China’s financial regulators have started 2026 with a flurry of activity.
On January 14, the securities regulator raised margin requirements on stock trading in a bid to cool investors’ exuberance.
The following day the central bank expanded a bunch of its relending facilities, and cut the interest rate on all of its structural lending tools.
Then on January 20, the finance ministry rolled out a bunch of measures designed to bolster investment by small, private sector firms.
What’s driving this hyperactive policymaking? That’s what Trivium Co-founder Andrew Polk and Dinny McMahon, Head of Markets Research, discuss on the first Trivium podcast of 2026.
They look at:
How weak Q4 economic data has lit a fire under regulators
Why authorities have settled on this particular combination of policies
And what signals Beijing is trying to send markets
But wait, there’s more! On the second half of the pod, Andrew is joined by a new guest to the podcast, Trivium’s lead AI and semiconductor analyst Linghao Bao. Linghao joins to discuss:
Beijing’s intervention in Meta’s recently announced acquisition of Chinese AI start-up Manus
The specific regulatory tools China is using to slow – or maybe even stop – the deal
The wider implications for China’s AI start-up ecosystem
The gents cover a lot of ground in this one – enjoy!
Transcript Follows
Andrew Polk: Hi, everybody, and 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 I am joined today for the first Trivium pod of 2026 by our regular guest and Head of Markets Research, Dinny McMahon. Dinny, how’re you doing, brother?
Dinny McMahon: I’m doing great, mate. Good to see you.
Andrew: Yeah, great to see you as always. And one of the key reasons we’re going to have Dinny on is because we’re already out of the box pretty hot on some moves, some macro policy moves out of China, specifically on the monetary side from the PBoC, as well as some adjustments to kind of rules around the stock market, which shows that Chinese regulators are trying to manage fluctuations in the stock market in this instance, seemingly trying to kind of pull things back a little bit, and not allow the market to move upward too quickly.
So, we’ll get into all of that. We’ll tie it a little bit into sort of the economic growth piece, why are these moves happening now as it relates to the economy? And then listeners will want to stick around — Dinny and I are going to keep this relatively tight — the second half of the pod, I speak with one of our Trivium colleagues, Linghao Bao, about the Manus-Meta deal, Meta’s acquisition of AI startup, Manus, kind of the dynamics behind that, and specifically why Chinese regulators are potentially intervening in that deal.
And then we opened up the conversation to sort of a wider look at some of the key tech dynamics, sort of dynamics behind the tech startup scene in China. So, really interesting discussion. You’ll want to stick around for after Dinny, and I wrap up on the economy. But before we get into all of that stuff, Dinny, 2026, first customary vibe check of the year. How’s your five met?
Dinny: Dude, my vibe is harried dad because the kids are home from school today. The Chicago schools are closed today because it’s cold, which I know sounds a little bit flippant, but really, I mean, Chicago deals with snow and cold weather better than you could possibly imagine, but with wind chill, it’s -35°F, so that’s about -38°C today. So yeah, they closed the schools. And so my kids are slowly going stir crazy inside. So, I managed to lock the door. So hopefully I’m getting in sort of a half hour’s respite now while I talk to you.
Andrew: Well, I’m glad I could offer you a little off-ramp here for a while. It’s funny how quickly things change, right? You just spent three weeks in balmy Australia. I mean, you’ve been back on a couple of weeks and already here... Sorry to hear that. I will say my vibe is disappointed. My 2026 New Year’s resolution was to not do any podcasts this year. And already here we are, and I’ve broken it. So, guess I was a dumb, dumb New Year’s resolution.
Dinny: Yeah. Addiction is tough, mate.
Andrew: Yeah, I know. All right, well, before we get into the rest of the content, we also have to do the usual housekeeping. First, a quick reminder we’re not just podcasts here. Trivium China is a strategic advisory firm that helps businesses and investors navigate the China policy landscape. That, of course, includes domestic policy in China, but also policy towards China out of Western capitals like DC, London, Brussels, and others. So, if you need any help on that front, please reach out to us hq@triviumchina.com. We’d love to have a conversation about how we can support your business or your fund. Otherwise, if you’re interested in receiving more Trivium content, check out our website, triviumchina.com, again, where we have a bunch of different subscription options, both free and paid.
You’ll definitely find the China policy intel option you need on our site. And finally, tell your friends and colleagues about Trivium and about the podcast. The word-of-mouth recommendation fans help us grow our audience and also help us grow the business so we can continue to bring some great free content to the people. While you’re downloading the podcast, please do leave us a rating or a comment on your favorite podcast platform. Helps us get more visibility for the podcast. All right, here we go, Dinny. off to the races on China Macro analysis for the year of 2026.
Dinny: Let’s do it.
Andrew: All right. It’s already been a big couple weeks for policy on the macro front. We’re just going to get straight into it. We’re not going to beat around the bush on kind of retrospectives to 2025, which I suggested, and Dinny quickly shut down, which I appreciate. All right, let’s get through this flurry of measures that have happened already this year. So, on January 14th, the stock exchanges raised margin requirements on new borrowing from 80% to 100%. So, that’s important in terms of managing liquidity into the stock market. We’ll go through that. Then the following day, so January 15th, the central bank increased a bunch of re-lending quotas and kept the interest rates on top of its structural lending facilities.
Again, quite sort of wonky and technical, but it should be impactful in terms of channeling credit to certain places in the economy. Then on January 20th, the Finance Ministry introduced a bunch of measures to boost investment, mostly by reducing the cost of borrowing for small firms. So, Dinny, of those three, where should we start? What do you think is the most important?
Dinny: You know, I think it’s the margin requirements for the stock market. So, I think that is perhaps the most interesting because it’s a pretty clear statement of intent from regulators about how they intend to manage and deliver their vision of a slow bull market. Now, this is something that we’ve talked about before that Beijing wants effectively what the United States has. If you look at the curve of the Nasdaq over the last 20 or 30 years, it’s sort of relentlessly moving up and to the right. I mean, sure, there’s corrections and the occasional crisis, but over time, it’s kind of these, you know, this ongoing march of rising valuations. Now, that has not been the case in China’s domestic stock markets.
And so, what China is trying to engineer is not just a permanent recovery of the stock market, but ongoing rising prices so that the stock market can be an engine for generating middle-class wealth. Now, as I said, we’ve talked about this a bunch. You know, Beijing is kind of got this four-point strategy. It is the foundation of which is trying to rebuild public trust in the market through corporate governance reforms and by cracking down on abuses and violations, and fraudulent behavior.
The second thing it’s trying to do is deliver returns, better returns from investors. And that’s not just from rising valuations, but it’s through pushing companies to pay out annual dividends. Ideally, pay out twice annual dividends, encouraging share buybacks, and that sort of stuff. Third point is they want a lot more what they call patient capital in the market, like long-term investors, insurance companies, pension funds. Have this idea that they’ve kind of looked at international experience and they find that, you know, at least that their conclusion has been that one of the reasons China’s markets don’t rebound as quickly as markets overseas when there’s a correction is because of this lack of patient capital that the market’s overly dominated by retail investors. And if they can inject more of this long-term mindset into the market, that they’ll be able to better deliver the slow bull market.
And then the fourth point is that they want more high-tech IPOs, not for their own sake necessarily, but because there’s a story that there’s something. That people can really get excited about and that this could provide a real fundamental foundation for rising stock valuations. So, they’ve been slightly putting in place all four of those conditions over the last couple of years at this point. But they’re also aware that it’s not enough just to push prices or to encourage prices to move higher, to kind of put in place the foundation for prices to move higher. Because what they have experienced in the past, ultimately, is sort of a boom-bust cycle. We saw this in 2005. We saw it in 2015. In both instances, prices accelerated upwards very quickly.
