China shifts AI focus to physical robotics, targeting $12T investment and margin expansion by 2030
Summary
Laura Wang and Robin Xing discuss China's strategic pivot from cloud-based AI to physical AI, including robotics and smart factories. They cite Morgan Stanley forecasts predicting $12 trillion in incremental Industry 5.0 investment over the next decade. Xing notes that while current bottlenecks in chips limit immediate macro impact, productivity gains are expected from 2028 onwards. Policy shifts aim to reduce overcapacity, potentially raising industrial profit margins from 5% to 8%. Wang forecasts Chinese equity earnings growth exceeding 10% by 2030 as the country leverages its manufacturing depth.
Key takeaways
- 1 The addressable market for broader frontier industries is projected to grow from $1.4 trillion currently to $3.2 trillion by 2030.
Unidentified speaker (speaker_0) · 0:47
- 2 Over half of the $12 trillion physical AI investment will focus on rewiring existing factories for proactive, adaptive production rather than greenfield expansion.
Robin Xing · 5:00
- 3 Industrial firm profit margins in China are expected to improve from 5% to 8% over the next three to five years due to these policy and investment changes.
Robin Xing · 6:27
Episode sections
- 0:00 – 1:41
Introduction and Market Forecast
Laura Wang introduces the topic of physical AI and Robin Xing, forecasting a $12 trillion incremental investment in Industry 5.0 over 10 years and a frontier addressable market expansion from $1.4 trillion to $3.2 trillion by 2030.
- 1:41 – 2:41
China's Physical AI Strategy and Bottlenecks
Robin Xing explains China's shift from cloud-based AI to physical AI (robotics, factories) to leverage manufacturing depth, noting current bottlenecks in chips and upstream equipment limit immediate macro impact.
- 2:41 – 4:40
Productivity Payoff Timeline
Xing predicts broader productivity gains of 0.3 percentage points and tripled capex contribution from 2028 onwards, as hardware innovations bypass single-chip limits, though current investment is muted regarding deflation.
- 4:40 – 6:57
Policy Shifts and Profit Margin Improvement
Xing cites policy changes reducing local government subsidies and export rebates to curb overcapacity, expecting industrial firm profit margins to rise from 5% to 8% over the next 3-5 years.
- 6:57 – 9:09
Household Consumption and Structural Reforms
Xing outlines three conditions for tech upgrades to boost household consumption: investment discipline, housing market stabilization by H2 2027, and social welfare reforms to reduce precautionary savings.
- 9:09 – 12:05
Equity Market Outlook and Global Role
Wang forecasts Chinese equity earnings growth rising from 6% to >10% by 2030 and ROE reaching 12-13%. Xing argues China's shift to 'made by China' via intelligent factories will sustain global market share despite trade tensions.
Claims worth verifying
Morgan Stanley forecasts $12 trillion of incremental Industry 5.0 investment over the next 10 years.
Unidentified speaker (speaker_0) ·0:28
How to verify: Morgan Stanley Research report 'Emerging Frontiers' or 'Physical AI' published prior to 2026-11-20; check Morgan Stanley investor relations site for the specific white paper. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.
Addressable market for frontier industries to expand from $1.4 trillion to $3.2 trillion by 2030.
Unidentified speaker (speaker_0) ·0:47
How to verify: Morgan Stanley Research report 'Emerging Frontiers'; verify against the specific market sizing methodology in the published report. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.
Broader productivity payoff from physical AI to become visible from 2028.
Unidentified speaker (speaker_0) ·1:07
How to verify: Morgan Stanley macro/equity research notes on China productivity trends; verify against the timeline stated in the 'Thoughts on the Market' transcript or associated research. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.
China's AI investment in bottleneck enablers (chips, GPUs) represents low single-digit percentage of total capex.
Robin Xing ·3:00
How to verify: China National Bureau of Statistics (NBS) data on total fixed asset investment vs. semiconductor/equipment investment data; compare with Morgan Stanley's estimate. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.
Physical AI will contribute about 0.3 percentage point in terms of productivity gain from 2028 onwards.
Robin Xing ·3:40
How to verify: Morgan Stanley China macro research; verify against the specific productivity model assumptions in their published outlooks. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.
Capex contribution from physical AI will triple from today's investment in bottlenecks.
Robin Xing ·3:40
How to verify: Morgan Stanley investment cycle analysis; verify against the projected capex figures in their Industry 5.0 report. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.
More than half of the $12 trillion incremental investment will be rewiring existing factories.
Robin Xing ·5:00
How to verify: Morgan Stanley 'Emerging Frontiers' report; verify against the breakdown of capex allocation between greenfield and brownfield projects. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.
Local governments have reduced export rebates and clawed back high-tech tax privileges.
Robin Xing ·5:40
How to verify: China Ministry of Finance announcements or State Taxation Administration notices regarding local subsidy reforms and tax privilege adjustments post-late 2026. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.
Industrial firm profit margins in China could improve from 5% to 8% over the next three to five years.
