China’s manufacturing output just grew at the fastest pace in six months, but retail sales barely moved last month — and the gap between the two growth rates just hit its widest point in over three years.
The upshot: Exports are still picking up the slack.
But an increasing chunk of China’s export growth is simply down to price increases, not rising volumes.
So even that stalwart economic prop is starting to wobble.
On this quick-reaction episode of the Trivium China Podcast, Andrew Polk sits down with Joe Peissel (Trivium’s Lead Macroeconomist) to unpack:
Why the gap between factory output and consumer spending just hit a three-year high, and why Beijing’s own policy playbook keeps worsening the imbalance
The hidden story in China’s “booming” exports: value is up 25%, but volumes for things like semiconductors and appliances are actually falling
The one semi-bright spot in the monthly macro-econ data: infrastructure investment is declining at a slower pace (hurrah!)
Why Beijing’s new interest-rate subsidy for consumers is aimed at the wrong problem entirely
Putting it all together: China’s economy will continue to decelerate into year-end.
Give it a listen and let us know what you think.
Transcript
Andrew Polk: Hi, everybody. Welcome to the latest Trivium China Podcast, a proud member of the Sinica Podcast Network. I’m your host, Trivium co-founder Andrew Polk. And today, I am joined once again by Trivium’s Lead Macro Economist, Joe Peissel. Joe, how are you doing, man?
Joe Peissel: Hey, Andrew. I’m good. Thanks, man. And happy to be here as always.
Andrew: It’s great to have you on. We are doing another of our shorter quick reaction podcasts. And this time it’s going to be a version of our monthly macro update with Joe. Part of the reason we’re keeping this one shorter is because not a ton has changed since last month. But as I just told Joe, doesn’t mean that everybody remembers what we talked about last month or that the story isn’t interesting. So, we’re still going to run down all the key themes, and Joe’s going to tell us the latest. So, are you ready to get into it?
Joe: Yeah. Macroeconomics is always interesting, Andrew. I don’t know what you’re talking about.
Andrew: Yes. This is my coaching from just before we started. You got to keep it fresh. Good work, Joe. All right, man. So, the title of your piece this month was Same Imbalances, Deeper Cracks. So, why don’t you talk to us about what got worse in August. So, this is the August data we’re talking about. It’s mid-September now, but the August data just came out a couple of days ago. So, if the cracks are deeper, which cracks are deeper and which ones matter?
Joe: Yeah, maybe I can just briefly start by saying what got better, which helps explain what got worse. So, what got better is China’s, surprise, surprise, China’s manufacturing output. I mean, it’s not just growing strongly, growth actually accelerated. So, manufacturing output grew by over 6% in August. That’s the fastest growth rate in six months. So, factories are producing output at an ever-increasing rate. The issue is Chinese consumers are consuming at an ever slower rate. So, retail sales of consumer goods, which we’re kind of taking as a proxy here for China’s domestic consumption, imperfect proxy. That increased by 0.4%, tiny, tiny increase. The previous month, it grew by 0.6%.
The month before that it actually declined. So, retail sales of consumer goods have effectively been flat over the past three months, even as manufacturing output grows at an accelerating rate. And this gap between manufacturing growth and retail sales growth, this discrepancy hit its highest or its widest mark in over three years in August. So, there’s a 5.7 percentage point difference between how much manufacturing grew by and how much retail sales of consumer goods grew by. 5.7 percentage point difference, that’s the largest divergence between the two series in, as I say, over three years.
And that’s really the central tension of the report that I’ve written. Production is racing ahead while the consumers that are meant to buy this stuff are actually pulling back, and this, of course, exacerbates China’s structural imbalances, this demand supplies mismatch that we’ve talked about.
Andrew: Yeah, I just want to pause on that for a second because, like we said, or we started the podcast with, the dynamics haven’t fundamentally changed, but they’re deepening, as you call them, deeper cracks. The widest gap between retail sales and manufacturing output in 40 months.
Joe: Yeah.
Andrew: That’s crazy. I mean, it just means these dynamics that have been evolving over the last three years are getting more and more and more intense. And it seems like policymakers are basically doing very little about it. I mean, we’ll get more to that point in a second. But is that how you see it?
