This year's "Golden September, Silver October", the past situation of heavily hyped hot sales is gone, this year's Golden September has become the quietest September for the auto market in recent years.
CPCA data shows, September national passenger car retail 1.618 million units, down 28% YoY, creating the largest single-month YoY drop this year. Jan to Sept cumulative retail 13.334 million units, down 21.7% YoY, at this critical juncture, JPMorgan released another industry report spanning 52 pages, report name is called "2027 Policy Headwinds, Profit Risks, and Exit Strategies".
Just from the name, it can be seen that this report is not optimistic about the expectation for China's auto market.

Its core judgment is 2026 domestic passenger car retail might settle at 18.96 million units, this data is 21% less than 2025, and 2027 drops another 5% to 18.01 million units. JPMorgan compares 2026 to 2018 — that year saw the first negative demand growth in China's auto history, followed by three consecutive years of decline.
But this time is more tricky than 2018, because from Q4 2025, low mood continues directly, plus previous years tax and fee benefits overdrawn purchasing power, domestic demand "directly became an empty well"; meanwhile, cost side Lithium Carbonate price rose 130% compared to early 2025, memory chips soared 180% in 3 months, adding thousands of RMB per vehicle cost.
JPMorgan downgraded profit forecasts for multiple automakers, but what's more unsettling might be that the industry logic operating for decades behind these numbers is collapsing.
#Auto Market Plummeted 21%, Just the Starting Line of Hard Times#
2026 auto market data decline doesn't need forecasting, it's a fact happening now.
Besides the coldest September report card in the "Golden September, Silver October" traditional peak season, more worth watching is structural change, new energy retail penetration rate hit 67.1%, but new energy retail sales actually dropped 12% YoY, past few years "New Energy Up, Fuel Cars Down" logic now turned into "New Energy Stagnating, Fuel Cars Still Falling", where comes incremental volume? This question has no answer in 2026 yet.
JPMorgan's judgment for second half is direct: Expect mainland auto market demand in second half won't see meaningful recovery, main reason consumer confidence insufficient, "If further policy stimulus measures are launched, effects might still be limited"; Morgan Stanley also predicts 2026 China passenger car sales down 5% YoY, domestic sales drop might reach 7%.
Why is this a foreseeable forecast? Because the forces causing decline are not temporary.
Past few years China auto market operation relied on a "Hidden Logic", Central government trade-in subsidies provide demand ammo, local matching subsidies amplify leverage, automakers use price cuts to承接 conversion, dealers use rebate mechanisms to maintain operation. This contract kept industry surface prosperity from 2023 to 2025, but in 2026, every link in this logic is breaking.

Subsidy side retreating, trade-in subsidies changed from "fixed amount" to "based on car price proportion", per vehicle subsidy strength substantively decreased; Price side out of control, past two years price wars broke all bottom lines, price cut can't sell, price cut more can't sell, consumers formed expectation "the more it drops the more they wait", price means failed.
Subsidy overdraws not only sales but confidence, taking a 200,000 RMB new energy car example, 2025 various subsidies total approx 41,700 RMB, subsidy amplitude accounts for about 21% of car price, in consumers accustomed to 21% subsidy, treat subsidy as taken-for-granted pricing benchmark, once subsidy retreats, bare car price actual rise is over 20%, this in consumers' eyes not recovering original price but price increase.
Deeper problem is, this logic lasted for years itself unsustainable, in 2025 to 2026 local areas continuously land auto consumption local subsidy, stacked with automaker whole series replacement subsidy dual policy window period, subsidy completely departed from temporary dividend attribute, became automaker pricing model fixed module, guiding price artificially high then fill flat with subsidies become industry tacit rule, at this time, subsidy no longer is demand stimulation lever, but poison destroying healthy price system.
#Channels, Profit, Finance, All Being "Pulled Bones and Sucked Marrow"?#
JPMorgan report lists three risk dimensions, but thinking carefully, current China auto market has many problems potential below water.
First, channels rapidly collapsing.
2025, auto dealers profit ratio narrowed from 39.3% to 23.5%, loss ratio rose to 55.7%. Entering 2026, situation further worsened. August, 78.1% dealers price inversion, new car average gross margin -21.4%, per car sold universally loss 20,000 to 30,000 RMB, same period auto dealer inventory coefficient reached 1.58, YoY up 20.6%, far exceeded warning line.
Whole industry new car sales gross margin fell to -25.5%, 81.9% dealers face new car price inversion, over half inversion amplitude over 15%. National nearly 15,000 4S stores closed network, past 12 months over 5,000 withdrew network, average close 13 per day. Top 100 dealer groups 2026 total revenue YoY down 8.82%, 5 years employees shrunk from 424,000 to 317,000, 110,000 left this industry.
Nanchang a 4S store head frankly admits, store cumulative loss already over 8 million RMB, but if formally stop business, will immediately trigger debt liquidation, tax make-up, contract default compensation and series rigid expenditure, actual loss might far exceed current book loss, so can only in continuous pressure "hard support". This "hard support" won't last forever.