At some point, regulators had to step in to slow it down. When they did, they burst the bubble, prices collapsed, and then the market moved sideways for a decade. And so, they don’t want that to happen again. So, they’re in this position where they want that market to rise, but they want to prevent a bubble, but they also don’t want to kill off the market’s exuberance. So, they’re really trying to thread a needle so to speak. So that’s kind of the context. So, they raised margin requirements, which effectively make it more expensive for people to borrow, to invest in stocks. From mid-December to about mid-January, China’s stock markets appreciated in value by about 4%, which is a pretty cracking pace over the space of a 30-day period.
And so, that said, at their peak in about the middle of January, they were still just back to the levels they hit back in October. So, it’s not that we were hitting new heights, it was just that the market had fallen and then rebounded very, very quickly. So, under those circumstances in the past, you would not expect regulators to intervene because, as I said, the overall level of the market isn’t particularly high. But by intervening in the way that they did now, not with a measure that’s going to shut down market enthusiasm, but just to kind of take away, to reduce the amount of liquidity flowing into the market, it’s a pretty clear signal from Beijing that, A, it thinks the market is appreciating too quickly. And ultimately, they’re setting the foundations, they’re sending a very clear signal to the market about under what circumstances it will intervene again in the future.
So ideally, by doing this, the market can go, oh, okay, a 4% increase over a month. I mean, that’s far too fast for the regulator’s standards, right? So, if we get close to that level again, then it’s probably a good time to take profit or whatever. So, ideally, regulators will never have to intervene this way again because they’re sending a very clear signal about how fast is too fast and under what circumstances they are willing to intervene. That’s why this margin requirement is really quite interesting, and perhaps the most significant of the interventions. Because it’s Beijing kind of laying down a marker about this is how we’re going to deliver this slow bull market. When things get too fast, and this, what we’ve seen over the last month, is too fast, we are going to step in to do something to calm things down?
Andrew: Yeah, it’s a good point. And I think probably a lot of listeners will be thinking, well, that’s not ideal, having the government be that interventionist in the market and determining how fast of price increases are too fast. But at the same time, the Chinese stock market in particular does swing wildly. And so, you can almost see it as not necessarily a market failure, but such an aggressive move upward is not sort of a typical market move in sort of a developed economy stock market.
And so, in that instance, it does sort of make sense for the authorities to step in and start sort of conditioning market actors as to what’s inbounds and what’s out of bounds. And I would say we see in a lot of places, especially financial regulators, act as what they call counterweights to the market. Right? We’ve seen that specifically over the years in the currency market, right? where the PBoC cares less about whether the currency is appreciating or depreciating based on, you know, structural factors, and really cares about the pace, doesn’t want things to become a one way bet because that encourages speculation, and that kind of counterweight activity by the US in the currency markets has actually, you know, really been quite effective so that it keeps us speculative bets from piling up.
So, we’ll see if that happens as well in the stock market. It may take a few iterations before people get the message. But it is interesting that they’re kind of intervening sort of early here to rein in the upside.
Dinny:
Absolutely. It is very early by Chinese standards.
Andrew:
All right. Well, let’s move on to the next big move. So, this is from the PBoC on the monetary side. On January 15th, the central bank said it will increase certain lending quotas specifically for rural and small businesses. That increase is going to be by 500 billion renminbi for rural and small business lending. Then the PBoC also said it will increase the facility for innovation and technological upgrading plans by 400 billion renminbi. That will take that overall facility to 1.2 trillion.
And additionally, on that innovation and technology program, the authorities are going to expand their remit to include small firms that spent heavily on R&D. And they are also looking at expanding the carbon emissions reduction tool to allow expanded lending for energy-saving renovations. The PBoC also said it would create an additional 1 trillion renminbi lending facility for private businesses, and furthermore, cut interest rates and its free lending and rediscovering facilities on the Pledge Supplementary Loan facility as well and other structured lending tools, all by 25 basis points.
So that’s a lot, Dinny. Probably, people fell asleep while I was talking about it, but basically, they expanded some targeted lending programs. They established a new targeted lending program specifically for private businesses, and then they cut the rates on many of these structural and targeted lending programs, making loans cheaper for specific economic activities. Walk us through why it matters and what’s going on.
Dinny: Well, to summarize all that, what they’re doing is they’re providing more credit more cheaply to key parts of the economy that Beijing deems important. That’s broadly what these measures do. But there’s other ways of doing this, right? You could just cut interest rates. So, the reason Beijing is doing this, instead of just a broad-based-
Andrew: Cut headline interest rates, you mean.
Dinny: Headline interest rates, yeah. Of course, I mean they’re cutting interest rates here, but it’s for very specific lending facilities from the PBoC say, you know, rather than cutting rates, broad benchmark interest rate cut across the economy. They haven’t done that. And the reason is they’re really trying to avoid that sort of benchmark policy rate cut. They’ve been trying to avoid it for the last couple of years at this point. It wants to cut rates, they want to reduce the cost of borrowing, but they want to do it without doing that broad-based cut. And the reason is the PBoC’s biggest constraint on cutting benchmark interest rates at the moment, sort of economy-wide, is they’re concerned about bank profits. Specifically, bank profits mainly come from the spread that they earn between borrowing money and lending money, the interest margins.
And they have been squeezed and squeezed and squeezed over the last few years. I think we’re now down to, across the banking system, the net interest margin in the last quarter of last year was 1.42%. And when the level gets below 1.8%, you start facing bank stability issues because banks find it more difficult to replenish their own capital. I mean, that’s how a healthy, functioning banking system works. Banks make profits, and they use a chunk of those profits to replenish their capital. And the capital there is effectively insurance against loans going bad. And when those profits get squeezed and squeezed and squeezed, it makes it more difficult for the banks to do it themselves. And in a pinch, they would need to rely on selling new shares, raising capital from elsewhere in the system and whatnot.
So, we’re at that position at the moment where the PBoC is incredibly reluctant to squeeze those net interest margins any further. And so, that’s why cutting benchmark interest rates so that the official loan rate that you can kind of get at a bank, it goes down, PBoC is incredibly reluctant to do that. And so, what these measures do by cutting the interest rate for things like the re-lending facility, rather than squeezing banks net interest margins, what you’re doing is you are reducing the funding costs of the banks, which the banks fund themselves at a lower cost.
Their source of borrowing goes down, and they can then pass those savings on to borrowers without squeezing the bank’s own profit margins. And so, this is kind of like a best-case scenario for the PBoC. You can kind of encourage investment. You reduce borrowing costs for those parts of the economy that you think are particularly important. And at the end of the day, you’re not eroding bank profits. So, at the same time of rolling out all these measures, the PBoC had a press conference, and it explicitly said, I mean, the thing which is preventing it from cutting interest rates further is bank profits.
However, interestingly enough, they said that there might be scope for cuts this year because bank net interest margins have stopped falling. They pretty much were stable for all of the second half of last year. PBoC thinks that actually might start rising this year. And the reason is because a lot of long-term deposits, sort of three-year, five-year deposits are due to mature this year. And when they do, the banks can issue new long-term deposits at significantly lower interest rates. And so that will reduce banks’ funding costs. I mean, there’s a whole lot of assumptions baked into that.
But the long and the short of it is the PBoC thinks there might be scope for interest rate cuts because bank profitability might improve this year. Now, of course, the thing to keep in mind is, given the state of the economy in any other place in the world, you would be expecting, you know, repeated rounds of interest rate cuts. I mean, domestic demand is so weak. Fixed asset investment declined aggressively in the last few months of last year. I mean, the domestic economy is not in great shape. And so, under those sorts of conditions, anywhere else in the world, you’d be expecting the central bank to cut rates. But the PBoC can’t and hasn’t because of bank profitability.
And even if bank profitability goes up this year, you wouldn’t imagine that it will improve enough to justify more than one rate cut, maybe ten basis points. If we’re lucky, 25 basis points. So, yet we might get a rate cut this year. But I think ultimately, it’ll be similar to last year. We had one rate cut in May. I think maybe we’ll see the same again this year, but that will be dependent on banks fortunes finally turning around this year.
Andrew: Well, walk us through kind of your view on how impactful all these moves are. I mean, I look at them and they are a pretty small interest rate cut, 25 basis points on a handful of pretty targeted lending facilities for specific economic activities — rural and agricultural businesses, some technological upgrading. So, in terms of scale, what do you think the impact is?
