E187 | The Tariff War Cannot Solve America’s Manufacturing Dilemma; the Old Order Is Collapsing
Summary
Above 125%, tariffs have no marginal economic meaning; the market is trading political signals and policy reversals. The U.S. raised additional tariffs on China to 125%, taking the total to 145% after adding the previous 20%; China also raised its rate to 125% during the recording and said it would ignore any further increases. 施展’s judgment was that “125% makes no more difference than 1250%” (“125%跟1250%没有区别”), suggesting the bad news may be fully priced and negotiating conditions are beginning to form, though volatility is not over.
Tariffs cannot save U.S. manufacturing because supply chains now cross borders, and political space no longer maps onto economic space. More than 70% of global trade in the 1990s consisted of finished goods; by 2018, more than 70% was intermediate goods. China’s main exports to Vietnam are the assembly “front end,” while the “middle office” of tier-two-and-below suppliers remains in China. Trade barriers will only change arbitrage and transshipment routes or pass costs on to U.S. consumers. “The shift of Chinese manufacturing to Vietnam is not a transfer; it is actually an overflow of China’s supply chain” (“不是转移，实际上是中国供应链的溢出”).
China’s manufacturing scale and network effects make simple substitution economically difficult. China accounts for roughly 31% of global manufacturing output by value and 35% by volume. Once a destination absorbs even modest additional capacity, land and labor costs rise; if China develops excess capacity, its costs fall instead. Land prices in Thailand and Mexico have already doubled, while industrial land rents around Hanoi and Ho Chi Minh City exceed those in Shanghai and Shenzhen, eroding a significant share of the tariff effect.
The core contradiction in the “Mar-a-Lago Accord” is that manufacturing reshoring, low inflation, and low Treasury yields cannot coexist. Miran’s vision is to use tariffs first, wage a currency war next, and finally deploy security leverage to restructure “America First” by adjusting the dollar, reviving manufacturing, and making allies share the cost. But reshoring low-efficiency manufacturing would inevitably raise costs and inflation, pushing up interest rates and Treasury yields. 施展 called this “fighting with both hands”: more than 20% of U.S. fiscal spending in 2024 already went to Treasury interest, and the program said roughly $6T of Treasuries would need to be rolled over around June 2025.
施展 believes the real restructuring may be America’s shift from a multilateral system to a bilateral, U.S.-centered hub-and-spoke globalization. The U.S. East and West Coasts once handled 0-to-1 innovation while the middle of the country handled 1-to-N diffusion; as production moved to China’s southeastern coast, the Rust Belt was left behind and economic fractures became social and political polarization. 施展 compares the old order to “networked Greek city-states” and the new one America seeks to a “Roman Empire,” but China—the industrial “bug” that cannot be forced into retreat—makes the system difficult to close.
In the short term, companies can only ship existing orders and wait; over time, they must add global nodes without mistaking “going abroad” for relocating the entire supply chain. Tariffs above 100% cannot be absorbed through cost cuts or automation, so companies rushed to ship orders before implementation and then largely froze, while any industrial-chain relocation takes at least 8 to 12 months. Capacity already in Southeast Asia and Mexico can still operate because of the 90-day pause and the temporary 10% increase. The next scouting wave may target Indonesia and similar countries with adequate infrastructure and smaller U.S. trade surpluses, but these are windows of opportunity, not permanent safe havens.
A key medium-term risk for China is that blocked U.S. exports flood third countries, triggering political defenses in Europe, India, and Southeast Asia. China’s goods surplus approached $1T in 2024, and “you earned all the money” is not sustainable. If receiving countries cannot rebalance through economic means, they will create friction through tariffs and regulation. 施展 sees the more viable path as companies genuinely going abroad and bringing growth and income to local economies, offsetting China’s surplus rather than merely rerouting goods.
Investors are confronting the failure of the old order, not a one-off trade dispute that can be bought on a historical template. 刘一鸣 cited Howard Marks’s “realm of the unknowable” and Berkshire Hathaway’s sharp reduction of its Apple holdings since Q1 2024 as evidence that large pools of capital must cut risk while markets are still hot. Later, exemptions for phones, computers, and some chips sent Apple shares down first and then higher. 施展 went further: if manufacturing cannot heal America’s fractures, AI, universal basic compute, and a commercial order led by multinational platforms could become the seedbed of a new system—but that remains a highly uncertain long-term possibility.
