Li Auto spent 4 billion yuan on a generational upgrade.

In the first half of 2026, Li Auto recorded a net loss of 3.981 billion yuan.

When the figure is spread across each day, over 20 million yuan vanishes the moment you open your eyes. Where did this money go? There are three mainstream interpretations in the market: the R&D spending narrative, the collapse narrative, and the model upgrade narrative.

First, rule out the technology spending account.

After the release of the interim report, the most intuitive explanation is that Li Auto poured 5.5 billion yuan into R&D in the first half of the year, and the 4 billion yuan loss is basically equivalent to R&D investment, which means the company is betting on the future and there is nothing embarrassing about it.

It sounds quite reasonable, but the books do not support this claim. The 5.5 billion yuan R&D expenditure only represents a 3.3% year-on-year increase, with actual incremental spending less than 200 million yuan. Li Auto was still making profits in the same period last year, and this small increment cannot fill the 4 billion yuan gap, which means the money has gone elsewhere.

Second, rule out the imminent collapse account.

To judge whether an automaker is on the verge of collapse, we first check two core indicators: cash on the books and operating cash flow. By the end of June, Li Auto had 87.5 billion yuan on its books, including cash, time deposits and short-term investments.

The quarterly operating cash flow turned positive to 15 million yuan from a negative 6.091 billion yuan in the previous quarter. The figure is so small that it can almost be ignored, but the trend has reversed: the revenue from car sales is now enough to cover daily operating expenses. With sufficient cash in hand and core business recovering, this is not the signal of a terminal crisis.

The third account is the real source of the loss: the model upgrade account, with money leaking out from three channels.

The all-electric i6, the cheapest model in the entire product line, accounted for nearly 70% of Li Auto's total deliveries in June alone. Even with its strong performance, Li Auto's total deliveries in the first half of the year still dropped by 5.1%, while the overall retail market fell by 20%.

The high-margin extended-range models that supported gross profit in the same period last year had a vehicle gross margin of 19.4%. The main sales force has shifted from the most profitable models to the cheapest ones, dragging the gross profit per vehicle down from just over 50,000 yuan to around 23,000 yuan.

In just one year, the profit per vehicle shrank by more than half.

For the full L-series product upgrade, the common industry practice is to sell old and new models at the same time: old models are sold at a discount to clear inventory, while new models are launched simultaneously to ensure uninterrupted deliveries.

Li Auto chose a different path: it stopped accepting orders for the entire L9 line, and also suspended orders for some variants of L7 and L8. Stores only sell existing inventory vehicles, and no new production will be scheduled after the inventory is sold out. The cost is that inventory vehicles have to be sold at discounted prices, but the benefit is that the price system remains stable and the residual value of used cars is guaranteed.

The brand's pricing posture is maintained, but the profit concessions are unavoidable: every slight relaxation of the existing car price will reduce the gross profit accordingly. Ma Donghui, the President of Li Auto, admitted on the earnings call that the phased disturbance was caused by old model inventory clearance, new product launch and policy transition during the model upgrade period.

The price of lithium carbonate has more than doubled in a year, rising from 70,000 to 80,000 yuan per ton a year ago to 169,000 yuan per ton now. The price of automotive-grade storage chips has also surged from 20 yuan per unit to nearly 100 yuan per unit.

Li Xiang said on the earnings call:

Li Auto's vehicles have a high level of intelligence, and the storage consumption is larger than that of its peers, so the impact of rising costs is more significant. Cost increase is a common challenge for the entire industry. While other automakers have their own coping strategies, Li Auto has not raised the price of any of its models. It is not the only one that has suffered from the cost pressure, but it is the only one that chose to absorb the pressure without raising prices.

The fourth account is the hidden one: production suspension itself costs money.

There are precedents for reference. When the 2021 Li Auto ONE was discontinued in 2022, the depreciation period of special molds and production equipment was directly shortened from 10 years to 3 years. The 10-year depreciation was fully recognized within 3 years, and this item alone added an extra 226 million yuan of depreciation to the books.

When a car model is discontinued, the production line assets do not stop depreciating. Every set of special equipment must be written down and settled in the financial report.

The corresponding amount of this L-series production suspension is not separately disclosed in the financial report, and the earnings call only mentioned that it was handled in accordance with strict standards.

With three explicit channels and one hidden account, the destination of this sum of money is basically fully explained.

