Are investors missing the most important macro shift in decades? In this must-watch interview, Julien Garran of the MacroStrategy Partnership joins Trey Reik to explain why the market has completely misread the economic agenda of the Trump 2.0 administration. Behind the headlines and media noise lies the “Real Trump Trade”, a radical restructuring of U.S. economic policy aimed squarely at reversing two generations of blue-collar decline.
Key Topics:
- Why markets are reacting incorrectly to Trump’s return and what they’re missing
- How four key forces: trade, government spending, energy policy, and antitrust are being weaponized to reshape the economy
- Why Julien believes we’re headed first into a deflationary bust, followed by a reflationary boom
- What this means for the U.S. dollar, commodities, gold stocks, and Big Tech
- Why Julien sees AI as a bursting bubble, with commercial viability hitting a wall
- How Trump’s economic team may be willing to endure a recession to trigger long-term change
This conversation offers powerful insights into what’s coming next and why the current investor playbook may be obsolete.
Investment Concerns? Get a free portfolio review with Wealthion’s endorsed financial advisors at https://bit.ly/3YoLTs2
Hard Assets Alliance – The Best Way to Invest in Gold and Silver: https://www.hardassetsalliance.com/?aff=WTH
Julien Garran 0:00
My central argument here is that, if we’re actually unwinding a process that has lasted two generations and kind of favored kind of financial capital over those two generations and kind of disfavored blue collar workers, do we think that three trading days is going to unwind all of that accumulated policy? And my answer is not a hope.
Trey Reik 0:33
Greetings and welcome to our wealthion show. My name is Trey Reich of Bristol Gold Group, and we’re here today with Julian Garen, partner of macro strategy partnership, a UK macro research boutique. Some viewers may remember, back in mid February, I interviewed Julian’s partner Andy Lee’s in a segment called Europe’s economic suicide. After a long career in commodities, Julian joined macro strategy about a decade ago, and his primary responsibilities are a study of global liquidity cycles and their implication for commodities markets and various asset classes. So Julian, thanks for joining us today.
Unknown Speaker 1:25
Thanks very much. Great to be here. Trey,
Trey Reik 1:28
terrific. So for those who are as not as familiar with macro strategy as I am, could you give us a brief rundown or description of your firm, your clients and how you share your duties among the firm’s three principles.
Julien Garran 1:45
Yes, sure thing. So we’re a boutique macro strategy outfit. It’s the writing partners and myself, James Ferguson and Andy lease, and we’re all independent global macro thinkers. We’ve all had about 25 years on the buy side and on the sell side, we have a wide array of clients, kind of largely, kind of buy side, hedge funds and long only, but also with a smattering of central banks, corporates, etc. And I think that the key thing that we do that differentiates ourselves from, if you like, the mainstream kind of economists is that we seek to think completely independently. We build up our views from the bottom up, using kind of widely
the macro and the key kind of areas and models that perhaps aren’t always accepted kind of by the mainstream at the Fed or at or at other central banks or at Treasury positions, etc. And what it means is that while we always kind of relay what we find, in many cases, because we’re kind of arguing from different premises, looking at different kinds of data, and kind of manipulating it in different ways. We often end up being quite contrarian, even though all the arguments that we make we believe are firmly grounded in economic truth. So that’s the and especially when we all agree with each other, which we do today, that tends to be when we we have the most powerful arguments for substantial changes in the global macro system. Excellent.
Trey Reik 3:35
Well, we scheduled this interview a couple of weeks ago, which was before the new Trump tariff world that we live in. But I want to go back to the reason I scheduled this interview with you was because I was really intrigued with your march essay the real Trump trade, reversing the blue collar squeeze, in which you point out that reflexive market reactions to Trump’s November re election may have been a little misguided. Can you give me a precis of your logic here?
Julien Garran 4:09
Yeah, sure thing. So in the immediate aftermath of Trump’s victory, we started seeing a number of kind of critical macro trades taking place, seeing the dollar strengthening dramatically. They were seeing a large risk on event, especially in kind of mag seven, kind of tech stocks and those types of those types of things. And it struck me that that was completely out of kilter with my understanding of what Trump had been saying on the campaign trail. And also my understanding of what the kind of brains unit, if you like, behind the Trump administration, had been pushing for, which includes kind of JD Vance and his backers, Scott minan, who’s the head of the Trump advisor, or the. Mac Advisory Committee, and, of course, Scott Besant. And what this really comes back to, and this is something that at macro strategy my colleagues, James and Andy and myself, have been discussing for really nearly a decade, is that many of the kind of policies in the West, the economic policies, the sort of liberal, globalist policies that have been put in place over the course of the past, kind of decade, which includes kind of liberal globalism itself, if you like, almost unilateral free trade with countries, even if they weren’t running free trade themselves, policies that led to very large government well over the optimum of 20% of GDP restriction on energy, and also the government allowing the sort of large monopolies to run free, like big tech, Big Pharma, big food, etc. And what all of those policies ended doing, whether intended or not, was that they started repressing the or they repressed the blue collar workers job opportunity shown no better shown than in this chart, which is the elephant from the World Bank. And what that does is it shows the incomes by percentile around the world, income growth from the point that the wall came down through to 20 years later or so. And what you can see there is that the back of the elephant is, is the kind of strong growth in the emerging market, middle classes, incomes, the tip of the elephant trunk. On the far right there is the powerful growth in the top earners incomes in the world. But then the base of the trunk, around the 75th percentile, that’s the stagnation and decline of the of the prospects and of the opportunities for the blue collar worker and for me and my colleagues, what we’ve been arguing is that this was going to cause a backlash, if you like, a populist backlash, that would then potentially cause a revolution against big government and against the very policies that had acted to do that. And what really struck me in the early days of the of the Trump administration was that everything that they were seeking to do was to reverse every single one of those policies that we believed had undermined blue collar workers, and that’s why we took what they were saying extremely seriously, and why we haven’t been surprised by events, because this is exactly what I think they’re attempting to do. And
Trey Reik 7:42
this makes it incredibly ironic that Trump is, you know, routinely portrayed as catering to the billionaire class. I mean, just about every Democratic statement that I’ve heard in the last month, the word billionaire is always in there, with respect to to musk and and to Trump. But what you’re suggesting is it really couldn’t be further from the truth.