Deep dive
1. The Tariff Timeline Has Become an Hour-by-Hour Policy Risk
何必’s timeline begins in January 2025: Trump imposed 25% tariffs on Canada and Mexico, then on April 3 introduced a 10% baseline tariff on all imports and added country-specific rates for more than 60 economies. On April 9, he suspended the “reciprocal tariffs” on nearly 60 countries for 90 days, excluding China.
China first raised tariffs on U.S. imports to 84%, after which the U.S. lifted its China rate to 125%. Including the previous 20% tied to fentanyl, the effective U.S. rate reached 145%. The episode was recorded on the afternoon of April 11, during which China announced another increase to 125%.
Financial markets directly rejected the policy’s stability. In the week before April 8, the Dow fell 10.85%, the S&P 500 fell 12.14%, and the Nasdaq fell 13.26%, while Treasury yields rose. The April 9 suspension triggered a sharp rally, followed by another retreat; 何必 described the market as a “roller coaster” and “flipping pancakes.”
During post-production, smartphones, laptops, hard drives, processors, and memory chips were again excluded from the reciprocal tariffs, explaining why Apple shares first fell and then rebounded. Whether a tariff applies is now enough to rewrite a company’s earnings outlook within a single day.
2. Ultra-High Rates Show That the Real Intent Is No Longer the Tariffs
施展’s first reaction was that “the diplomacy of the 17th and 18th centuries may be back”: diplomacy is once again a contest of imagination, courage, ultimatums, and maximum pressure. The aim is no longer to redistribute a fixed pie within an existing framework, but to overturn the pie itself at any moment.
His threshold judgment was blunt: “Once you are above 50%, raising the tariff from 100% to 500% makes no difference.” Rates are still rising because the contest is no longer about import prices, but political will, bargaining chips, and who blinks first.
刘一鸣 said the defining word this time is “fast.” Trump’s first term took 14 months and four rounds to expand the scope; this time, nearly all goods shipped to the U.S. were covered within days, with the rate pushed in one shot to 145%. Companies had no time to adjust.
Even an island inhabited only by penguins was taxed after someone used its name to register companies and import luxury goods and seafood, exposing the roughness of the calculation. 刘一鸣 therefore revised his initial question: tariffs may be “both a means and an end.” 施展’s summary was that “doing it so unseriously means the intent is fundamentally not the tariff.”
3. Vietnam Is Taking China’s Supply-Chain Front End, Not the Entire Chain
施展 corrected the common description of “supply-chain relocation”: what moves to Vietnam is mainly final assembly and a small number of tier-one suppliers, while tier-two, tier-three, and tier-four suppliers remain in China. “It is not a transfer; it is actually an overflow of China’s supply chain” (“不是转移，实际上是中国供应链的溢出”).
This structure explains why Trump imposed a 46% tariff on Vietnam: Vietnam runs a large surplus with the U.S. and a large deficit with China, with Chinese intermediate goods processed locally before entering the U.S. This is not simply “washing through Yangcheng Lake”; real processing and value-added activity occur locally, allowing the overflow to keep expanding.
If the U.S. blocks both China and every receiving country, the after-tax difference between assembling in China and assembling in Vietnam disappears, leaving U.S. consumers to pay higher prices. If it blocks only China, more front-end assembly will move abroad, but the middle office will still be difficult to relocate.
4. Intermediate-Goods Trade Has Reversed the Power of Political Borders Over the Economy
施展 defines “economic space” as the physical space in which a complex product is made. In the 1990s, more than 70% of global trade was finished goods and less than 30% intermediate goods, so political and economic space roughly coincided. By 2018, the proportions had exactly reversed.
The U.S., Europe, and Japan fought multiple trade wars in the 1960s and 1970s, when a country could usually complete most of a product domestically and a tariff wall could keep out a rival. Today, most products are made across multiple countries, allowing companies to adjust production and logistics nodes; theories that worked 30 years ago no longer apply.
The result can even run against the policy’s objective. 施展 said China’s goods trade surplus may have been around $351.76B in 2018, versus roughly $299B for Germany; by 2024, China’s surplus was close to $1T. The more intense the trade war, the larger China’s export surplus became, while ASEAN replaced the U.S. as China’s largest trading partner.
5. China’s Manufacturing Scale Creates a Cost Inversion Between Sender and Receiver
China accounts for roughly 31% of global manufacturing output by value and about 35% by volume because its value added is relatively low—more than one-third of global production. Citing the view that a share above 20% can begin to affect prices, 施展 said this is no longer a situation in which other countries can change the balance by taking on a little more capacity.