And then secondly, 25 basis points cut is normal size cut in terms of what an average central bank would do, usually in a series of cuts. But the big challenge is in terms of the cost of credit in China, it’s not the cost of credit in China, is that companies don’t want to borrow because they don’t see much demand in the market, right? And so, we don’t see too much demand. It doesn’t really matter what the price of credit is. You’re not going to borrow to invest for the future. It’s the whole balance sheet recession thing. So, anyway, just what do you think the impact of these moves is going to be?
Dinny:
Yeah, I mean, it’s a really good point. These measures, the bigger re-lending facilities and a reduction on the interest rate for the re-lending facilities, these kind of just exist in isolation. Then I wouldn’t imagine this would result in a particularly meaningful increase in either borrowing or investment for the reasons that you just said then. I mean, just the domestic demand is so weak in China at the moment that you kind of need more than marginally cheaper credit to really convince people to go out and invest in their businesses.
What would be interesting, of course, is to see if this is paired with other incentives. And we’re already getting a new quota for fiscal support for firms to upgrade their existing plant and factories and machinery. So, you pair that together with lower borrowing costs. That might create an incentive for more investment among some firms. Or does it get paired with other policies that create an incentive for people to invest?
Beijing is incredibly good at this sort of stuff, shaping policies that create domestic demand for specific industries, like telling industrial firms that they have to source a certain percentage of energy from renewable sources. Ipso facto, you create overnight additional demand for solar and wind, and hydro, and whatnot. So, does this cheaper credit get sort of paired with sort of niche policies like that? I mean, that might have a bigger impact as their combined forces to create demand. But on the face of it, I don’t necessarily think this is going to really move the needle in terms of addressing China’s big problems, which are increasing domestic demand and reviving fixed asset investment.
Andrew: Well, yeah, thanks for that. I think that all makes sense. Just trying to get listeners a sense of like how big these policy moves are. Let’s move on now to kind of last big move over the past few days, which is kind of complimentary, as you said to the PBoC, the monetary moves or credit moves is a better way to characterize that. And these were announcements on January 20th from the Ministry of Finance to support demand in the economy, both through the investment channels and consumption channels. So, on the investment side, Ministry of Finance announced a two-year 1.5% interest rate rebate on fixed asset loans. So, loans for investment for small and micro, and medium-sized enterprises in 14 key industries.
MoF also announced 500 billion renminbi of funding for a government-backed loan guarantee quota to support expansion and facility improvements among medium, micro, and small enterprises as well, and then a central fund to provide credit enhancement for similar companies. The Ministry of Finance also extend its industrial equipment interest rate subsidy, which was introduced in 2024. It extended it from the end of 2025, now through the end of 2026, and expanded kind of the types of coverage that it used to include businesses in the consumer industries, infrastructure industries, AI, and elderly care.
So, all of that’s kind of the investment side of things. And then on the consumption side, the Ministry of Finance extended its 1% consumer loans subsidy through in 2026, removed certain restrictions, and then extended its 1% interest rate subsidy for service sector firms in particular also through in 2026, and expanded some eligibility that we won’t necessarily go into the details. So, a bit of support for consumption, bit of support for investments, kind of like the PBoC moves seemingly kind of marginal. Talk us through, you know the implications here and what you think about it all, Dinny.
Dinny: Yeah. Well, I just realized just how technical and wonky this podcast has become.
Andrew: I’m trying to summarize it because I can see the listeners falling asleep behind the wheel, and I just don’t want to be responsible for a car crash or anything.
Dinny: Yeah, this podcast should come with a driving warning — operating heavy machinery. Okay, so as you said, a bunch of stuff going on in there. The consumer support aspect of this is really a non-event. It extends a program that was launched in August, but it didn’t really seem to get much take up. They’ve extended it a bit here. They’ve expanded it, maybe including credit card debt to stimulate spending. But I don’t think it’s the cost of borrowing that’s really constraining consumers from taking up the facility. I think it’s really household confidence, and you need to fix that before people are going to start borrowing again.
Now, all the investment support stuff is more interesting. So, the fact that the Ministry of Finance is providing interest rate subsidies, this is kind of the same rationale that the PBoC has for leaning into the re-lending facilities and cutting the interest rate on the re-lending facilities. With the re-lending facilities, the PBoC is reducing banks funding costs. So, the central bank is effectively taking upon itself the cost of reducing borrowing costs, right? Rather than pushing the burden onto the banks. With the Ministry of Finance, it’s doing a similar thing. It’s using interest rate subsidies to reduce the cost of borrowing for people, for companies.
But rather than that burden falling on the banks, it’s become a fiscal burden. It’s falling upon the government. Exactly the same rationale, slightly different tool. And then on top of that, you’ve got the guarantees, the credit enhancements. You take all these measures together, and it’s very clear that this is about trying to increase investment by the private sector, not just by reducing costs, but, as much as anything, by convincing the financial sector that it’s worth taking the risk on private sector firms.
That’s what the guarantees are about. That’s what the credit enhancement is about. It’s assuming that it’s not just enough to give private sector firms an incentive to borrow. It is by reducing costs. It is also convincing the financial sector, that’s banks, and the bond market, that lending to the private sector is a good credit risk. So, it’s those two things coming together in the hope that it will stimulate investment. Now, clearly, investment is a worry. Now, I touched on this a minute ago, but fixed asset investment fell 15.1% year on year in December. It was down 12% flat in November, and it was down 12.2% in October. Now, that’s incredible. You know, that ongoing sustained contraction in investment, which has always been a driver of the Chinese economy, I mean, it’s kind of mind-blowing.
And frankly, I think I said this on the podcast, I really thought things would improve in December. And that’s because the policy banks mobilized 500 billion renminbi. I think they distribute it as capital, you know, injected it as capital into local government infrastructure projects in October. I kind of figured it would take November for those projects to raise additional funding. And then the construction, the actual investment would start in December. And additional to that, the local governments got an additional 200 billion renminbi in special purpose bonds towards the end of the year, which they’d be able to use for infrastructure at least partially.
I thought all of this would come together. And that real contraction in investment that we saw in October and November wouldn’t turn around in December, but it just it’d be better, you know, rather than a 12% contraction, maybe it’d be five or three or something. So, the fact that things got even worse in December, it really kind of caught me off guard. And so, I think Beijing is looking at small firms and it’s looking at the service sector in particular as potential new sources of investment. I mean, really, what they’re trying to do is come up with creative new ways to boost investment, so they don’t have to rely on invest infrastructure, and they don’t end up further exacerbating overcapacity in the manufacturing sector and industrial sector.
And so, that really leaves the service sector as much as anything. I mean, Beijing, we’ve seen pretty robust growth this year. I think it was 5.5% in consumer spending on services. It’s a reasonable source of growth for the economy. Maybe we can encourage private firms and service sector firms to invest more. I think that’s what’s going on here. There’s a real feel that Beijing is sort of casting around, trying to work out where might be a reasonable source of additional investment growth, and that’s what they’ve settled on. Now, I can’t imagine that any of this would be a silver bullet, but I think it does give at least some insight into the regulators’ frame of mind as to how they’re trying to deal with this issue of what now is starting to feel like chronically weak investment.
Andrew: Yeah, thanks for going through all that. I think that makes sense. I think I see regulators, officials, policymakers as kind of tweaking around the edges on this investment piece. We will get Joe P. back on the pod probably next week to talk sort of about how all this plays into the broader macro environment, macro growth trajectory. And we’ve touched, as you mentioned, on the investment piece before. I will note that, I mean, the profile of property investment just over the last quarter of the year just got worse and worse and worse.
