Where Do You Hide From an AI Bubble?

Bloomberg Audio Studios, podcasts, radio, news.

Okay, so just to be absolutely clear on this, in order to justify the cost of the investment going into this, in the sense that to generate what we think or you think is a reasonable return on capital, by 2030, users will have to be spending close to 15% of US GDP on AI services. And that seems completely implausible.

Absolutely.

Welcome to Meron Talks Money, the podcast in which people who know the markets explain the markets. I am Meron Somerset Webb, and this week I am speaking with Chris Clothier, co-chief investment officer and co-manager of CG Asset Management. Chris has recently co-authored a white paper examining the reality of the AI capex boom, something we talk about quite a lot on this podcast. and what its fallout means for the global economy. I wanted to get Chris on today to talk about that, to talk about what has happened since he wrote that white paper, and also to get his views on the general trends in markets and economies.

We also sideswipe the UK fiscal situation and budget, although just be clear, we are recording this episode online. on Monday, 26th of October. So before the budget. And so we're going to talk about all that. We're going to talk about gold. We're going to talk about silver, a little on the history of inflation. We're going to talk about what risks are worth taking right now and what risks are definitely not worth taking. Chris, welcome to Merrin Talks Money.

Hi, Merrin. Good morning. Thanks very much for having me.

Yeah, pleasure. Looking forward to it. Now, listen, I wanted to start by talking about this white paper that you wrote. And I know it was a good month ago now, and I know things are moving at the speed of light in AI, or rather in the perception of AI. But I just wanted to ask you to talk a little bit about that paper, and then we can talk about how it is panning out since. Your thought was, and again, as I say, we do talk about this quite a lot on the podcast, but you put some fairly hefty numbers on the extent to which the AI companies have to produce revenues and profits to justify this level of capex, right?

Yeah, absolutely. It was quite a simple analysis, really. We said to ourselves, okay, the forecast is that hyperscalers are estimated to spend about $ 1 trillion a year. But it's kind of amazing that we just throw numbers like that out without even thinking about it. But anyway, fine, $ 1 trillion a year over the next five years. So in five years' time, there'll be $ 5 trillion worth of kit in the ground. What's the expected life of that? Don't know.

But I put a number of seven years on it. Frankly, it could be higher, it could be lower.

So all this kit is fully depreciated after seven years out of date?

Yeah, that seems to me to be a reasonable guess.

And that's just because GPUs are developing so quickly, everything is changing so fast, there are better versions. It doesn't mean that these versions will be unusable. It just means there'll be better versions.

I think that's a fair assumption. And particularly given that power is such a large component of the the cost base of these things and and subsequent gpus are so much more power power efficient than their predecessors that means that they come out of use relatively quickly the cost of gpus is roughly 50 of a data center what sort of gpus plus the kind of the immediate infrastructure around them and then everything else you know the the server racks the cooling the hvac the power all the rest of it that's that's the other 50 percent So perhaps we might say that, I don't know, four years is a correct appreciation period for the GPUs themselves. Maybe you put 10 on everything else and that gets you

back to seven. But the funny thing is that actually, whether it's seven or whether it's 10 doesn't really change the outcome of the analysis that we went into. Not in any meaningful sense. Okay.

Got that sorted. Onwards. So what does the analysis tell us?

So that means that you've got an annual depreciation bill of about $ 780, $ 790 billion per year, just for the hyperscalers. If we assume that they've got an EBITDA margin of 55%, and again, we got to that by poking around in the segmental reporting of the hyperscalers themselves. They don't, of course, break it out as neatly as you would want to.