Joe: Yeah, I think if anything, China’s short slash medium term stimulus policies probably exacerbate it because China traditionally relies on manufacturing or supply-side measures to support the economy, right? Even short-term stimulus. So, you can think of things like… well, channeling economic support through manufacturers is like Beijing’s playbook. And so the economy slows. Beijing supports manufacturing growth even more, but exacerbates this divergence between manufacturing and domestic consumption, which is what we’re seeing, which ultimately leads to kind of that ends up weighing on economic activity at some point, this discrepancy. And that causes Beijing to provide even more supply-side stimulus.
In the report, we graph this discrepancy between manufacturing and retail sales. And it’s pretty interesting because you see this percentage point difference, which is what we’re tracking, has just steadily grown over the past few years, which is a really nice way of trying to visualize this demand supply imbalance.
Andrew: Yeah. I’ve got some, what I think are interesting observations from my trip to Shanghai last week that I think are relevant, but they’re most relevant to the next element of this that we’re going to talk about, which is the export piece. So, I’ll kind of bring some of those in in a minute, but obviously the natural outcome of this supply-demand imbalance is basically ever-growing exports. So, exports grew 25% in USD terms year on year in August. So, that’s even faster than July. Normally, you’d say that’s good news, right? Is that actually good news in this instance?
Joe: It’s a symptom of China’s domestic issues that manufacturers have to rely on exports to offload this plus production. What’s quite interesting is in value terms, as you say, 25% growth, really strong growth, and it has been the whole year, really strong export growth. But when we look at China’s exports in volume terms, it’s a much more sobering picture. So, some good examples from August data: semiconductor exports in value terms, more than doubled year on year.
Huge growth. In volume terms, they actually fell. So that’s a price story. It’s not a volume story. It’s all price. We see the same pattern in mobile phone exports. The price is heavily influenced by memory, by memory chips, which have soared. So input costs go up. The price exporters charge goes up. In value terms, mobile phone sales have surged. In volume terms, they’re actually exporting less mobile phones now than they were this time last year. Same idea with household appliances.
We see the same. And so exports in nominal terms, yes, growing really strongly. And that matters. That’s still good. That’s good for the economy. That’s a positive contribution to GDP growth. That can still benefit exporters’ bottom lines. But in real terms, if we strip out the export inflation and really look at things in volume terms, it’s not as positive a story as it seems.
And that matters because although nominal export growth is important, we also really care about export volumes because manufacturers need rise in export volumes to absorb this output, this growing output we just talked about that domestic consumers won’t buy.
Andrew: Well, I want to get into some of the drivers of this. This is what I teased earlier, which is some of these discussions I was having on the ground in China. And one of the key observations that one of the guys I was presenting alongside in Shanghai made in his presentation deck was, he kind of had this chart of all the breakdown of key export areas, key export goods from China over sort of, say, the five-year period of 2019 and then the post-pandemic period, so 2020 to 2025.
And they were pretty strong, but then also had year-to-date growth of 2026 and all those. And for almost every single category, they exploded in 2026. It’s just like the speed with which the exports have accelerated this year in particular was striking by this chart. And I hadn’t quite put that together. The only category that was really seeing kind of very rapid growth before was NEVs. So NEV exports have been growing for a while. And so, my question was sort of what to him was what’s driving this. And it was, he was kind of unclear. He didn’t exactly know. But, you know, as part of that question, I said, is part of what’s going on here just the global and, in particular U.S., but AI build out, particularly data centers, right?
Chips are coming out of China. Other AI data center inputs are coming out of China. And that’s a big part of it and would explain kind of, at least partly, the accelerated export pattern in 2026. And that kind of makes sense to me, that there’s this demand component that just simply wasn’t there prior to this year. But then he also said it doesn’t explain all of it. So, like refrigerators and washing machines are also going gangbusters. And the question is sort of why did exporters all of a sudden of those segments start to accelerate even further this year?
So, it’s a two-part question. What is driving this? Three-part question. What is driving it overall, you think? How much of what is driving it is the AI piece? And then the last part is for these other categories that aren’t related to AI, do you think it’s just that domestic producers suddenly realized that they could get such a better margin externally? Or what’s driving that last piece? A lot of different threads there. So, what do you think?