So that is still the number one thing dragging down this overall investment, you know, the overall investment numbers year on year, making such a large contraction. So, a bunch of that is just property. And policymakers seem sort of resigned to their fate for the most part on property. So, what are they trying to do instead is take the other parts of investment and tweak them a little bit. The biggest overall part of the investment, fixed asset investment, the biggest single chunk is manufacturing investment. And it’s growing in I think like sub-1% year on year right now, like 0.9%, 0.6% or something. So, even if you can take these, nudge that up at the edge through some of these moves, that’s probably a third of the overall FAI, I believe. So, that starts to pick up the overall numbers if you get the private sector manufacturing or private manufacturing sector investing, even just a little bit more again.
So, I think that’s probably part of the thinking here. Interesting point on services. We all need to get Joe to dig into the exact number. I don’t know what the number is in terms of the proportion of investment in the service sector in terms of overall in FAI, but that would be an interesting thing to dig into because some of these moves are also incentivizing investment in services, and that could sort of help to boost FAI at the margin as well. But then the last big one which you touched on is infrastructure. And it’s kind of been stuck, but it looks like, you know, that’s the lever they can really kind of pull more aggressively by adding more central funds if they need.
Partly infrastructure investment has been weak because local governments have been using the funds that they’ve been raising to pay down previous debt. So, that’s the variable I think policymakers have the most to play with. And it seems important to me that they’re, again, not panicking. We’ve seen three, four months, or I guess 4 or 5 months now of contracting FAI, very weak FAI. So, they’re not hitting the panic button. But I will kind of throw it back to you, Dinny, on one last point is that the Ministry of Finance, I think the vice finance minister, Liao Min, a few days ago mentioned that basically that overall government borrowing and spending is going to remain elevated, like just kind of for the foreseeable, but for 2026 in particular.
And so maybe just walk people, to wrap us up, through what they should be thinking about that broader fiscal stance as the NPC, the Lianghui, comes up in March, and kind of what you think the borrowing stance might look like overall from the central government and particularly around supporting infrastructure.
Dinny: Yeah, it’s a good question because I think it’s got the potential to be a real mess this year in the sense that comparing apples to oranges, comparing 2025 to 2026 will be more difficult because I can only assume there’s going to be a bit more of a rearrangement about what local government special purpose bonds, what the money from that is spent on. Does the central government take a greater responsibility for funding infrastructure, and acknowledgment that local government finances are still overstretched? We started to see a little bit of this in 2025, but I think the process will continue in 2026. So, at the bare minimum, I would expect that the target budget deficit will be 4% again this year, like it was last year.
I mean, that was a big deal. I think it was the first time in 30 years that it had been anything over 3%. So, I think another 4% budget deficit. Then on top of that, when it comes to the special purpose bond quota, once upon a time, that was all about almost entirely about funds that would be deployed towards the investment in infrastructure. But these days, some of it’s getting used to pay down local government areas, their unpaid bills. A big chunk of it is being used for debt swaps. So, local governments effectively acquiring the hidden debts of the local government financing vehicles. We had a piece in yesterday’s Markets Note, which kind of broke down, why infrastructure investment declined 2.2% last year.
It’s because these traditional sources of funding for infrastructure have been diverted into really dealing with the local government debt crisis. So that will be difficult to watch, to see whether we get clear numbers about how the local government debt quotas are supposed to be spent this year. And then the other thing to watch will be special treasury bond quota. Do we have another one of those? What is the central government’s commitment to investing in infrastructure?
Yeah, watching how much was going into infrastructure and local government building, it used to be fairly straightforward a few years ago, but now at the beginning of the year, when all these numbers come at you at the NPC, it’s difficult to get as clear of a vision about exactly what it’s going to mean for growth versus cleaning up the financial mess at a local level versus other government priorities. So, yeah, look, I wish I had a clear answer, but I think the thing to keep in mind is that there’s going to be a lot of moving parts when the numbers finally come out in a couple of months.
Andrew: Yeah. And I’ll just also reiterate that people make a lot of the monthly lending numbers, which used to be a very reliable indicator of kind of the health of the economy and the pace of growth. But similarly, that indicator has become much less reliable because much of the funding that’s raised each month goes to paying down previous loans. So, the net credit creation ends up being very weak, and because a lot of government investment vehicles in particular, are issuing bonds to pay down previous bank loans. And so, it’s all kind of a mess. And in a way, like I said earlier, that’s sort of the definition of a balance sheet recession like companies and local governments are prioritizing balance sheet repair over new investment.
And that’s just going to be a process that plays out for a while. And so, we will continue to track this. We’ll get more data and projections on the fiscal side of things. I know everybody’s, you know, favorite Trivium China podcast topic is fiscal policy. It’s the most interesting thing in the world, right? But it is so important, which is why we spend a lot of time on it. And it’s kind of become the main lever of managing the economy when it was credit monetary policy for so long. I guess, you know, well, that leaves a lot of questions unanswered. But thanks for walking us through the specifics of the moves out of the gate here on the fiscal, monetary, and stock market side as we’ve started January, Dinny. It’s good to be back in the saddle. Thanks for joining me today.
Dinny: Yeah, absolutely, mate. Great talking to you.
Andrew: Thanks so much. And stick around, everybody, for my conversation coming up now with Linghao Bao, our analyst on semiconductors and AI, talking about the Manus-Meta deal and the wider kind of dynamics behind what’s happening in the Chinese AI startup scene. It’s great conversation. So, take a listen.
I’m now joined by a brand-new guest to the Trivium China Podcast, but a longtime colleague of mine, and that is our Lead Semiconductor and AI analyst based in Shanghai, Linghao. How are you doing, Linghao?
Linghao Bao: I’m good. I’m glad to be here.
Andrew: Yeah. It’s great to have you on, man. Glad to have you on this pod for the first time and to talk about-
Linghao: Yeah, I was wondering why it was taking so long for me to be on the pod.
Andrew: It’s the time zone issue. It’s the time zone. It’s so hard to record across the world, late nights, early mornings. Just to timestamp, this part of the pod, we are recording at 10:30 P.M. U.S. East Coast time on January 22nd. So, this is our first start of 2026. We took a couple weeks. So, I’m glad to have you on the first part of the year, man. Welcome.
Linghao: Yeah, I’m glad to be here.
Andrew: So, today with Linghao, I’m going to get into something that’s been sort of developing over the past few weeks in the tech space, and that is specifically the acquisition or the mooted acquisition of the Chinese AI startup called Manus by Meta, which was announced back in late December, yet in the early days of 2026, January 2026, it became clear that China’s regulatory apparatus, Chinese officials have taken an interest in this potential acquisition and are, at the very least, looking into it, and potentially likely to intervene. So, we’re going to talk a little bit about why this matters, sort of for the AI space generally, what it means for kind of U.S.-China tech competition, what it means for kind of China’s levers in terms of how it can control and influence its AI companies. We’ll get into all of that, but I just want to start with a little bit of the background here.
So, December 29th, 2025, Meta announces that it’s going to acquire the Chinese AI startup Manus AI. The reports were that the acquisition would be for USD 2.5 billion. Now, inside China, this deal was widely applauded as a total home run, a blockbuster exit for startup founders who built a breakout agentic AI product, and a vindication for early venture investors who place right bets on this company. But then, just a few days later, the bad news came, which was that Beijing is getting involved. So, on January 8th, the Ministry of Commerce spokesman He Yadong said this, “The Ministry of Commerce will, together with relevant departments, conduct an assessment and investigation into whether the Meta-Manus deal is consistent with relevant laws and regulations, including those on export controls, technology import and export, and outbound investment.”
So, a lot of different angles that they’re looking at this deal from. So, at this point, this is not just a tech acquisition any longer. It’s officially morphed into sort of the U.S.-China geopolitical zero sum game, similar to the forced sale of TikTok with the Chinese state wanting to have some influence over how that went down, which actually, I guess that deal is timely because that official transfer of TikTok to U.S. investors was announced today, earlier today. So, there’s going to be more back and forth anytime there’s some kind of tie up between U.S. investors and U.S. companies and Chinese tech companies, specifically in the AI space. To understand why Beijing is stepping in and what it could mean for the deal’s outcome, we first need to take a look at Manus’s origin story. So, Linghao, I’m going to bring you in to start here.