So there's a few assumptions that we've made to get there, but fine, 55% seems fairly reasonable. Then in order to get yourself back to a 13% return on invested capital pre-tax, which is admittedly a number that I plucked out of the air, but seems like a reasonable target, especially when you bear in mind that, say, Alphabet's return on equity and return on invested capital is well north of 20% and has been for the last five years. So 13% is a much, much asset heavier business model. Then in order for that all to make sense, you need revenues of $ 2. 75 trillion a year in 2030. Real money. We're starting to get to real numbers here, right? But of course, that's just the cost of the compute. So what is

the revenues that, as it might be the frontier labs, the anthropics, the making wild assumptions here, but their gross margins today, open AIs are in the sort of 35 to 45 area. Anthropic claims a much, much higher gross margin, but also inevitably it's an adjusted figure. I guess we'll have to wait until we see their S1 before we really know what it's like. But anyway, let's just say for the sake of argument that they can operate better than open AI is today. So at a 50% gross margin. Well, that points to $ 5. 5 trillion of revenue. So putting that into context, GDP in the US today is about $ 33 trillion.

Let's imagine that we roll that forward to 2030. It's probably going to be nominal GDP might be about $ 39 trillion, let's say. And so, you know, math isn't very difficult to do, $ 5. 5 trillion. is about 14% of GDP, which, to put it in context, is a little bit less than what Americans spend on food today.

Okay, so just to be absolutely clear on this, in order to justify the cost of the investment going into this, in the sense that to generate what we think or you think is a reasonable return on capital by 2030, users will have to be spending close to 15% of US GDP on AI services. And that seems completely implausible.

Absolutely. Now, why don't we immediately turn to the objections really quickly. So, of course, I'm not quite comparing apples for apples, because I'm saying that all of this spend needs to be serviced, as it were, by American consumers. And of course, as we know, the hyperscalers are global businesses.

But then equally, it's also true that There are lots of other people that are investing in AI infrastructure around the world. By the way, there's lots of other ways in which I might be wrong about this as it happens. Joachim Clements, who's an analyst who I respect hugely, he was very reluctant to put a central figure on it.

He bookended it at $ 2 trillion to $ 12 trillion. So, you know, we're sort of in the ballpark here.

Although he's very much on your side when it comes to this as a bubble, isn't he? Because he's very suspicious of the idea that LLMs are going to be the long-term answer in the AI world. And he's a big believer in the rise of SLMs and different types of competition to the what's beginning to look almost old-fashioned data center LLM model.

Yeah, absolutely. But in some senses, I don't think that's actually an objection to this because the sums of money that anthropic and open AI are spending and then in turn that the hyperscalers are spending on their behalf are so enormous. And then there is that very interesting question of what if all of this spend gets swiped off at the knees by the SLMs or indeed the Chinese model's or the lower cost open weight models and so on and so forth. And that's a whole separate question, which again, I think is really interesting. And admittedly, I'm probably not qualified to talk about.

So Chris, since you wrote this, skepticism has really risen in the market, hasn't it? And we're moving very fast into an era where other people are looking at all the numbers and looking at your numbers and saying, gosh, can they really make this revenue? And all the hyperspecialists beginning to go cash flow negative, which is such a big change from even a couple of years ago. There is a changing perception in the market of how this boom or bubble, whatever you like to think of it is, is going to unfold?

Yes, 100%. And the first thing that we have been seeing over the past two to three weeks, which I think is the obvious thing, is, as it were, the winnowing out of the weakest members of the herd. Because the conclusion that we come to is that it's unlikely that all of this spend is likely to be systemic for the hyperscalers themselves. Because one, they start with such pristine balance sheets. And two, they've got these extraordinary core franchises, the likes of which the world has never seen before.

But on the other hand, either the neoclouds.

Define a neocloud for us, please.

Sorry, yes, I beg your pardon. So the neoclouds are essentially businesses that do nothing else but provide compute to AI labs. And many of them are repurposed crypto miners. And so the most prominent of these is a business called CoreWeave. And the thing that we've seen over the last couple of weeks is the cost of its bonds falling in price and therefore the cost of its debt rising up to sort of in the region of 15%.

And I think it's highly unlikely that at that cost of capital that it can make any sense for them to continue investing.

But these are interesting businesses, the near clouds, aren't they? I mean, they're opportunistic upstarts. They saw an opportunity where there wasn't enough compute. They dive in, borrow enough money to do it. I mean, you know, admirable, very admirable, hugely entrepreneurial.