Julien Garran 8:07
Yes, I think it couldn’t be further from the truth. And I think that’s that’s the difficulty that the market and particularly the press and some kind of fund managers are struggling with, is that they’re getting so distracted by the noise and the circus around the administration, they fail to actually grasp that the real purpose of this administration is to unwind those globalist, liberal policies that have suppressed an entire, you know, two generations of blue collar workers and Increasingly now are also kind of repressing the middle classes as well. And so I think that that’s that’s a major fail. And given the size of this kind of effort to turn around these kind of powerful forces that have been going on for two generations, my view this is going to take a long time to price in, not just three days.
Trey Reik 9:03
So we could spend a couple of hours on this theme and those four components, liberal globalization, big government, energy and anti trust, sort of monopoly. But can you just give us, like three sentences on each one of those. For an example of how the system has been stacked against the average guy for the past, you know, 1520, years.
Julien Garran 9:29
Yeah, national thing. So well starting with liberal globalization. And this is, this is the sort of Hot Topic, so I might do a few more than three sentences on this one. Okay, so this year four, just started with, with David Ricardo and his famous theory of comparative advantage from the early 19th century. It was, it was a time, and he used the example of Britain trading cloth with Portugal that would then trade port wine back with us. And if you see, that’s the original David. Cardo chart there that he used. I’ve got another one showing how when you change, when you specialize in trade, it’s a win win for everybody. But the problem with that is that that’s too simplified, because that’s not really how the world really works. The way the world really works is that civilization and progress is driven not by things like tech. It’s driven by hard work and thrift and using that surplus to build physical capital, the houses, the offices, the restaurants, the roads, the ports, the factories, etc, which are the literal foundation of civilization upon which we can reach higher as economic agents than our ancestors did because of that accumulation of capital. And so the question is, if you trade freely with a country like China, does that improve your the US kind of capital base? And the answer is, given that China has tariffs, given that China has non tariff, huge numbers of non tariff barriers, they use slave labor, they kind of have they pollute more than the US does they kind of hold very different kind of views on on kind of human rights, etc. Does that the US capital base? And the answer is, of course not. It doesn’t help at all. It accelerates the decline that takes place in the Midwest. And using that example, kind of Ricardo example, of cloth in the cloth being traded from the UK when he was writing that a town called Paisley, which is just outside of Glasgow, accounted for 99% of the world’s cloth production, and I went to visit there now accounts for 0% I went to visit there last summer. And while it’s got kind of beautiful kind of kind of government buildings and Georgian housing and merchants houses and Muslims and all of that kind of thing. The place is completely run down with with rundown shops. It’s actually so depressed, depressed that the McDonald’s in the center of town closed down. That’s how that’s how bad that was. And that experience of free trade, which Ricardo completely missed, has been replayed over 1000 times across the past 40 years, across the Midwest, as factories closed down too quickly, that meant people then lost their jobs, became kind of addicted to opiates, kind of became kind of dependent on social welfare as children stopped growing up in a working, working class environment and where you’ve effectively seen a destruction of physical capital, and with it a destruction of human capital constantly interrupt
Trey Reik 12:55
you to ask you one question, yeah, how long Will it take, even if everything goes right to turn that around, are we talking? That’s the
Julien Garran 13:05
difficult question. That’s the million dollar question. Okay, so I think what happens is that it the rot stops quickly. So the speed, the speed of further de industrialization, that ends pretty quickly. Okay, it takes a bit more to actually start reversing the process. I think you can start to see kind of domestic manufacturers who still in business, still competing, providing car parts and what have you. Well, they’ll start ramping up to full capacity. If they’ve got, if they’ve got an easy expansion they can do, they’ll start doing that, compared to car makers, for instance, importing from overseas. So we actually heard GM planning to bring some of their supply chain back. So those would be the easy wins. What’s difficult to do is to persuade a company that’s got its kind of operations offshore, say foxcom, to come and produce Apple parts in in the US, because then, you know, you have to, you have to beat a whole bunch of kind of cost of capital hurdles to justify doing that. And given the uncertainty in the system, and you don’t know if these policies are still going to be in place in four years time, let alone in four months time. That bit’s more difficult, but it is a step in the right direction, if that’s your focus, to try and to reinvigorate the blue collar worker. All right,
Trey Reik 14:31
how about a few sentences, maybe a few fewer sentences on big government and big tax and how that’s sort of worked against, you know, 70% of the middle of the economic spectrum.