Robin Xing ·6:27
How to verify: Morgan Stanley China equity research; verify against the margin forecast in their '2026 New Year Outlook' or similar industry reports. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.
Chinese equity market earnings growth can jump from 6% to above 10% by 2030, with ROE reaching 12-13%.
Unidentified speaker (speaker_0) ·9:28
How to verify: Morgan Stanley China equity strategy report; verify against the earnings growth and ROE projections in their published market outlooks. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.
Do your own research
- Verify Morgan Stanley's Industry 5.0 investment forecast against their original research report.
- Check current Chinese industrial profit margins (5%) and capex data from National Bureau of Statistics of China.
- Review local government subsidy policies and 'anti-involution' KPIs for officials in key manufacturing provinces.
- Assess the timeline for housing market stabilization mentioned for H2 2027 against real estate sales data.
- Compare current AI capex as a percentage of total capex with official Chinese investment statistics.
- Analyze historical correlation between Chinese industrial margin expansion and equity ROE performance.
- Look for counter-arguments regarding the feasibility of bypassing single-chip limits through hardware innovation.
- Check for recent trade tension impacts on Chinese exports to validate the 'made by China' global share claim.
Full transcriptexpand
0:00Unidentified speaker (speaker_0): Welcome to Thoughts on the Market. I'm Laura Wang, Chief China Equity Strategist at Morgan Stanley Research. And on this very special episode of the podcast, we're coming to you directly from Morgan Stanley's Asia Pacific Summit in Singapore to discuss the physical AI revolution, how China is bringing AI to the real world.
0:21Unidentified speaker (UNKNOWN): I'm
0:21Unidentified speaker (speaker_0): joined by Robin Xing, Chief China Economist. Let's get started.
0:27Unidentified speaker (UNKNOWN): So
0:28Unidentified speaker (speaker_0): AI's next chapter is not just about what it says to us. It's about what it can do in the physical world, from vehicles on our roads to robots on factory floors. Our emerging frontiers research points roughly at 12 trillion US dollars of incremental industry 5.0 investment over the next 10 years across the broader frontier industries. we actually forecast addressable market to expand from the current 1.4 trillion US dollars to 3.2 trillion by 2030. And China's manufacturing depth and deployment scale could give it a distinctive advantage. But we all know that more investment is not necessarily more value. We expect the broader productivity payoff to become more visible from 2028, and today we want to explore what turns smarter machines into better economics, stronger profits, and higher incomes. So how China can build physical AI and actually make it pay off.
1:26Unidentified speaker (UNKNOWN): So
1:27Unidentified speaker (speaker_0): I want to start with the big picture. Much of the AI debate focuses on who builds the best model. When AI moves into the physical world, what changes in the competitive equation, and what do you think will give China an edge?
1:41Robin Xing: Thank you, Laura. So phase one of AI revolution is probably some people will call it a chat box helping you writing emails, right? But when we call physical AI, that's the second chapter of AI revolution. It's industry-wide adoption of AI. It's AI shifting from just staying in the clouds, going into the real world, factories, ports, logistics, and that could rewire the existing industry and try to unleash productivity for all these scenarios China is good at. China is largely a manufacturing-based economy. So simply working from the cloud It's is not so helpful for their productivity. Instead, they are betting on in the next five years it would be robotics, it would be industrial factories upgrade, it would be these drones and cars helping them to unleash productivity. But on its macro impact, I see a two-stage macro equation, because at this moment China still faces of a bottleneck. That's chips, that's upstream equipment, semiconductor GPUs. So they are still investing in these bottleneck enablers, which you call it peaks and shovels. That's booming, but that's only a small portion of the Chinese
3:00economy, about low single digit of their total capex. So its contribution to China's potential growth is still limited. But from 2028 onwards, we are expecting a series of moments in China's AI computing power, in China's hardware, be it uh GPU, memory, or the superpod interconnector, pulling them together to use some engineering innovation to bypass single chip limits. So that would give China the wider opportunity of using this, using these AI brands into robotics and the industries, upgrading their factories. That's why we do see much broader implication for the economy from 2028 onwards, contributing about 0.3 percentage point in terms of productivity gain, and its capex contribution will also triple from today's investment in those bottlenecks. But in your term, it's still muted in terms of helping China to break out the deflation trap because right now it's still about investing in upstream bottlenecking neighborhoods.
4:10Unidentified speaker (speaker_0): Thanks, Robin. We understand that the investment spending doesn't necessarily take place together with the productivity gains. And as you and I just published our 2027 New Year Outlook, we both understand that the Chinese macroeconomy. will continue to be challenged by some of the macro concerns. So when do you think we would start to see some of that real impact on the macroeconomy and what would need to happen for that to take place?