Absolutely. And in all investment booms, it's perfectly rational to try and jump on the front of it and see if you can fashion something valuable. I think the other thing that we're starting to see, though, is that more established businesses are showing some signs of stress. So I think Oracle is probably the pivot point in all of this.

It doesn't have the same kind of core franchise that the hyperscalers have. It has always run its business with a much higher level of indebtedness. And so even that is starting to show signs of a little bit of stress.

So all in, it looks like it might end sooner than you thought, even when you wrote this piece, given the doubts creeping into the market as a whole. Do you think that is the case, that we might see this boom bubble coming to its end maybe even this year? And if it does, is there systemic risk in that or is it okay?

I think it has to be right that there is a chance of this comes to an end. And we're seeing all sorts of different elements of that. So partly it's the political resistance to the build out of these data centers, which may be the thing that causes this whole thing to slow down. And then on the other hand, it's the availability of credit to do this. And by the way, it is just worth touching on this because the numbers are absolutely wild. So I was talking to a credit trader and we've seen some of the issuance that's been going on in the bond market this year.

And so in round numbers, Google and Amazon both came to the UK market and the Euro market and issued sort of four to five billion sterling and 10 or 11 billion euros each. He reported that some large fund managers in London were getting sounded out by syndicate desks of investment banks about the prospect of raising 100 billion euros for Meta in a single transaction. And those numbers are just absolutely wild. And so I think it is possible that if the availability of credit just isn't there, then all of this does slow down. And then actually, I think the fallout at that point Setting aside what might happen to the neo clouds and perhaps some of the weaker players. I think the fallout would be relatively

Limited because actually what AI compute has been deployed is actually likely to be pretty valuable Because as we've seen the demand for this stuff is very high so whenever you listen to the quarterly earnings reports of Microsoft of Google their CFOs are saying, look, we've got these huge backlogs of demand. And so all of the investment that's going on today is entirely rational. But it's this sort of runway of kind of unlimited investment that strikes me as being kind of economically insane, frankly. Yeah.

But with that rising cost of credit, with all the doubts, you have to look at the equity market as well and say that if this comes to a juddering halt, we may also see, because lots of the rises in the AI-related companies and the industries have been based on very sharply rising earnings. And if we say, well, do you know what? Maybe this wasn't a valuation bubble. Maybe this is kind of an earnings bubble. And then you do begin to see prices fall. And if that happens, there's a massive wealth effect in the US, right? In that people's wealth is more tied up in the equity market.

This is a historical high. So if you do start to see an AI and AI-adjacent equity-led fall in the stock market, that's going to feed into the economy relatively quickly, or not?

100%. And so I think that that's the... Perhaps we don't have a financial crisis, but it's entirely possible that we have a nasty recession. And just to put some numbers on that, we tried to come up with an estimate of the profit pool that would arise from this $ 1 trillion capex spend. And we came up with a number of about $ 400 billion. because of the fact that particularly the chip companies have such high gross margins. Chip companies' proportion of S & P earnings this year is 10%, and their estimated share of S & P earnings next year is 17%. And so you're absolutely right that any major slowdown as a result of availability of credit would hit them. So the first order impact on S & P earnings would be very great. The second order effect

is that so much of the fantastic earnings that we've seen of the banks, for instance, has been wound up in trading revenues and their fixed income origination, but their DCM businesses and so on and so forth. And as you say, the US economy has never been more tied to the stock market. And also that the savings rate in the US is really very low at the moment.

And so we're at this very interesting point where the three major pillars of society are are all essentially on a spending splurge. So obviously, the federal government is running a 6% budget deficit. Corporates, as we've seen, are investing on a scale that we've not seen since, I don't know, perhaps the railway boom. And then the private sector is Actually, as it happens, the savings ratio just got written up by the BLS. But depressingly, that's just because it turns out that the rich were even richer than we thought they were. And obviously, they have a tendency to save.

