Summary

李峰’s core judgment is that China’s shift from old growth to new growth has reached an inflection point, or may even have passed it. Markets read May’s data as a “K-shaped divergence”—strong tech growth, weak consumption—but he says that recognizing a K shape itself means recognizing that the new and old are at least roughly offsetting each other, unlike the deflationary period from 22 through early 2026, when “the US had high blood pressure and China had low blood sugar” and new growth could not pull up old decline.
Keeping the renminbi relatively strong over the medium term is an active policy tool. Emerging-market currencies broadly fell against the dollar over the past 1.5 months, while the renminbi did not depreciate and even edged higher; appreciation expectations are being used to increase import capacity and bring the trade surplus toward balance over the medium term, alongside 6 major banks opening offshore renminbi trading. The share of local-currency settlement should exceed 1/2 this year. But do not expect it to replace the dollar—the dollar took 50 years and 2 world wars to replace sterling, and “for a very long time, the renminbi will not be able to replace the dollar as No. 1.”
Chips are the physical embodiment of the K’s upper line. In 2026, China’s chip imports should reach $500B-plus at a minimum—effectively buying 1/2 of the global market—while exports exceed $300B and account for 1/3 of the global total; compared with before the trade war, imports have doubled and exports have grown 5x. Incremental growth is driven mainly by global AI capex, and more than 1/2 may be price-led rather than volume-led; if the bubble bursts, high growth will take a hit, but “everyone will care more about cost structure… over the medium term, that is not a bad thing for China.”
The RMB2T deposit migration definitely did not mainly go into equities. The RMB20-30T that rolled into 3-year time deposits at rates close to 3% in 22-23 is maturing in stages; a combined RMB40-50T will mature from this year through H1 next year. Facing the psychological gap between rates beginning with 1 and CPI also in the 1s, funds will keep reallocating over several months to 1 year among bank wealth-management products, savings-type insurance, equity products and early mortgage repayment—also explaining the paradox of first-tier home sales rising while household mortgage balances are flat to falling.
Consumption is not as weak as the data suggest. May retail sales fell 0.6%, with autos alone subtracting RMB70B (last year’s national subsidies pulled demand forward); excluding autos, growth was +1.1%. More important is the slow-moving variable: as the economy shifts toward services, wages’ share of GDP will rise by at least 10-plus percentage points—manufacturing puts roughly RMB400 of every RMB10,000 created into wages, versus about RMB3,000 in services; even if GDP grows only marginally, income per worker should rise gradually.
The shift in local-government finance from land to equity is the biggest institutional change. All provinces and cities ran fiscal deficits in Q1, and government debt crossed RMB100T for the first time, strengthening the central government’s hand. State capital that backed “politically aligned, technically sophisticated” projects during the coldest 3 years already shows 10x book returns; 1 fund of more than RMB3B now shows more than RMB30B on paper. Document No. 54, issued 2 weeks ago, hit the brakes by barring counties and districts from setting up new funds; the Social Security Fund is entering as an LP, while the consumption tax will arrive “sooner or later, perhaps within 1 year.”

Deep dive

1. The opening paradox: everyone calls it K-shaped divergence, but old and new growth have reached an inflection point

It began with a viral post on X by an American economist about the May data: China manufactures goods for the entire world, yet its own people have little desire to consume—trade data are very strong while retail sales are weak. 李峰 takes the popular “K-shaped divergence/recovery” framing 1 step further: the upper arm represents rapid growth in investment and import/export activity in technology and high-value-added industries; the lower arm is household consumption, including real estate.
His key judgment is that the deflationary period from 22 through early 2026 (“the US had high blood pressure and China had low blood sugar”（美国是高血压，中国是低血糖）) was one in which new growth could not pull up old decline. When people talk about a K-shaped economy today, “the implication is that, at least from what we can see today, the new and old are at least roughly able to offset each other”—the replacement of old growth with new has nearly reached an inflection point, or may already have passed it.

2. The trade-surplus signal inside double-digit external-trade growth: imports and exports are heading toward balance

In recent months, dollar-denominated imports and exports have continued to grow in the teens, with imports in the 20%-plus range (about 6 percentage points lower in renminbi terms because of the exchange rate). Machinery and electronics account for more than 2/3 of China’s exports, up several percentage points from before the trade war; the main drivers are data-processing equipment (computers), communications equipment (phones) and integrated circuits.
Imports are growing nearly 10 percentage points faster than exports, even as May’s trade surplus was still up 2%-plus year on year. Extend the trend and imports stay strong while exports grow at a reasonable pace: the surplus first stops growing, then may decline, and eventually moves close to balanced trade—an outcome for the medium term.