If supply chains leave China on a large scale, China would see unemployment, bankruptcies, and lower land, labor, and capital costs. Receiving countries with smaller industrial bases would be “stuffed” by even modest additional capacity, sending factor prices sharply higher. The resulting fall on one side and rise on the other would create a cost inversion that offsets a significant share of the tariffs.
High tariffs also turn illegal channels into a high-return service industry: “The tariff rate is effectively the profit margin for smuggling” (“关税的税率，实际上就是等于走私的利润率”). The episode used a joke about New York’s higher egg prices to illustrate that when the profit from transporting eggs across borders approaches that of transporting drugs, barriers may stimulate smuggling rather than manufacturing.
6. Manufacturing’s Middle Office Is Harder to Replicate Across Borders Than Its Front End
施展 uses the internet as an analogy for manufacturing: the back end is infrastructure and skilled labor; the middle office is a supplier network capable of rapid iteration and efficient component production; the front end is final assembly. “As long as your middle office is strong enough, the front end can be extremely light.”
His cross-border research found that almost everything going abroad is front-end capacity; he sees no prospect of the middle office moving out as a whole for now. The three economic roles—consumer, producer, and resource supplier—are also unlikely to change fundamentally: the West, especially the U.S., consumes; East Asia, with China at its center, produces, alongside Germany and parts of Eastern Europe; resource economies supply energy and minerals.
If consumer economies rebuild manufacturing, they must pull resources out of sectors where they have comparative advantages and push them into weaker areas such as manufacturing. 施展 believes a purely market-driven process will not allocate resources this way, while the West lacks the government capacity to force the transition over the long term.
When 施展 visited an OPPO factory in Indonesia in 2023, only 8 of several hundred Chinese suppliers followed, all of them technologically simple tier-one suppliers; not a single tier-two supplier went. The local manager saw little chance of a domestic supply chain developing in the foreseeable future because the missing conditions were “not one or two.”
7. New-Energy Vehicle Competition Is First a Contest of Organization and Scale
Chinese automakers told 施展 that new-energy and combustion-engine vehicles “are two completely different species”: the former is a software-and-hardware product developed in coordination, while the latter is a mechanical product. Europe’s century-old automakers have rigid organizational structures; transformation requires changing both departmental boundaries and R&D processes, naturally reducing efficiency.
BYD originally made batteries, while Li Auto and Xiaomi came from internet or technology backgrounds; all could build organizations around new-product logic from a blank sheet. 施展 believes Chinese new-energy vehicles lead Europe in user experience, driving and riding comfort, and intelligence.
Germany and Japan once relied on “hidden champions” making valves, gears, and other components to occupy irreplaceable positions in the ecosystem. As combustion engines give way to new-energy vehicles, those suppliers are struggling. During his research in September and October 2024, 施展 heard that many companies would be “fighting for survival at the edge of life and death” over the next 24 months.
The EU is nominally one large market, but actual sales are divided by language, dealer systems, and regulation into more than a dozen small markets, making high-frequency iteration difficult. China’s large market continues to support new models and suppliers. The BBA system, which once rejected Chinese suppliers, is now “actively coming to us for quotes.” 施展 is not optimistic: if China’s EV advantage fully translates into competition in Europe, Europe’s auto industry could disappear.
8. Miran’s Roadmap Seeks to Rewrite Trade, Currency, and Security at Once
刘一鸣 called Stephen Miran’s “A User’s Guide to Reconstructing the Global Trading System” essential reading for investors and said it was once viewed as the blueprint for this trade war. Miran identifies three U.S. problems: an overvalued dollar, declining manufacturing competitiveness, and the fact that the U.S. bears global security responsibilities without receiving sufficient economic returns.
The sequence is to use tariffs first, launch a currency adjustment next, and deploy security tools last. The goals are to reduce the deficit, revive manufacturing, shift the cost of the dollar’s reserve-currency status outward, and ultimately establish an “America First” trade and financial order.
刘一鸣 retained three key variables: whether allies such as Europe, Canada, and Australia will accept “security in exchange for economic benefits”; whether China can respond effectively; and whether the U.S. itself can withstand inflation, debt, and financial-market stress. The rise in European defense stocks and retaliatory tariffs already shows that allies have not fully complied.
9. Manufacturing Reshoring and Low Rates Negate Each Other Within the Same Plan
何必 laid out the mechanism behind an overvalued dollar: the dollar is not only a trade currency but also a safe-haven asset, and it is linked to oil. Countries hold dollars beyond their goods-trade needs, pushing up the exchange rate. The U.S. also wants foreign capital to buy Treasuries and flow back into the domestic economy, so Treasuries must remain sufficiently attractive and affordable to service.
施展 said reviving manufacturing first requires low financing costs. But the fact that the U.S. became a consumer economy itself shows that its manufacturing efficiency is relatively low; forcing production back would inevitably raise costs and inflation. If inflation rises, Treasuries will not sell without higher yields, making it impossible to keep rates genuinely low.
This is the “fighting with both hands” embedded in the Mar-a-Lago vision: reshoring manufacturing requires accepting inflation, while controlling fiscal costs requires suppressing both inflation and yields. If higher tariffs also stimulate smuggling, the profits could pull people away from legitimate production.
The pressure points cited by the episode are stark: more than 20% of U.S. fiscal spending in 2024 already went to Treasury interest, exceeding military spending; around June 2025, roughly $6T of Treasuries would need to be rolled over. If other countries sell Treasuries, the entire financing chain could break.
10. U.S. Trade Policy Comes From a Geographic Break Between 0-to-1 and 1-to-N
施展 summarizes America’s old division of labor this way: East Coast capital combined with West Coast technology to complete 0-to-1 innovation, while the central regions handled 1-to-N, spreading innovation into jobs and broader wealth and containing inequality.
After China’s southeastern manufacturing network emerged, its overall efficiency remained irreplaceable even as land and labor costs rose, thanks to the network effects. The more the two U.S. coasts innovated, the more China was pulled along; 1-to-N no longer happened in the American middle, and the Rust Belt was steadily left behind.
Economic fracture then became social fracture and political polarization. As society split further, candidates who moved toward the center were rejected by both sides, ultimately producing more radical domestic and foreign policies. 刘一鸣 said: “If this problem cannot be solved, there will definitely be a second Trump after Trump.”
11. 施展 Believes Trump Seeks Hub-and-Spoke Globalization, Not Simple Deglobalization
施展 believes that when the U.S. held overwhelming advantages, multilateral globalization was not very different from a bilateral system. But as America’s relative advantage declined over time, the existing rules began to “tie its hands,” leading it, in his view, to consider overturning the multilateral system.
The alternative would be for the U.S. to negotiate separately with every country except China and, by virtue of its scale, always act as the principal. A large number of bilateral agreements would form a U.S.-centered hub-and-spoke structure. 施展’s metaphor is that the old system resembled “networked Greek city-states,” while the new vision resembles a “Roman Empire.”
刘一鸣 added that Trump’s simplified formula does not count European purchases of advertising and virtual services from Google and other U.S. internet companies; Americans could also point to European fines on U.S. technology companies. This mismatch in accounting is itself a new source of conflict in bilateral negotiations.
12. China Is the “Big Bug” America’s Reconstruction Plan Cannot Avoid
施展 believes the Mar-a-Lago objective can close only if China’s industrial position is pushed back to a weaker state—a form of cooperation China cannot provide. 何必 noted that Germany and Japan were sovereignly constrained countries during their postwar rise, making them fundamentally different from China; China’s sovereignty and industrial capacity mean old negotiation templates cannot be copied directly.
China can “stand and fight head-on,” but it will also bear unemployment and bankruptcies. The U.S., meanwhile, would face higher inflation and fiscal costs without Chinese supply. Both sides need each other, so 施展 interprets the extreme rates as a political step that had to be taken, not an economic state that can be maintained indefinitely.
His negotiating logic is that “the bad news is fully priced.” U.S. tariffs of 125% and 1250%, or China’s 84% and 840%, make no substantive difference. Once policy cannot get worse but life must continue, the value of further escalation falls and room for negotiation begins to appear.
At the end of the episode, China raised its tariff rate on the U.S. to 125% and said it would “not respond” if Washington continued raising rates. 施展 reiterated that “the atmosphere has been fully built up to this point” (“氛围已经都烘托到这儿了”), so the next step should be finding an exit into negotiations; this was a forecast, not a firm commitment on timing.
13. Apple Can Reroute Arbitrage, but It Cannot Reproduce China’s Manufacturing Capacity
Apple has simultaneously benefited from Ireland’s low income-tax rates, low manufacturing costs in China, Southeast Asia, and India, and consumer markets in the U.S. and China—a classic beneficiary of the mismatch between political and economic space. Tariffs do not eliminate arbitrage; they force multinationals to change how they arbitrage.
As the effective date approached, Apple urgently flew phones from China and India back to the U.S. on several aircraft and pledged to increase U.S. investment. But 刘一鸣 believes large companies will eventually have to address the wealth divide between the coasts and the middle of the country, or policy shocks will keep recurring.
施展 recalled that Obama once asked Jobs whether Apple could bring production back to the U.S.; the answer was that there were not enough engineers and workers. That condition has not fundamentally changed. If reshoring is achieved through mass robotics, capacity may return, but employment would “have little to do with these MAGA brothers.”
何必 noted that redistributing income through social policy does not fit the MAGA conception of the American dream; it looks more like “big government” and the “European dream.” 施展 acknowledged that the transition would be painful, but argued that when conditions change so much that the old methods can no longer escape the trap, political traditions themselves will be forced to change.
14. AI and Multinational Platforms Could Incubate a New Order Parallel to the Nation-State
施展 believes another path for healing America’s fractures is for AI to “explode into” a sufficiently large new economic form, then use incremental growth, fiscal capacity, and social policy to support people who cannot enter the new economy, rather than expecting traditional manufacturing to absorb them again.
Borrowing Harari’s framework, he speculated that AI could create a small “godlike class” and a large “useless class.” If the latter were genuinely excluded, society would see unprecedented class conflict, making UBI—or universal basic income—a potential institutional necessity.
Altman’s proposal that “everyone should have basic compute” is, in 施展’s view, a corporate version of UBI: individuals could use or trade compute to generate income. In the future, ByteDance, Facebook, Google, and other platforms could offer competing forms of UBI, compute, or tokens, competing to attract users rather than leaving all payments to the state.
施展 calls this “a return to medievalization”: states would resemble rival lords, while cross-border commercial networks operate in parallel. Large platforms could become the economic world’s “Senate,” with small and midsize supply-chain companies forming the “House,” using competition among states to seek institutional and arbitrage opportunities. This is a long-range hypothesis, not an established order.
15. Third-Country Pressure and Corporate Expansion Abroad Will Drive a Difficult Rebalancing
何必 identified the hardest spillover: the U.S. could push its tariff border into Vietnam, Europe, and Japan and demand that these countries build walls against China. Products China can no longer sell to the U.S. could then enter third countries at lower prices, forcing local governments to protect domestic industries even without wanting to cooperate with Washington.
施展 used Vietnam to illustrate the dilemma. If Hanoi builds a wall for the U.S., higher Chinese intermediate-goods costs will make Vietnamese assembly products less competitive in America. Vietnam is also an RCEP member, so blocking China alone would affect its relations with other ASEAN countries. Small countries ultimately can only “take hits from both sides.”
China’s nearly $1T surplus in 2024 is also unsustainable: “If you earn all the money, then you cannot keep earning it.” 施展 argues that Chinese companies going abroad should genuinely help receiving countries grow and create income, offsetting China’s excessive surplus. Otherwise, other countries will use political tools to “create trouble for you.”
Real-world costs are already constraining relocation. Land prices in Thailand doubled in under 2 years, Mexico has also seen a doubling, and industrial land rents around Hanoi and Ho Chi Minh City exceed those in Shanghai and Shenzhen. Vietnamese land prices have risen much faster than wages, pushing capital toward real estate. If moving only a small share of the front end has already produced this outcome, relocating the middle office as a whole is even less plausible.
16. Companies Are Shipping, Diversifying Nodes, and Cutting Risk Through the “Realm of the Unknowable”
刘一鸣 observes that a 34% tariff can still be met with cost cuts or higher automation, but rates above 100% cannot be handled by companies themselves; further cost reductions cannot preserve margins. Chinese manufacturers rushed to ship orders before the effective date and then largely froze, because relocating one part of a supply chain takes at least 8 to 12 months while policy changes daily.
Companies already operating in Southeast Asia or Mexico can continue because of the 90-day pause and the temporary 10% increase. Longer-term scouting will turn toward countries such as Indonesia, with adequate infrastructure and smaller U.S. trade surpluses, to secure a window outside the main line of fire. That is a temporary opportunity, not a permanent safe haven.
刘一鸣 cited Howard Marks’s “realm of the unknowable,” calling this potentially the largest economic upheaval of their lifetimes. Berkshire Hathaway’s sharp reduction of Apple holdings since Q1 2024 likewise shows that very large pools of capital cannot wait to sell after risk materializes; they must contract exposure while markets are still hot.
何必’s closing observation was that people became accustomed to the postwar peace, multilateralism, and free trade, making it easy to see Trump as a “mad king” off the rails and expect the world to return to normal. 施展’s conclusion was colder: “There is no way back to the old order” (“不可能回到旧秩序了”). The next step is to judge which new arrangements are structurally impossible based on production, consumption, resources, fiscal capacity, and other fundamentals.