Joe: So the monthly growth we’re seeing, the main driver is the AI buildout. That’s the main driver. And you see this if you look at disaggregated export data, the semiconductors, computer hardware, AI-related hardware, is growing the fastest of all China’s export categories and accounting for an increasing share of China’s exports. But it’s true, it’s not just an AI story here. So, the growth in things like household appliances. Well, I think a lot of this, as we just talked about, is export inflation.
And that’s driven by increases in memory chips and commodity prices. And so, exporters are forced to raise prices because their input prices are rising. The other thing we’ve noticed, and we talked about this actually earlier, Andrew, before we started recording, is we’re seeing a lot of companies relocate production to China, right? So EU companies manufacturing in China and then export into the EU. That’s a growing trend as well. And that’s kind of another tailwind for China’s exports. I mean, it’s very hard to pin down a single reason. More generally, I’d just say China’s got an ultra-competitive industrial manufacturing base across all areas of industry.
So, kind of a good anecdotal example is Volkswagen, their largest R&D center outside of Germany is in China. And that’s not necessarily to do with Chinese manufacturing per se, but to do with China’s entire industrial base and all the industrial services, producer services that support it as well, kind of the whole ecosystem. And so, that’s kind of another part of the story is there’s lots of manufacturers relocating to China because China is so competitive.
The irony being, even as certain Chinese manufacturers actually offshore to other countries to try and evade tariffs. So, I think this all speaks to kind of the structural advantages in China’s manufacturing export base.
Andrew: Yeah. Okay. We have to stay on this for a second because, I mean, the reason I’m asking these questions is sort of everyone kind of goes back to, “Well, China can’t maintain this level of exports because trading partners are going to push back. And also, trading partners are complaining more and more and more about the deluge of exports from China.” But if you look at the data, a lot of, as you just said, the biggest chunk of growth in Chinese exports is due to this basically brand new demand segment, right? the AI build-out, the data center build-out. So, the margin of the biggest chunk of marginal growth is not Chinese exporters just slashing prices and finding new markets.
It is that there is a new demand driver built around AI. And so, from that standpoint, from a policy standpoint, Western policymakers have much less to complain about. I won’t say they have nothing to complain about, but we’re buying a bunch of stuff from China to build the data centers to input into the AI build-out, the data center build-out. And that happens to be the one segment of the economy that is basically driving U.S. growth at the moment, for example. So just throwing that out there, that Chinese producers are just meeting this surge in demand, and I think that undercuts a lot of our complaints about the current export dynamics from China.
It doesn’t totally undercut them, but undercuts them to some extent. What do you think about that? What’s your reaction to that?
Joe: Yeah, I mean, that narrative should help to reduce some trade friction, right? Because these are essential inputs into AI build-out that Western economies need. That’s the point trying to make more broadly, is that China’s exports, even though they’re growing fast and displacing domestic industry for other countries, they’re also essential inputs. So, the EU’s green transition can’t happen without cheap Chinese solar. The U.S. AI build-out can’t happen without competitive AI inputs, so on, and so forth. Of course, that’s a somewhat legitimate argument that China’s consistently made. Its export is almost a service to the global economy. And to an extent it is. Of course, the counterpart that Western economies will argue is, well, actually they want to be manufacturing these inputs themselves. The issue being they can’t do it as competitively.
Andrew: Yeah. Well, that’s the second point, which is you alluded to it already, but a lot of my conversations in Shanghai, so I was talking to a bunch of different companies, and in particular, the European companies would kind of hushedly say, “We may or may not be doing this now, but our company is increasingly considering exporting more from China back to our home market because basically it doesn’t make economic sense for us to produce XYZ industrial good in Europe.” And I think from my standpoint, that is not a well understood point for EU policymakers that a bunch of the “problem” is coming from their own companies producing out of China and exporting to Europe.
And I would say that this seems to be sort of an increasing trend in the multinational space, which is China for the world. So for years and years and years when I was there, everyone was increasingly localizing. It was China for China. And now, more and more companies are thinking, how can we use our manufacturing and industrial that we’ve built up in China to service global markets. And so, also, I’m not saying again that it undercuts Western arguments, but it certainly complicates them. And I would say at least it makes the approach from Western policymakers broadly, but the EU in particular, more complicated because you can’t just say, “Stop exporting to us.” I think the only option in that case is to say to China, “Fine, we’ll keep buying your exports, but you need to buy more of our exports as well.” And then you get to the question of what does China need that it doesn’t produce? And of course, there’s import substitution, all that stuff.
