For stock investors, the golden Qilin analyst research reports offer authority, professionalism, timeliness, and comprehensiveness, helping you uncover potential thematic opportunities.
On November 9, Pony.ai (HKEX: 02026) closed at HK$34.80 in Hong Kong, down 8.42%, with a market capitalization of HK$15.144 billion.
Recently, Pony.ai's Hong Kong listing marked its first anniversary. Looking back to November 6, 2025, the company completed its issuance at HK$139 per share and broke below its offer price on the very first day of trading, meaning its share price has fallen more than 70% from the offer price within a year.
According to Tonghuashun data, Pony.ai generated revenue of US$70.47 million in the first half of 2026, up 98.9% year-on-year, but its net loss for the same period reached US$98.861 million, already exceeding its full-year 2025 loss of US$76.758 million. It is evident that Pony.ai's product expansion and widening losses are occurring simultaneously, and scale growth has yet to translate into profitability.
On the business side, Pony.ai has made intensive progress on the product front. Its three business lines—Robotaxi, Robotruck, and smart solutions—share the virtual driver full-stack technology and the PonyWorld world model, and can be adapted to multiple vehicle models, forming a reusable technology foundation. Based on this technology foundation, Pony.ai has leveraged its Robotaxi operating experience into heavy trucks, domain controllers, and overseas co-built fleets. Pony.ai also disclosed that as of the end of the first half of 2026, its cumulative autonomous driving mileage had exceeded 100 million kilometers, of which fully driverless mileage exceeded 40 million kilometers, validating the transferability of its technology.
Overall, Pony.ai's revenue nearly doubled in the first half of 2026, reaching US$70.47 million, with gross profit of US$11.911 million and gross margin edging up from 16.3% to 16.9%, a slight improvement in profit quality. However, the revenue structure reveals a mismatch in growth drivers. Looking at Pony.ai's segment data for the first half of 2026, Robotaxi revenue was US$20.643 million, up 534% year-on-year, yet it accounted for only 29.3% of total revenue. In contrast, Robotruck revenue was US$23.524 million and smart solutions revenue was US$26.303 million, with the two businesses together accounting for about 70% of total revenue. In other words, it is not Robotaxi but the two thinner-margin businesses that support 70% of revenue. This further explains the quality of the gross margin improvement. Pony.ai says the gross margin improvement is precisely due to a higher contribution from the higher-margin Robotaxi business, but on the other hand, the two businesses supporting 70% of revenue carry thinner margins. Therefore, this round of growth is currently driven mainly by non-core businesses rather than the high-margin Robotaxi.
More noteworthy is the divergence in growth rates. Pony.ai generated total revenue of about US$36.2 million in the second quarter of 2026, up 68.8% year-on-year, but up only about 6% quarter-on-quarter, a clear slowdown in expansion. In the second quarter, Robotaxi revenue grew 691.2% year-on-year and passenger fare revenue grew 849.3% year-on-year, a standout performance. Yet in the second quarter, the smart solutions business grew only 3.9%, due to fluctuations in the delivery pace of domain controllers, showing that the high growth of non-core businesses is not stable.
Scale expansion has yet to translate into profit. In the first half of 2026, Pony.ai's research and development investment was about US$104 million, with a net loss margin of 140.3%. Under non-GAAP measures, the company's loss was US$85.906 million, widening 26% year-on-year, and losses are still deepening.
Cash burn is also accelerating. In the first half of 2026, Pony.ai's net cash outflow from operating activities was US$118 million, and free cash flow net outflow was US$163 million, widening 73.2% year-on-year, while capital expenditure was US$44.62 million, up 208.5% year-on-year, a significant increase in investment intensity.
The pace of delivery remains the main variable determining valuation. Management says Pony.ai needs to deploy a cumulative 40,000 to 50,000 Robotaxis to turn free cash flow positive, a target more than 20 times its current fleet of 1,975 vehicles. Positive per-vehicle profitability is currently limited to specific cities and vehicle models, and whether depreciation, remote assistance, and compliance costs can be amortized remains unanswered.
Overseas deployment and annual targets likewise need delivery volumes to be tested. Most of Pony.ai's planned overseas deployment of more than 4,000 vehicles is in the nature of plans and memoranda, and whether co-built fleet revenue comes from one-time licensing, vehicle sales, or ongoing revenue sharing will also determine the quality of growth. Pony.ai's targets of "3,500 vehicles by the end of 2026 and full-year Robotaxi revenue reaching 3.5 times that of 2025" still need to be verified by actual delivery volumes.
Finally, one capital account needs to be clarified. Pony.ai's net proceeds from its Hong Kong global offering in November 2025 were about HK$6.454 billion. As of the end of June 2026, Pony.ai's cash and wealth management products totaled RMB9.435 billion, equivalent to about US$1.3 billion. If statically extrapolated based on the annualized scale of first-half operating net outflow, these funds could support the company's operations for more than five years, but only on the premise that the cash burn rate does not accelerate further. In fact, second-quarter capital expenditure was already about 3.4 times that of the same period last year. The interim report has written "prudently managing capital expenditure and prioritizing capital efficiency and returns" into its outlook, and after the share price fell more than 70% from the offer price, Pony.ai's room for subsequent equity financing has also narrowed.
In summary, from Pony.ai's product progress to its actual income statement, there are three barriers in between: fleet density, order volume, and cash flow. What can reprice this company is still verifiable operating data, not yet another cooperation agreement.
Comments