On the evening of September 1, I listened to the full NIO second-quarter earnings call.
Li Bin spoke for nearly an hour on a single theme: quality of growth. He and Qu Yu answered 11 questions in total:
From inquiries about ES8 orders to discussions about next year's new models from Lixiang, the conversation kept coming back to these four words.
The data looks very impressive, with sales rising 67% in the first half of the year, revenue increasing 86%, and gross profit soaring 282%, with each successive layer growing faster. In the second quarter, revenue reached 32.1 billion yuan, up 70% year-over-year.
Losses narrowed by 90%, with a gross margin of 18.4%, up from 10% at the same time last year.
The words are quite encouraging, but the market performance is completely disappointing.
I checked, and during the day, Hong Kong stocks fell by 6 points, closing at 31 Hong Kong dollars, with the lowest intraday level at 29.44, touching a new 52-week low. At that time, the financial report had not been released, and the decline was due to the August delivery data.
After the earnings report was released in the evening, US stocks continued to fall by 4 percentage points, closing at just over $4. This time, the decline was due to the sequential comparison in the earnings report. Even the best performing stocks were not spared, with two declines and not a single increase.
It all started on March 10:
On the evening of March 10, 2026, NIO released its Q4 2025 earnings report, achieving quarterly profitability for the first time in its over-a-decade-long history. The net profit attributable to the parent company was 120 million yuan, a modest amount but of great significance.
That evening, Li Bin spoke in a lively tone, saying that NIO had officially entered the third stage of its development, embarking on a new round of high-speed growth, with the goal of achieving full-year profitability in 2026.
The market had waited for over a decade for this moment, and on the day the US stock market rose by 15 points, with market capitalization increasing by HK$126 billion in a single day. The following day, the Hong Kong stock market rose by another 14 points, and by late March, its market capitalization had rebounded to over HK$1 trillion.
One profitable quarter has pulled NIO out of the "loss-making narrative".
From that day on, the market's perspective on NIO changed. Previously, reducing losses, from a loss of 490 million to 50 million, was considered a good performance. Now it's different, as the company has turned a profit, and the focus has shifted to whether profits can increase quarter by quarter. The criteria for evaluation have completely changed.
Why is that?
Because a company that has been losing money for over a decade is willing to tell a story, the market is willing to listen. Battery swapping, research and development, multi-branding, and high-end development - with quarterly losses of several billion, each story is worth some patience.
Once profitability is achieved, the story is over, and from the day NIO turns a profit, it will no longer have the protection of being a "loss-making company", leaving only numbers. The market looks at numbers, and it's all about acceleration.
With the new rules applied to the second-quarter report, all three key figures are in the red.
Net loss was 528 million yuan, a 59% increase from the previous period, while adjusted net profit was 26.1 million yuan, a 40% decrease from the previous period, with more sales but lower earnings.
The picture for net profit attributable to parent shareholders was even worse, with a loss of 722 million yuan, widening 45% quarter over quarter.
What does 26.1 million mean?
With quarterly revenue of over 30 billion, the profit left for ordinary shareholders is only more than 200 million. Before Deutsche Bank released its report, the estimate was 1.8 billion, which was not even close.
On May 21, NIO released its quarterly report, achieving adjusted profitability for the second consecutive quarter. US-listed shares rose 5% in pre-market trading, but closed with a gain of only 0.18%; the next day, they plummeted 7%. During that period, the short-selling ratio of NIO's Hong Kong-listed shares surged to 56%, with short sellers placing bets with real money.
Today's rout was foreshadowed back in May.
What's more interesting is that institutions, such as BofA, have raised their forecast for NIO's full-year adjusted net profit by 593%, yet maintained a neutral rating, with numbers going up but attitude remaining unchanged. Analysts are also using double speak.
In March, they believed it. By May, doubts crept in. By September, they stopped listening altogether. The market's faith in "NIO profitability" has worn thinner with each passing quarter.
The stock price accounted for these fluctuations, reaching $6.87 in April and closing at $4.06 on September 1, a decline of 40% in five months.
Li Bin spoke about year-over-year comparisons during the conference call, but the market is focused on month-over-month comparisons.
No one is wrong, the acceleration of growth is real, and the month-on-month decline is also real, when combined, that's the situation we saw yesterday.
This is where NIO is currently facing a dilemma, with revenue increasing by 70% and the growth momentum still ongoing. The market wants to see acceleration, but can it continue to speed up in the next quarter?
No text to translate.