Julien Garran 14:45
Sure thing. So the best paper out there on this subject is by Bergen Henriksen. And what they found was that for every 10% of extra government spending and taxation above an optimal. Rate of about 20% of GDP, you took away a percentage point of growth. And in the States, it’s not quite so bad. You’ve got sort of total governments spend at about 35% or so. In Europe, it’s toxic. We’re about 4045 in the UK, 53% in France, and I. And the thing is that this isn’t just a one off. This is cumulative. So I calculated that had we been running kind of optimal government sizes since 1970 disposable incomes in the UK would be four times higher today than if than with the current regime, and in the States, they’d be about three times higher. So these this is a huge kind of opportunity lost to bring up everybody in society through kind of effectively allowing a smaller government less crowding out much more productive use of capital, but because the government keeps on wanting to solve every problem and doesn’t have the respect, if you like, for the market and for everybody’s kind of individual kind of incentives to do that. Because of that, we’re actually in a substantially worse place than we were before, and a large part of that falls upon the kind of blue collar worker. And I think particularly the issue is that whenever there’s a problem, part of the government, the Fed, chucks liquidity at the situation, and what that’s ended up doing is creating a huge divergence between physical assets kind of which should be priced, you know, just a little bit over replacement cost and financial assets, which end up getting priced hugely above and there’s one measure of that called Tobin’s Q, my colleague Andy lease pointed out that’s until very recently, been at its highest ever levels. And so that is the that is the way in which large government has taken away opportunities, crowded out opportunities for blue collar workers.
Trey Reik 17:04
And how about a few words on energy restriction and those things in the last
Julien Garran 17:10
on the energy side, kind of the way that most economics works is that they use an accounting system for adding up government spending and consumer spending, corporate spending, taking away the current account deficit, and that gives you GDP. And they look at all the components and energy, 7% and so on. That way of thinking, if you just stopped the use of energy, then you take 7% off GDP. But that isn’t correct. If you stop the use of energy, you take 90% off GDP. In a sense, energy is the foundation at which everybody converts to more productive uses. And so it’s absolutely imperative that the foundation keeps growing in order to allow a bigger pyramid of growth and economic welfare upon on the top of it, and it’s just as an example. This is a picture of the Aztec champers. So what happened then was that the Aztecs lived on the side of these shallow lakes. They worked out that if they pushed all the silt together, bound it by kind of tree branches, they could create much more productive fields. And Falco Smith, who’s an expert on this great book, energy and civilization, calculated that the food it costs to feed the workers to build these things was repaid within 12 months by increased crop yields. So effectively, that’s an investment that creates 1,000% return over 10 years, and that’s something that you can base the civilization upon. Whereas, when you look at something like renewables, the problem with renewables, particularly wind and solar, is they’re intermittent, so if you want to use them, whereas the grid needs the six nines, 99.9999% kind of availability. So if you use it in intermittent you need to have backup power. This effectively a redundant backup power system, which is spinning gas reserves. So you basically get those spinning turbine spinning, wait until the sun stops shining, then plug them in. That’s incredibly inefficient. My colleague Andy estimates that its costs about every percentage point of renewables that you add to your electricity mix raises the prices to consumers by 5% so the European move, for instance, to go from 20 to 40% renewables over kind of from 2019 through to 2030 should double the cost of electricity to customers, and that takes a huge amount of way from consumer welfare and the ability to be kind of competitive. And my estimate of the return on renewable investment in was that if that return. Is minus 600% over a decade. So if you compare that to the chinapas, which were the basis of building a civilization, in this case, this is the basis of eroding a civilization, especially in Europe, but also to a degree, to the degree which it’s been done in states. And that’s what the kind of the New Energy Secretary Chris Wright, has been immediately day one, actively moving against
Trey Reik 20:26
all right, in the fourth category is this sort of monopolistic big tech antitrust type development. How has that impacted the average guy?
Julien Garran 20:37
Well, in that case, so economics is quite useful on this because, because what economics shows is that once you start running as a monopoly, restricting supply, etc, then you create not just a dead loss, but also a major redistribution of capital, or of rate of economic returns, away from consumers and towards the owners of the capital. And because the owners of the capital are all top kind of decile or top quintile people, and the consumers include the entire kind of blue collar worker, kind of glasses, that is just straightforward, kind of a redistribution of the gains of capital from kind of blue collar workers towards towards higher kind of income individuals. And the interesting thing about this was that in the early stages of the new Trump administration, the markets were buying the heck out of the big tech companies with the view that Trump would be transactional, that they would see, that we’d see, that we’d see sort of big tech lobbyists run rings around him. What’s very clearly happened is the opposite. In fact, he’s continued with a very clear antitrust kind of measure. And I think if people had been watching more closely, especially listening to the speeches of JD Vance, and understood their history, populist movements have a very large kind of anti, kind of monopoly and anti kind of big, big business, kind of populist approach. And very clearly this administration is tapping into that, and we now face a huge number of problems. So Google is, is already, kind of has already lost cases on its advertising business and on its kind of kind of Google Play or App Store Business, Apple’s lost, lost kind of cases on its app store business. Remedies are going to be coming in through all these things. There’s a kind of case pending which is going to seek to break up, kind of Facebook Amazon is going to is under pressure, because it’s about to face a case where the Amazon tax on suppliers is going to be under the microscope, and I think that markets have completely failed to price in the potential risk to growth from a populist regime Seeking to push back against some of these monopoly abuses. And
Trey Reik 23:03
all of this feeds into an observation that I’m going to make. Obviously, markets in the past two days have voiced their overwhelming displeasure with the Trump tariff structure. But of course, these are the 10% of Americans that own 88% of the equity market. You know, the next 40% owns 12% and the bottom 50 owns 0% so this may explain why, despite this calamitous reaction in markets, Trump’s approval rating in the weekly Daily Mail poll actually went up four points last week from 49 to 53 this week, and his popularity rating with 18 to 29 year olds increased 13% and his popularity rating among black voters increased 17% so it’s, it’s clear that Trump tariffs are popular with a large group, if not a majority, of Americans. They just don’t happen to be the ones running $5 billion hedge funds.