3. Europe’s 3 imperatives—and China’s new posture of immediate countermeasures

The EU’s brewing tariffs on China, including autos and other goods, reflect China’s cumulative surplus with Europe rising month by month. Europe is simultaneously doing 3 things: rebuilding its energy system—China has an edge in new-energy-related areas, driving some import growth; pursuing military autonomy—after the Middle East war, “every country in the world is quietly thinking about developing its own nuclear weapons,” while also guarding against a Hormuz-style energy choke point; and figuring out how to pay for massive military spending by fixing its own industry and growth. That is the backdrop to the tariff barriers.
China’s posture was different this time: it temporarily canceled 2 high-level meetings with Europe, and immediately signaled countermeasures when the EU raised tariffs. At the subsequent G7 and European meetings, Europe as a whole did not again issue particularly hard-line statements targeting China, nor did it deliberately avoid statements and positions aimed primarily at China.

4. The renminbi’s non-depreciation is deliberate: using strong-currency expectations to rebalance trade

Over the past 1.5 months, global capital flowed back into dollar assets: the Indonesian rupiah, Malaysian ringgit, Indian rupee and Japanese yen depreciated materially against the dollar, while the renminbi barely weakened and even appreciated modestly in certain windows. The EU summit even described the renminbi as 30% undervalued, rather than 16%.
李峰’s read is that maintaining an appreciation expectation “substantially raises the potential for imports” and should play an important role in adjusting China’s medium- and long-term trade structure and balance. Bottom line: even with a strong dollar—a short-term phenomenon, since the US itself faces an AI challenge—the renminbi is expected to remain relatively strong over the medium term.

5. Offshore renminbi opens up: from “something you can only hold” to usable tools

After the Lujiazui Forum, 6 banks—中行、农行、工行、建行、交行 and 中信—opened offshore renminbi trading for overseas institutions, countries and central banks holding renminbi. The renminbi is no longer only a transaction currency; it can also be used as a reserve asset, an exchange-rate hedge and for diversified financial purposes. China accounts for roughly 1/6 of global trade; local-currency settlement represented more than 30% last year and should exceed 1/2 this year.
A second objective is to address the large China-US official interest-rate differential: the central bank previously had to provide “points” in offshore forwards to bridge the gap. As the amount of offshore renminbi held and the currency’s toolkit expand, that subsidy should shrink—or, as 李峰 puts it, “we have much more control.” This also highlights Shanghai’s role as a financial-market hub in renminbi internationalization.

6. A decade-long chip transformation: buying half the world, exports up 5x

Before the trade war, the global chip market was about $500B; China imported more than $200B a year and exported more than $60B—“I assume those exports were mainly from foreign companies setting up plants in China.” In 2026, the global market is just over $1T; China’s imports should reach $500B-plus at a minimum—“we bought 1/2 the world”（我们买了全球一半）—while exports exceed $300B, or 1/3 of the global total. China ranks No. 1 worldwide by both import and export share.
This is the source of the K’s upper sloping line: 8-9 years of sustained catch-up in technology since 2017 have produced a qualitative shift. As the global market doubled, China’s imports doubled and its exports grew 5x. Chip exports could account for 10-plus percentage points of the global total this year; combined with data-processing and communications equipment, the share is about 30%.

7. AI capex is the main driver; a bubble burst would not necessarily be bad for China

Most incremental external trade is coming from this global wave of AI-driven capex, and more than 1/2 may reflect price rather than volume growth. Memory has seen the sharpest price gains, with foreign-owned plants also contributing. China’s supply chain is capturing the price upside because it can scale production quickly; “when everything is expensive, we may have a relative cost advantage.”
If the bubble bursts, this high growth will certainly take a hit. But “people will care more about cost structure. If they care more about cost structure, over the medium term that is not a bad thing for China.” The other side is supply-chain diversification: apparel, footwear and hat exports have indeed declined year on year, while key materials and design remain in China, are shipped to ASEAN for processing and then re-exported to developed markets—“the clothes you buy in China… may not be made in China.”
李翔 added an observation from 单伟健’s podcast to explain the gap between the data and lived experience: much of GDP growth is driven by imports and exports, especially chip-related trade, which “does not have much to do with ordinary people.” Ordinary people can feel only consumption—the data look good, but the lived experience does not.