You’re the expert. Why don’t you just kick us off by telling us a little bit about Manus AI, kind of what the operating model is, why is it such a cool company, why would Meta be looking to acquire it. W hat’s the deal with this company?
Linghao: Yeah. So, the company was founded in 2022, in Wuhan, China. So, it’s a pretty young company. And their first breakout product is called Manus. They launched a AI agent product in March 2025. For the audience who don’t know, this product kind of looks like ChatGPT. It has chat interface, but instead of just answering questions, it can do things for you, right? It can be connected to various external environment like browsers, Google Workspace, slack, etc. So, it’s kind of like giving an AI a computer, and Manus just using a computer to do things for you, like send a slack message for you or send emails for you. Right? It does things for you. I mean today, like this concept is not really new.
But a year ago, there wasn’t like any product like that, or I should have said like the demo just sort of looked really good. They made it looked like the thing actually works. So, the product launch was an immediate hit. Every tech media bloggers inside China were covering the news, and they had so much demand coming in. Like, on day one, their model providers couldn’t service them because they just don’t have enough GPUs for them. And we’re not talking about like Alibaba and ByteDance. We’re talking about like U.S. hyperscalers like Google, Amazon and Azure because they use overseas models. And the reason is because these type of products just consume a ton of compute, right? Instead of just spitting out the answer in the few seconds, they go out and do things like it might take like several minutes. It just consumes a ton of compute and GPU resources, right?
But anyway, the product launch was an overnight success. And the growth rate just wasn’t up to the right. The product went from zero revenue to $100 million ARR in just three quarters.
Andrew: Oh my gosh, that’s insane.
Linghao: It’s absolutely insane. I mean, that kind of growth is really staggering, right? I think it was probably the fastest growing Chinese AI product in terms for revenue ramp. And even if you compare it to some of the popular U.S. AI apps, it’s really up there in terms of revenue velocity. I mean, it’s not as crazy as the ChatGPT and Claude Code of the world, but it’s definitely in the same league of Cursor and Lovable. So, it was really a success story for a Chinese startup like you don’t hear these kind of breakout products anymore from China, right?
Andrew: Yeah. And I mean, you make a good point that I guess now, fast forward almost a year to 2026, that more and more of these type agentic products are starting to pop up, but it’s still a unique product. It was an early mover. And also, it makes sense to me just hearing you say that, that if it’s sort of competing with the likes of, you know, OpenAI or Anthropic, that of course, Meta is going to be a natural suitor, right? Because it doesn’t have the same in-house capability like Llama is not used so widespread as other products. And also Meta has been on this huge AI binge like many of the big tech companies, but to really, really an acquisitive both in terms of personnel hiring people at astronomical rates but also buying up small companies to really kind of get in on the AI game.
So, that all makes sense to me. But there was a particular aspect of Manus that made it ripe for a non-Chinese kind of acquisition, specifically U.S. acquisition. And that was kind of how it not only structured itself in the market, but also like how it raised capital, which was actually to get out of China quite early on in its journey. So, walk us through kind of that aspect of sort of how Manus is structured and sort of which markets they’ve been trying to focus on.
Linghao: Yeah, yeah. So, one of the key decisions they made in the early days is that they went after the global market on day one. The Manus founder, Xiao Hong, explained like the rationale like pretty explicitly in a podcast. It was even before like the Manus product launch. He goes like overseas customers are like five times more willing to pay for software subscription than Chinese users. That’s number one. Number two, you get paid in dollars. So, that’s another multiple of seven. And five times seven is 35. So, the TAM is at least like 35 times larger. So that’s a pretty compelling argument, right?
Andrew: Totally.
Linghao: Yeah. So, they stuck to that strategy. And also, like if you look at the product launch, it wasn’t like launched during… I think it was like 10 P.M. China time. But that’s the like a time like people go to work in the U.S. And if you do a product launch, if you’re targeting the Chinese market, you wouldn’t do that at 10 P.M. China time. And the co-founder was speaking English the entire time in the launch video. So, this is a definite a US market first product from day one because like that’s where the paying customers are. There are also a few things they did to execute that strategy. Like, as I mentioned, after the product launch, they got super popular in China.
And there are rumors that Wuhan local government actually approached them and they wanted to invest in them, but they said no. Yeah. And then they turned around and the raise some money from Benchmark, which is a very famous U.S. VC firm. And then I think that was in April. Right? And then in June, they moved completely out of China and relocated to Singapore.
And had they not done these serious steps to remove their Chinese identity as much as possible as a company, I don’t think they would even get a acquisition offer from Meta because there’s a piece of U.S. regulation which is called the reverse CFIUS, which some listeners would know, like these rules don’t allow U.S. investors to invest in Chinese AI companies, right? It also had very strict restrictions over other sectors like quantum and semiconductors. And apparently, like Benchmark drew some scrutiny from U.S. regulators for that Manus investment, according to some reports.
Andrew: Because that’s still when they were located in China, or was that already when they were in Singapore?
Linghao: Yes. That investment, that fund raising happened when they still had operations in China.
Andrew: Okay.
Linghao: But there are also reports that that investigation faded after Manus relocated to Singapore because right now, they don’t have any employees in China anymore. So, these serious steps, it basically paved the way for later acquisition by Meta.
Andrew: Right. And so, you can see kind of obviously, I guess, why the U.S. government would have concerns around a major investment by a U.S. VC firm in the Chinese AI startup for geopolitical calculations and for competitive calculations vis-à-vis China. But it sounds like, you know, through this sort of divestiture of the Chinese market, moving operations to Singapore, they were able to allay those fears. I mean, it’s pretty remarkable that you say that they turned down money from the Wuhan government in order to raise foreign capital. I mean, that’s like the way you win in China is raising government money half the time, right? Or if not more than half the time.
Linghao: Yeah, but if you want to keep your option open for later acquisition by a U.S., big tech, you would never want to accept money from local governments, right?
Andrew: Yeah.
Linghao: I mean, if Meta like look at Manus’ cap table and say, “Hey, what is this Wuhan Industry Investment Fund?” They would never like make the offer in the first place.
Andrew: Totally, totally. Well, that just goes to show that the founders of this company kind of had that vision in mind the whole time, and we’ll kind of get into sort of some of the reasons behind that later in our discussion. But it’s unique among Chinese companies to be intentionally or globally focused from the jump, wouldn’t say?
Linghao: I think it’s not so unique anymore. Yeah, I think this generation of AI companies and talent, a lot of them are actually targeting the global market from day one, because I think it’s the same reason that I just mentioned that the market is so much bigger, like overseas markets. And it’s just so hard to make money in China anymore. It’s a pretty stark contrast from how it is, what it was 10 years ago.
Andrew: I mean, that’s like sort of a binary choice, right? You either service the global market or you service the China market for the most part, unless you’re like one of the, as a startup company, if you’re like one of the Chinese major tech companies you can service, at least other markets, maybe not Western markets, but you can do global and China. But for the most part, it sounds like you’re saying you’re either going to choose China or you’re going to choose the rest of the world, right? I mean, that at least is how I read it from what the actions Manus took.
Linghao: Well, I think that only applies to a few sectors, right? Because the reverse CFIUS only have restrictions over AI, semiconductors, and quantum. So, I think, you know, it basically only applies to AI, because semiconductors, you can sell your product. And like semiconductor companies, they will raise money from Chinese government. I mean, most company investors are Chinese government. I think it only, actually, only applies to this cohort of AI startups. But that’s really important, right? You don’t want to lose your AI ecosystem.
Andrew: Well, that was actually my point. That was my point was I was actually referring to AI companies. I mean, I’m not talking about tech companies generally. I’m saying like, if you’re a Chinese AI startup, like you kind of have to make a choice. You’re either all in on China or you’re all out on China.
Linghao: Yeah. I mean, I think Manus is sort of extreme case. They sort of really committed to it. I think one reason, you know, they moved so fast is because I think it’s about like keep your options open for a later acquisition, right?