Julien Garran 24:15
Yeah, I think that. I think that’s fascinating, and it just shows how powerful a proper understanding of populism is in that you can argue that this, you know, this was a great mistake because we all these policies were great mistake because we saw equities go down, but actually, what they were intending to do was to stand up for the blue collar worker. And I think the blue collar worker understands that, and that’s why these pop, these policies, are proving to be popular in the first instance. And that’s why also, I think that there’s a much stronger spine, kind of much steeliar spine, to this administration and its willingness to stick with this stuff. Then I think. Think that the commentariat are aware of even close to being aware of. So
Trey Reik 25:04
taking this to its logical conclusion, how can people invest in markets? If your thinking is accurate, you’ve identified four parts of the market that you think will result from this process. One is a weakening dollar outperformance of commodities, relative outperformance of US manufacturing, and a severe underperformance in tech land. Can you just address each of those for for a few sentences? Yeah, initial
Julien Garran 25:39
thing, and so I think we have to get our timing kind of clear on this. Okay, so the first thing that we’re going to see is we’re going into a deflationary burst. Tariffs aren’t inflationary. I can talk about that in more detail later, and especially if the Fed is pushing up against them. And that was the clear message that we got from Powell, kind of on Friday night. What he said was that our obligation is to make certain that one time increase in the price level does not does not become an ongoing inflation problem. So he’s going to stay too tight too long, until he’s sure that the price rises that come from the terms of trade shock from the tariffs aren’t going to turn into increased inflation expectations at the same time, that means he’s going to be too tight for the real economy that’s going to deteriorate, not just because of the shock from tariffs, which hits kind of real incomes, but also because of the shrinkage of government, which in the short term hits real incomes because of the rise in savings, because of other factors that are causing a slowdown, etc. So the first instance is that this is going to be deflationary. It’s going to be bad for most, almost all risk assets. But the next question is, how do the authorities then react to that? And the big question that I think everyone needs to answer to be able to invest over the next few years is, is what kind of president is Trump? Is he a strong dollar President like Reagan was, or is he a weak dollar President like Nixon was, and I believe categorically he’s a weak dollar President like Nixon was. He was tweeting on Friday night that he wanted Powell to lower interest rates immediately. He will be putting huge amounts of pressure on Powell over the course of the year to do that, there’s then going to be the entire discussion about who Powell’s success will be in May of 2026 and clearly they’re going to be someone who does Trump’s bidding. And then we’ll have the transition to the new kind of to the new Fed chair. And in my view, what they’re going to do is that they’re going to run extremely easy monetary policy to try and get us out of recession by that time. And as they do that, they weaken the dollar. They also have a lot of policies and or policies that they’re setting up to deliberately try and shrink the capital account surplus. This is a paper that Stephen minan did while he was still at a hedge fund before he became chair of economic advisers that will have the kind of effect of pushing money offshore via the self sovereign wealth fund. That’s why they’re talking about buying the Panama Canal, Tiktok, Greenland and all those things. It’s the money out of country, and what that does, in our kind of framework is that that creates a huge reflationary boom as we go forward, it’s obviously bearish the dollar. It’s very, very bullish for commodities, because when dollars move offshore, they’re used as the sort of base, the deposit base for lending, for commodity finance, trade finance and emerging market finance. So you get a boom offshore, and that’s very resource intensive and very good for emerging selected emerging markets, I think, particularly the emerging markets that are getting into the with a burgeoning middle class, getting into that 10 to $15,000 income bracket. And the two that I would highlight are India and Vietnam. Vietnam is obviously under under pressure on its but is reacting to pressure in terms of tariffs. But those are the types of economies and markets that could become very, very positive as we go through it’s in terms of resources. It tends to be bullish for precious metals in the very early stages, or the earliest stages. And one thing that I would highlight is that once we get away from this tight fed, and we get into a Fed that’s attempting to reflate, what tends to happen, like in 2001 is it’s gold equities that get re rated. So we’ve had an unusual cycle where gold has risen sharply when it would wouldn’t normally in this cycle because of the fracture of the geopolitical system. But I think the same trade that worked in 2001 where the best performing stock in the world. Was goldfields, a South African gold miner that no one in 99 would touch with the budget. I think that same potential re rating takes place, and then later you see that broaden out to a broader resources category as well.
Trey Reik 30:14
Interesting and US manufacturing and high tech,
Julien Garran 30:19
yeah, on the US manufacturing side. And I think that, I think they’re relative winners. Asked that so, so when you, whenever you get a global reflationary boom, what tends to happen is that it’s relatively good for blue collar workers in kind of resource type areas, manufacturing type areas, and so, yeah, the best example, kind of, going back to kind of, kind of the early 2000s was the shale patch, which just exploded at that point. And then you were seeing kind of, you know, wages going through the roof. It was, it was a very tight market. You’re getting kind of guys who drive those, or, in fact, mainly women who drive the really big iron ore trucks, kind of in Australia, they were ending up with kind of salaries of 150 200 grand and and so that’s the type of area, that’s what does relatively well. And I would expect that to happen again. And then on the tech side, the difficulty in the reflation, and this may even turn into a new inflation, depending on whether the trump of the new Trump fed reacts to this reflation or just continues to accommodate it like it did under Nixon. If we get that, then what you tend to do is you de rate equities that are basically sold on a promise about some far future income, and that’s exactly what tech is. And so I think that on a kind of ratings basis, this would be extremely negative for the winners, the big tech winners of the past few years.
Trey Reik 31:58
So last week, the Trump tariff reciprocal structure took the average Fitch says that it took the average tariff rate from two and a half percent to 22% which Fitch quotes as the highest since 1910 is that what you expected?