8. The renminbi replacing the dollar? First consider that the dollar took 50 years to replace sterling

Renminbi internationalization will ultimately proceed step by step and phase by phase toward opening the renminbi capital account, but 李峰 poured cold water on the idea: the US overtook Britain in 1900, yet the dollar had to wait for Bretton Woods to be established—and for 2 world wars to weaken Europe—before replacing sterling as the main trade and reserve currency 50 years later. “And it was still on the gold standard then.”
So even if internationalization and capital-account opening are achieved, China would probably not replace the dollar as the largest reserve currency; it might only secure the share appropriate to the renminbi. The renminbi’s share of global financial reserves is currently below its share of GDP, just as the US was from 1900 to 1940. “China becoming No. 1 is still a long way off.”

9. Financial structure is changing: slower loan growth is itself evidence of transformation

At the Lujiazui Forum, policy officials put it bluntly: since 2000, China’s growth has stood on 3 legs—infrastructure, real estate and manufacturing-capacity expansion—and the underlying financial demand for all 3 was loans. As the economy shifts toward consumption and high-value-added industries, companies no longer need ever-growing loans to raise money. The investment-project-led model is changing; loans now account for around 1/2 of total social financing.
In the podcast’s shorthand, financing is moving from indirect to direct finance, or from loans to “higher-quality loans”: ticket sizes that once ran from several billion to more than RMB10B are now much smaller for each company; investors must select the competitive ones and use different risk-assessment methods. The fact that loans in total social financing cannot grow rapidly is precisely what “shows the economic structure has indeed changed.”

10. RMB175T ($25T) in household deposits: money no economy has ever seen

Household deposits fell by just over RMB2T in April-May, prompting everyone to ask whether the money went into equities. 李峰 said, “I should say responsibly that it definitely did not mainly go into the stock market.” Start with the scale: China’s RMB175T ($25T) in household deposits “is probably something no country or economy in world history has ever seen”; the previous high was the $16-17T accumulated in the US after stimulus payments in 22-23.
Once this money starts moving, “it is an enormously consequential event”: without sensible, effective guidance, a mass flow into any single destination would create an asset-price bubble. For comparison, when China entered its biggest real-estate cycle in 2009, household deposits were only a little over RMB20T; over the following 10 years, they still helped drive roughly RMB400T in total real-estate volume.

11. The psychological gap from 3% to 1%: where RMB40-50T in time deposits will go

The backstory: 2 rounds of wealth-management products breaking below net asset value, in H1 and H2 of 22, triggered roughly RMB2T in redemptions; in 22-23, 3-year time-deposit rates were close to 3%, and many banks even offered rates above their 5-year rates; combined with the pessimism of the time, RMB20-30T rolled into 3-year time deposits. Smaller banks have now scrapped large-denomination certificates of deposit and 3- and 5-year time-deposit products, while the big banks’ 3-year rates are below 2%; rollovers now all begin with a 1-handle, and CPI has also been in the 1.1-1.3 range since March. During the earlier deflation, putting money at 3% was “earning on both sides”; now it barely offsets inflation and leaves a psychological gap.
As a result, roughly RMB40-50T in deposits maturing from this year through H1 next year face 4 destinations: bank wealth-management products—the underlying assets are bonds, and savers are willing to accept them if expected returns are only a little above a 1-year time deposit; savings-type insurance (“for upgrading my consumption in old age”); higher-risk equity products such as stocks and ETFs; and early loan repayment. Funds will move back and forth among the 4 according to individual risk appetite and the prevailing social mood—“perhaps 1 quarter, perhaps 6 months, perhaps even 1 year.”

12. The answer to the mortgage puzzle—and banks’ fee-income business

The most puzzling April-May data were that first-tier city home sales rose both month on month and year on year, while household medium- and long-term loans showed no corresponding increase and even declined; short-term loans also fell. The answer remains deposit migration: people with loans priced above 4% and no safe, attractive place to earn a return “take maturing time deposits and use them to repay short- or long-term loans.” “You cannot calculate it that way economically, but you can psychologically.” Short-term loans first, long-term loans second.
The related subtext: bank net interest margins are at a historic low, so banks are turning to fee income from distributing wealth-management products. The new red line capping private-lending rates at 4x LPR has put the assisted-lending industry under tight regulation since the Spring Festival. 李峰 was careful to clarify: “I did not say ordinary people’s deposits were being eyed; I am only saying that, as a purely economic phenomenon,” the time-deposit behavior of that period now shows up as deposits maturing without being rolled over.

