No. of Recommendations: 6
That's the title of a recent analysis by Cathie Wood's Ark Investments estimating the value of the 2030 robotaxi market. That prediction would certainly be an interesting trend, so I looked into it a bit more. The analysis by itself was a little hard to follow without the footnotes, so I linked one of those below and combined the two a bit.
ARK estimates that by 2030 the autonomous ride-hailing market could be worth $10 trillion, and that AV ride-hailing could generate $1.2 trillion a year for Tesla alone, with 63% of that ($756 billion) from robotaxi rides. Google tells me that the current ride hailing market is about $300 billion, so ARk is anticipating enormous growth in this market.
On June 22 in Austin, Tesla launched its robotaxi service to a limited group of users, beginning the shift in its model from one-time, low-margin hardware sales to recurring, high margin revenues. Now that it is scaling, Tesla should be able to meet urban Austin Vehicle Miles Traveled (VMT) with only ~200,000 vehicles, an opportunity that far exceeds today’s ride-hail market.
According to our research, Tesla’s robotaxi business could represent ~90% of its enterprise value by 2029, capturing a significant share of ARK’s projected ~$10 trillion global robotaxi market...It is a little hard to follow some of the assumptions they are making, but essentially it seems to come down to the idea that if ride hailing were cheaper, people would use it a lot more. Which makes a certain amount of sense. Now if Tesla can capture the Austin market with "only" 200,000 vehicles, at say, $35K per vehicle that's $7 billion in capex in the next four years. Ark notes to serve any of the top 10 US cities it would require about one to two million vehicles per city, which is roughly same as global production as models Y and 3. Or roughly $35-70 billion in capex per city. Capex numbers are mine, by the way. My point here is that it will require enormous capital investment to even approach the numbers Ark is talking about.
Now, number of current ride hailing vehicles per city currently is typically in the thousands or tens of thousands in really big cities. So they are anticipating the ride hailing market will grow by orders of magnitude.
From the article:
Now that its Austin and San Francisco launches are underway, Tesla is beginning to enjoy three key competitive advantages—vertically integrated manufacturing, data, and cost per mile—that should position it to expand rapidly across other US regions.
The main flaw I see in this analysis is that Tesla doesn't actually have AV ride hailing. Tesla's ride share program requires geofencing and a safety operator. This is not conducive for rapid expansion, and the cost of a safety operator in each car certainly eliminates whatever cost per mile Ark advantage thinks Tesla will recover by operating cheaper vehicles. Ark's data argument is that Tesla has so many more autonomous miles than Waymo that Tesla simply has more information to improve their models. A problem with this thinking is that even with that advantage Tesla still doesn't have AV ride hailing, and the number of robotaxi miles is miniscule compared to Waymo. I think we can safely cross off rapid expansion as something that won't be happening soon.
In this video, the robotaxi carrying the Ark Investment analysts takes a left turn into a parking lot, but winds up blocking the road while waiting for a car to clear the driveway ahead of them. The robotaxi then inexplicably fails to drop them off at their destination, exits the parking lot onto the public road, re-enters the parking lot, and drops them off as intended.
https://www.youtube.com/watch?v=anAYGAwOnWQ&t=1945...In this video, a different robotaxi with a different Ark Investment analyst gets confused trying to enter the same parking lot. After a quick check-in with HQ, the safety operator gets out and moves into the driver's seat, and drives the car manually in the parking lot.
https://youtu.be/RnrgVkoj334?si=vxb-htZw5V7sjDWU&t...I think Ark Investments is drinking their own Kool-aid. I think they got the trends and timelines wrong.
https://www.ark-invest.com/articles/analyst-resear...https://www.ark-invest.com/articles/valuation-mode... 