Julien Garran 32:19
What? Yes, I wasn’t trying. I wasn’t trying to be a slave to fortune if I’d actually put numbers to what I was expecting. But I was expecting them to go reciprocal, because that’s that was all the discussion in the background. Now, to be honest, like even a week before, things weren’t settled in the background, Besant was going for 10% across the board, kind of Latin, and others were going for the reciprocal numbers. And so, you know, no one, no one actually knew quite where it would fall, but, but my view the whole time was that this, the whole idea is this, this is reciprocal. And a lot of people have kind of criticized the methodology they’ve used, saying that the administration kind of got it off chat, GPT and all that, kind of haven’t applied it properly, and all that goes but that’s kind of by the by the real issue here is that it’s not just the tariff level that someone’s got against you, it’s everything else they’re doing. It’s subsidies, it’s kind of, it’s manipulating your currency. It’s, it’s the non tariff barriers. LUT Nick talked about non tariff barriers in South Korea that made it almost impossible for some US firms to export there. And so you can’t just take them at their word. These are going to be the tariffs will lower them. So therefore you should lower yours to zero. You have to say, No, we’re going to we’re going to look at the actual surplus you have with us, and we’re going to impose tariffs until that surplus goes down to zero. And that provides the incentive for that regime, the regime overseas, to actually take down all the tariffs in order to consequently level playing field. So while there’s a lot of economic theories sort of showing that the tariffs never quite work as well as you want them to, because they’re always offset by falling margins, falling real incomes. Kind of currency moves and what have you, what really does work is if you threaten tariffs and then your trading partner, then kind of accepts and negotiates and lowers all forms of tariffs that they do. That’s a complete win win. And while that’s clearly not going to happen with China, and I don’t think the administration is too upset about that, it looks like it’s going to happen in a lot of other places, with Vietnam, apparently, kind of stepping up and kind of offering, kind of all sorts of negotiations and various offers to try and get the Trump administration.
Trey Reik 34:50
Yes, I’ve read, in the last you know, 24 hours that up to 50 countries have you know, the White House phone is ringing off the hook. As people are calling up to 50 countries floating reciprocal tariff structures as low as 0% so I know you can’t you don’t have a crystal ball, but what are the chances that 60 to 90 days from now, Trump’s imposition of tariffs have actually catalyzed a new secular trend of reducing tariffs in global trade. Yeah, I
Julien Garran 35:24
think, I think there’s, I think kind of put it this way, I think China made a major mistake coming back and calling his bluff. I think that’s going to have almost the opposite effect. And technically, if you’re putting up reciprocal tariffs, and then your trading partner puts their tariffs up again, you’re going to put them up again. And I wouldn’t be at all surprised if we hear that on the China side, or Trump’s attitude to China, I think they could easily raise tariffs back at home again. However, for these smaller countries, that’s potentially very, very positive. And so I think, yeah, I don’t think we’re going to get a new free trade environment. But what I do think is we’re going to get an increased bifurcation of the world where there will be countries within the US sphere of influence and open trading system, and countries that aren’t. So if we were entering a bipolar world to some degree already. I think this just accelerates that process and but however, that’s going to be, you know, that’s going to be a significant positive for both sides, if all of those tariffs then come down. So yes, it’s, it’s not quite the complete calamity that the markets might have been suspecting over the last three days, there is definitely going to be upside. But given the size of China, I do think that there are significant problems that are going to remain all the standard analysis of tariffs and how they lower real incomes that’s still going to apply because China is such a large part of the trade deficit that the US runs.
Trey Reik 37:04
So all in all, are you a fan of Trump and besson’s framework? Do you think they’re doing it the right way? Are we sure we even know what they’re trying to accomplish? How would you what grade would you give them in their approach. Okay? So
Julien Garran 37:23
what I would say is, I think, and I’ve believed this for a long time, that economic progress cannot just be progress for the few. Economic Progress has to bring everybody along with it. And so I am hugely in favor of policies that seek to reverse this squeeze on the blue collar worker that we’ve seen for two generations now. I think the difficulty that the markets and the commentariat have is that you can’t choose your revolutionaries and and the and the way that they’ve carried some of these things out, some of the doge cuts, for instance, the sort of surprise element of the tariffs, those types of things, what they tend to do is they create a lot of uncertainty. And that’s one thing that markets hate, but it’s also something that the C suite takes if you’re making investment decisions and you just don’t know what the tariff is going to be in three weeks time. It’s actually very, very difficult to do it. And so I think that while I fully support the aim to improve the lot of the blue collar worker, I think some of the way that they’ve done it has probably substantially increased likelihood that we’re going to go into a really quite difficult recession first, before we get to see those benefits come through.
Trey Reik 38:48
And we both agree, I think that a major plank in the Trump economic platform is a weaker dollar. But how do you think say Scott Besant deals with the trade off between recession and lower 10 year yields. Do you think he sort of looks at that as an even trade off?
Julien Garran 39:12
I don’t think I’m convinced that Scott Besant wouldn’t want a recession. I think he might, you might start. You might be getting sort of to a point where he realizes it’s going to happen. Whatever he does, I think he probably is getting to that point. They Yes, they do want to have lower yields that is, or lower kind of yields that that is one of the sort of stated aims of the strategy, I think that, and we may see some policies down the track that try and achieve that with like, for instance, yield curve control, etc. At the moment, I don’t think that even within the administration, they fully articulated those so what, I think, what, what, the way that Besant put it, was that, that, you know, we. We’re trying to make a generational change in the way the global trading system works, in the way that blue collar workers have been supported that can’t do that in, you know, in one kind of in one year or in one month. And so you have to accept that there’s going to be some pain first, before the benefits come through. And and a good example is a sort of, if you like, the spiritual kind of guru for the for the Trump administration, Xavier millet from from Argentina. He handed his his chainsaw over to Musk when Musk started his his Doge, kind of Doge plans. And, millet had to go through a very difficult year where he was slashing government spending and the economy was falling to pieces. However, within 15 months, a recovery was becoming increasingly visible, and now that recovery is in good shape, and so I think what they’re hoping is that the downturn related to these actions will be relatively Swift, and they’re hoping that they can then fire out kind of the blocks as they go through the final couple of years of the administration into the end of 2028 and I think that that’s I think they’ve accepted that that’s probably going to happen, and that’s why, because they’ve accepted it. That’s why I don’t think they buckle and step back. And
Trey Reik 41:28
we talked about this a little earlier. The weather tariffs are inherently inflationary. The market certainly thinks they are. But I’m getting the sense that you disagree with that assessment,
Julien Garran 41:42
yeah. So they’re basically what tariffs do. And my colleague James Ferguson wrote a nice piece about this. Is that there are terms of trade shock you get. Suddenly the price of one thing increases relative to everything else. Now, the thing about that is that if everything else is equal, so if policy doesn’t change, and that policy is relatively fed, policy is relatively restrictive at the time, all that that does is that it means that people buy a bit less of the thing that’s got more expensive, and at the same time, they have less funds to buy other things, And consequently, you get an overall kind of decline, or overall downward pressure on aggregate demand and overall downward pressure on the general price level. So as long as the Fed doesn’t change what it’s going to do and doesn’t accommodate these things, then that terms of trade shock is not inflation. However, if you’re like Arthur burns, for instance, back in the 70s, and you do accommodate in terms of trade shock, in that case, the rise in the price of oil, then what that then does is it means that your real GDP kind of holds up. But then that all that extra liquidity in the system then start means there’s going to be more aggregate demand chasing the same amount of fewer goods, and that pushes up prices. So it all depends on the Fed, and in the first instance, the Fed’s response has been, powers. Response has been, we’re going to remain extremely cautious, and we’re going to make sure that this price increase doesn’t turn into an actual increase in general inflation over time, but that may all change once you get to sort of later in the year, with more difficult growth, with higher unemployment and with the potential for a new Fed chair. And I think after we get to that point, then I think markets might start to price in a higher inflationary regime later in the Trump administration. But for now, I don’t think this is inflation
Trey Reik 43:48
interesting. So in what I refer to as the modern era of unconventional central banking, stocks have mostly gone up because of all the liquidity that’s been created. But one thing that’s for sure is that on the few occasions where stocks go down, they seem to be going down quicker and harder than they did in past corrections. So whether it’s COVID or the fourth quarter of 2018 or the first half of you know 2022 stocks go down because there’s so much excess liquidity that’s been out there pumping them higher. So the S P was down 10 and a half percent on Thursday and Friday. And I would argue that the biggest reason they went down that quickly is they were far too high to begin with. So whether it is valuation concentration, sentiment, us, dominance of global averages on every possible measure, stocks were obviously at the very, very tip top of historical valuation. So my question to you. Is this 10 and a half percent move? Do you think this market move is justified? Or, you know what to what do you attribute the downward pressure?
Julien Garran 45:12
Yeah, and if you, if you’re watching, of course, you were watching the market before the tariff announcements, and it wasn’t behaving very well, and you know, we were below the 200 day on the s, p, before the tariff announcements, and we’re now kind of a long way below. And my, my central argument here is that if we’re actually unwinding a process that has lasted two generations and kind of favored kind of financial capital over those two generations, and kind of disfavored blue collar workers. Do you? Do we think that three trading days is going to unwind all of that accumulated policy? And my answer is not a hope, and this thing could go on for a long way, and and I think that, as you mentioned, that
Trey Reik 46:03
suggesting that the correction could go on for a long way, or the process,
Julien Garran 46:09
yeah, the process, and the process could be, could be very negative for stocks, particularly over the course of the next kind of 12 months before the Fed manages to put enough reflation into the system, and that catches a spark so between, but between now and then, and especially over, you know, through this, through the course of this calendar year, I think markets could get into all sorts of trouble, because not only, as you said, were valuations at kind of huge levels, not only was kind of private money highly levered, private money kind of extremely exposed and that, and that’s part of why this, this downturn over the past three days have been so aggressive. But then on top of that, those high valuations are on the top of earnings in as a share of GDP that were at the top of their historical level. And so if you’re if you try and unwind this, policies that are friendly to blue collar workers, which are more, perhaps reflationary or inflationary, that are kind of anti big, big kind of corporates or anti tech and and that seek to redistribute kind of profits back towards kind of blue collar wages, that means that you’re potentially going to be derating falling earnings. And when that happens, that’s when you can get into all sorts of trouble. And broadly speaking, there’s a there’s a mathematical there’s an economic kind of identity, which was discovered by one of my kind of professors at Cambridge, which was win godley’s dynamic sector balances approach basically says that if you add up the change in government savings or deficit, the change in consumer savings, the change in corporate profits, It equals change in the balance of payments deficit, and what this administration is trying to do is it’s trying to shrink the budget deficit. It’s trying to shrink the balance of payments deficit. We’re highly likely, as a result of those actions, to see savings rates go up, and if that happens, you’ve got a triple threat to profitability, and that could be very dangerous. And that’s not just a three week issue. That’s kind of, that’s going to be, that’s going to be playing out for at least a year, and probably into the next cycle as well. And so I think that what we’ve had is the first aggressive move in what is going to be a longer term bear market, before that reflation kicks in. Once that kicks in, and that may be as early as q1 or q2 of next year, then something very different will happen. It’s then going to be back to the races for resources and emerging markets. But until that point, we’re in a deflationary bust, and I think investors need to trade that way. And
Trey Reik 49:02
