recently shut down, meaning within the last couple of years, because the decommissioning process hasn't started yet. If you start the decommissioning process, it's way too expensive to then reverse that and get that plant back online.
So at maximum, I can think of maybe two or three other shuttered plants that would be eligible for restart. The next gen designs-- everybody likes to talk about them-- sodium cooling, gas cooling, molten salt. And according to the DOE's nuclear people, at least a decade away from commercialization in the US.
And then small modular reactors-- very impressive pipeline. 6.5 gigawatts, $200 billion in orders. They're still lottery tickets. And the power commitments are not take or pay contracts. They're all contingent on final cost, as was the case with the NuScale plants that were abandoned in Utah last year. So let's take some reality pills on this nuclear renaissance. Take a look at this section. It gets into the details.
(DESCRIPTION)
A painting appears of two men on horseback in a desert setting. Title: Don Quixote and Sancho Panza, Daumier, 1865 to 1867
(SPEECH)
We also have a section on DOGE Quixote. I am envisioning Vivek Ramaswamy and Elon Musk as Don Quixote and Sancho Panza from Don Quixote. The thing to remember is that one of the underlying themes of the model is the pragmatism and humility are required for them to complete their quest. And I would question whether Vivek or Elon qualify for each of those, or either of them. Anyway.
Vaya con Dios, DOGE Quixote. I think they'll obviously get some things done, but they're going to run into a number of different buzz saws. And one of my top 10 predictions is they're not going to make it anywhere near some of the targets that they've mentioned. There are savings to be had from reversing some of Biden's recent executive actions, but those wouldn't count towards DOGE targets on reducing current spending. That would just be reducing potential future spending that hasn't really even been penciled in yet.
Non-defense discretionary spending has already been cut. Budget Control Act of 2011. It's at its lowest level in decades. So therefore they're going to have to focus on entitlements. And the DOGE billionaires are now finding out that that's a political third rail. Good luck with that.
The SEC, the Department of Labor, the Department of Education, and the EPA combined represent less than 2% of all federal workers. So fine, start firing them. But that's not going to get you very much. The vast majority of federal workers are the Postal Service and the Department of Defense.
There will be efforts underway to expand-- the CMS's is maybe expand weight loss coverage for weight loss drugs. And those are going to offset some of the DOGE savings, which takes place elsewhere.
Trump has talked about impounding or redirecting congressionally approved and congressionally appropriated spending. That's going to face judicial challenges. There was an Impoundment Control Act passed in 1974 in response to something Nixon did. A president can't, according to the Supreme Court and according to this legislation, can't just say, well, I'm going to take the money that was appropriated for the Department of Education and spend that on something else, which is what he's announced that he wants to do.
Most defense spending is people and operations on the order of 70% to 80% rather than weapons procurements. And then with respect to weapons procurements, good luck. The number of federal contractors has fallen from 51 in the early '90s to five. So it's going to be very difficult to cut those procurement costs with that much concentration in the defense contracting.
And then we also get into some of the things on improper and fraudulent payments that the government makes, sometimes accidentally. You can try to reduce them, but it's going to cost a lot of the money in terms of investing in systems and people to do that. So we get into the details here. It's interesting if you want to understand it. And good luck to all the billionaires that they've recruited onto the team, along with Vivek and Elon, to try and get this done.
(DESCRIPTION)
Line graph, text: China economic activity monitor (Bloomberg). The x axis shows 2022 to 2024 and y axis shows 1% to 9%. Text: Exports, coal and electricity consumption, steel production, motor vehicle sales, real estate investment, medicine consumption, telecom equipment production, S O E output private enterprise output. Source: Bloomberg, November 30, 2024. The line shows a decrease from 2022 to 2024.
(SPEECH)
We have a section on China that looks at both the liquidities trap and the Thucydides trap, which is something we wrote about a couple of months ago. The bottom line is that China is experiencing some kind of liquidity trap, shares some kind of characteristics with Japan in the early 1990s. As you can see here, Bluebird maintains an economic activity monitor-- exports, coal, electricity consumption, steel production, car sales, real estate investment, consumption of medicine, the kind of things that are for the Chinese government to paper over and obscure.
And the momentum on this thing is fairly negative, which is why China announced that stimulus package.
(DESCRIPTION)
Bar graph. Text: Impact on growth of major policies China has announced since September 2024, Annualized impact on growth, %. Source: Bridgewater November 2024. Rough estimate of drag shows negative 2%.