But nevertheless, as you say, the savings rate is well below its long-term average. And so you could entirely see that the savings rate would rise very dramatically if there were a major earnings slowdown in the stock market. And I don't see where the kind of classic fiscal stabilizers can come from. I can't see that the federal government can start spending even more. And so, yes, that could be very unpleasant.

And would that be very deflationary? I mean, your main worry, isn't it, is inflation. A lot of the portfolio is in index-linked bonds, for example. So when you think about your major concerns for the future, looking at the portfolio anyway, and from our previous conversations, your main worry is always inflation.

But this kind of scenario... would give us more of a deflationary environment.

I think you're absolutely right. That could well be deflationary. I guess, on the other hand, we've got a whole bunch of other inflationary things going on in the economy. So quite how those two things would play out against one another, I don't know. But yes, a sharp slowdown in the AI boom would definitely reduce inflation.

Okay, but your core case remains medium, longer term inflation globally.

Yes, I think so. In some senses, it's not a terribly sophisticated piece of analysis. If you go back to the Second World War, since then, every major surge in inflation has come from some combination of an energy price shock or a war. And presently, we've got both of those situations in play. And then there's just the general fact that we're seeing the sort of process of deglobalization starting to take hold. And when you combine that with rising wages in emerging economies, that just makes it very difficult to see the level of goods deflation that we've seen over the past 25 years persisting. And that's before we start getting into the impacts that war has on the physical destruction of capital and then the realignment of

trade flows And all of that is very inflationary. Yeah.

And then I suppose there's what all countries have to do to protect themselves from the consequences of wars and supply chain changes. So obviously there's this conversation in the UK about defence spending, which also has to take in defence from cyber attacks, defence of our cables under the sea, etc.

All these things are extremely expensive.

Yeah, I guess the point about defence spending is that it is economically non-productive. Now, it's clearly desirable. But it doesn't actually do anything useful. It doesn't increase the productive capacity of the economy. And so, yeah, all money that is diverted away towards that. And obviously, and it consumes resources.

So put all of that together and that's an inflationary backdrop. Absolutely.

So if we look at the Cabell Gearing Trust, which I think probably quite a lot of our listeners hold, how much of that portfolio is now in assets that are specifically designed to protect from inflation?

Just under half. So we've got about 48% in inflation-linked bonds today that split roughly 50-50, but tilted towards the U.S. into U.S. tips and U.K. index-linked bonds. By the way, actually, coming back to your point around whether or not it does turn out to be a deeply inflationary environment, the interesting thing about inflationary bonds is that they've massively outperformed nominal bonds since they've been in existence. Tips came into place in the late 1990s. So since 2000, TIPS, which are Treasury Inflation Protective Securities, US inflationary bonds, they've outperformed their nominal counterparts by 1.2% per year. So that's cumulative outperformance of over 90%. So that's been an era when

inflation has actually been really well-mannered and under control. And yet even in that environment, they've done so much better. So we don't necessarily even need for it to be this deeply inflationary environment for them to be a sensible place to have some of your money.

And what about gold? Is that part of your inflation protection?

So we currently don't own any gold.

No gold?

No gold at all. Not one coin. Sharp.

Intake of breath from all my listeners, by the way. Absolutely.

Now, obviously, as it turns out, we should have owned a lot more gold than we did.

On the way into the mini bubble.

On the way into the mini bubble, absolutely. But when it got so, and I suppose we sold this probably in about February of this year at about 5,100, what we did have, which was a little over 1% of the portfolio. And the reason we did that was that we saw that gold was starting to behave like a speculative asset. that just felt weird and fundamentally wrong to us.

And so we were happy to take it off the table there.

Let's look then at the equity part of the portfolio, which is pretty small, right? If you've got under 15% of the portfolio now in equity-related stuff, a lot of it is infrastructure, real estate. You talk us through that.

Yeah, absolutely. Essentially, we felt for some time that equity markets were stretched. And obviously, we were concerned recently about the AI boom or bubble or whatever it might be. And so we were looking for, as it were, places to hide. And I guess infrastructure has been, we think, a really good home for us. And so this is some of the listed investment trusts. So Hickel, IMPP, 3i Infrastructure.