Why can't we accelerate growth? Breaking it down, there are only two issues: quantity isn't increasing, and costs are rising.
NIO's delivery guidance for Q3 is 108,000 to 111,000 vehicles, up 0.28% to 3% quarter-over-quarter, which is essentially flat when rounded.
It will become clear by September.
It sold 35,900 units in July and 35,800 units in August. To meet the lower end of its guidance, it needs to sell 36,200 units in September, an increase of only 400 units from August. This sounds achievable, but August was already the second consecutive month of decline, and stopping the decline is a challenge in itself.
In the August new energy vehicle sales rankings, Zerop ran sold 103,000 units, XPeng sold 39,000 units, Li Auto sold 37,000 units, and NIO sold 35,000 units, with the latter ranking last among the four. Zerop's sales volume for one month is close to NIO's sales volume for three months.
In addition to Liuli, there is also Ledao.
At this time last year, Leado accounted for half of NIO's deliveries. In August this year, Leado sold 8,810 units, a year-on-year drop of nearly half, and its share was reduced to a quarter. It was still rising in the previous months, but has fallen behind recently.
Li Bin said that the weakness of Lidao is brand recognition, and the problem is that brand recognition is something that cannot be rushed.
Li Bin is more aware than anyone of the challenges facing Lido, likening it to NIO five or six years ago, saying that once consumers understand the product, the conversion rate is not weak, implying that what Lido lacks is not products, but time.
What the market lacks most is time, as Lido will only launch its strategic new product next year, and it cannot be relied upon for the remaining three months of this year.
Let's talk about costs again.
Qu Yu reported a figure during the conference call, stating that in Q2, each vehicle is 14,000 yuan more expensive than at the end of last year, and is expected to increase by another 2,000 to 3,000 yuan in the second half of the year. As a result, the annual cost per vehicle will increase by 16,000 to 17,000 yuan.
Where did the money go?
Automotive-grade memory chips are seeing price increases as AI absorbs production capacity. Lithium prices are rebounding, and industrial metals like copper and aluminum are also rising. All of these are raw materials — and all of them are in the hands of suppliers.
The chip sector is the most troublesome, as storage manufacturers allocate production lines to more profitable AI chips, and car manufacturers wait in line with cash in hand. The hotter AI gets, the more expensive cars become, and NIO cannot avoid this issue.
NIO can raise prices for customers, with average prices even higher than those of BBA, and customers accept it. However, NIO cannot force suppliers to lower prices, and it has to absorb the cost when chip prices rise. It has pricing power over customers but not over suppliers, leaving NIO with only one hand to exert its pricing power.
On the cost side, a warning was given in March. Qu Yu said during the first-quarter earnings call that the cost per vehicle would increase by more than 10,000 yuan, with some potentially rising to 15,000 yuan.
The September figures came out at 14,000, which falls right within that range. It can't be considered a black swan, as the bad news was already out in the open.
Facing these two issues, NIO used three hands to plug the holes.
First, the product structure.
NIO's average price is 406,000 yuan, rising to over 430,000 yuan in July, with three-quarters of ES9 users coming from outside the ecosystem, while cumulative ES8 sales have exceeded 140,000 units and are expected to reach 150,000 in September.
Relying on high-priced models to support profit margins is currently effective.
What's effective is effective, and the cycle is what it is. In April, Li Bin said at a forum that new cars are selling well as soon as they are launched, but by the time production capacity ramps up, demand has already fallen, which he calls the "new car death valley effect".
The ES8 has been on the market for almost a year and is still holding on, but the next ES8 is still unknown, and we will have to wait for new products from the 5 series and 6 series.
It's visible from the ceiling that the premise of structural optimization is an increase in quantity, but the quantity in Q3 has basically remained stagnant, causing the margin of this business to thin.
Second-hand, caliber.
The company claims three consecutive quarters of profitability, based on adjusted operating profit, which was 207 million yuan in Q2. However, after deducting 554 million yuan in equity incentives, net profit was only over 20 million yuan. The numbers add up, but listeners are becoming increasingly skeptical.
The third hand, saves money.
Research and development expenses decreased by nearly 30% year-over-year, Li Bin said the company's R&D expenditure is around 2 billion to 2.5 billion yuan per quarter, with efficiency no lower than others' 3.5 billion, and the saved profit is also profit, but with limited flexibility.
Li Bin has done the math himself: if bicycles are cheaper by 10,000 yuan, from a profit perspective, they would need to sell three to four times more to make up for it, so he would rather not lower prices. This calculation is correct, assuming the quantity can still be maintained.
