Li Bin said in April that NIO has to keep tightening costs after the electric-vehicle maker extended its profit-making momentum into the first quarter of this year. The message landed as a warning as much as a victory lap: NIO had finally stayed in the black again, but Li made clear that the company cannot treat that as proof the hard part is over.

That is why investors are searching NIO now. The company was close to the edge last year, when it posted a net loss of more than 6 billion yuan in the first quarter and its cash reserve slipped below 26 billion yuan. Sales were still hovering around 10,000 vehicles a month, a level that underscored how fragile the recovery looked even before profit returned.

Li’s own comments show how unlikely that turnaround once seemed. He later said fewer than 1% of people in the world believed NIO could make a profit at that time, even as he pushed the full rollout of NIO’s CBU mechanism and promised the company, which had been in the red for ten consecutive years, had to turn a profit in the fourth quarter of that year. In the same period, NIO launched two new models, LeDao and Firefly, while trying to manage a difficult ramp-up for the LeDao L60.

The turnaround mattered because the company was trying to climb out of a stretch when weak sales, heavy losses and strained cash left little room for error. Firefly was originally aimed at overseas markets but had to pivot to China because of tariff issues, and that change added more pressure to a launch plan already testing the company’s execution. NIO’s profit run into this year suggests the business has stabilized, but it has not erased the conditions that made the recovery so difficult.

Li framed that challenge more broadly at the Hundred-Person Conference in April, where he introduced the idea of the Valley of Death Effect for New Cars. In his view, some new cars now sell strongly at launch and then lose momentum once production scales and deliveries settle, with most models’ shelf lives compressed to less than a year. He contrasted that with the fuel-vehicle era, when a facelift usually came only after two to three years and a full replacement could take five to seven years. For NIO, the question now is not whether it can survive one difficult stretch, but whether it can keep making money in a market where even a fast start can fade almost as quickly as it arrives.