in that intervening, you know, 12 month period, you still believe that commodities, generally, and precious metals, perhaps specifically, have the opportunity to provide a little portfolio protection and and perhaps appreciate,
Julien Garran 49:20
yeah, and I think it’s more, it’s more likely that precious metals can appreciate in absolute terms. I think resources are going to do very well in relative terms, compared to the big loser from all of this, which is big tech. So a bit like, I think it’s a bit like trading through 2000 like, imagine today is, is September, October of 2000 and what you wanted at that point was to be short of tech, long of resources, within resources. What went first was the gold stocks in 2001 but then in 2002 when we were still a year and a third. Away from the bottom in the S P 97 of Goldman Sachs broad commodity index actually went up, and miners substantially outperformed, and some emerging markets substantially outperformed tech at that point. So I think more of the performance of my recommendation, which is on one side to be short of tech. I think more of that performance comes from the short tech part over the next 12 months. But I think we’re going to get a gradually increasing performance coming out from the emerging market and resources part probably starting. I would say it’s going to take, probably take a few more months to get to the point where you can predict that the Fed is going to have to start cutting more aggressively than it currently wants to. But once we start getting into that cycle, especially with the prospect of a Trump nominated kind of Fed Chair in 2026 that’s when I think you start getting the gold stock rerating really kicking in, and then, with a lag six to nine months, you’ll get the absolute performance of resources and emerging markets, adding on to that,
Trey Reik 51:09
because many of our viewers are interested, as is everybody in the mag seven and the AI revolution, You have done quite a bit of work on AI, and we could probably set up another hour discussion about this sometime in the future. But why don’t we close out with a little pre seed from you on what you see in the AI bubble and how that is likely to pop over, say, this 12 month period.
Julien Garran 51:42
Yeah, I’ll focus in on two key issues, and the first one is that large language model training has hit a wall. Now it’s quite funny, because Sam Altman tweeted in November last year, there is no wall, even though he’d had internal proof that his attempt his Orion project, which is meant to build chatgpt Five, had actually already hit a wall over a year before. Now, the reason that it’s hit a wall is because the way they built it. So they’ve built these large language models using vectors for every word, and for you and me, a vector would be say, you can say exactly where I am, but putting my longitude, my latitude and my height above sea level, and you can say the same thing exactly where you are by quoting yours. And if we compared our two vectors, we’d know exactly how far apart we were. Now, for large language models, they have these vectors for every single word except they have 12,288 numbers in each vector. That’s just chat GPT three, where we know the numbers. And what they’re doing is they’re trying to work out how words are connected to each other. So, you know, cat, dog, cat and running, cat and running in the garden or garden running in the park, and using that, they can predict the most likely sequence of words in the sentence, and it can come across as coherent. Now, the first problem with that is that if you want to make it better at what it’s doing, and say you want to add one number to the word vector, make it 12,289 you then have to reformulate all the other numbers in that word vector so that they properly correlate with that new number, and then you have to do that for all the other words as well. So the costs go asymptotic, and at the same time, all you’ve done is increased your vector by nought point, naught 8% naught, naught, 8% so the benefits go sigmoid, and that’s where the problem is. And this chart, just here, by Andrew Ord, shows that kind of sharp increase required for compute for very little improvement in accuracy. Now, if that was true, if that’s a correct representation of what’s going on with these things, what we’d expect to see is major AI trainers like open AI, invest 10 times more training the new version than their old one, and then have nothing much to show for it, no real improvement, and that’s exactly what’s happened. So when chatgpt Five was released, they called it 4.5 because they hadn’t shown enough improvement. There’s a good Wall Street sort of journal, insider kind of commentary on that. It was judged to be the sixth best kind of coding. AI, well, you’ve just spent kind of approximately 80 times more, or you’ve got 80 times more compute in something that’s not as good as something that already exists. The Columbia Journalism School found that when you kind of tested AIS on kind of new searches, there were problems with the search 62% of the time. Grok three, which is Musk’s new one that uses 15 times more compute than the previous one. That’s. That’s wrong 94% of the time. And so the problem with this is that you build these things for 10 times more than the last one you build, and you’ve got no advantage over ones that already exist. Now, to anyone who’s commercially minded, that immediately says there is no commercial reason for building another large language model ever, especially as it’s possible to effectively copy them. What deep seek and Ernie have both done is copy using synthetic data from Best in Class models to create a more efficient but just as good model that can operate with a 10th to a 100th of the inference cost. And so there really is no point in training one of these models. And so the danger here is that we’ve built all of this compute and actually we don’t need it for training anymore, and the inference costs have collapsed. So we have years of spare capacity already in the system, and that’s starting to show through the economics of the major players. So open AI spent, or had revenues of 5 billion last year and losses of 8 billion. The data center operators were all heavily loss making. And if you’re a data center operator who buys an NVIDIA chip and then rents it out in the in the sort of gold rush back in q2 of 2023 you could, you could buy an h1 100 for 50,000 bucks, all in fully installed rent it for eight bucks an hour. You make your money back, plus a small return within nine months, and everything after that would be gravy. Latest prices for h1 100 in the states were about a Bucha. 35 you’d make a 72% loss over a year if you bought one of those. And if you want to buy the new Blackwells instead, they’re meant to be, well, they’re probably about twice as powerful, but given that they’re 70,000 all in you’d need to be renting those for 960 per hour, and that’s before tariffs, which are still up in the air on semis. But what’s the point of renting something for 960, an hour, if you could get the same compute for two bucks, 70 an hour by renting 2h one hundreds? And there isn’t any. And so there’s a huge issue here that this is now just purely a sort of keeping the plates kind of kind of circling in the air so that you can say that the principals can get their money out, rather than any kind of commercial operation at all. And I think as soon as Satya Nadella probably had the best inside understanding of what’s going on in open AI out of anyone on the planet, as soon as he said he was going to start to step away from investment, and we saw reports that Microsoft has abandoned plans for two gigawatts of kind of data centers. Two gigawatts is the entire IT load of London and Tokyo combined, right? So this is a massive slashing of investment. As soon as we saw that, basically, that’s, that’s the dominoes starting to fall, and we’re going to see that. When was that? When was that? So that that was, there was a, I think it was it. There’s a broker that did a channel check, Carlson. I think the name it was, and and they reported, and they’ve been basically reporting sequentially since January on the number of cutbacks that that Microsoft is making. And that’s not everything they were doing things before then as well. So this is this is already heavily underway, and that process tells you that we are now going into a we are already in a massive bear market for compute and my so if
Trey Reik 58:49
you can figure this out with all due respect, why doesn’t the market figure it out like there’s some really smart people in this area of the market and industry, and we’re talking about billions and billions of dollars. Why hasn’t the market figured this out yet?
