Community Contributor
Member since 2024
Views
0 Total number of views
Fair Values Set
0 Total number of fair values set
Comments
1 Total number of comments
Followers
0 Total number of followers

No bio added yet

No link added
Checked it. The arithmetic is fine; the inputs aren't, and the conclusion doesn't follow even if they were.1. Fare. $7 is a US robotaxi price. Pony's record peak day in Shenzhen (22 March 2026) was RMB394 net revenue per Gen-7 vehicle on 25 orders — about US$2.20 per order. You're roughly 3x high.2. Utilisation. 25 orders/day is Pony's all-time single-day high, not an average, and you then run it 365 days with zero downtime for charging, cleaning, maintenance, weather or geofence interruption.Corrected, the best day Pony has ever recorded yields ~US$55/day. On $43k of hardware that's ~26 months of gross revenue, before any operating cost. The reported actuals agree: FY2025 robotaxi services revenue US$16.6m on a fleet just past 1,000 units; Q1 2026 US$8.6m with the fleet above 1,700 — call it US$20–25k per vehicle per year against your $63,875.3. The caveat is the whole argument. You flag "not including operational costs (people costs)" and then set it aside. Remote safety operators, platform commissions, charging, insurance, cleaning, depot and maintenance are what determine whether a robotaxi contributes anything at all.4. Payback isn't profitability. Q1 2026: 16.2% gross margin on US$34.3m revenue, US$63.9m of opex, US$53.5m net loss. Marginal hardware payback says nothing about R&D, mapping, licensing or overhead.5. Falling BOM cuts both ways. Pony targets sub-RMB230k (~US$34k) total vehicle cost for 2027. Great for new units, bad for the residual value of fleets already deployed on a five-year depreciation schedule.What you get right: the cost trajectory is real, and city-wide UE breakeven in Guangzhou (Nov 2025) and Shenzhen (Feb 2026) is a genuine milestone. But that is contribution-margin breakeven per trip — not "cracked the per-unit cost," and not an 8-month payback. Your post predates all of it; the data has since landed, and it's less favourable on revenue per vehicle than the model assumed.
0
|
0