13. May retail sales: remove autos and it’s not so bad

Retail sales fell 0.6% year on year, with the RMB4T-plus total down by more than RMB20B; autos alone accounted for a RMB70B drop. Excluding autos, growth was +1.1%—“certainly still not good,” but not as bleak as the headline suggests. The negative contributors were all major categories that benefited from last year’s national subsidies—phones, communications equipment, appliances and home furnishings—so demand was pulled forward; people even joked that “every marriage-related spending category collapsed in May this year.” Carmakers face “very, very” severe pressure this year, even though auto exports are growing rapidly.
Consumption also has a K-shape internally: 李峰 suspects above-designated-size firms are squeezing out small and micro businesses. Firms with brand strength and word of mouth—“whether you are expensive or cheap, I still want to go to you”—continue to draw spending; small and micro firms face a bigger challenge, and they likely account for most employment, which is why social-media pressure is greater. Services consumption grew about 5% from January through May, showing that the shift from goods to services is already visible even in the downbeat data.

14. The hidden dividend of the shift to services: wages’ share of GDP could rise by 10-plus points

China’s services economy is being driven by 2 engines at once: life services—food, drink, leisure, travel, shopping, entertainment and transport—and what 黄奇帆 calls “producer services.” Among new listings in Hong Kong and A shares, including STAR Market companies, only a few such as 长新 have large-scale production capacity; whether the business is a chip or a new drug, it is essentially a service business. Of the technology sectors being developed at full speed, “80%-90% is service-related.”
The math is straightforward: labor costs in appliance manufacturing are a low-single-digit share of sales—only about RMB400 of every RMB10,000 created goes to wages. In services, the 2 categories average roughly 1/3, or RMB3,000 flowing to wages. Once the restructuring takes hold, even if aggregate GDP grows only marginally, wage income’s share of GDP should rise by at least 10-plus percentage points; without much growth in employment, income per worker will still increase gradually. This is the most certain medium- to long-term slow-moving variable in his view, with a footnote to Phase 1 of the 鲍莫尔 effect.

15. 300M flexible workers: where they live, where they pay social security, and 10 years of urban divergence

Flexible employment exceeds 300M people, accounting for roughly 40% of the workforce (no official statistic has been published). The State Council has required cities to provide public pension, medical and education services to their resident populations, raising several hard questions. First, urban-village and shantytown redevelopment is demolishing homes used by food-delivery and courier workers: “once you tear them down, they have nowhere to live.” Like Singapore, China may need to build public-rental and low-rent housing. Second, where social-security and medical-insurance contributions are paid—and whether they are paid at all: workers cannot pay into their registered hometown and still expect their city of residence to provide the services. Since April, policies on flexible-worker contributions have been issued in quick succession, covering contribution tiers, back payments and pilots for sharing costs between platforms and individuals.
Third is the 10-year question: the registered-residence urbanization rate is 47% versus 67% by resident population, leaving 300M people living away from their registered homes. Which cities can retain both workers in life services and workers in high-value-added services? “You cannot have only high-tech value added and nobody to provide services.” That will determine urban divergence over the next 10 years; “basically, the medium- and long-term real-estate problem is solved as well.” The short-term variable is the wealth effect from tech listings: 智谱 has just broken a RMB1T market cap, and nearly half of its few-hundred-person team now has paper wealth above RMB100M (李翔 joked, “Can we eat something nice tonight?” 李峰 replied, “They will buy houses”); the person behind the 长新 listing distributed nearly RMB1B in equity among about 1,000 people.

16. Everyone in deficit and government debt above RMB100T: the third turning point in central-local relations

In Q1, every province and city spent more than it collected—even Shanghai, historically a surplus city, ran a deficit. Outstanding government debt crossed RMB100T for the first time, equal to roughly 70% of RMB140T GDP. 李峰’s read is that the central government’s effort to help local governments defuse hidden debt has “more or less reached the end”; in financial shorthand, it has finally brought off-balance-sheet debt onto the balance sheet. The ratio is not high by international standards: many G20/G7 economies are above 100%, Japan is above 2x, and the average is 1.1-1.2x GDP. The central government still has room to borrow.
The history of central-local relations: before the 94 tax-sharing reform, “the center was poor and local governments were rich”; Beijing had to borrow from Guangdong to fund its operations, and the premier toured provinces seeking money. After the reform, central fiscal power strengthened, while 100% of land-related taxes stayed local, creating the land-finance model from 95 through 23. Local governments now need the center to coordinate borrowing quotas: “you all have to come to me for money, so you have to listen to what I say.” At a turning point, greater central control is useful; the cost is that local governments have little cash and may lose initiative.