Andrew: Sure.
Linghao: If you don’t think you’ll be ultimately acquired by a U.S. big tech, then you don’t really have to move out of China. You can still have your operations in China and still target the overseas place, right? So, it only applies to those startups that want to keep the option open for later acquisition.
Andrew: Well, speaking of which, it seems like the actions that Manus by moving to Singapore and mostly banning the China market allayed the concerns of U.S. regulators if you were reviewing this. But, of course, it ended up getting the attention of Chinese regulators. So, talk to us about sort of why China is looking at this, so we can kind of get into sort of, I don’t know, I’m feeling, maybe I’m wrong here, but you’re sort of damned if you do, damned if you don’t. If you’re too tied to China, or tied to China at all, the U.S. is going to have a problem. If you’re a Chinese company, even operating outside of China, as we’ve seen here with Manus, then the Chinese government’s going to take an interest. So, talk to us about why you think China is looking to intervene here. And we can talk about whether we think they will actually try to stop the deal or if it’s just a review, and maybe they’ll try to extract some concessions from the U.S. in the process.
Linghao: Yeah. I mean, you never know what’s on their mind, right? But I’m guessing like they’re afraid that this madness playbook becomes a trend. We just mentioned like this acquisition was applauded as a homerun success inside China. And what if other VCs started telling startups, “Hey, you should do the same, right? You should move out too able and so we can prepare you for a U.S. big tech acquisition.” So, now you’re talking about a risk of outflow of talent in the startup ecosystem. And also, what if next time it’s some company that’s more important? What if it’s DeepSeek? I’m not saying DeepSeek wants to do it, but if you’re a Chinese government, you probably want to have some oversight in case some sort of core technology got transferred out of China under their radar.
So, perhaps they’re using this investigation to let the companies know, hey, if you’re relocating outside of China and bring your technology with it, we need to take a look first. I think that’s number one. Number two, I feel like maybe they feel like they are being outmaneuvered by the U.S. government.
Andrew: The Chinese government feels that.
Linghao: The Chinese government. And they want to do something about it. I mean, because like you just said, like one core reason why Manus tried so hard to move all of China is keep the option open for later acquisition. And these serious actions are directly downstream of U.S. regulation because they’re investors have to comply with these rules. I think China maybe just fanned out like that piece of regulation may be doing more damage than they thought originally.
Andrew: The reverse CFIUS regulation, you mean, specifically?
Linghao: Yeah, the reverse CFIUS regulation. Because when it was initially released, there was not much reaction from the Chinese side. If you think about it, what they would always do was sort of pull the U.S. LPs from China’s private market funding. But if you look at within the grand scheme of things, that’s only like single-digit percentage point of the total private market funding. I mean, the number might be more significant if you just count early venture investment because local government don’t do that. But that’s not an issue that China can really solve internally.
They can sort of push their source of funding towards more early venture. And it’s not a huge deal at first glance. But now, if you’re a Chinese policymaker and you found out, you just found out it’s now causing some of your best AI startups to leave China and maybe more to follow, I think your alarm just goes off immediately. It’s an entirely different issue right now. So, you probably want to tap on the break to see like what’s going on here. And in fact, like there was a piece written by a university professor called Cui Fan. Cui is a figure, what we call as a policy advisor. He’s those think tank type of figure that has the ears of policymakers.
And his piece was that I published just a few days after the deal announced, and he specifically said that he underestimated the impact of the reverse CFIUS rules. So, I think that what we’re seeing here is a knee-jerk reaction from the Chinese government. And if you look at the timeline, like everything happened so fast. First, we saw the acquisition announcement, then a few days later, we saw Cui Fan’s piece, and then a few days later, we saw the official statement from the MOFCOM saying they’re going to investigate this deal. So, it happened so fast. That made me think they’re actually not even sure what they’re going to do. But they feel like they need to do something totally.
Andrew: Well, that makes sense, actually, as I think about it. I can understand from the Chinese policy perspective. And I mean, I would say I kind of had this thought as well that the reverse CFIUS kind of role is not that big of a deal. If I’m a Chinese policymaker, I’m thinking, all right, whatever, we don’t need American money investing in our high-tech ecosystem, in quantum, in AI, and chips, and all this stuff. So, fine, keep your money. Don’t let your investors invest in our promising companies. But then if one of the outcomes is some of your high-profile startups start moving overseas so that they can keep their options open, then absolutely, the reverse CFIUS rules are starting to have what sounds to me like sort of, I think, unintended impacts, right?
I don’t think that was what U.S. policymakers had in their mind when they were saying, “We don’t want U.S. companies investing in China.” They weren’t saying, like, “We’re hoping these companies move out of China.” So that’s an interesting wrinkle that I think probably neither sides are coming. So, you can totally understand why China would say, “Okay, wait, wait, wait. We do not want this to become a trend.”
Linghao: Yeah, exactly.
Andrew: So, talk to us then a little bit about, so China’s obviously, Chinese authorities, MOFCOM in particular, has expressed the intention to review this thing to investigate. We don’t know the level of intervention. But talk to us about the sort of mechanisms through which they might intervene. What are the lines of investigation they are sort of claiming authority over when it comes to this company that ostensibly is a Singaporean company now, right? How are they claiming jurisdiction, and what are they trying to look Manus for in terms of potential, I don’t know, what’s the kind of spirit of the law they’re trying to put forth here?
Linghao:
Yeah. So, to be honest, I totally did not see that coming when the deal was announced. Because, normally, like for Chinese a regulator to review a mergers and acquisitions, there are some thresholds that need to be triggered firs. And then it’s normally reviewed under the antitrust laws. It’s under SAMR. And for SAMR to initiate a review, the companies involved needs to have some presence or some revenue inside China. But neither Meta or Manus do. So normally, it wouldn’t even cross your mind like the Chinese government has anything to do with it.
Andrew: In theory, they shouldn’t have jurisdiction, or at least SAMR, the market regulator, wouldn’t have jurisdiction over this deal, right? In theory.
Linghao: Exactly. So, that just speaks to, you know, how creative the Chinese government has become to get themselves involved in this, right?
Andrew: Uh-huh.
Linghao: And the specific regulatory toolkit they’re going to use on Manus is actually the same one we saw it used to block the sales of TikTok. It’s not revealed through the antitrust lens but through a technology transfer angle. So, MOFCOM has a regulation in place that require companies to seek approval when they transfer certain technologies outside of China, and the scope of the technology is vague enough that it’s applicable to TikTok and Manus.
And if you think about it, the TikTok deal was kind of similar because TikTok doesn’t generate any revenue in China. It’s totally facing overseas market ex-China. So, normally, the Chinese government doesn’t have jurisdiction. And Manus was similar, right? The product and the technology was created inside China. So, when Manus relocated to Singapore, it brought its technology with it. But they didn’t seek government approval for that tech transfer. So now the government says, “Hey, you skip this legal process, even though you know the relocation is done, we still have to go through this process.” In other words, they’re not reviewing the acquisition deal per se, they’re reviewing the tech transfer from Manus China to Manus Singapore as an indirect way to exert influence. That’s pretty creative.
Andrew: Absolutely. I mean, that it’s very creative. And I guess we’ll talk a little bit around your expectation kind of how this plays out, what Chinese government’s endgame might be. So that’s the main angle that they’re going after. I don’t know if you have thoughts on this, but our pal over at GeoPolitics Substack, great Substack on Chinese Tech wrote that he also had heard that potentially there was an issue around Chinese citizens owning equity in non-listed overseas companies. Is that a part of potential investigation, or do you think that’s not going to be a line of inquiry for authorities on this?
Linghao: I don’t think so. I think there are probably many cases that Chinese citizens owning overseas assets.
Andrew: Yeah. So maybe that’s too technical. It does seem like that law is pretty hard to enforce and not use. So, okay, I just wanted to kind of get check that to see if there might be multiple avenues through which they might kind of try to slow things down here.