By the way, there is another loss not included in the 4 billion yuan: the poor launch performance of the all-electric i8, which only accounted for 4.51% of total deliveries in the first half of the year. This is an execution issue and not part of the model upgrade related loss.

After breaking down all the accounts, there is still a remaining point to address. Li Xiang once said in the spring of 2023 that gross profit margin is the blood of the enterprise, and a gross margin above 20% is considered healthy. In the second quarter of this year, Li Auto's vehicle gross margin was 9.4%, meaning its "blood" has been reduced by half.

How could an automaker that earned nearly 20 billion yuan in 2023 and 2024 drain its profit level to such a state? What went wrong?

......

The answer lies in the market battlefield. The SUV market above 200,000 yuan is the very market where Li Auto built its success.

In 2023, Li Auto made a net profit of 30,000 yuan per vehicle in this market, with total annual deliveries of 376,000 units, ranking first among all new Chinese EV makers.

In 2024, Li Auto took 15.3% of the new energy vehicle market above 200,000 yuan, ranking first among all Chinese brands. It also remained among the top three players in the SUV market above 200,000 yuan last year.

This loss was generated right in this very market.

The competitors on the other side of the market have changed, and the most thorough change is in the high-end extended-range segment. All the profits of 2023 and 2024 came from the L-series extended-range models.

In the first half of this year, in the extended-range vehicle market above 300,000 yuan, AITO M9 and M8 sold 43,000 units in total, while the combined sales of Li Auto L7 and L9 were only 26,000 units, meaning AITO's sales were nearly 70% higher.

In 2025, most of the large luxury SUV market above 400,000 yuan was occupied by AITO M8 and M9. The high-end market that once belonged to Li Auto was taken over by AITO.

The large SUV segment was also completely overtaken by competitors. NIO ES8 recorded 78,618 retail sales in the first half of the year, followed closely by Zeekr 9X, while Li Auto's most expensive model L9 only sold 12,807 units, a gap of 6 times.

What does this mean? Li Auto L9 sells just over 2,000 units per month, while NIO ES8 sells 13,000 units per month.

The overall market trend leaves no time to hesitate. According to data from the China Passenger Car Association:

The new energy vehicle market above 400,000 yuan increased by 46% in the first half of the year. The highest price segment of the market is still growing, and in the top 5 best-selling SUVs above 400,000 yuan in June, Li Auto only ranked fourth, with the top three positions all taken by AITO and NIO.

Pushed into such a difficult situation, other automakers might have chosen to cut prices sharply to fight back.

Li Auto chose a more costly path: to temporarily withdraw from the market and reshuffle its product lineup.

The delivery gap caused by model upgrade lasted until July, when Li Auto delivered 30,468 units, a year-on-year decrease of 0.86%. In August, the new L6 took over, with 37,679 units delivered, representing a 32.1% year-on-year increase, and the monthly delivery volume quickly recovered.

The company effectively paused its revenue growth for a full quarter, a move that passive automakers would never make. In return, the all-new L8 launched in June, the new L6 launched in July, and the entire L-series product line was fully updated. Ma Donghui set a monthly sales target of 10,000 units for the new L6. The company also adjusted its headcount, with 3,680 fewer employees in the first half of the year.

Li Xiang said at the shareholders' meeting in June that gross profit margin represents all the funds and ammunition that the enterprise can use.

It can be understood that maintaining gross profit margin is more important than maintaining sales volume. Besides, the wholesale volume of the entire extended-range market dropped by 13.1% in the first half of the year. Instead of clinging to a shrinking market, it is better to allocate production capacity to all-electric models.

The model that supported Li Auto's market position is the i6, Li Auto's first mass-market all-electric model. Less than 7 months after its launch, the 100,000th unit rolled off the production line, and it has ranked among the top three best-selling models above 200,000 yuan for 6 consecutive months.

In the first half of the year, Li Auto still ranked first in total sales among all brands in the SUV market above 200,000 yuan, outperforming NIO and Zeekr. Its market position remains, but the profit margin has visibly shrunk significantly.

Looking at the bigger picture, the company has delivered a total of over 1.8 million vehicles. Looking at its performance in this market over the past four years, it recorded a net profit of 11.8 billion yuan in 2023, 8 billion yuan in 2024, 1.1 billion yuan in 2025, and a net loss of 4 billion yuan in the first half of this year.