Dealer system large area collapse, automaker channel reach ability will appear systematic gap, and this gap, not rely on direct sale model short term can fill, first tier cities core business district new energy direct sale showroom year operating cost generally over 4 million RMB, as customer flow drop and conversion rate low, some brands started close inefficient direct sale stores, turn to agent, franchise mode, channel transformation cost and pain just started.
Second, profit intercepted by upstream industry, automakers go from "Revenue increase not profit increase" to "Production increase then loss".
2026 Jan to Aug, auto industry revenue 7.0062 trillion RMB, YoY up 2.9%; Profit 253.4 billion RMB, YoY down 16%; Sales profit margin only 3.6%. Jan to Feb data even worse: Industry profit margin 2.9%, relative to downstream industrial enterprise profit margin 5.8% average level, auto industry only half.
More intuitive comparison comes from profit distribution. 2026 first half, 8 listed power battery enterprises total net profit approx 50.8 billion RMB, while 23 mainstream listed automaker enterprises total net profit approx 25.1 billion RMB. Eight battery enterprises earned money, is twenty-three whole vehicle enterprises two times.
Whole vehicle manufacturing link profit margin even fell to historical ice point. 2026 first half, domestic whole vehicle manufacturing link average profit margin fell to 1.5%, create near 10 year new low. One machine average price 200,000 RMB car, whole vehicle factory can only leave approx 3,000 RMB net profit. This is not "thin profit", this is "no profit".
Again, finance gap become time bomb hanging over industry head.
Many people not noticed a structural problem: New energy vehicles long term exempt consumption tax and maintenance fees, while traditionally these fees internalized in fuel cars fuel tax. When new energy penetration rate reaches 67%, means two thirds of new added vehicles no longer contribute this part fiscal revenue, national tens of millions electric vehicles actually formed a huge road use and maintenance fund gap.

At same time, local governments on auto industry "blood transfusion" ability also declining. Past few years, local state capital is auto industry project main force, but 2024 since, local state capital on auto project investment willingness obviously declined. When land fiscal revenue shrinks, local debt pressure rise, local governments neither has ability continue subsidize car buying, nor has ability continue for loss automakers guarantee. Subsidy two ends simultaneously invalid, consumer end demand stimulation retreating, production end capacity protection shrinking.
Finally, domestic whole vehicle total capacity over 40 million vehicles, full year domestic demand insufficient 22 million vehicles, capacity utilization insufficient seven tenths; Parts enterprises "escaping" auto industry; "Fast-formed car" chaos press consumer trust last straw, above etc combined factors promote current China auto market sales decline status.
#Give Up Fantasy! Era of Going Out to Make Money Already Ended#
When domestic market become meat grinder, going overseas viewed by many automakers as "avoiding vortex" optimal solution.
JPMorgan report also admits, overseas market expected can account for mainstream automaker profit 20% to 50%. 2026 first 7 months, China passenger car export cumulative 5.35 million units, YoY up 72.5%, data looks very beautiful, but current on going overseas expectation, far not like before so active.
If time retreated 3 years back, then going overseas indeed automaker give themselves find new profit curve path, but now, not 2023 that "go out can make money" going overseas anymore.
Europe Market: Tariff after is quota, quota after is localization rate.
EU imposed anti-subsidy tax on China pure EVs, BYD, Geely, SAIC additional tax rates respectively 17%, 18.8%, 35.3%, stacked 10% base tariff, total tariff highest reach 45.3%; China automakers turn to plug-in hybrids open gap, 2026 first half China brands Europe plug-in hybrid market share reached 25%.