Linghao: Yeah. I think, you know, there are multiple avenues as sort of mentioned by the MOFCOM press release. So, they mentioned the technology transfer. They also mentioned the export controls. They also mentioned all banned investments. I’m not sure if they kind of use all of these toolkits. So, my view, I think the technology transfer angle is the most likely because they’re familiar with it. They already used it on TikTok, and Manus is the same type of deal.
Andrew: But that makes sense. Well, you’ve laid out what Manus is, why it’s important, why we care about it, why Meta’s interested in it. We’ve talked about sort of the unique structure of the company and how it sort of set itself up for this deal, why the Chinese government got interested and the paths it might take to sort of slow things down. So, what’s your best guess on sort of what’s next? Will Chinese regulators try to stop this in the same way that they sort of tried to stop the TikTok deal? I mean, that was obviously a fundamentally different thing because it was under duress, right? It was a forced acquisition by the U.S. government. But would they want to keep this from happening?
Will they try to extract some kind of concession from the U.S. or from Meta in as part of the deal or? I don’t know, how do you think this plays out? What do you think Beijing’s main sort of calculus is in how they proceed?
Linghao: Yeah, I think the dream scenario for them is actually extra, some concession from the U.S. I’m not sure if they can. It would be great if they can pack this little issue into the broader negotiation with the U.S. But does Trump care if Meta acquires Manus or not? I mean, he definitely seems to care about buying TikTok. Right?
Andrew: Yeah. But that’s an order of magnitude sort of bigger deal, right? In terms of U.S. users and the sort of reputation of the company and all that stuff. Right?
Linghao: Exactly. That’s exactly the point. So that’s why I feel like China doesn’t have all the privilege to actually pull U.S. government to the table. So, I’m actually not sure what their end game is, or they do have an end game in mind right now. As I said, like everything has happened so fast. Maybe it’s just a national panic. They feel like they need to do something.
Andrew: Yeah. Tap the brakes.
Linghao: Yeah. But I think there’s a risk of overplaying their hand because if they just say, “Hey, we’re going to stop this, you can’t acquire Manus,” I think that would probably backfire. Right?
Andrew: How would it backfire?
Linghao: Yeah, I think it would undo a lot of efforts of the Chinese government’s own policy initiative. One, they want to encourage entrepreneurship. So, punishing success is not the way to do it. It sends the opposite message. And number two, the government is trying to help the Chinese VC industry to get back on its feet, which hasn’t been in a quite a good shape for a few years now. And now is probably not a time to pour cold water on them. And third, like the Chinese government doesn’t want to keep Chinese companies confined to domestic markets. They want Chinese companies to go out and compete and succeed in global markets. So, in many ways, like Manus is kind of the canonical example of success that the Chinese government wants to promote.
But now, because of geopolitics, they feel like they need to take control of the situation. That’s where we are right now. But I would argue, if you punish Manus, that will actually hurt China, actually not protect China. And yeah, I feel like, at this stage, I think, this deal is still negotiable. I think they’re still in the process of assessing their options. And then the first thing they’re thinking about, whether we can pull U.S. to the table, if we can, what are we going to do with the deal? I think it’s still in the very early stage of deliberation.
Andrew: Yeah. Well, I mean, it’s not even clear to me that the U.S. government’s really involved at this stage. I mean, I’m sure there’s like some level of, I mean, to the extent that the two sides are talking, which they obviously are because of the trade negotiations and all that stuff. I’m sure at some working level, this has been raised as an issue, but wouldn’t strike me as a central issue. So, it’s not clear to me like how involved, as you said, Donald Trump obviously, this may not be on his radar. He probably doesn’t care. The wider USG, how much do they care about this deal? And if there’s no real interest there, there’s certainly not going to be an opportunity for China to extract any kind of concession if the USG is not even at the table.
So, that’s just kind of me thinking out loud on that side. But then the second thing I’ve been thinking is, you know, if Chinese regulators want to just stop this as… I mean, I guess the one thing that they could try to get a concession on is some kind of watering down of the reverse CFIUS rule, which if that somehow or to wind itself into the issue that the U.S. and China are negotiating over, and maybe it’s something the U.S. could give in response to some other concession China’s giving in the broader negotiation. So, I could kind of envision that happening. And we’ve seen it, you know, those kinds of things, like the loosening of export controls are on the table for the Trump administration. But that’s just kind of thinking out loud.
The real question I would ask is, what if Chinese regulators decide we just don’t want this to be an ongoing trend? As you point out, you’ve just said that China is where you say they want their companies competing and thriving abroad. But they may not want a wholesale exit of their AI startups who are trying to keep their options open for potential acquisition by foreign, especially Western companies or investors. So, do you think there’s any chance they just stop it because they want to sort of make an example out of Manus? And even though you just said it should be the poster child of success, there’s also this flipside of, yeah, but we also want to open the floodgates of everyone leaving the domestic market.
Linghao: Yeah, I guess that’s possible. Well, I think, by and large, they’re still rational — famous last words. But sometimes we’ve seen a few examples of those irrational decisions. But I mean it’s possible they’re trying to make an example out of the Manus, to send a signal that all of these type of relocations requires government review. But I’m not sure that’s the only way to send the message. You don’t have to kill the deal to send a message. I think there are other ways to do it. Just give Manus a fine or something. That will probably do it too. I don’t know.
Andrew: Yeah, now that’s a great point. Well, so let me just nail you down on this. I mean, it’s early days, the acquisition was announced less than a month ago. The investigation was announced just a couple of weeks ago. But if you had to put a percentage chance on it that this acquisition does move forward and the Chinese sort of investigate, but at the end of the day say, “Okay, this is okay, what’s your percentage that it moves forward?”
Linghao: I’ll give 70%.
Andrew: Okay. So, you think it’s high. You think this is a hiccup or a road bump, not a genuine like extended turtle.
Linghao: Yeah, that’s my real situation right now. I mean, if you look at the Cui Fan’s piece, he also mentioned that these types of success should be encouraged. But he also stressed that companies need to comply with Chinese laws and regulations. I think, broadly, Chinese regulators understand how much damage has been done during the tech crackdown. I think they’ve learned the lesson of not overdoing it. And the confidence has just recently been rebuilt to some extent. And that confidence is pretty fragile. It takes time to build it. But it only takes one event to completely destroy it. So, I think they totally understand that. So, that’s why I think I give 70% because it knows the impact of their decisions.
Andrew: Yeah, that’s good to know. I think this is like hugely impactful. It’s good to know that you’re sort of outlook is that this is probably going to go through. It’s still instructive as to the fact that the Chinese government’s taking an issue at all, and that it’s sort of led by MOFCOM, and they’re finding these creative ways to sort of gum up the works and make sure that their interests are asserted, the states interests are asserted in this deal. We’ll be interested, though, if it has any long-term sort of reverberations on China’s thinking or approach towards the reverse CFIUS rule, right? Because that never sits well with Chinese authorities, the fact that some U.S. regulation is impacting the way in which its companies are operating, both at home and now potentially abroad.
So, my guess would be that there will be kind of lasting reverberations that Beijing kind of thinks through, all right, maybe they’re bigger implications here because of this regulation than we thought before. What those are, we won’t know. But you’ve walked us through everything so eloquently and really appreciate it. I just wanted to wrap up with one other question, which is kind of a step back on the context, again, of why Manus sort of wanted to keep its options open aside from getting acquired by U.S. tech company, but more like some of the structural deficiencies on Manus would be sort of less interested in zeroing in on the China market in the long-term.
You write that, “The Manus story also exposes a deeper structural issue of the Chinese economy, which is weak demand. If the Chinese market was as monetizable as U.S. markets, perhaps Manus would never have left its own market in the first place.” So, you argue that people sort of underestimate how big of an advantage the U.S. has over China, and in that respect, by having the strongest consumption-led economy in the world, which helps to kind of pull in talent, capital, and all the leverage that comes with that. So just talk to me about how you think about that and how that sort of plays into this story, but also how that might impact the bigger market for AI startups in China more generally.