The real turning point was in the third quarter of 2025, when the company recorded a net loss of 630 million yuan, with operating cash flow outflow of 7.4 billion yuan in that single quarter.

Looking at the profit list of the industry, Leapmotor earned 210 million yuan in the first half of the year, achieving consecutive semi-annual profits. NIO also recorded profits for two consecutive quarters after excluding share-based payment, with a comprehensive gross margin of 19%.

Li Auto and XPeng are both on the loss list, two companies in the same difficult situation. The title of "the most profitable new EV maker" has a new owner. Li Bin of NIO said at Fudan University in August that China's automotive industry has entered the most brutal final stage, and the final surviving players will be determined within three to five years.

The 20% pass line on the market was drawn by Li Xiang himself, but on last month's earnings call, he personally lowered the standard by one level.

Once this line is adjusted, the company's growth story will have to be told in a completely different way.

......

But why is the capital market so calm about this? Because half of the bad news in this financial report is reflected in the books, and the other half is reflected in the forward guidance.

The Q2 financial results themselves are not bad, with revenue of 25.667 billion yuan, exceeding the upper limit of the company's own guidance. The negative part is the Q3 guidance, which is set at 26.6 to 28 billion yuan, far below the market expectation of 32 to 35 billion yuan, a 10% to 20% gap. The adjusted loss is also 30% higher than the market consensus expectation. The financial report was released after the market closed on August 26. The US stock price of Li Auto fell by 1% that day, and the Hong Kong stock price rose by just over 1% the next day. Goldman Sachs lowered its target price on August 28, and the Hong Kong stock price once fell by 3% during intraday trading, but recovered by the close. After three days of trading, neither market saw a breakdown in the stock price trend.

XPeng just went through a similar scenario. It released its financial report on August 24, with Q3 guidance failing to meet market expectations. Its Hong Kong stock price fell by 9.19% the next day, hitting a new 52-week low, erasing more than 8 billion Hong Kong dollars of market value in a single day, and dragging the entire sector down. Li Auto's stock price also fell by 3.6% on the same day.

Li Auto's guidance was also 10% to 20% lower than market expectations, but its stock price barely moved. In the same month, in the same sector, both companies failed to meet expectations, but one was severely punished by the market while the other was not affected.

First, look at market expectations. Its US stock market value has shrunk from 21 billion US dollars last November to around 12.5 billion US dollars now, a 40% drop in 8 months. The price-to-sales ratio is less than 0.9, meaning the market is only willing to pay 0.88 yuan for every 1 yuan of revenue. The stock price is trading near its 52-week low. The bad news has been priced in for a whole year, and the market has already paid for the negative factors in this financial report in advance.

Second, look at market trust. The company has a habit of setting its guidance at a relatively conservative level.

In Q1, it guided deliveries of 85,000 to 90,000 units, but the actual delivery reached 95,100 units. In Q2, it guided a gross margin of around 10%, but the actual gross margin reached 11%. For three consecutive conservative guidance, the company either met or exceeded the target. The market will automatically discount its negative guidance.

Finally, look at pricing power. The market no longer prices Li Auto based on its single-quarter financial results.

No matter how bad the Q2 results are, they are already in the past. The pricing focus is on September: the MEGA will be launched on September 2, the i9 will be launched in mid-September, the intelligent driving platform will be updated in the same month, and the company will also hold a launch event in Dubai to expand into the Middle East market.

The market is waiting for new performance reports. Li Xiang made it clear on the earnings call:

Whether the full-year cash flow can turn positive depends on Q4. He also added that the overall cash flow this year will be stronger than that of last year.

There is also a background to note: the market value of the entire automotive sector has shrunk by more than 1.1 trillion Hong Kong dollars in half a year, with all 13 mainstream automakers seeing their stock prices decline, and the worst-performing ones have even lost more than half of their market value.

All the bad news in the industry share the same source: the purchase tax exemption was reduced from 100% to 50%, the demand of last year was overdrawn, and the entire industry is paying the price for last year's overdrawn consumption. Li Auto's stock price has fallen by 28% this year, which is at the medium level of the entire sector.

Horizontally, it still ranks first among all new EV makers in terms of market value, at 97.7 billion Hong Kong dollars, 12 billion Hong Kong dollars higher than the second-ranked NIO.