While SoftBank funds France, China funds its chips. And the Chinese consumer waits.
Two funding models, two destinies for consumption
May 30, 2026, Versailles
At the Choose France 2026 summit, Emmanuel Macron announced €93 billion in commitments from foreign investors in France. The headline move came from Japan’s SoftBank, which committed €75 billion to developing 5 GW of AI-dedicated data center capacity in France. Phase one: €45 billion by 2031 for 3.1 GW in Hauts-de-France, at the Dunkirk, Bosquel and Bouchain sites. Leading French partners: EDF, Schneider Electric.
It is the largest AI investment ever announced in Europe by a non-European group. Foreign money flowing into French soil. Industrial jobs. Local suppliers. A direct leverage effect on the ecosystem.
Move from analysis to action
This report is your foothold. To dig deeper into these signals, audit your strategic position or secure your operations on the ground, start a conversation with our experts.
May 27, 2026, Shanghai
That same week, the Shanghai Stock Exchange approved the IPO of ChangXin Memory Technologies (CXMT, 长鑫存储), China’s leader in DRAM chips. The planned raise: 29.5 billion yuan, or about €3.7 billion, on the STAR Market. It is the largest IPO in mainland China since CNOOC in 2022, and the largest in Asia since CATL in May 2025.
CXMT’s stated target valuation is 300 billion yuan (about €37 billion). But according to Caixin, the Chinese market projects a post-listing valuation that could reach 2 to 3 trillion yuan, or €250 to €375 billion. CXMT’s revenue in the first quarter of 2026 jumped 719% year on year to 50.8 billion yuan (€6.4 billion), with net profit up 1,688%.
Eight days earlier, on May 19, its twin YMTC (Yangtze Memory Technologies, 长江存储), China’s leader in NAND chips, launched its own pre-listing process. A formal filing is expected in mid-June. Projected valuation, once again: 2 to 3 trillion yuan, or €250 to €375 billion.
To put these orders of magnitude in perspective: Tencent is worth about €500 billion today, Alibaba about €295 billion. If CXMT and YMTC reach the top of their range on the stock market, the two Chinese chipmakers will approach the combined value of Tencent and Alibaba. Unheard of for industrial companies.
But this capital will be almost exclusively Chinese.
The Party’s dilemma
Beijing is caught between two promises it made at the same time, which pull in opposite directions when it comes to funding.
Promise (a): “common prosperity” (共同富裕). Keep making the people richer, as the five-year plans have hammered home for twenty years. The 15th Plan reaffirmed it: improve social coverage, raise pensions, broaden access to public services, support employment. Invest in people, not just in machines. Funding: the state budget.
Promise (b): sovereign technological self-reliance. Target number one: semiconductors, the core of US sanctions and China’s industrial Achilles’ heel. The 15th Plan makes it the absolute priority: according to the analysis of France’s High Commission for Strategy and Planning, 90% of the plan’s implicit priorities go to productive power and economic sovereignty. Funding: state funds + the domestic financial market (institutional capital and Chinese private savings).
The two promises can coexist if, and only if, each finds its own funding channel. That is where the system seizes up.
In theory, (b) could be funded by capital from all over the world. That is what Nvidia has done for twenty years. Listed on Nasdaq and the world’s largest market capitalization, Nvidia siphons American, European and Asian capital, redistributes it across its ecosystem, and thereby funds American AI innovation at an ultra-low cost of capital. The model long worked for Alibaba, Tencent, JD, Baidu and BYD too: a listing in New York or Hong Kong, capture of foreign funds, domestic financing of their growth.
For Chinese chips, this channel is doubly blocked.
First by Beijing itself, which demands 100% control. The two chip champions are listed exclusively on Shanghai’s STAR Market. Not in Hong Kong. Still less in New York. Beijing refuses to let foreign capital take substantial stakes in sovereign assets. See my previous notes on how Chinese financial markets work: China Equity 1 and China Equity 2.
Then by Washington. CXMT and YMTC both appear on the US Department of Defense list identifying “Chinese military companies operating in the US”. This listing de facto bars American institutional investors from investing in them. European institutions, out of caution or alignment, largely self-censor.
The most brutal signal came on June 3, 2026. Tiger Brokers (UP Fintech, listed on Nasdaq) announced that from June 12, mainland investors would no longer be able to open new positions or deposit funds. The same restrictions apply to Futu Securities and Longbridge Securities. All of this is part of a coordinated crackdown by the CSRC and seven other agencies, announced on May 22, which aims to eradicate within two years the cross-border brokers operating illegally in mainland China. Scale: 570,000 to 630,000 mainland accounts affected, €25 to €27 billion in assets frozen or repatriated.