17. A 10x state-capital return in the coldest 3 years: from land finance to equity finance

23 through early 25 was the worst 3 years in 李峰’s 20-year investment career—“I have never seen it this cold.” Dollar capital was unavailable because of LP and US-policy constraints, renminbi PE funds could not raise money, and the only capital left was state-owned money, which backed projects with a “politically aligned, technically sophisticated” profile that nobody would fund at the time: large models, major chips and AI infrastructure. 智谱’s last several rounds were almost entirely state capital; apparently the same was true of 壁仞, 摩尔线程, 沐曦 and 燧原. By 25, these had become the capital market’s most sought-after companies, and the state-capital exemplars had enormous book gains—智谱’s last few rounds moved from a RMB3-5B valuation to more than RMB20B; “even if you value it at RMB500B… that’s still 20x.”
The top performers singled out at Lujiazui were Suzhou, Beijing’s Haidian district, Shenzhen and Hangzhou (perhaps Shanghai as well). One fund established at the end of 22, fully invested by December 25, with more than RMB3B already showed more than RMB30B on paper—about 10x—with more than RMB10B of gains coming from 1 major chip company and 1 large-model company. 李翔 asked when it could exit—“good question”; but fiscal money has no pressure to return capital to LPs and can theoretically be held long term.

18. Document No. 54 hits the brakes, the Social Security Fund becomes an LP, and the consumption tax could arrive within 1 year

The example effect sharply increased the enthusiasm of local governments nationwide, from provinces and cities down to counties, so Document No. 54, issued 2 weeks ago, hit the brakes: “counties and districts may not set up new funds.” New funds are to be concentrated on sub-provincial, provincial and provincial-capital platforms, echoing the lesson of the failed grassroots expansion of village and township banks in the banking reform 20 years ago, when they were ultimately consolidated into city commercial banks. The aim is to prevent “excessive, rushed expansion and waste.” Over 3 years, local governments have shifted at least 1/2—or 1/3—of their land-finance attention to equity, because equity investment creates a chance to attract new industries.
The central government is opening its deepest pocket: the Social Security Fund—“the most important lifeline money of ordinary people.” Its 3 large and 3 small funds are entering government fund-of-funds as LPs in major regions, alongside 3 RMB50B equity funds from the NDRC. The implication is that China’s capital-market scale, asset mix and investee base will undergo a qualitative shift (“of course, today’s market is too one-sided, with everything crowded into one place”). 李峰 guesses the ultimate answer for local fiscal autonomy is the consumption tax: collection at the point of sale would make consumption a local tax base and an incentive to develop the economy; property tax can be left aside for now—“sooner or later, perhaps within 1 year.” Especially strict individual-income-tax audits follow the same logic: once land no longer supplies tax revenue, local governments have a greater incentive to collect taxes.

19. The lesson of 建行’s new high: “touching interests is more effective than touching souls”

When bank stocks were a mess, only 建行 hit a new high; analysts almost universally pointed to 建银投资’s direct investments in giant unicorns. The irony is that 建行 had the largest mortgage book in the last cycle and should theoretically have been most exposed to property. Bank transformation had been pushed for 10 years without moving—the 14 financial reform, 18 asset-management rules and 24 AIC, with 6 intervening years lost to the trade war and pandemic. This time the pivot came because AIC provided matching capital for government funds to invest in unicorns that later listed: “once you have eaten the dividend and made money, you have confidence, capability and motivation.” “Apparently, touching interests is more effective than touching souls”（触动利益要比触动灵魂更有效）.
李峰 called his own retrospective Monday-morning quarterbacking: had he understood the macro picture earlier, he should have invested in late-stage unicorns in 24-25—“put down RMB50M almost anywhere then, and today it would probably be worth RMB300-500M.” Outside state capital, the only investors willing to act were 蚂蚁 and 阿里, which happened to have begun investing aggressively. The entire structural adjustment is “half pull and half push”: the pull puts people into a positive feedback loop and gives them motivation; the push constrains them until they can only move in that direction. His closing thought was that, from government and banks down to ordinary people shifting their deposits, China’s ability to drive the structural shift “in hindsight, is really quite impressive.”

(Preview) A Summer Break Mailbag: Memory Mania, Vibe Coding, Mafia PR, Caffeine Intake, Garages, and How to Fix Soccer E241 | The Evolution of Running-Shoe Technology: Breaking 2 in the Marathon—Human or Shoe?