Why isn't Xiaoma Zhixing's commercialization being accepted yet?

Pony.ai-W（02026.HK）（PONY.US） 2025 annual financial report released as scheduled.

From the surface-level data, this is a “positive” performance report: revenue has grown steadily, losses have narrowed significantly, commercialization continues to advance, and the autonomous driving industry business even shows signs of accelerating into a breakthrough. But what stands in contrast with the underlying fundamentals is that after the financial report was released, the stock price pulled back.

As of March 27, Pony.ai’s Hong Kong-listed share price closed at HK$74.6, down nearly 18% compared with before the financial report release, while the U.S. stock market fell 14.66% on the day.

Behind this is not a simple case of “missing expectations,” but a recalibration by the market of the valuation logic for autonomous driving companies.

Commercialization starts “running”

First, look at the core financial indicators. In 2025, Pony.ai achieved total revenue of $90.01 million, up 20% year over year; gross margin improved to 15.7%, with a better profitability structure. But what’s even more worth attention is the loss side. Under GAAP, the company posted a net loss of $76.758 million for the full year, narrowing 72.1% year over year, significantly easing overall operating pressure. Although losses under non-GAAP still widened, this was more driven by continued ramp-up in R&D and commercial expansion spending—essentially still a typical growth-company path.

What truly widens the gap is the change in business structure. Autonomous driving mobility services are the biggest highlight: full-year revenue grew 128.6% year over year; passenger fare revenue nearly quadrupled; order scale and user penetration increased in tandem. This means Pony.ai is no longer just “a technology proof,” but is securing sustained demand in real markets. At the same time, the trucking business continues to advance steadily: technology licensing and domain controller delivery volumes have increased noticeably, and multiple business lines are starting to form synergy.

Judging from the financials and business structure, Pony.ai is transitioning from a “technology company” to an “operational technology company.”

The L4 business model starts to be validated

If the financial data reflects trends, then operational metrics carry even more “turning point” significance.

As of March 2026, Pony.ai’s autonomous driving taxi fleet size has already surpassed 1,400 vehicles, placing it among the global leaders, and it has become the only company currently achieving fully autonomous, chargeable operations—without drivers—in all four first-tier city clusters of Beijing, Shanghai, Guangzhou, and Shenzhen.

More importantly, there is a breakthrough in the profitability model. The company has achieved “full-city single-vehicle profitability turning positive” in both Guangzhou and Shenzhen. Especially in Shenzhen, monthly order volume expanded rapidly, and average daily revenue per vehicle hit a new high. What does this mean? It means that the most core question for L4 autonomous driving—“can it make money?”—has a real-world answer.

In the past, the industry mostly stayed at the stage of testing, subsidies, and demonstrations, while Pony.ai is proving that under specific cities and density conditions, autonomous driving has the possibility of an independent commercial closed loop.

On this basis, the company has begun to accelerate expansion. Domestically, new first-tier cities such as Hangzhou and Changsha continue to roll out. Internationally, the rollout is even more diversified: markets such as Singapore and Croatia are advancing in parallel. The goal is to cover 20+ cities globally by the end of 2026, with fleet size surpassing 3,000 vehicles. In other words, Pony.ai has moved from the “model validation” phase into the “model replication” phase.

What the market is worried about

The business model is gradually taking shape—but the issue is precisely here: the closer it gets to commercialization, the more cautious the market becomes. With share weakness as a result, there are mainly several reasons behind it.

First is controversy over earnings quality. The large narrowing of losses under GAAP is partly due to fair value changes in transaction financial assets, not driven entirely by improvements in core operations; meanwhile, core operating losses under non-GAAP are still expanding.

This creates disagreement in the market over whether a “true earnings inflection point” has genuinely arrived.

Second is downward pressure on valuation across the industry. Autonomous driving is still in a heavy investment cycle, making it difficult to achieve industry-wide profitability in the short term. In the broader environment of a pullback in global tech stocks and reduced risk appetite, companies with “high investment + long cycle” are more likely to be repriced.

Third, liquidity and sentiment factors are amplifying. As a company with a relatively short time since going public, market pricing has not stabilized; the expectation gap after the financial report release is easier to amplify, leading to concentrated trading and thereby increasing volatility.

Finally, it comes down to a trade-off game between expansion and the timing of profitability realization. With rapid fleet expansion and synchronized progress both at home and abroad, it means ongoing capital expenditures. The market has started to focus on a practical question: as scale grows, will profitability realization be delayed?

If you look only at short-term share price, this financial report seems “not impressive enough.” But if you extend the time horizon, its significance becomes clearer. Pony.ai is accomplishing something more critical: turning autonomous driving from “can operate” into “can make money.”

With breakthroughs all at once in the three core metrics—fleet size, city coverage, and per-vehicle profitability ability—this essentially explains the industry’s commercial logic. The capital market’s short-term adjustment is more like a reassessment of the pace, not a denial of the direction.

From the mid- to long-term perspective, once the scale effect is gradually released and operating efficiency keeps improving, the valuation logic for autonomous driving companies will shift from “telling stories” to “looking at cash flows.” And perhaps that is the true value of Pony.ai’s financial report.

(Editor: Wang Zhiqiang HF013)

     【Disclaimer】This article only represents the author’s personal views and is not related to Hexun. Hexun and the site remain neutral regarding the statements and judgments in the text, and provide no express or implied guarantee regarding the accuracy, reliability, or completeness of any information contained herein. Readers are for reference only and should bear all responsibility themselves. Email: news_center@staff.hexun.com

Report