The market believes in one thing: the answer is written in the stock price. It only believes in one thing, cash.
The first item is the account for market value and cash.
NIO's market capitalization is HK$76.9 billion, based on the closing price on September 1, which translates to over RMB 70 billion. Of this, HK$56.7 billion is cash. The brand, battery-swapping network, three product lines, and autonomous driving team, all non-cash assets, are valued by the market at only HK$14 billion.
What does 14 billion mean? NIO's quarterly revenue is 32.1 billion. The market prices the entire company at less than half of its quarterly revenue.
The second entry is the account for cash inflows and outflows.
In the last four quarters, operating cash flow saw a net inflow of 3 billion, turning positive, which is a fact.
The bad news is that capital expenditures reached 6 billion yuan for the year, with the bulk going towards battery swap stations. NIO has cumulatively invested 20 billion yuan in charging and swapping, with 4,123 stations nationwide, and is still expanding at a rate of 1,000 stations per year.
Incoming funds are 30 billion, outgoing funds are 60 billion, with a shortfall of 30 billion over the course of a year.
This gap has been filled through financing, with over 10 billion yuan raised from share issuances alone in the past year. Cash reserves have increased from 45.9 billion yuan at the end of 2025 to 56.7 billion yuan, and the more they rise, the more it indicates that operations themselves are still unable to fill the hole.
In other words, this network has been successfully connected.
The cumulative number of battery swaps has exceeded 120 million, with the fifth-generation stations reducing construction costs to 1.4 million yuan per station. "Running through" means that operations are smooth, but profits have not been recovered yet.
Bullish investors don't think so, NIO's battery swap network is a unique asset that competitors can't replicate, and once electricity trading is liberalized, these 4,000 stations will become 4,000 energy storage nodes. There's a gap between making sense and being valuable, and that gap is filled with cash flow.
A moat and a noose are the same thing—during growth periods it's called a moat, but in a market downturn it's a cash guzzler. Right now, the market is pricing it as the latter.
Third, the balance sheet account.
NIO's shareholder equity has dwindled to 4 billion yuan, while the redeemable non-controlling interests that weigh on it amount to 10.2 billion yuan, which is 2.5 times thicker than the common stock.
How did the layered structure come about? Zhixing Chip completed two rounds of financing in June and August, with a post-investment valuation of 12.25 billion yuan, and NIO's subsidiary holds 59.95%. The external funds that came in were invested in redeemable preferred shares, which have a redemption obligation that takes priority over ordinary shareholders.
These mezzanine equity interests also require an annual amortization charge, with Q2's 192 million directly borne by ordinary shareholders. The 722 million net loss attributable to parent exceeds the total loss by 200 million, entirely due to this item.
Issuing shares to fill the holes in the cash flow statement, while the middlemen erode the profits in the income statement, with the same financing deal being recorded in two separate sets of accounts.
Ordinary shareholders are at the bottom, with money going into the middlemen first, and only the remainder belonging to the shareholders.
The last stroke, the market waits for the next number.
NIO has its strengths. It has achieved positive operating cash flow for four consecutive quarters, and the cost of its fifth-generation battery swap stations has been reduced by 100,000 yuan per station, which is true. William Li said the company aims to deliver an average of 40,000 vehicles per month in Q4.
It sold 35,800 units in August, 4,000 more than in August, an increase of 4,200 vehicles, and also higher than the upper limit of Q3 guidance.
Taking this calculation a step further: at the start of the year, Li Bin set a full-year growth target of 40% to 50%. Based on last year's 326,000 vehicles, that would mean selling between 456,000 and 489,000 units for the full year.
In the first eight months, 262,900 units were sold. With four months remaining, sales of 190,000 to 220,000 units are expected, with a monthly average of at least 48,000 units.
In other words, even Li Bin's own projection of a 40,000-unit monthly average in Q4 would fall short of the lower bound of the annual target. There's a notable gap between the goal and the guidance.
On March 10, the market priced NIO entirely at a premium—one profitable day bought a 15-point rally. By late March, its market cap had crossed HK$100 billion. At that point, the market believed in the second half of the story: full-year profitability, a new growth cycle. The conviction was genuine.
On September 1, with a market capitalization of 76.9 billion, the market was only willing to value the company based on its cash holdings. The same company, within a span of five months, had two vastly different valuations, with a gap of over 1 trillion to 76.9 billion.
This difference is the depreciation of trust. Looking ahead, there is only one variable, and the market is waiting for the next number, which will not be clear until the next quarter.
Source of data: