It's Friday the 10th of July and this is your capital economics weekly briefing. I'm David Wilder coming up. What are clients asking about AI and the global economy? An exclusive clip from our recent online briefings. First, though, kneels out this week but I'm happy to say that in the hot seat, literally given the weather in London, it's Jennifer McEun, our chief global economist. Hi Jenny. Hi David. We're asking first things first, the US and Iran. We had some, what we were calling adverse scenarios dating back to the start of this conflict back at the end of February around what the fighting could do to the global economy. We decided to shell those scenarios when the ceasefire became a reality last month, but given the week's events, given what we've been seeing in the news, should we be getting those adverse scenarios down off the shelf? Yeah, it's a good question and one that some of our clients have asked us, but the answer is no, not yet. We'd always assumed that the past for the recovery in energy flows for the Middle East and therefore for oil prices would be a pretty bumpy one. And that was partly about the fragility of the truce and the risk of renewed tensions. But also, even once flows resume, several economies need to rebuild their inventories, for example, and that's going to keep some upward pressure on oil and more generally on energy prices for a world to come. As we speak, the oil price is around $76 per barrel and that's actually pretty much in line with our end year forecast. And it's certainly not comparable to the peak of $120 per barrel reached at the heights of the conflict earlier this year or the higher prices that we've modeled in our adverse scenario. So no, not time to dust off the adverse scenario yet. Just assuming that we stay at around current levels for the oil prices, which is what is in our baseline forecast, then fuel effects will knock about one and a half percentage points of headline inflation, average headline inflation across the advanced economies in the next few months. So central banks should draw some relief from that. Now, of course, there are risks around that view if the conflict escalates further, but for now, we're comfortable without baseline forecasts. OK, holding on the baseline, I do want to talk about the disinflationy impacts and central banks. Before we do, you mentioned that oil around $76 now can ties in with how we see oil at the end of 2026 with all this volatility going on, but that is still much higher than we were forecasting on February 27th. Since then, we've had all manner of energy disruption, but look at how the global economies performed over the past four or so months. There's not been too bad. Has it? Why do you think that activities held up so well? Yeah, that's right. That's certainly true. This week, we've had data from the Eurozone on industrial production for Germany and retail sales for the Eurozone as a whole. And those both look pretty good. German industrial production raised by 0.9% in May. Eurozone retail sales raised by 0.2%. And that kind of mirrors the picture that we're getting from the other advanced economies. You're right, the oil and energy prices are higher than they would otherwise have been. And there is a drag to come from that. And we've seen something of a drag already. But I think that inflation has probably already peaked in the Eurozone at least. So that drag should start to ease going forwards. Those retail sales numbers in particular. I mean, you think about the impact of the conflict and we are, you know, hopefully we're moving out of that phase. But it's still, it's a headline inflation spike. So lower real incomes, lower consumer, we've seen the hit to consumer confidence in the surveys. But actually, it seems like the consumption story is actually pretty solid and actually we've seen similar happening in the US as well, right? Yeah. Yeah, that's right. Consumption is still growing or seems to be despite this pretty significant hit to real dispose of the incomes from the rise in inflation. It's hard to say exactly why that is and it's possible that there's more pain to come. But I think broadly speaking, there are two different stories in Europe, both the Eurozone and the UK. It seems like households are dipping into their savings a bit. Household saving rates are still relatively high. And I think that probably like financial markets and like most of us, households are assuming that this renewed escalation is going to prove temporary so they're not too concerned about. They're permanently higher inflation or a prolonged hit to their real incomes. So they're willing to do that to dip into their savings. There have been some temporary factors supporting spending as well, the World Cup, the warm weather, has encouraged people to go out and make some purchases. I'm sure I'm not the only person who's bought a paddling pool in the past few weeks. So the US though, I think it's more a story of strength in the labour market despite a bit of a wobble in the June data. Employment is growing fairly strongly in the US and actually we've seen incomes growth matching as a spending growth. So there are really good fundamental reasons for the US consumer to still be fairing pretty well. Just looking at the US consumer very quickly, what about this idea of a K-shaped economy where the strength that you're seeing on the consumer side is actually down to the activity of a relatively small group of consumers who've done very well out of things like, you know, this is extraordinarily rise in stock prices, rest of the US consumer not do so well, struggling with, you know, credit card debt and all manner of things like that. That's a reasonable concern. The top 30% of households bought by income now accounts over half of spending in the US. So it's clear that we do have that balance in that skew towards higher income people supporting spending. Once the equity market corrects, which we believe it will before too long, they will suffer some adverse wealth effects and it's likely that they'll reign in their spending to some extent. Once that happens. But there are a few sources of comfort, loan delinquencies have stabilized. They were rising fairly sharply and pointing to some strain among lower income households, but that seems to have stabilized now, unskilled jobs are rising again. So the bottom end might start to catch up. And in the meantime, the top end should benefit from further equity price gains in the near term. We don't think the corrections coming just yet. You spoke about the strength in or the relative strength in US labour markets. And they're all that playing in supporting consumption, but talk a bit more about what's happening in Europe. The European labour markets are not as strong as the US labour market. The UK in particular, its labour market looks pretty weak. It's had some headwinds from rising national insurance for employers from a higher minimum wage. And those have really limited hiring and kept the UK labour market pretty soft. And now with this extra headwind from higher energy costs, it seems fairly likely that we'll see further job cuts in the UK and further rise in the unemployment rate. The US own labour market is fairing. The unemployment rate is still edging down there, but wage growth is much softer than it is in the US. And if we look at the survey measures of the outlook for employment, they're not as positive. So I certainly wouldn't say that the outlook for European consumers and for the labour market is as positive as that in the US. But all feed into the policy question which lights going to. Before I do though, it has been a fairly light week in terms of the economic data release that you and the rest of the team are tracking. Coming week, looking quite a bit busier. What are the key releases that you think investors should be watching? What are we expecting? How do we see the narrative evolving in the coming week from the releases? Yeah, I think the really key release in the week ahead is going to be the US CPI release. We're expecting a fall in the headline rate of inflation in June from 4.2% to 3.9%. That's driven pretty much entirely by energy prices and energy inflation easing off in a very predictable way. If we look at core inflation, we actually think that might have edged up in June, which would certainly give the Fed more food for thawed and the hawks in the Fed a bit more ammo. That increase will reflect several factors, but including some indirect effects from energy prices such as on air fairs and also perhaps a boost to hotel prices from the World Cup. We've got July meetings in a couple of weeks. I think broadly speaking, not much is expected to come out of those. I'm thinking sort of Fed, but also ECB and Bank of England. It does seem as though the attention is already focusing on what these banks are going to do come September. They're September meetings. Where are we expecting policy makers to? Let's start with the Fed. Kevin Warsh, he's made it clear he's not happy with the level of prices in the US. Other movements on the FOMC suggesting they're not happy. They're not comfortable with what's happening on the price for an even if headline inflation is coming off. How does that then translate into policy moves through the end of this year and into 27? Well, you're absolutely right. I'd characterize these upcoming meetings as communication events rather than action events. Nobody's expecting any change in the policy rates unless there's a major surprise in the incoming inflation or labor market data. The focus is very much going to be on guidance for the author.
rather than on immediate policy changes. Now, so far as the UK and the Eurozone are concerned, that the fragility of their economies, the ongoing fragility of their economies, means that assuming their energy prices behave, I don't think we're going to see any rate hikes or any indication that those are coming. So the Fed is more interesting. The minutes showed, which we've got this week too, showed that all participants thought tightening was warranted in the coming months. So that's interesting. Employment risks have clearly moderated in the US and AI demand is keeping inflation up. And that's something that the FOMC is explicitly acknowledging now. So I think the question isn't whether they're high at the when and by how much. And there are clearly some major divisions on the FOMC about those questions. Those CPI data will be really important in helping them to make that decision and financial markets will certainly be looking at the detail in that data release to judge how the Fed will respond to it. Our expectation that we'll see a slight rise in core inflation will provide some support for the Hawks and we think that we'll see ongoing labor market resilience too, which I think will mean the vote tips as we head into the autumn. We're actually anticipating two rate hikes this year and then another one next year, which is more than markets are currently pricing in. I mean, I know that you talked about the UCB, but we have this case, and some of them pretty solid case, we seem to be not moving in September, but that's it's about market pricing and most analysts feel differently that actually they're not one and done that June wasn't the end of their tightening and that we will get another move in September. So why do you have a different take on whatever it is saying? Yeah, I'm a bit surprised by the consensus on this. I think it probably relates to the fact that at the start of the conflict, the ECB really came out with some quite hawkish rhetoric. I think it was burned by the experience of 2022. It has historically been a relatively hawkish central bank, so with inflation still significantly about target. Perhaps that's why there's this assumption that the ECB will need to hike a bit further, but we think that inflation has already peaked, assuming that we're right that oil prices will remain at around their current levels will be it with some volatility in the coming months. Then I think the ECB will have every reason to focus not so much on the inflation data, which is coming down, albeit from a high starting point, and much more on underlying economic activity. And I think their things like German industrial production will continue to show fundamental weakness, which will limit inflation pressures in the Arizona. Okay, I wanted to just finally, you touched on it before this idea of heat waves. You've got a new paddling pool. I've got fans in every room in the house. Seems like Catholic economics clients in Europe are suffering as well by the volume of questions we've had about the heat waves, not how to cope with the heat, but what do temperatures like this, how do they affect economies? Now scientists say, get used to heat waves in Europe of greater frequency, greater intensity. Frankly, the continent is not built to deal with temperatures like this. So how does a structural shift like we're seeing in the temperature change the outlook for economies? Well, heat waves matter for economies, mainly because of what they do over time, rather than in any single year. So a short period of extreme temperatures, like we're experiencing now, unlikely to have a major impact on GDP. And actually, I mentioned earlier that in the UK data, for example, we're already seeing some boost from the warm weather when people go out and buy things like paddling pools and sands. So you can see some very near term positive impact on the GDP data. But going forwards and more frequent heat waves are likely to reduce productivity, disrupt agriculture in particular, and potentially damage infrastructure too. There are some offsets even over the medium to long term, spending on rebuilding and adapting economies through a warmer climate can support economic activity. But our work at capital economics suggests that in a world where temperatures rise by more than three degrees above pre-industrial levels, global GDP would be about 10% smaller than it otherwise would have been in 50 years' time because of those adverse effects that I've already mentioned. The biggest hit slightly to be in emerging markets, of course, particularly hotter economies in South Asia and Africa, where agriculture and outdoor work are more important. For advanced economies, the bigger issue may be inflation and public finances as governments face rising costs to upgrade infrastructure and respond to extreme weather events. So the real economic story isn't this year's GDP number. It's the gradual drag on growth and increasing fiscal pressures over time. Jennifer McEun on the state of the big DM economies and how policymakers are likely to respond. Jenny touched on a broad range of issues in short-corder. If you're not already receiving our global US Eurozone or UK coverage and want access, do get in touch at