But essentially, making reasonable assumptions, we think that these things offer nominal returns of the order of 9% per annum, and they've got pretty good inflation protection. So they're going to deliver something like 5% to 6% real. And that feels like a really nice place to be and something that we think our investors would appreciate.

So it's been the wrong call for the last few years to have so little in equities across the board, right?

I mean, it absolutely has. But the trouble is, of course, is that we have a We have a dual mandate, much like a central bank. And the first part of our mandate is to avoid losing money. We define that as avoiding losing money over 12 months as being fundamental to what we do. And then the second leg of our mandate is to generate returns above the rate of inflation.

Something that's made it weirdly more difficult to own equities for us is that we have increasingly felt, because of our concerns about fiscal pressures... we've been very reluctant to own long duration bonds. And so our portfolio duration is very short today. And that's broadly been borne out over the past six to nine months or so. But of course, what that means is that if you don't think that you can find a good hedge to a portfolio of equities, you need to have commensurately fewer of them if you're trying to avoid losing money.

When do you get to the point where you would be prepared to buy more longer dated bonds? bonds. I mean, obviously yields have gone up a lot, but there's still a sense there might be fiscal crisis ahead in the likes of France, possibly the UK. What's the kind of yield that bring you back in? Or is there, given the appalling fiscal situations of some of these countries, is there no number that would make it okay?

There has to be some number where we would start dipping our toe in the water, where the returns were sufficiently attractive. But essentially, we're not interested in owning long-dated bonds until the fiscal zeitgeist changes. And we think that the only thing that is going to bring about a change in the fiscal zeitgeist is a bond market crisis.

And so we're waiting until there is that crisis before we're prepared to go. I mean, sure, absolutely. If there's a slowdown in the economy, I'm sure that yields will come down. But we think we're in a secular bear market for long-dated bonds until this changes. Obviously, there are several poster childs for this kind of fiscal irresponsibility, France being but one.

We are talking before the UK budget, but this podcast is going to be released after the UK budget. So I'm almost certain that whatever you say in response to this question will not have happened. But for the purposes of the conversation, what would have to happen... in this budget for you to think to yourself, yeah, all right, then I'll buy those long-dated bonds in the UK. That'd be absolutely fine. Those 30-year gilts, I'm all over them.

To be honest, there's nothing that the Chancellor could do that would make me want to own 30-year gilts. And the reason for that is that the UK is, as we know, a small, open economy, and increasingly few people need to own long-dated gilts. As we know, all of the defined benefit pension funds are closed and in healthy surplus.

And so what that means is that the single biggest thing that drives UK yields is what happens to American yields. And equally, I'm absolutely sure that if we have a bond market crisis in France, I'd be, well, absolutely sure, I'd be surprised if the UK were seen as a safe haven against that backdrop. And I think, I mean, I'm sure that yields would outperform, but I think that the risk of rising yields, because these things just trade in such sympathy with one another these days.

Okay, so... In some ways, the UK isn't really investable.

I find the conversations around fiscal matters in the UK incredibly depressing. And obviously, I hope that I'm going to have been proven wrong by what comes in the budget this week. But I think that's fairly unlikely.

There's one thing I wanted to go back to that we were talking about earlier, which is more a comment than a question, but it's an extraordinary situation, isn't it, that we have got ourselves into globally where there is now suddenly worries about the availability of credit. We've gone through decade after decade after decade, where pretty much anyone can have as much money as they wanted at any point.

And we interviewed a while back the author of 1872, 1873, whoops, suddenly lost all my years, when there was this sudden realization as a result of a massive fundraising in France that credit was available. There was money everywhere. You could have as much money as you wanted to invest in the railways, et cetera. Of course, that ended in tears.