(SPEECH)
Now the stimulus package was substantial, impressive. I believe there's more fiscal stimulus coming. Bridgewater did an interesting analysis a couple of months ago that show that the total impulse from monetary and fiscal stimulus would offset roughly half of the economic drag that's taking place in China right now. That's a good effort. And the domestic A-share market kind of soared after the announcement. And it has retained its gains whereas the MSCI China index, which is a broader index of Chinese-exposed stocks, has since given up about half of the gains on the stimulus announcement.
I'm not going to go through the details here. You can read the piece.
(DESCRIPTION)
Line graph, text: M S C I China valuations, Forward price to earnings ratio. x axis shows 2006 to 2021 in 5 year increments. The y axis shows 5x to 30x in increments of 5. Source: Bloomberg J P M A M, December 27, 2024. The line spikes and dips below 12x after 2021.
(SPEECH)
I just think the days of 14 to 18 PEs on MSCI China are a thing of the past. And for geopolitical reasons, they're going to be capped at about 12x, which doesn't leave a lot of room for a rally from here. So you could take a look at the piece. We go through the details of some of the tariff and espionage issues at the core of some of the conflict issues between the US and China.
(DESCRIPTION)
Annotated line graph, text: Publicly reported Chinese espionage incidents sourced from open source materials. Events span 2000 to 2021 with a dip in 2015 then a sharp increase to 2021.
(SPEECH)
I'll just say that the increasing intensity of Chinese espionage spreading across almost every industrial sector in the US and other Western countries, it makes the idea of a rapprochement that I've read about highly unlikely. And the same goes for some of the Naval blockade exercises that China has been conducting with regards to Taiwan. So you can read more about that.
(DESCRIPTION)
A graph shows creation of new public companies in the 21st century with the US leading Europe. US companies listed include Meta, Alphabet, Tesla, Visa, Mastercard, Netflix. And Europe includes Airbus, Porsche, Equinor
(SPEECH)
We have a section on Europe called Dr. Seuss goes to Europe. It's just two pages with some Dr. Seuss style language and some charts. I just want to show you one of the charts here. We have charts in comparisons of labor productivity, and energy prices, and deindustrialization in Europe, and problems with monetary policy being applied to countries that are too different compared to the US, low labor mobility, very poor ROA and ROE numbers in Europe other than in health care, and then very negative earnings revisions.
The chart I really wanted to show you was this new one, which is the creation of new public companies on the 21st century compared the US to Europe. This kind of tells you all you need to know about the outperformance of the US versus Europe over the last 20 years. It's kind of an amazing chart to see. I can't describe it verbally. You just kind of have to look at it.
(DESCRIPTION)
Line graph, text: Crypto coin prices by segment, market cap weighted. Lines represent Bitcoin, Transactions Focused Basket, Decentralized Finance Basket. The x axis shows 2022 to 2025 and y axis shows 25 to 250 in increments of 25. The lines for bitcoin and transactions focused basket reach above 225 at the end of 2024. Source: Bloomberg, J P M A M, December 27, 2024
(SPEECH)
The crypto presidency-- I wrote a piece called "The Maltese Falcoin" a couple of years ago. It was February 2022. And I was skeptical of Bitcoin as a store of value and then even more skeptical on the transactions-focused coins and the DeFi coins. And if you're not familiar with the jargon, you can look at the Eye on the Market on this section and it gets into the details here.
I said I would take another look at Bitcoin if it fell. And it fell. It fell hard. It fell 65% after I wrote the piece. And my mistake for not looking at it again. It has since rallied pretty sharply. And a lot of the surveys of institutional investors are fairly positive on the prospects for increased digital ownership.
As it relates to the transactions-focused coins and the DeFi coins, I don't believe it. I was right about those coins for three years. Nothing much has happened in terms of actual utilization of either of those kinds of crypto coins. There's been this spectacular rally in a handful of them, Ripple in particular, but it looks highly speculative to me.
And then specifically, I've got this chart here looking at the volumes of utilization of Ripple, Litecoin, Monero, and Stellar.
(DESCRIPTION)
A line graph shows transactions per day on blockchains from 2022 to 2025.
(SPEECH)
None of them have changed since two years ago. And so if the value proposition of these coins is that people are going to use them more, you'd expect to see that happening and you don't. So that means that this is a very technical and speculative rally.