Julien Garran 59:08
I think, I think a significant part of it is that there’s a promoters, kind of behind AI from Sam Altman through to Sati and Adela have been very, very effective at promoting these things. I think the second thing is that when you go and use a large language model, it’s quite easy to get fooled when the model is quite a long way away from reality and what I define reality as is operating in the physical world to survive, basically what what life has been doing for the past 4.2 billion years, language is quite a long way away from that. And so if you read a coherent sentence, you can go, oh, that seems intelligent. Or if you get a summary, you go, Well, that seems. Reasonable, but to actually identify that there’s a problem with a summary, for instance, you have to do the work of reading all the underlying kind of evidence and working out whether the summary is actually accurate. Now, the Australian Government did that and said that AI summaries are in every way worse than humans at summarizing kind of memorandums and emails. And I actually did, I do a quarterly kind of survey of us corporates and and we thought we’d test our our survey or our summary Against what chatgpt could do. And I asked chatgpt to come up with a killer quote, and it said it gave a quote from Jamie Dimon saying something along the lines of the economy isn’t still in good shape, but there are some patches of weakness or something like that. And I didn’t really trust that, because that doesn’t sound like Jamie Dimon. And so I went and reread with a fine tooth comb every word of the conference call that he gave, and he never said that. And so I think it’s when you start going closer to reality, the it starts becoming more obvious that this thing is, in fact, deeply flawed. It can’t do mathematics. It can only regurgitate mathematical problems that have been on its training set. So there was a paper out just just this week where they asked questions from the US maths Olympiad, and the average LLM got about 2% of the questions right. It was from the latest one, so it hadn’t managed to get onto the training set. Yeah, they got about 2% right. And and, for instance, when you try and get it to do an image. This picture here, I asked it to do a picture of the United States, clearly, labeling every single state and spelling each one correctly. And as you can see from that picture, it’s not for 50 so absolutely terrible. But the thing is, you wouldn’t necessarily notice that if you just asked it a kind of boilerplate question on its language bit, the closer you get to it actually trying to operate in real world, the more you discover that this thing is is, in a sense, they’ve hit the wall before they’ve become useful. And that’s why we haven’t seen a mass market application yet that can actually make a profit. And I think that in the way that they do large language models, kind of with all this training, etc, using that system, I don’t think they’ll ever get a mass market product. And the danger is that you’d expect this thing was starting to go south, kind of as people started to realize some of the things we were saying was were actually a real problem. This was going south before this tariff announcement, the chances are that companies are going to step away from this even more quickly in this environment, and so I think this is just going to accelerate the downside.
Trey Reik 1:02:53
Excellent. Well, thanks for that update. So wrapping this all up so you can get back to your day. We’re two and a half months into the Trump administration. You were one of the most vocal writers about the Trump trade before he was inaugurated, and you have a pretty clear framework of what you thought was going to happen. So in this first two and a half months, have things sort of matched or lived up to your expectations, or what, what parts of what’s happened have been a surprise.
Julien Garran 1:03:29
I knew kind of at the start of the administration that there was a that there was a brains trust behind Trump who were very clear on what they wanted to achieve in terms of improving the lot in the blue collar worker. And knew how they were going to do it. They were going to do it via tariffs. They’re going to do it potentially by manipulating the capital account. They were going to do it with energy policy. They were going to do it by shrinking the government. And they were going to do it by continuing to push a an antitrust agenda, kind of against the major, kind of big tech and other companies. What I didn’t know before the administration started was how strong their spine would be in the face of pressure, either from markets or from kind of the broader public, etc. And what I think is really interesting is that they’re not feeling the pressure from the markets, and to your quote from earlier, they’re actually becoming more popular amongst the base that they’re seeking to whose lot they’re seeking to improve. And so for me, I think this gives me even greater conviction that they’re going to follow through with this. Now, it’s going to take some time to work out whether it works, but I think we should be getting a good idea sort of in the last couple of years of the administration, but, but I’m convinced that they that the real Trump trade unwinding the blue collar squeeze. I’m convinced that they’re all in on this, and they’re going. Stay the course.
Trey Reik 1:05:01
Terrific. Well, Julian, this has been fascinating, and I appreciate your generous commitment of time to our conversation. And what we’ll do is we’ll schedule a get together sometime in the fall so that we can compare notes on everything that’s transpired in the in the meantime, see how this all plays out. That sounds great. So thank you very much. Julian, have a great day.
Unknown Speaker 1:05:27
Thanks, Trey. Lovely talking to you.