But EU quickly made response, October China-EU consultation reached on hybrid car export "up arrangement", based on 2025's 302,000 vehicles set incremental space, by first seven months monthly arrival speed conversion, first year 350,000 vehicle quota compared to full year benchmark shrink 46.3%.
Tariff is hard knife, quota is soft rope, knife chop down will stir rebound, rope tie up can only slowly tighten, for subsequent China cars in Europe expectation, perhaps only those few complete Europe localization construction automakers can reduce strike.
North America Market: Door already welded closed.
2024 US raised China EV tariffs to 100%, 2025 Jan used "national security" reason prohibit China smart connected cars enter US market, 2026 further push this ban permanent legislation. This not trade policy, this is tech blockade.
Southeast Asia Market: Becoming second "China Auto Market".
Some China automakers concentrated pour into A-level new energy passenger car track, price high overlap in 100,000 to 150,000 RMB zone, put domestic "exchange price for volume"打法平移 past. Take Thailand market example, China brand EV price from 2023 average approx 1.2 million Thai Baht dropped to 2026's 800,000 Thai Baht zone, single vehicle gross margin compress approx 15% to 20%. Price war this China automaker always want to overseas bypass topic, in Southeast Asia again appear.
Middle East Market: Geopolitical risk from expectation become reality.
2026 March, Middle East situation sudden change, already Chinese enterprises suspend Dubai Robotaxi fleet operation. Strait of Hormuz if blockaded, international oil price, sea transport freight, insurance fee will all large rise, directly push high full chain cost. Calculation shows, 2026 Middle East market export involved scale approx 300,000 vehicles.
Going overseas encounter not only tariff and geopolitical risk, after-sales network construction, parts supply localization, insurance product adaptation, residual value management system construction, every item is slow work. In Middle East, parts import tariff high promote repair cost and car insurance rate; In Southeast Asia, lack specific for new energy vehicle exclusive insurance products, car insurance renewal price high, directly hit terminal sales conversion.
JPMorgan points out, China auto market going overseas logic already welcomed quality shift, by early period "pure pursuit shipment scale" fully move to "overseas market share sustainability and profitability expansion", translate is, before going overseas is "sell how much as much", now going overseas is "can live or not".
#Most Automakers, Won't Reach Final Round#
For subsequent forecast, picture should be like this domestic demand continues shrinking, export becomes only growth engine. But export market being tariff, quota, geopolitical conflict and localization requirements layer increase. When export growth rate 2027 from 61% suddenly drop to 20%, inside and outside two engines simultaneously lost speed.
At that time, industry faced not "growth slowing" problem, but "total demand can accommodate how much capacity" problem. 2026 first half auto industry capacity utilization 70.57% already lit yellow light, if demand continues down probe, capacity utilization break 65% only time problem. When a industry has one-third capacity idle, price war not "competition strategy", but "survival instinct".
JPMorgan compares 2026 to 2018, but 2018 after followed three years decline, industry still has five automaker stock price doubled, rely on solid profit surprise and new models. 2026? When BYD retail YoY down 28.9%, FAW-Volkswagen down 44.6%, you hard imagine who still can rely on "profit surprise" reverse rise.
Worst moment hasn't come yet, not because JPMorgan forecast not pessimistic enough, but because earlier model collapse just started. JPMorgan defines 2027 as "stock grab stock, weak eliminated final battle year", this final battle might have already started in 2026.
China auto industry past five years enjoyed growth, essence is three dividend superposition: Policy subsidy driven pre-demand release, New energy tech substitute fuel car structural dividend, and global supply chain reconstruction bring export window period. These three dividends now simultaneously enter decline period. Subsidy retreating, new energy penetration rate approaching 70% ceiling, going overseas from blue sea turn red sea.
Worst moment hasn't come yet, not because 2027 will be worse than 2026 — because most automakers haven't prepared for "no dividend" competition. When subsidy no longer pricing fixed module, when going overseas no longer can digest excess capacity, when channel no longer can rely on rebate mechanism maintain, industry will forced return to most original competition logic: Whose product truly worth this price, whose brand truly has people trust, whose system truly can make money.
This logic, never had change. Just past five years, too many people covered by growth dividend hide its existence.