But I talked to a bunch of my friends in China, and they’re like, “I would never buy anything from overseas because everything I have, I can easily get on Taobao from a local producer.” Talk to me about those dynamics.
Joe: I think having foreign companies embedded in China’s economy actually provides Beijing with structural leverage when trying to push back against any sort of Western complaints or efforts to decouple. It makes decoupling kind of more politically costly for Western governments. And I think this trend of foreign companies relocating to China, one can argue from a macroeconomic perspective, it’s bad for these domestic economies. Where it gets politically tricky is from a shareholder perspective, it’s massively beneficial, right? Because companies are utilizing more efficient supply chains, cheaper manufacturing inputs increase in the bottom line.
So, there’s a trade-off, I think, from a Western policymakers’ perspective, what may, kind of, in the medium to long term, be damaging to one’s domestic economy by this de-industrialization, having your manufacturing base hollowed out as companies relocate to China. Certainly in the short term is beneficial for a lot of shareholders. And who are the shareholders? It could be suffering wealth funds. It’s almost certainly going to be pension funds for one’s citizens. So yeah, this is why I kind of said, I kind of think of almost a structural leverage for China when entering these trade negotiations. If there’s already foreign companies embedded in China’s economy, China has a lot more clout, a lot more weight in trying to push back against any sort of trade measures against Chinese exports.
Andrew: Yes, those are all good points. And I will say the other piece of leverage it has, should Chinese officials choose to employ it in a full-blown trade war with the EU, is that because there are so many EU companies with large domestic presences in the Chinese market, that’s also a point of leverage that they can apply pressure to these companies or in various ways mess with them. And I, unfortunately, that was one of my messages to the company that I was talking to was you potentially have a lot to lose. You’re very exposed here if things go really south in EU-China trade relations, which increasingly it seems like they might.
Mostly what I’m just trying to say is this export picture is more complicated than just China selling too much to the world. But we’ll talk more about this. We don’t have to dwell further on this. I do think it’s important for listeners to think about. Last thing on the export piece is you’ve talked about how there are signs that the export juggernaut might actually be weakening, at least on a sort of a cyclical basis. What are you seeing there?
Joe: Yeah, I mean that goes back to the volume story I mentioned earlier so export volumes in certain categories have declined. This is very tentative. I’m not arguing China’s exports are now on the down and down at all. But I think it’s definitely worth readers and listeners being aware that there’s very much a huge difference in China’s exports when you look at price versus quantity. Quantity of certain exports, as I mentioned earlier, home appliances or mobile phones or semiconductors, they declined in August in volume terms.
The caveat being, it’s from an incredibly high base. Exports in volume terms are already very high. And so, like a 5% or 6% year-on-year decline still means exports in volume terms are very high. But I just think it’s important to distinguish between value versus volume. It’s definitely anyone who’s following China’s economy should be tracking that discrepancy very closely because it does matter.
Andrew: Yeah. Well, and something we’ve talked about, this is kind of, I think, the reverse that we’ve talked about maybe in previous discussions on the macro economy, but there’s a macro versus a micro. Again, I’ll go back to part of the reason prices are surging for components related to the data center build-out is because China can’t make them fast enough. And so, there are shortages, at least in relation to the surging demand, which makes the price go up. And while, like you said, from a macro standpoint, you would like to see volumes growing, from a business standpoint, if you’re selling the same amount for a lot more money, that’s totally fine. Right?
Joe: Yeah. We talked about this last time actually. Yeah, from an individual company, I don’t think I necessarily care, right? That their profits go up, their profits go up. But yeah, from a macro perspective, China needs factories to continue producing lots of goods to employ lots of workers to sustain the unemployment rate, and try put up a pressure on wage growth and all this sort of stuff.
Andrew: Yeah. Well, it’s a dynamic picture, so we, of course, will talk about this a lot more, again, at the risk of repeating myself too many times. I think the export picture is more complex than people often boil it down to. But let’s go to the next piece of the economy that you highlighted, which is actually if you squint, maybe the one sort of bright spot here, which is that you flagged infrastructure investment saw a genuine improvement last month. Talk to us about that. What’s happening there?