[email protected] and we can arrange for you to be set up with a trial. Look out for our online drop-in briefing that we'll be holding shortly after the Bank of England meeting on July 30th to discuss the month's big interest rate decisions. If you're a capital economics client, you'll be getting details of that session soon. We hold several drop-ins each week across a broad range of macro and market issue. The goal is to have them wrapped up in 30 minutes or less so the clients can get the answers they need and get on with their day. Just this last week, we did two sessions on AI's macro and market impact. Capital economics began regular coverage of AI in 2023 with a series of then-provocative calls on how the technology was going to change the global economy and deliver outsized market returns. While many of those calls have since become part of mainstream thinking, a lot of the questions that we posed in that original research continue to feel conversations with clients. So the briefings this past week were designed to cover off our latest thinking on AI and to address those questions. Here's a short edited clip from one of those briefings in which you'll hear Vicki Redwood, our chief economics adviser for macro, and it starts with her observations of where AI's impact is visible in the current data. The key point is that up until now, the main economic effect of AI has been via the infrastructure boom, the investment spending. And so that's giving a boost to short-term demand in the economy. So that's where we're really seeing the effects come through at the moment. And those effects are concentrated in the US where much of the investment spending is happening, but also in those big AI hardware tech explorers, like Taiwan or Mexico, who are exporting a lot of this hardware to the US and some other countries. In terms of the productivity gains, these are starting to come through, but it's going to be a long, slow process. And so far, the sort of most concrete evidence we've got is in the US, as you might expect, where, although the aggregate productivity data is still a bit up and down, if we look into the sector or breakdown, we can see that those parts of the economy where AI usage has picked up the most are giving the strongest contributions to overall productivity growth. So that's encouraging. Other countries were not really seeing anything in the hard data yet, but I'm not too surprised by that because other countries are generally further behind the US in terms of the diffusion of AI throughout the economy. And we always said that it's going to be quite a slow progress for these productivity gains to actually materialize. Just to put some numbers on that, I mean, over what period would we expect to see hard evidence? I know you said maybe the aggregate level in the US not so much so far. If you look at the sectoral level, perhaps, that we can see the signs of productivity improvement, when are we going to have much clearer evidence of that perhaps at the aggregate level, when are we going to see it outside of the US? And also actually how large are these gains that we're expecting? Well, in the US, the forefront of all this, I'd expect within the next year or two to see firmer evidence in some sort of aggregate productivity impact coming through. In terms of the magnitude there, we think that AI will boost productivity growth by approximately a percent a year, maybe over the course of about 10 years, that would be akin to what we saw during the ICT boom in the US. Elsewhere in other countries where diffusion will take a bit longer, I think we're looking at more of our late 2020s and going into 2030s story for those productivity gains to materialize. And the magnitude of those gains will different quite a lot by country, according to how well placed they are to adapt to AI and to use it. So in the countries that have perhaps been a place, we won't see productivity increase as much as in the US, but you might still be looking at a boost if they half a percent a year, a percentage point a year, maybe in countries like the UK. In other countries where diffusion is a bit harder, then you might be seeing a much smaller boost, maybe 0.1%, 0.2%, maybe in some of those European countries, which are a bit further behind in terms of digital readiness. So quite a spread, I think, in terms of how much of the boost countries will see and when they will see it. I want to come under the global picture in a bit. Before I do very quick, if someone's asking which sectors we can expect to see these productivity gains, and we very quickly, what are we seeing at the sub aggregate level at the moment, is that the kind of trends that we would expect to develop? Yeah, I think the surveys of business usage of AI are generally quite useful in this regard. So, unsurprisingly, the text.