Same
(DESCRIPTION)
Line graph, text: Ethereum daily verified contracts, Total verified contracts per day, 7 day average. x axis shows 2022 to 2025, and the y axis shows 100 to 1000. The line peaks at 1,000 in 2023, dropping by 2025 to about 200.
(SPEECH)
thing for Ethereum. The number of daily verified contracts in Ethereum being executed on a blockchain are now unchanged versus what they were in early 2022. So to me, what's really happening is that people are earning a sky high return on their political donations. Crypto PACs received almost $250 million in donations, around half of all corporate donations made during the election. And Coinbase gave $75 million to a Super PAC called Fair Shake. Their candidates won almost every race they got involved in.
Look, they're playing the game and participating in the system in ways that they're allowed to do. I'm not criticizing it. I'm just saying that this, to me, is a better our explanation of the pop in crypto than the presumed proposition that their acceptance and utilization has been in changing. All that's really been changing is the politics. And the GOP Majority Leader of the House is pledging a crypto bill in the first 100 days. Trump has announced a crypto council and a crypto czar. The new SEC head will probably roll back some of the crypto accounting disclosure rules. So this is a shift in sentiment rather than a shift in utilization. And you could take a closer look.
And then just to close out here-- and thank you all for listening.
(DESCRIPTION)
A table appears. Text: 2024 predictions recap, Scorecard. Rankings appear from 1 to 10, January 1, 2024 Prediction. Highlighted green, 1, Despite all the hand-wringing over risks to reserve currency status and BRIC efforts to develop a competitive reserve currency, the US dollar remains stable, within a +/- 1 std deviation band of 7%. Highlighted green, 2, The DOJ/FTC win a big antitrust case. Highlighted green, 3, President Biden withdraws between Super Tuesday and the election, citing health reasons. Highlighted green, 4, Little to no progress on self driving cars in the US, no recovery on Lidar stock basket. Highlighted yellow, 5, Broadly syndicated leveraged losses rise above credit losses for the first time. Highlighted yellow, 6, Argentine dollarization will fail if implemented. Highlighted green, 7, Russian invasion of Ukraine drags on with no ceasefire in 2024. Highlighted green, 8, US regional bank stocks will do well in 2024 with stable or rising price to book values. Highlighted red, 9, Major US cities will face electricity outages and/or natural gas outages due to outdated grid infrastructure. Highlighted yellow, 10, Researchers will complete work on an inhaled COVID vaccine that will reduce transmission.
(SPEECH)
I started a top 10 list last year in honor of Byron Wien, who had passed away. He did this every year forever.
I made a bunch of predictions last year. Six of them were right. Three of them were kind of either too soon to tell or hard to say. And then one was wrong. I've made a new list for this year. I will tick through them very quickly and then sign off.
At least three confirmed Trump cabinet members will be gone by the end of the year. Renewables will underperform traditional energy again in 2025, like they've done in the last three years. MSCI Japan will outperform China despite China's higher PE multiple.
The discontinuation rates, which are approaching 50%, of weight loss drugs are going to hurt some of the GLP stocks, specifically Novo Nordisk. I think we'll end the year below consensus analysts price targets. Despite the order book referenced earlier, I don't think any small modular reactors in the United States will even begin construction in 2025.
I think DOGE cost cuts are going to be no more than $150 billion a year, excluding reversal of executive actions compared to targets of $2 trillion. Hedge funds I expect to continue to outperform the stock bond benchmarks. And we explain what we mean by that. But they've been doing better than expected. Housing inflation, I think because of completions, both single family and multifamily, will start to finally bring housing inflation down below 4% the way that the Fed computes it.
Blue state outmigration will continue. Over the last few years, 70% of all movers in the country in the US are moving from either blue to red or purple states. I think that's going to continue. And then lastly, I think there's a chance that the Trump administration launches a targeted military action against drug cartels in Mexico.
So that's my top 10 list. Thank you for listening. Take a look at the outlook. There's a lot in there that we can't get into in a podcast like this. And it's been a privilege. Thank you very much. Bye.
(DESCRIPTION)
Logo: J.P. Morgan
Eye on the Market Prospettive per il 2025 : Gli alchimisti
Altri Eye on the Market
Economy & Markets
Nov 6, 2024
"Kamilton": the 2024 election and who tells your story
Economy & Markets
Oct 17, 2024
