Trade wars, shooting wars and the AI investment boom

It's the 24th of July and this is your capital economics weekly briefing. I'm David Wilder and I'm joined by Neil Shearing Group Chief Economist once again to wade through a very busy week in global macro. Hi Neil. Hi David. Hi yeah, I thought we'd start this week. I'd give you a choice. Do you want to talk about first the the shooting war or the trade war? Where do you want to start? One an unenviable choice that that is. And somewhat with some point we're going to talk about the productivity miracle, but I don't think it's going to be this week is it? Let's talk with the trade war because we've talked at length haven't we abaid the struggle for most. We can obviously oil back through a hundred. But let's talk about the trade walks. I think we're going to hear more about trade in tariffs over the coming days. Yeah, I mean, the headlines are saying more than 80 countries hit with tariffs, US tariffs of at least 10%. And it turns out the Trump administration is very concerned about forced labour in in supply chains. What's the context here? How much of a problem is this for the global outlook? We're going to get into an alphabet soup of different trade and tariff laws soon. So brace yourself. But if you remember, go all the way back to liberation day. We had tariffs put in place on a range of different countries under the International Emergency Powers Act, IEPA. That was the billboard of tariffs that Trump had that he unveiled in April 2025. The Supreme Court struck down those IEPA tariffs in February. Now, at that point, tariffs were bringing in revenues equivalent to about 1% of GDP on an annualized basis to the US Treasury. So this, of course, in the context of a federal budget deficit in the order of about 6% of GDP. Potentially facing quite a wide expansion of the budget deficit from an already large level. So rather than allowing that tariff revenue to complete the atrophy, the administration imposed so-called section 1 to 2 tariffs that was a global tariff 10% on all countries levied for 150 days. That's due to expire today, as it happens. So now that gave them some breathing room essentially to resurrect the previous IEPA tariff regime, but on a more legally durable basis, hence the so-called section of 3 or 1 tariffs that were seeing coming into place. Now, those have been levied supposedly on the grounds, as you say, pushing back against countries not doing enough to combat forced labour. As it happens, most countries are getting 10% tariffs, some are getting 12.5%, we don't have the full list yet. That probably returns the US average tariff rate to about 9.5%, which is where it was before the Supreme Court warning currently it's about 7%. Is this a threat to global trade? Is this more friction in the global trade machine? Well, a couple of points I think worth stressing. The first is that although the average tariff rate to the moment's just over 7%, it's going to head back, we think, to about 9.5%. As I say, there's some pretty big differences between average tariff rates on different countries. So if you look at China's average tariff rate at the moment, it's just over 20%. If you look at Canada's in the news this week with renewed tariff rates from the US, Canada actually the average tariff rate now is around 3%. So big difference there between the media is potentially reporting this, which is China's off the hook. There's a repression between Trump and Xi, but actually the average tariff rate on China is well above any other country, whereas Canada, which is in the hitting the headlines for all the wrong reasons when it comes to tariffs, the tariff rate is still comparably low. So big differences between countries, and that gets the second point, which is that global trade has been pretty resilient through these tariffs. That's for a couple of reasons. One is that the differentiation between tariff regimes has allowed some re-routing of trade to circumvent the very, very high tariffs, for example, on China and other countries. So we've seen some re-routing going on, but also other countries are not retaliated in a meaningful way to US tariffs. And of course, this tariffs are not the only game in time. We've also got an AI boom and an AI investment boom going on that is incredibly trade intensive. And so we've seen a surge in particularly high tech exports, semi-conductives and hardware around AI over the past 12 months that continue to explode despite the imposition of US tariffs. What is the goal here of the Trump administration? We've been talking over the months, haven't we, about the motivations for building up this tariff all in the first place. You kept mentioning revenues there. Revenues come into federal coffers from these tariffs. Is it any clearer what the goal is for the Trump administration? And can we expect tariffs to rise much further from here to the point that perhaps global trade is harmed? As ever, it's as clear as mud, I'm afraid. There were three broad justifications from different parts of the administration that were given for the imposition of tariffs. The first was that these tariffs were supposed to go into help to combat global trade imbalances. We were always very skeptical on that and so it has proven to be the case. The second is that the imposition of a tariff wall would help to resure manufacturing back to the US equally. We were skeptical on that basis. We said this perhaps scope for some limited reshoring in certain sectors like semiconductors and high-end advanced manufacturing. Some evidence that it's happening in those sectors, but certainly not on a broad and widespread level. So second objective and justification certainly not met. And then the third one was that it would bring, as you say, some revenues into the Treasury coffers. And on that front, I think the administration could claim to success. As I say, revenues are equivalent to about 1% of GDP on an annualized basis before the Supreme Court decision. But of course, that's just attacks on really attacks on US consumers. It's not attacks that's paid for by the exporters to the US. So it was bringing some revenues into the coffers of the Treasury, but those revenues were really being paid by US consumers. Now if we look ahead and think, well, what's the objective here and where are we going? We don't have much in the way of a kind of current intellectual framework from which the administration is working from. But Treasury Secretary Besson has suggested that the ultimate goal here is to try to resurrect the tariff regime that was in place at the start of this year, that is to say before mourning against the Ipa tariffs. Now if that's the case, it would suggest, as I say, the average tariff rate going back to around 9 to 10%. But also some, again, some big dispersions between different tariffs on different countries. So China at the start of the year had a tariff rate of close to 30%. India was much higher to currently its under 10. At the start of the year, it was close to 20. And other Asian economies have seen the biggest falls, some of the biggest falls in tariffs following the Supreme Court decision. So if you take Secretary Bessons at his word, I think the objective is to try to get the average tariff rate back up to where it was at the start of this year. And potentially most of the heavy lifting on that front will be done by raising tariffs on countries in Asia. So we'll see how that plays out. But that is as clear an indication as we have as to what the ultimate goal is here. In terms of the inflationary implications for the years of all these tariff rates, has the inflation from that already passed through, is the Fed going to be looking at this with any degree of concern or are there bigger fish to fry here? Well, I think there's bigger fish to fry at this stage. And there is wrong to say that there was no inflationary effects of tariffs. We've had, obviously, had people in the administration, including the president himself saying there's no evidence of higher inflation as a result of tariffs. That's not the case. We did see a rising goods inflation last year. It just happened to be offset by a slowdown in services inflation. So that there was a rising goods inflation in the US as a result of tariffs. But that has now broadly washed through. So I wouldn't expect to see a big surge in inflation as a result of these new tariffs. No. It's also fair to say, however, that new avenues for imposed tariffs are opening up all the time. So we have these section three or one tariffs that are coming in at the moment. There's some additional three or one tariffs coming down the track based on structural excess capacity in industrial sectors. Those investigations have yet to conclude. But when they do, as I say, I suspect it's going to be countries in Asia and China in particular, they're in the firing line. And if that's not enough to keep on top of, we've got now into the mix. We've got section three, three, eight tariffs, which was the piece of legislation that the administration used to threaten those tariffs on Canada earlier this week. That's actually part of the the Snoop Hawley legislation from the 30s. Now, if that has a legal footing, many scholars or some scholars suggested that has been superseded by the IEP legislation. But if the administration can impose tariffs using this three-to-rate mechanism, then that could open up a whole new avenue for imposing tariffs where the president potentially does have the authority to impose tariffs on individual countries at his whim. So we've not heard the last of tariffs. We're going to hear more on this over the summer and over the second half of this year and into 2027, I suspect. But so long as other countries don't retaliate, and so long as we see a continuation of this AI investment boom, I suspect the global trading system will just about hold up. Alright, let's move on to those bigger fish that the furs half have to fry. I'm looking at headline here, Trump says he's considering ordering an around attack that's quote unquote bigger than ever before. Oil was back up above $100. Yes, you're on Thursday. It's a little bit below that now, but gas prices hit $4 again in the US this week. When we spoke last week, oil was at around $85. At what point should we start getting worried? Well, this goes back to the conversation that we had last week, doesn't it? I think very rarely are there tipping points in markets or economies where suddenly you start to become worried. Last week when we spoke oil was at $85, we discussed how given the rundown in inventories, how there was a much capacity within global markets to absorb another supply shark and that we could see as a result, quite sharp increases in price if the straight remains closed and if there's renewed threats against shipping in the red sea. Since then, we've seen another $15 a barrel on oil. I think it's quite plausible that if that continues over the next week or so, we see another $10, $15 on the price of oil. I think it's not like someone's going to happen on Saturday. We reach a tipping point and suddenly there's $30 on the price of oil. I think we're in the phase where we see a gradual ratcheting up of pressures and constraints in the physical market and we start to see prices continue to grind higher and that's what we've seen over the past week. I suspect that's what we'll see over the coming week if both sides can't find an off-ramp. We've got Fed meeting in the coming week, Bagvingham as well. We just had an ECB meeting yesterday on Thursday. We had the statement, we had a press conference from Christine