And it slightly feels like we've gone through something similar, a period quite long and then quite short when you can have as much money as you want for anything. And now suddenly we're talking about governments being crowded out by AI and AI companies not being able to get their hands on the money they want and the cost of credit for neoclass going up to 14, 15%. I mean, it feels like such an enormous and such a quick change.

It does, doesn't it? It's kind of all happening so quickly that I don't know if this is just one of those sort of brief little blips that markets have. or if it's a kind of a fundamental realization. But I think it comes back to that thing that I was saying earlier, which is that we have this strange situation when all branches of society are essentially going on a spending splurge. And so the competition for capital is real.

I've just been on, I took my children to Greece for half term last week, and I finished.

Perfect time.

It was absolutely delightful. And I finished Liaquat Ahmed's book. It And I'm going to get it wrong as well. Is it 1872 or 1873? Anyway.

Now we're both confusing ourselves. I think 1873.

I think so too, but I couldn't say for sure. I mean, I think it's an absolutely brilliant book. And of course, in that sense, it's very redolent, isn't it? Because you had, as you say, France raising huge sums of money in the bond market to pay an indemnity to Germany. You had Egypt and Turkey raising money essentially to hand it over to their despotic rulers. And then you had all of this money being raised to finance the railroads. Now, I guess the thing that was probably different then was that the savings rate in countries like the UK was probably quite high and was, as it were, the source of capital for all of this. And whereas I think the thing that's interesting is that the savings rate,

you know, at least across the world as a whole, is pretty low at the moment. And so, yeah, I think this could all come to a shuddering halt. But of course, as ever, it's impossible to predict as to timing.

So what risks are worth taking at the moment? What do you think? Look at us working ourselves up into this misery of bearishness. There must be risks worth taking.

I do think that real yields in government bond markets are fundamentally very attractive compared to history, compared to long-term growth rates. And so while we are very concerned about a fiscal crisis and what effect that might have on the long end of the yield curve, presumably if there is a fiscal crisis, and or if there is the sort of the AI slowdown that you and I have been discussing, is that short-term interest rates will get cut.

And so we are very constructive on the yield curve, both nominal and real, going out about as far as five, six years. The reason for that is partly just that here in the UK, we've got, what, four rate hikes penciled in over the next 12 months or so, even with the recent elevated inflation print that we got in the UK, I still think that looks overdone. And obviously there's a limit to how far the UK curve can steepen if short rates are being brought down. So I do think that you can own, say, five-year nominal or index-linked gilts with a huge amount of comfort. And then by the same token, while it's true that credit spreads are very tight at the moment and i think that there is a reckoning to come in in credit markets

of course if you're buying relatively short duration credit the price risk that's involved in that if spreads widen it's relatively modest so that's a perfectly sensible place um to be um but yeah i'm afraid i'm i i i find it hard to be excited about anything um at the moment if i'm okay how.

About this how about this i think you can do this what what price would you buy gold against.

Three and a half.

Three and a half? You think it'll go that far?

I have absolutely no idea where it'll go.

But that's where you buy it again.

Yeah, absolutely. I mean, you know, the trouble with gold is that, of course, I mean, I just, I hate it as an asset.

No income, hard to value, prone to little bubbly things.

Yeah, absolutely.

But, you know, it's good enough for central banks, right?

Absolutely. Le Coran's book, 1873, illustrates the danger with these things, which is that silver was, as it were, as good as gold. And then all of a sudden it wasn't.

And then it wasn't anymore.

And then its price fell by, I can't remember, but it was sort of over the order of 75%. I mean, admittedly, that's not a prediction that I'm making. I see no reason why central banks will give up on gold, not least because it would have the effect of bankrupting them, I suppose. Okay.

What could go right? What could go right? Globally, geopolitics, AI bubble, what could suddenly make you and your colleagues wake up in a month and go, do you know what? Let's take that equity exposure up to 17%.