4:40Robin Xing: That's a good question. In the past cycles, China was able to make a lot of technology progress, winning global market share, but that does not translate to profits for corporates or income for households. But three structural factors are helping us to break this deflation trap, making this new fiscal AI investment different from previous cycles. Number one is the nature of investment. Out of this $12 trillion dollar incremental investment in Industry 5.0 or physical AI in China, more than half of that will be rewiring existing factories about making them more intelligent, making this more passive automation into proactive, adaptive production. And that means robotic industrial software. The second factor here is policy. 80% of China's industrial policy, this subsidy for capacity, was done by local government in the past. But since late last year, there has been two initiatives. One is so-called anti-involution, changing the KPI, the promotion metrics for local authorities. So now the local mayors don't have the mandate to subsidize all these factories. Number two, they are also running out of money given the land net revenue shrink shrank a lot. So
6:00local government basically reduced the export. rebates and also clawed back some of these high-tech tax privileges for firms. In your term, that's probably negative for certain firms. But over the medium to long run, that's encouraging firms to consolidate factories to not just keep expanding, but try to keep efficiency and profit the margins. That's why we think with both policy and the different nature of industry investment, over the next three to five years, the profit margin of industrial firms in China could improve from today's 5% to 8%.
6:37Unidentified speaker (speaker_0): Robin, as you and I worked through the blue paper on emerging frontiers, I think one conclusion we have reached is that the corporate earnings actually can improve ahead of the household effect when the benefit can trickle down and spill over to household consumption. So, um, how do you think about this? When can we and how can we start to see that positive feed into the household consumption?
7:02Robin Xing: That's a very good question because um robotics they don't drink latte, they don't wear sneakers, so it's not directly creating consumption demand. And in the past, when China created so much supply, the spillover to household income was limited partially because of the excess supply. So we need to see three things happening before consumers can really benefit from the technology upgrade. One is investment discipline. The thing we mentioned, they have to spend this money on rewiring existing factories, making them more intelligent, cost reduced. reduction, more efficient, rather than doing all this greenfield capacity buildup to create more redundant factories. That's something they need to avoid. Number two is China's housing problem. Part of this weak consumption is due to the deflation triggered by housing adjustment. Recently they have further tightened the rule, which is a negative for growth in your term, but if it's also limiting new supply. And by second half next year, our property team believes housing prices in large cities were stabilized. So the well-effect channel for consumers were largely normalized. Finally, it's also domestic reform on social welfare. And the Chinese
8:22some consumers save a lot, partially because the lack of a proper social welfare system for migrant workers and farmers. If they can make a pivot towards consumers with a social safety net reform, that's going to really unlock this precautionary saving. So we are looking for these three factors. But 2027 is still tough. They are still in a stage of very slowly healing from the balance sheet, be it local government failure. fiscal consolidation, the household deleveraging, and the anti-involution in your term hurt capex. But by 2028 and onwards, we do believe they may have healed from this balance sheet repair, and that spillover to household income will be stronger onwards.
9:09Unidentified speaker (speaker_0): Thanks very much. Uh And hopefully, partly of that household income incomes spillover, be over, in fact, can come from the equity market investment as an opportunity. When we look at how the equity market could potentially benefit from this industry 5.0 and the frontier emerging frontier space development, we actually think the earnings growth for Chinese equity market can actually jump from the current six percent on an annual basis to above ten 10% percent by twenty 2030. thirty. And we actually think the ROE level for investors. who invest in Chinese equity market, the ROE can jump from the current 10% ish level back to above 12%, close to 13% level. So hopefully more investment opportunities and the wealth effect can be created also in the capital market together with the direct effect on household consumption and the wage growth. Robin, I want to bring this back to you and rapidly. wrap it up. So far we've been talking about all the great changes that are happening within China on technology, on hardware, on software, on deployment in different industries. How will this change China's role on a global basis?
10:17Robin Xing: Yes, that's also an important debate. If China is so good at the industrial revolution, it's is gaining more market share, will that trigger more geopolitical or trade barriers from other trading partners and that could derail China's rise in manufacturing? We don't think so, because number one, China is shifting from made in China to made by China. China is working on upgrading with intelligent factories, with more industrial software, the ecosystem. So with that super cycle happening globally, on rejoining reshowing, on reindustrialization, other countries are also building more and more factories. But the foreign factories are running based on Chinese components and machinery. That's actually helping China to gain further market share. But end of the day, China is facing an unbalanced growth model. Too much supply, too to few consumer demand. They have to work on domestic reform on social welfare. to rebalance that growth model. That could create more demand opportunity for global firms as well. Only by doing that, I think it could try to manage the trade tensions in a better way.
11:27Unidentified speaker (speaker_0): Thank you for a great discussion.
11:29Robin Xing: Thank you.
11:30Unidentified speaker (speaker_0): For those at home, if you enjoy thoughts on the market, feel free to share with a friend or colleague today.
11:36Robin Xing: Thank you. Thank you.
11:38Unidentified speaker (UNKNOWN): You've
11:38Unidentified speaker (speaker_2): been listening to part one of a special two-part episode from Morgan Stanley's 25th Annual Asia Pacific Summit. Join us Monday for part two of the conversation.
11:48Unidentified speaker (UNKNOWN): The
11:48Unidentified speaker (speaker_2): preceding content is informational only and based on information available when created. It is not an offer or solicitation, nor is it tax or legal advice. It does not consider your financial circumstances and objectives and may not be suitable for you.