Lagarde. Did we learn anything from that in terms of how central bankers are working through the uncertainties around what's happening in the Middle East? Almost nothing. Bit of a snooze fest from the ECB. They were expected to keep rates on whole. They did keep rates on whole. They gave almost nothing away in the press statements and nothing in the press conference either. The ECB is packed up. They've gone to the beach for the summer. There's a sense, I think, that central bankers are just bracing themselves through the summer months and we'll see where we are in September. We have over the coming week, Bank of England, the Fed. I'd be very surprised if either moved, particularly the Bank of England, the markets at the time that we're talking about a 30% chance of a hike by the Fed, which is a big turnaround in market pricing over the past week or so. I think it does speak to the fact that there are stronger underlying inflation pressures in the US that we've discussed before. Even so, my base case would be the Fed, if things stay where they are, the Fed hiking in September. Even then, I think it's a higher bar for the Bank of England and the ECB to start hiking. We've seen big moves in Bondields across developed markets over the last few days. Given all that you've said, do those moves seem reasonable to you? This idea that investors are repricing the risks of higher rates? Does that fit in with the capital economy? Yes, and no, as ever, for an economist. Where does it seem justified to me? The US, we've argued for a while that we think that the Fed's not done. It has a bit more work to do. The market had obviously entered this year to expect to rate cuts. It's now started to factor in hikes, but we've got a few more hikes and we've got three hikes in our profile over the next 12 months. I think higher yields in the US justify Japan, too. We've been consistently hawkish on the Bank of Japan. We've argued for a while that the market was not pricing sufficient tightening by the Bank of Japan. I think the rising yields there, which is to some extent, crept under the radar because everyone's focused on the weakness of the end, but the rising yields, I think, is justified. Perhaps it has a bit further to go. Where I think it's more questionable is in Europe, would the ECB in the Bank of England? Because economies are weaker, particularly in the case of the UK. Now, as it happens, Q2 data for both the eurozone and the UK look okay. It might be the case that both economies grew by 0.3.4% Q1Q in the second quarter, but I'm not sure that that pace of expansion can be sustained going into the second half of this year. Labor markets are weaker, particularly in the UK. There's no indication yet of the indirect effects of high-energy prices becoming an issue. Certainly no evidence yet of second-round effects through wages and broader prices becoming an issue. My sense is that there, the bar to raising interest rates again is higher in the UK and in the eurozone. Now, it might be the case that if oil stays at $100 a barrel, maybe even crux a bit higher, it stays there through September. DCB comes back from the summer break and raises rates in September. That wouldn't be a massive surprise if oil stays where it is. On the basis that there'll be some off-front find, oil prices kind of start to edge back down, my sense would be that both the bank of England and the ECB try to model through and don't raise rates. But the Fed is on its wall path, it seems. The inflation issues that you talk about in the US oil is part of it, but it's certainly not the whole story. And regardless of what's happening in the Middle East, the team has been forecasting that we're going to get these rates hikes coming through into early next year. Exactly. The US economy is in a very different place, so I would argue, for a couple of reasons. One is that it's a net energy exporter, a small net energy exporter, so actually it's terms of trade improve when energy prices increase. So that's a boost to the real economy in a way that it's not in Europe. The second is that there's this enormous AI investment boom going on in the US that's not happening in Europe. And actually at the margin that it looks either that inflationary rather than disinflationary for the time being. And the third is that physical policy is just much more supportive in the US, the physical deficit, 6% of GDP, potentially a bit higher, depending on what happens to tariffs and tariffs refunds. And that requires monetary policy to be correspondingly tighter. So the US is in a different place to the eurozone in the UK and therefore the Fed needs to be in a slightly different place too. How important is this AI investment boom in terms of what's going on in global trade, in terms of the US growth story, in terms of the China growth story as well? Well, it's interesting you mentioned China because we focus a lot on the difference between the US and Europe. We've just focused in our conversation on the difference between US and Europe. But actually the investment boom around AI is providing a significant prop to growth in China now. Now, so to put this into perspective, it's very difficult to say exactly how much growth over the past 12 months has been due to AI. This is more than art than a science. But to put some kind of numbers on it, the US economy over the past year or so has grown by just over 2%. We think about a third of that can be attributable one way or another to the investment boom around AI. So a significant amount of the wedge between growth in the US and in Europe can be explained by AI, the investment boom around AI as I say, about AI investment US growth in the kind of mid ones. At the moment, the eurozone is growing up about 1% a year, something like that. And an AI's bet, barely making contributions