sector itself is a very heavy user. But now we're starting to see pretty high usage rates coming through in professional business services. So again, that's what you would have expected places where AI can in particular automate a lot of routine tasks, but also augment some of their tasks further up the value chain. The sectors where we're still seeing very low usage rates, again, probably not too surprising, ones like manufacturing, construction, where AI should over time be able to be applied, but perhaps the applications are not so immediately obvious. Very quickly, you mentioned the top about this investment boom that's gripped the global economy. Obviously that's feeding into the debate about what's happening with inflation. We've had lots of questions about whether inflation or how inflationary this AI boom is. But also we've had longer-term questions centered around this idea that Kevin Worsh is advocating. AI is going to deliver a big disinflationary boost with the economy down the line. So very quickly price impact near-term long term. Yeah, also so far, I'd say the net impact of AI is mildly inflationary because we've had this quite big boost aggregate demand from the front loading of investment. And that'll be outweighing for now any disinflationary impact on the supply side. That said, the positive impact on inflation is pretty isolated in a few sectors and bottlenecks of the economy, such as chip prices or electricity prices. And certainly in the last few months any inflationary impact would have been dwarfed by the other factors affecting inflation. Most obviously the fallout from the Iran war. Further ahead, as the productivity gains increase and that should give firms scope to cut their prices then that disinflationary impact could increase. But I think we need to remember that if prices do fall, that's going to be boosting consumers' real income and boosting their spending power, which will be boosting aggregate demand. So over the long run you've got this rise and demand rise and supply. I don't know obvious reason why either should dominate. What is possible though is that if the gains from AI are concentrated in those groups that have a relatively low marginal intensity to consumers, we call it, then that means that maybe the rise and demand will be a bit less than the rise and supply. But that's, yeah, we're looking quite far ahead. For now, the, I'd say the net impact is multi-positive on inflation. There's a question that we did a session. One of these sessions this morning and it came up during this session, Vicki. And it's about the share of national income between capital and labor, this idea that the benefits of AI are going to increasingly flow to capital over labor. It very much ties in with this note that was produced by Citrini, this sub-stack investment newsletter at the beginning of the year that gripped markets of a very doomsday-ish scenario for AI's impact. Can you address this idea that is going to be the capital is where the benefits are going to agree? Yeah, so I suppose it's various reasons why we might expect capital to disproportionately benefit. One is if employment does end up permanently reduced from AI, that's not something we expect as I said earlier. So I think if the capital share worth increase, it would probably be more likely because AI reduced work is bargaining power generally, even if employment held up. Or if just the gains from the productivity gains from AI managed to be captured by capital more than labor. Now, don't think it's kind of a foregone conclusion that the labor share is going to drop and the profit share is going to rise. If you look back at past technological breakthroughs, actually the labor share doesn't seem to have suffered in the long term. And even if the labor share did suffer a bit in the short term, the profit share rose, you might expect that to be competed away over time. But I think if there was a drop in the labor share, I think then policy would step in and come to the rescue in a sense. And we would see a much bigger role for government in undertaking redistributive fiscal policy to make sure those gains from AI were shared more equally amongst the urban and households in the economy. - I do still have questions coming in, but we are up against time. But I have to ask you this question. What if we have this idea of AI as an economically transformative technology, think steam engines, think electrification, all of these sort of general purpose technologies, AI is another one of those. But what if, just in terms of thinking about the global economy, what if it all turns out to be rather disappointing in its payoff? - Well, in that case, I think, you know, all is not lost. AI, even if it doesn't turn out to be this hugely transformative technology, we can see already that it's generated a huge number of applications and uses in the economy is already being used to automate a lot of low-level tasks. So I think, you know, there will still be productivity gains to be had from it. And equally, although some of the investment in AI may turn out to have been wasted, if it's one for a better word, a lot of it will be able to be repurposed. A lot of capital will be able to be repurposed and still use productively. So I think it's partly just a case of relative expectations that AI might disappoint. But, you know, let's suppose that actually the productivity gains are not as big as that as we are hoping for, what's the main implication of that? Well, I think it's mainly disappointing for the countries of whom demographics is going to be a big drag over the next few years. And we're hoping that AI will generate productivity gains to offset that, but if it doesn't, or doesn't generate sufficiently big productivity gains and they're the ones that will probably see the worst outlook as a result. Vicki Redwood on AI's transformative potential, but also what happens if it disappoints? I'll link to the floor recording in the podcast notes. It's part of a large and growing body of analysis about this technology that's all housed on a dedicated page on our website, which I will also link to. But that's it for this week. We'll be back next week with more for the world of macro and markets, including a review of the Q2 and June China activity data. But until then, goodbye.
PodcastTranscript AI