China's EV industry wasn't planned into existence — local credit bought the right of way
This episode takes half the credit for the rise of China's carmakers away from central industrial policy: the real engine was local governments trading loans and equity for factories, and entrepreneurs using relationships to route around SOE licensing — at the cost of an overcapacity nobody can now unwind.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
One number — 150,000 units — decided which carmakers were allowed
In the 1990s China sold only about 500,000 passenger cars a year, a market the size of Poland then or Vietnam today, yet every province except Tibet had its own car plant; a count in the mid-1990s found roughly 130 carmakers. The reason was not absurd: under a system of planned allocation, if you wanted cars you had to build them yourself. Volkswagen and other foreign partners told Chinese regulators that a car plant needs at least 150,000 units a year to be efficient, and so ‘150,000’ became the magic number in policy circles — with a market under 1 million units, it could hold only four to six companies. That simple piece of arithmetic produced the ‘three bigs and three smalls’ list of designated licensees directly.
— Fan MingluChery was never a private firm; it was one prefecture city's SOE
Chery was not a state firm belonging to a big province, a big city, or the central government — it belonged to Wuhu, a city in Anhui. In the early 1990s Wuhu could not find a pillar industry, then discovered a township and village enterprise on the city's outskirts turning out three to five thousand cars a year using engineers from a Hunan arms plant and Korean KD kits, priced at roughly $10,000 (70,000-80,000 yuan), far below the joint-venture cars of the three bigs and three smalls; it became the first company in Wuhu to pass 100 million yuan in revenue. The city concluded from this that ‘carmaking can work’, first tried to partner with FAW and failed, then took the 200 million yuan from selling its Saiwen plant (later Conch) plus bank loans arranged by the province and bought an engine production line in Wales, in Britain — because the lesson FAW had taught them was that without your own engine you cannot build your own car.
— Fan MingluThe barrier then was not technology or capital but the license
For its first few years Chery could not obtain a central production license, so it could sell only inside Anhui province, and one province's market cannot support a carmaker. Their solution was to sell cars in Chengdu while having buyers accept Anhui plates. The first party that could not tolerate this was not consumers but the Chengdu public security bureau, which notified buyers directly that they had to register locally. The detail shows where the barrier sat in those years: not in technology and not in capital, but in administrative permission itself — the cost of routing around a license was high enough to make people invent a transaction structure this ridiculous.
— Fan MingluGeely did not pay for Volvo; local banks did
Ford's condition was ‘if you can raise the money, we'll sell’, so Geely's core work through all of 2009 was assembling about $2 billion. An investment banker involved in the deal says Geely talked to as many as 60 cities — out of roughly 300 prefecture-level cities nationwide, one in five. The pitch was to show local officials Volvo's next-generation product line (the cars that later became the XC60 and XC90): I will let you build some of these models, and in exchange you put up ‘five fingers’ — 500 million yuan, then about $70-80 million, as the ticket to ride, in the form of credit from local banks or equity injected into an SPV. Line up four or five cities and the system turns. The biggest early backer was in fact Beijing, and there was even a letter from a Beijing municipal leader to Ford's CEO serving as an endorsement, but in late 2009 BAIC came out against the city supporting a competitor at home, and Beijing withdrew.
— Fan MingluThe NDRC handed out permission, not funding
Geely did not force its way through: it waited until it had the NDRC's unwritten approval before it moved. In 2007-08 the earliest candidates to buy Volvo were state firms like Changan, and all of them ultimately said they did not have the money or that it was not a priority; only after the NDRC had confirmed that no SOE wanted to buy did it let Geely proceed. The contrasting case is another private Sichuan company that wanted to buy Hummer, announced the deal itself without approval, and left regulators in Beijing to learn about it from newspapers and television — the deal was strangled outright. After approval the center's attitude was: financially, you sort it out yourself. Policy banks including the Export-Import Bank refused to lend at first, and only once Geely had raised the entire amount was CDB willing to lend for the plants that came next.
— Fan MingluBYD was selling local governments a megaproject, not cars
The ‘10 city thousand vehicle’ program handed the targets to local governments, and BYD found the arbitrage inside it: localities had to deliver a record of achievement, BYD needed volume. In partner cities like Shenzhen and Changsha it converted entire taxi fleets to its own electric cars (in 2008-09 the Changsha government sold a bus plant to BYD outright), and it built its own finance company to help taxi drivers and bus operators buy the vehicles; Shenzhen had no Olympics and no World Expo, so it used the Universiade to run its own demonstration project, and BYD accounted for a quarter of the 2,011 vehicles in service. The more important chip was a dedicated rail transit division it set up — it could tell a local government: if you want more, I can build you a light rail line, so you can hand your superiors a ‘big project’. Through the 2010s it sold only about 30,000 cars a month, which earned it the nickname ‘30,000 units’; this state-business exchange is what kept it alive.