It's difficult because the things that would make us much more positive about the world are likely to involve short-term pain. So, for instance, if you saw that Western governments really trying to get a grip of their fiscal situation, well, of course, that's going to start with a nasty slowdown. But for me, that's probably the precondition the feeling that we're on a more sustainable footing. I'm sure you're very familiar with the Kaletsky equation, which essentially says that large government deficits show up in corporate profits. So corporate profits are artificially elevated by the fact that essentially we have all of this government spending. Now, of course, that's a situation that has gone on for a very long time, and I've no

reason to suppose it won't carry on going for a very long period of time. And yet, if it does come to a crashing halt, then that's very nasty for equity markets.

And one day it has to, one day it has to come to a crashing halt. I mean, government spending can't go on forever. Something has to end at some point.

I would agree.

Just when, just when.

Absolutely, absolutely. It's really difficult when you are looking after people's money, as we are, and it's something that, it's a job that we take very seriously. It's a great responsibility. I wouldn't be able to look as it might be my mother in the eye, because she has a lot of her money in capital gearing trust, and say, well, you know, mother, that I was going on about the fact that these governments were being completely profligate, and that I knew that it would come to an end one day, but I couldn't figure out when. So I thought, ha, what the hell, I'll just, you know, I'll just pretend like it's not happening. That's not good enough.

I bet your mother's got a sneaky position in Scottish Mortgage. It's got a secret Hogwarts Ransom account you don't know about. Fully invested in tech.

Do you know what? I have not asked her that question. I will phone her up after this one.

Ask her. Ask her.

And I will report back to you, Maren. Absolutely.

Chris, just one more thing on inflation. I note that you often make some excellent historical comparisons. And in a webinar you did recently, you had talked about the price of sesame seeds in Babylon. Why is that relevant to today? Babylon in 242, by the way.

Yes, I have a sneaky suspicion I might have got the date wrong.

We're getting all sorts of dates wrong today. It doesn't matter. We're after the ideas, not the dates.

So that was part of the first, at least recorded, inflationary boom. And I suppose it would definitely be grist to the mill of Milton Friedman, who said that inflation is always and everywhere a monetary phenomenon. Because when Alexander returned to Babylon... After his semi-successful military campaigns on the Indian subcontinent, he returned there and, of course, ultimately died there.

And that coincided with this massive rise in the price of sesame seeds, wheat, and barley. I think sesame seeds, I think they went up tenfold over a short few years. which presumably was disastrous for the price of people's tahini and yogurt breakfasts. And I suppose while we don't actually know why that happened, I think we can probably guess that it was just a massive influx of silver into Babylon at that time. And so anyway, yes, that's how that all came to pass.

Yeah, it always happens in the same way, doesn't it? It does.

It does, absolutely.

Chris, last question. question. We've just been talking about your Greek holiday and what you've been reading there. What are you reading now or about to read?

I just finished Evan Moscovich's memoir. He was the American journalist who was falsely imprisoned in Russia, essentially to provide currency for prisoner exchanges with Putin wanting to get back people who have been doing his dirty work for him in the West. So I read that, which is about the sort of creeping autocracy in Russia. So that was fascinating.

And this morning, I have just started a book on Bayesian economics, which is something that I've always struggled with. This is a book by Tom Chivers, which I think will hopefully give me a better sense of how to predict the future, which might be useful.

Okay, predicting the future would be so handy, wouldn't it?

Absolutely.

It would really work out well for all of us. Chris, thank you so much for joining us today. I really appreciate it. It's been fun.

Thanks very much, Mary. Lovely to see you.

Thanks for listening to this week's Merrin Talks Money. If you like our show, rate, review, and subscribe wherever you listen to your podcasts. And keep sending questions or comments to merrinmoney at bloomberg.net. You can also follow me and John on Twitter or X. I'm at Merrin S.W. And John is John underscore Stepek. This episode was hosted by me, Merrin Somerset Webb. It was produced by Sam Asadi and Moses Andam. Sound design by Blake Meebles. A special thanks to Chris Clothier.

Where Do You Hide From an AI Bubble?

Kapitelklicken springt direkt hin

Shownotes

Transkriptvom Podcast bereitgestellt · Zeitmarken springen hin