Joe: Yeah, you’ve really got to squint to frame this is a positive. Infrastructure investment fell almost 7% in August year on year. Huge drop. The reason why it’s kind of a positive is that there’s a huge slowdown from what we’ve seen over the past few months. So, infrastructure investment in May, June, July declined by double digits.
Andrew: A slowdown in the decline.
Joe: Yeah. Decelerating decline. A double D, if you like.
Andrew: Yeah, you know we’re in bad shape when we’re saying the decline was only single digits and we’re happy about that.
Joe: Yeah, in and of itself is a grim print, but I think relative to what’s been happening over the past few months, it’s an improvement in the decline. And it coincides with a bunch of policy support towards infrastructure. And so, that suggests, well, maybe policy support is really starting to gain momentum and is going to have an effect. So, we’ve talked about this, and I’m sure you’ve talked about this with other guests on the podcast. There’s China’s Six Networks Initiative, this huge infrastructure investment pipeline of projects in a bunch of different networks, like telecoms and water and gas pipelines and stuff. That’s gathering pace.
Over the next four months, China’s policy banks are going to issue $800 billion in seed funding for infrastructure projects. There’s going to be a whole new fresh wave of funding to support infrastructure spending. And over the last quarter, so from October onwards, there’s really favorable base effects infrastructure collapsed in the final quarter of 2025, these huge double-digit declines. So, kind of putting that together, This Six Networks Initiative, plus a bunch of policy bank seed funding, plus favorable base effects, I’m confident that that infrastructure print is actually going to start growing again across Q4.
And so that the decline slowed significantly this month, I think, also kind of backs up this idea that things are changing. There’s kind of an inflection point in infrastructure spending. Now, the caveat here is, firstly, even if there’s strong growth in Q4, infrastructure spending full year is still going to be down. And other parts of investment spending, so manufacturing investment or property investment, that continues to slide, continues to decline. So, the overall picture is still negative, but there’s kind of one potential tailwind in the coming months. That is infrastructure FAI.
Andrew: Yeah. And for listeners, Joe just wrote an excellent note for our readership on the dynamics within overall investment. The three main categories being infrastructure, manufacturing, and property, and sort of walking us through the three narratives, the three dynamics in each category, and also kind of doing an outlook for each category and thus for overall investment going forward. So, certainly, encourage people to check that out. If you’re not a market subscriber, you can sign up free trial on our website. So, little plug there.
Joe: Great plug. So organic as well. Yeah, the report has, and it wasn’t just me, it’s me, David Zhang, and Wenye Sun. So, it’s very much other colleagues as well that I should mention here. So I’m not taking credit for all their work. As much as I’d like to, they won’t let me get away with it. But yeah, we provide full-year forecasts as well for FAI growth, which is from a macro perspective is so important, right? Fixed asset investment is still one of the key drivers of China’s economy.
And so, understanding or having forecasts of what’s going to happen to investment can really help investors kind of gauge what’s going to happen to China’s overall GDP trajectory.
Andrew: Yeah, and I’m actually glad that you highlighted the six Networks Initiative in terms of helping to stabilize infrastructure investment. We actually haven’t really talked about that much, if at all, in the pod. So that might be something that we need to revisit with potentially you and Dinny and some others on the team. But it is a very important element of what’s happening in that space. So something to keep an eye on. Let’s now end with the consumer. So, there was a small move from Beijing, from policymakers to support consumption over the past few weeks, which was an increased interest rate subsidy for consumers. But you’re pretty skeptical it won’t help much. Talk to us about the dynamics the consumer is facing and what’s happening on the policy front there and whether or not it’ll matter.
Joe: Yeah. So there’s two things on the policy front. The first is this expansion of interest rate subsidies for consumers. So, the Ministry of Finance pays a portion of the interest rate consumers pay on loans. Idea being this lowers the cost of credit. And so, consumers, it’s going to increase their willingness to borrow money and spend. The reason I’m skeptical about it, and I talk about this in the report that went out today, September 17th, I guess the podcast is going to go out a couple of days later, is because China’s weak consumption isn’t a function of expensive credit.