to similar numbers for the UK. The interesting development I think over the past week, and I wrote about this in my notes on Monday, is that we've learned that AI is also providing a significant prop to growth in China. So we had Q2 GDP growth from China last week, lots of focus on the weakness of the headline number, drop in headline number. Actually, we wouldn't read too much into that, not least because all it does is bring it closer to our own in high measure of economic growth in China, the China activity proxy. Instead, I think the interesting detail came later when we got the breakdown of the contributions to GDP growth in Q2. And a lot of that seemed to be at least from the production side, from sectors that are benefiting from this investment boom, both domestically and globally around AI. So a similar degree of growth in China, the similar share of growth in China, it can be attributable to AI, as is the case in the US. And I think it takes that back, that makes sense, because these are the two countries that are at the forefront of the AI investment boom. So you'd expect that to be driving a more significant amount of growth in those countries. So big picture, AI and investment around that appears to be responsible for about a third of GDP growth in the US, perhaps a similar amount in China, but barely anything at this stage in the eurozone and in the UK. Final point to say is that all of this so far is just about the booster activity around the investment in infrastructure and frontier models. We're not seeing much evidence, we've seen some evidence, but not much evidence of kind of a widespread pick up in total factor productivity, these wide productivity gains that you might expect from AI. That I suspect will come, but either in the US and particularly in China, we're not seeing much evidence of that happening just yet. And what happens if this investment boom turns to bust? Well, these are important to separate two things out. The first is what's happening in financial markets, and the second is what would that mean for the real economy? Now, when it comes to financial markets, we've long hour. you that yes AI is a general purpose technology. In our view, these are the technologies that do eventually lead to a transformational effect on productivity and boost to productivity, but they're also very fertile ground for bubbles to develop because investors try to capture the benefits of that technology ahead of them actually crystallizing in the real economy. And for that reason, we've been very bullish on US stocks and in particular US tech stocks. Now, that's played out, but actually it's happened because earnings growth has been extremely strong. Now, your question was, well, what happens if that investment boom turns to bust? I think the interesting development over the past week or so has been the rapid development of frontier models in China, which of course are open source. And that I think is starting to feed concerns that actually if China's catching up to the US, indeed, in some cases, may even be overtaking some US frontier models in terms of performance. What does that mean for the revenue streams of the likes of OpenAI and Anthropic that are at the forefront of large language models in the US? Potentially means a bit less for the hyperscaders because we're going to need compute power and data centers and what have you come up with, I suspect. But that I think is the interesting development. I think that is something that is starting to weigh on in investors' minds. Now, for the real economy, what happens if boom turns to bust? Well, my sense is that we're going to need in the real economy anyway, an enormous amount of computing power, data center capacity, come what may. Yes, it might be the case that ultimately, some of the firms that are developing large language models, maybe they have kind of overvesting and maybe some of that investment doesn't produce the gains that are being anticipated, the returns that are being anticipated that crystallize some losses. But for the real economy, actually the roll-out of technology often involves massive investment, the oversupply of that technology in capital that depresses the prices, you can then get real investment going up even if the nominal price of the capital goods goes down. So I'm slightly less concerned about the impact in the real economy other than some financial fallout if boom were turned to bust. Clearly, there would be big implications for the companies involved, but a real economy level, it might be manageable. Neil's sharing there on a jam-packed week in Global macro and another busy week to come. I'll add our Fed and Bank of England previews to the podcast notes, as well as our latest analysis on Trump tariffs and their global implications. We've got a drop-in about these new tariffs this coming Monday at 11 a.m. New York, that's 4 p.m. London. Drop-ins are our short form online briefings and we run a packed schedule of them for capital economics clients. There are five in the coming week alone, alongside that tariff briefing we're holding drop-ins on everything from the outlook for construction in the UK to the threat that those Chinese frontier models that Neil was talking about posed to US tech firms. Details of all of those drop-ins are on our events page, capitaleconomics.com/events. And if you're not yet a capital economics subscriber and want access to these alongside our written analysis, our data tools and our global economist team, drop us a line at [email protected] and we can set you up with a trial account pronto. But that's it for this week. We will be back next week with a review of what the Fed and Bank of England did, as well as the latest from the Middle East and its implications for the global economic outlook. Until then, goodbye.