The message leaves no room for ambiguity. Chinese capital must stay in China. Not out of hostility toward foreign markets, but out of domestic necessity. The chip IPOs are going to soak up tens of billions of euros in local liquidity. Bloomberg notes, moreover, that “CXMT’s IPO revives memories of past market peaks”, a reference to the historic mega-IPOs that absorbed so much liquidity that they marked cycle tops.
The impact on the Chinese economy: K-shaped growth
Citi’s diagnosis: a K-shaped economy
In its report China Economics: 2026 Outlook, Mind the Gap (January 2026), Citi Research delivers the clearest diagnosis:
“China’s K-shaped growth model is showing signs of becoming entrenched, with the new economy and the supply side leading, and the old economy and demand lagging. It would take major catalysts to change this, and we do not expect any in 2026.”
A K-shaped economy is one in which the upper branch of the K (hard tech, AI, EVs, batteries, solar panels, chips) accelerates, captures capital, attracts talent and gains margin. And in which the lower branch (consumption, real estate, services, traditional SMEs, household wages) stagnates, struggles, or even declines.
The figures converge. Goldman Sachs expects Chinese growth of 4.8% in 2026, CPI inflation below 1%, and persistent PPI deflation until a return to positive territory only in early 2027. On the 15th Plan published in March, Goldman Sachs writes that “this plan focuses entirely on the supply side” and that “sustainable consumption growth will require policy support for job creation and incomes”. Support that is not in the plan.
Three takeaways
1. Domestic consumption will have to wait
I was reasonably optimistic about a recovery in Chinese consumption by 2027. I am revising that view.
The massive absorption of domestic liquidity by the chip IPOs, compounded by the cascade of suppliers listing at the same time (AMEC, Naura, Piotech, Nanda Optoelectronics), drains the capital that could have fed consumption, service SMEs and real estate. The cross-border lockdown confirms that the CCP has no alternative channel. PPI deflation has persisted for three years and is not expected to reverse before early 2027. The 15th Plan does not correct this trajectory; it amplifies it.
Chinese consumption is not collapsing. It remains contained, under structural pressure. And it will stay contained longer than expected: probably no real recovery before 2028, perhaps 2029.
2. Chinese exports as the last resort
The direct operational consequence: Chinese manufacturers are condemned to export to survive.
If they could live comfortably off their domestic market with acceptable margins, they would not bother conquering foreign markets where they are not welcome: 100% US tariffs on EVs, European anti-subsidy investigations into solar panels, German investment screening, British scrutiny of telecoms. No sensible manufacturer would voluntarily choose this obstacle course.
But they have no choice. As long as deflation persists at home and domestic demand remains sluggish, internationalization becomes a matter of life or death. For makers of EVs, batteries, solar panels and industrial equipment, and soon pharmaceutical and consumer goods companies, exporting is no longer a strategy, it is a condition of survival.
Europe will have to negotiate these arrivals with clear eyes, rather than contemplate them as an abstract threat. See my previous note on the reverse JV playbook.
3. For foreign brands in China: perseverance and radical plasticity
For foreign brands already present in China, the message is different, and probably the most important.
The Chinese market is not going to boom as some had hoped for 2026-2027. It will remain under pressure, segmented, demanding and deflationary. But it is not going to disappear either. Chinese consumption remains the second-largest in the world, and the rebound potential remains intact: a middle class of 400 million, massive accumulated savings (the highest of any major economy), an appetite for quality that is not waning in categories where the brand creates real differentiation; see my previous note on Gen Z consumption.
The foreign leaders holding up in 2026 (L’Oréal bouncing back, Aldi expanding, Sam’s Club accelerating, Adidas reinventing itself) share one trait: they have accepted the new reality of the Chinese market rather than waiting for a providential rebound. Perseverance and radical plasticity, not defensive retreat. See my previous notes on Aldi’s adaptation in China, and Adidas’s China for China, China for Global strategy.
The Final Word
The whole world is waiting for Chinese consumption to recover. This is not wishful thinking. It is the scenario that would rebalance global trade, ease the pressure on European and American industries, and defuse a large share of today’s geopolitical tensions.
A China that consumes more at home is a China that exports less out of necessity.
Do you have a matter that could benefit from an outside perspective?