— Fan MingluHefei made three to four times its money; the imitators got only capacity
NIO was at first extremely unwilling to take local government money, because it came attached to a mass of conditions (how much you must invest locally, and so on). Once its cash ran out and its US-listed shares collapsed at the end of 2019, it had no other option. Beijing again withdrew at the last moment because BAIC objected; Hefei took over, injecting about $1 billion (7 billion yuan) of equity, and exited completely on the stock market roughly a year later at three to four times its money. The problem came after that: the Hefei model was treated nationwide as a template, at exactly the moment the NDRC had already stopped approving new vehicle capacity in 2020-21. Guangdong province and Guangzhou injected about 2 billion yuan into XPeng, and Ningbo about 10 billion yuan into Geely's Zeekr, all of them counting on reproducing that ‘three-to-four-times exit’ — and after 2021, with Chinese listings abroad and EV valuations crushed together, the exit route was gone while the money had already been turned into capacity.
— Fan MingluThe Changan-Dongfeng merger has stalled a year on two host cities
The other way to digest the surplus is SOE consolidation, but that road is harder than it looks. Since last year the center has gone beyond rumor to something more concrete, requiring the relatively successful Changan and Dongfeng to merge; nearly 12 months of pushing has produced no result, because Chongqing and Wuhan, the two home cities, will neither of them give ground, and it is now basically shelved. This stands in contrast to more than a decade of centrally driven SOE mergers that succeeded in other industries, and it explains why autos has never been an industry the center administers directly: in the 1980s China tried establishing a national company to run all of the country's car plants, and it survived only a few years before dissolving, because localities were unwilling to give up their own plants.
— Fan MingluExports are near twice Japan's peak, and that curve cannot hold
Extrapolating from this year's pace, China's vehicle exports for the full year could reach about 12 million units. The reference point is Japan's export peak (around 1986) of roughly 6 to 7 million units, which set off the US-Japan and Europe-Japan trade wars — China is already close to twice that. So the guest argues the curve is unsustainable: you cannot imagine exports of 15 million next year. The EU is plugging the gaps, and a policy proposal extending high tariffs from pure EVs to plug-in hybrids has already appeared. The real way out is local production, and the opportunity sits precisely in European carmakers' own idle capacity: Europe's automakers have been hit both by Chinese cars (not only electric ones, combustion cars too) and by over-expanding after overestimating demand during the pandemic, and with young people buying fewer cars and even getting driver's licenses less often, what the European market faces may not be stagnation but decline.
— Fan MingluIn their own words · checked verbatim
Jennifer's pitch was that well although it was a mouse trying to swallow an elephant but it was backed by a gigantic dragon.
Fan Minglu33:34
I'll show you the next generation lineup from Volvo cars… And well I can manufacture I can make you manufacture some models of those but in exchange well five fingers. Five fingers meaning well 500 million Chinese yen
Fan Minglu35:34
But the thing is that well we need to we are not providing you any well financial support you need to sort out financial support on your own.
Fan Minglu43:47
in the US well unless well you reach uh your capital uh market cap reaches somewhere like what 10 billion or maybe 100 billion you need to you need to learn how to deal with the government. But well in China well from the beginning… you need to know how to deal with the government first
Fan Minglu1:23:25
it will take them one year to repay their part suppliers so basically they are borrowing at the expense of their part suppliers well It's definitely not a very sustainable way of making the whole industry work.
Fan Minglu1:43:50
Figures
| Number of Chinese carmakers in the mid-1990s | about 130 | 5:12 |
| China's auto tariffs before WTO accession | cut from 150-200% to about 80-220% | 9:17 |
| Brilliance's 1992 NYSE IPO proceeds | about $70 million (five years before TSMC) | 27:32 |
| Cities Geely approached to fund the Volvo purchase | about 60 (out of roughly 300 prefecture-level cities nationwide) | 35:34 |
| Number of Chinese EV startups after 2015 | 60-70 (five or six in the US over the same period) | 1:00:06 |
| Hefei's equity investment in NIO and its return | about $1 billion / 7 billion yuan, exited a year later at three to four times | 1:05:07 |
Glossary
- three bigs and three smalls
- The list of designated passenger-car producers fixed in the late 1980s; anyone not on it had no production license.
- township and village enterprise (TVE)
- Collectively owned firms that proliferated in the 1980s; Chery's predecessor began as a TVE building cars.
- 10 city thousand vehicle project
- A demonstration program from about 2011 requiring pilot cities to put a thousand electric buses or taxis into service within two or three years.
- Hefei model
- A local government rescuing an EV startup with an equity injection, then exiting on the stock market at a profit.
- special purpose vehicle (SPV)
- A shell company created for a single acquisition; Geely used one to receive the equity money injected by each city.
How to listen
Analysts who study industrial policy and local government behavior, founders who need a local government partner in China, and investors sizing up Chinese carmakers' overseas expansion and European capacity.
The chat about China's podcast ecosystem and ad monetization after 1:44:51.