Linghao: Yeah. So, if we just take a step back from geopolitics, the fundamental reason that Manus wanted to move abroad was not the reverse CFIUS rules, right? The rule cost is weak domestic demand of the Chinese economy. I think that is the deeper issue causing these Chinese startups to choose U.S. or China, to pick a side, right? I mean, like you just said, if the Chinese market were just lucrative as the US market, would Manus actually have moved out? I mean, you never hear a U.S. tech company say, “Hey, we’re going to go after the European market first,” right? they’re in the US, right? Naturally, you want to target your home country first. And it will be nice if it happens to be the biggest market in the world.
But China doesn’t have that anymore. It’s so hard to make money here. So, Manus basically decided, we just don’t want to try. We don’t even bother trying. And like we just mentioned, Manus is not alone. A lot of Chinese companies and startups are going after a overseas market right off the bat. And that’s really different from 10 years ago during the mobile internet era. Right? People always say China is the biggest market. We got 1 billion internet users. No other countries give you that many potential customers. But the problem is these 1 billion customers don’t spend money. So, that’s why I say like people underestimate like how big an advantage the U.S. has over China just by having the strongest consumer-led economy in the world.
When you have the biggest market, it’s like a magnetic field, they’re posing the talent, the capital, and the innovation that comes with it. But when people talk about U.S.-China competition in AI, the only words you hear is chips, chips, chips, and power infrastructure. I mean, and don’t get me wrong, these issues are really important, but having a economy with the appetite to consume those technology is also really important because market is ultimately what drives revenue and profits, and more revenue allows you to spend more money on R&D, and it drives innovation.
So, it’s a flywheel. And actually, if you make more revenue, it allows you to buy more chips. If you look at the Capex of the U.S. hyperscalers and Chinese hyperscalers, the U.S. companies spend five times more, give or take, and then you can just look at their stock price, right? The max 7 has been going up to the right since forever. But the Chinese tech stocks haven’t recovered from their 2021 high. Why? Because the revenue has not really been growing. Their profits have been flat. I think that ultimately has something to do with the economy. Anyway, I think a shift to a consumer-led economy is pretty urgent at this moment because it’s not just a macroeconomic picture anymore. It’s starting to affect your tech competitiveness.
Andrew: Yeah, that’s a really good point. So, you actually make a ton of great points there. And there’s a bunch I’d love to follow up on. I think each of them would be a podcast in themselves. One is sort of the profitability of Chinese tech companies, why that’s been lagging really since 2021, ever since the tech crackdown. That’s a whole thing we could get into. But your point about the consumption story, the weak and someone’s story becoming a liability, not just for the macro economy, but for some of these tech startups is also interesting, I guess the one, and I’d love to dig into that more, we don’t have time to go into it fully now, but the one thing I I’m just thinking is I wonder if some of these sort of consumer-facing I startups in particular are in a sort of weird zone in China.
Because the overarching, and Kendra and I have talked about this on the podcast a lot, the overarching policy support mechanism or policy framework for AI in China is the AI plus policy, right? It’s all about industrial diffusion of AI technology, right? How can we make manufacturing more efficient and industry more efficient? It’s not the ChatGPT model of, “Hey, I want to look at, you know, a recipe for dinner tonight or I want a new workout regime.” It’s not really consumer-facing as far as the government’s concerned. They have much less sort of focus on supporting AI companies that are kind of building that type of model.
So we talk a lot about how China’s approach is very different from the U.S. in that it’s a less focus on AGI specifically and more about like let’s get the AI technologies we currently have and the ones that are going to develop quickly in the manufacturing sector, in the industrial sector so we can boost the productivity of those firms. Does that leave these consumer-facing AI tech companies in a, I mean, just a weird regulatory gray zone in your view, or am I just totally off base there?
Linghao: I think, yeah, the state wants AI technology to be diffused in every part of the economy, but it hasn’t really happened yet. We’re just talking about like the consumer demand has been lacking, but China’s enterprise market is notoriously difficult to serve. There’s a reason why SAAS had never taken off in the Chinese market. Right. In the U.S. market, you have all these hundreds different enterprise software that is subscription based, right?
You build the software once and you just copy them endlessly and distribute them at a zero marginal cost, and that gives you a very hard margin. But that model doesn’t work in China because every single Chinese customers want their own customizations. It turns your software companies into a consulting company, basically. So that’s why a lot of software companies in China, they’re only thinking about the consumer market. They’re not even thinking about like the enterprise market because it has been a very terrible market to serve, to crack. But yeah, even the consumer markets in China right now, these consumer apps are not really making any money. They don’t have an advertising model yet on these apps. I mean, they do have subscriptions, but nobody’s paying for them.
So, the business models of these AI companies still haven’t been figured out. When the ChatGPT moment happened, it was basically an accident, right? It was a AI research lab in the beginning, then ChatGPT happened, and Sam Altman was like, oh shit, we might have a consumer tech company at our hand that seized the opportunity, and they started immediately charging subscriptions. And that revenue can pay for your compute so we can scale right from now on. And then you can even use that business model to raise more money, which gives you more compute. But Chinese companies can’t really do that, right?
Andrew: Well, if you don’t have the subscribers, yeah.
Linghao: Yeah, yeah. I mean, this DeepSeek moment was kind of similar. I mean Liang Wenfeng could turn that company into the ChatGPT of China. But he didn’t really do it. I mean, you could say like he was like staying true to his mission that he just wanted to be AI research lab, his goal is AGI, but there’s also this reality, like, if he wanted to, he doesn’t have that luxury that Sam Ohman has to turn it into a consumer tech company immediately.
Andrew: That’s a great point. Again, you’re just raising so many issues that I want to go down a bunch rabbit holes on. One is I hadn’t even thought about the software enterprise kind of realities in the SAAS model, and why that doesn’t really work in China. You make a compelling case that making an AI company or any kind of software company that’s consumer facing is a very difficult road in China. We’ve got to pick up all of these on more podcasts. I got to have you on again. This is great conversation starter of the year. But one final, final question is I was thinking actually earlier while you were talking is why wouldn’t an advertising model work? You can still have the billion in that case for these companies, it’s still just a user acquisition game, right? And you still do have the billion potential users in China.
And then if you’re making money through ads, why not do it that way? So, you’re not asking people to build money out of their pockets. I mean, that’s a tried and true methodology, and you’re B2B business, but your focus is on, in terms of user acquisition, it means you’re just trying to get your product out there to as many eyeballs as you can. Why is that not a viable model?
Linghao: Yeah, I think it’s a viable model, to be honest. I’m not sure why they haven’t done it. But the thing is, at the end of the day, if you are launching a advertising model, you’re also competing with existing incumbents because the advertising market is not really growing. Again, like it’s sort of capped by the growth of your economy, right? If your economies aren’t really growing, then these companies don’t really increase their advertising budget so the pie is not really growing, and you are competing with the incumbents. Do you really have a chance to succeed or not? I think doubt is a big question. So, I think that ultimately, I guess my point is there are a lot of challenges for these companies to figure out their business model. And I think one thing that can really fix it is actually you get the economy back on its foot first.
Andrew: All comes back to the economy, which, as a macro economist, makes my heart sing. So, I’m glad you think that’s the key to everything. Well, awesome, man. We’ve done it longer than I intended, but this has been a great conversation. We veered from specifically the Chinese government, MOFCOM intervention into the Manus deal and its implications into much wider implications of kind of how mantis operates and the challenges for consumer-facing AI companies.
There’s a ton more threads I want to pull on here over time, so we’ll definitely have you on to talk about a ton of these issues because I know this is, as we’ve talked about, this is kind of a space that not a lot of Western observers, policymakers, Western companies, even the tech companies really know that much about. The Chinese sort of AI ecosystem, especially in the startup space, is not well understood. So, I appreciate you walk me through all this, man. It’s great to see you. Thanks so much for the time. And we’ll definitely have you on again soon. So, I appreciate it.
Linghao: Yeah, thanks for having me.
Andrew: And we appreciate everybody listening. We’ll see you next time, everybody. Bye.