It’s not a credit problem. Rates are already low. The reason this weak consumption is far more structural, it’s to do with slowing wage growth, kind of this structural long-term slowdown in consumer confidence and their willingness to spend. And of course, all of that is tied to the negative wealth effect. So, property prices down almost 30% from their peak five years ago. Interest rates already low, meaning fixed-term deposit, the return on fixed-term deposits or wealth management products is low. Return on government bonds is low. China’s stock market is super sluggish.
And think about that in contrast to Western equities, which this year have absolutely soared. So cheap credit doesn’t fix any of those things. It doesn’t repair household confidence. It doesn’t somehow translate into higher wage growth. So, without addressing the structural problems, interest rate subsidies are going to do very little for consumption. The second policy we saw in August was a bunch of county-level supply-side measures. So, what that means when I say supply side, consumption supply side measures are really all about trying to either improve the quality or increase the quantity of goods and services that exist in the market.
The idea being if there’s more things that consumers can buy, then they will buy them. And so, supply side measures that were released last month include increasing the amount of charging points for electric vehicles. So consumers maybe want to buy electric vehicles because they’re easier to charge. Building new shopping centers, upgrading existing local consumption facilities that can be like storage warehouses or upgrading shops, things like this. But we talked about this at the beginning of the pod. Beijing always relies on supply-side initiatives to try and boost demand. And it ain’t going to work, Andrew.
Andrew: Yeah. Well, you know, your typical rational consumer, not that consumers are always rational, is not going to look at their financial situation and say, “My employment prospects are somewhat shaky. My wages are growing more slowly. My confidence in my economic future is not particularly high. Maybe now’s the time to take on a bunch of new credit to buy stuff I don’t necessarily need.” Now-
Joe: Yeah, at this new shopping center.
Andrew: Yeah, exactly. Now, I was going to say, there might be some American consumers that think that way. I do think that maybe we’re built a little different over here, and not necessarily always a good way, but Chinese consumers and most consumers in the world, and even most American consumers do not think that way. And so, it’s just making it easier to borrow as a consumer in these circumstances. I agree with you. It’s not going to have much impact.
Joe: Yeah.
Andrew: Let’s wrap it up. This was a good breakdown. Thank you for that. I think to wrap it up, I mean, the really $10,000 question that everybody wants to know is, okay, put it all together for us. And what are the prospects for the economy? What’s the trajectory? We’ve talked before, we don’t really do GDP forecasts per se because the official GDP numbers are a political target, blah, blah, blah. But so you got weak consumption, as always, or not, you know, as has been for several years now. You’ve got booming exports, but maybe a little of the shine coming off. You’ve got weak overall investment, but maybe infrastructure turning a bit from a low base.
What does it mean for the next six months? Is this an economy that’s decelerating or accelerating or staying flat? I mean, we understand the supply-demand mismatch dynamics, but what about the overall kind of momentum of the economy?
Joe: It’s an economy that’s decelerating. It’s an economy that’s trapped in a, I guess I could say, like a narrowing growth model. So, it’s an economy that relies on manufacturing and exports. And as you said, some of the shine on the exports is starting to fade. So the growth drivers just get narrower and narrower. All the while, domestic demand, so whether that’s consumption or property, or private investment, just keeps deteriorating. So, if one’s interested in the six-month outlook economy, we need to ask ourselves, what changes this trajectory? And the answer is some sort of demand-side reset.
And there’s nothing on the policy front that suggests that’s coming. So without any sort of demand-side reset, the economy keeps drifting towards slower growth, propped up by manufacturing and exports. I mean, particularly exports, that’s also cyclical in some respects, driven by this AI build-out. And so even that’s not necessarily a structural strength. It’s a cyclical one. All of that means it’s an economy which is decelerating and faces a growing number of challenges.
Andrew: Well, on that pessimistic note, I mean, it’s hard to talk about the Chinese economy and not really be pessimistic at the moment. And yet, more and more foreign companies are doing more and more production there. So, it’s an interesting dynamic. But we will continue to cover these details, these dynamics on the Chinese economy going forward. Joe, really appreciate this breakdown this month, looking at the most recent data. Thanks a bunch for the time.
Joe: Yeah. Thanks for having me on, Andrew. It’s good fun. As I said, macroeconomics is always interesting.
Andrew: There it is. Love it. And thanks, everybody, for listening. We’ll see you next time, everybody. Bye.
