The Rocky Mountain Institute just published an insightful deep dive into the economics of self-driving cars, entitled Peak Car Ownership. It presents granular data on the drivers of the shift to autonomy in a way that is immediately understandable. You can download it here:
I’m such a fan of this analysis that I wanted to highlight a few graphs and discuss them here.
First, why do you think Uber and other transportation network companies (TNCs) are pushing so hard to roll out autonomous vehicles (AVs)? Why did Uber CEO Travis Kalanick call them “basically existential for us”?
This graph makes it obvious:
TNCs cost more than 2x the operating costs of a personal vehicle, and more than half of that cost is associated with the driver. Removing the driver, and adding autonomous hardware, creates only a slight net cost increase ($0.82/mile to $0.86/mile). That assumes that TNC revenue doesn’t deflate, which is probably will. All in all, AVs should be cheaper than personal car ownership, especially if you value your time. If Uber doesn’t get there first, its flagship product will be priced double its competition, even if it chooses to give 100% of the proceeds to the driver! That’s the definition of an existential crisis.
The authors also believe that electric vehicles (EVs) priced at $35,000 (like the Tesla 3 or Chevy Bolt) represent a tipping point for the technology. For the first time, in 2018 the total cost of ownership for an EV should be lower than that of a comparable gasoline vehicle without subsidies:
The authors estimate the incremental savings of using EVs are ~$0.07/mile by 2030 — a meaningful 10% reduction in the cost stack.
While Silicon Valley may be the center of AV innovation, it won’t likely be the first market to see AVs at scale. For that, the authors identify Austin, which has the perfect combination of urban density, nice weather, market size, and friendly regulation. The authors calculate Austin to be a larger market than Uber’s estimated 2015 revenue of $1.5B:
The other early markets are mostly in the south: Dallas, Houston, Atlanta, Miami, etc. California, with some potential showstopper regulations, and snowy northeast markets will be a big challenge.
Perhaps putting regulatory challenges aside, however, the penetration rate of AVs could rise above 50% by 2030 — a ramp similar to color TV, but slower than smart phones. Baked into that assumption is deflation from around $0.86/mile today to around $0.30/mile in the future:
The transition to 50%+ AVs, most of which will be EVs, will devastate the auto industry if it doesn’t evolve its business model. The authors believe peak car ownership will occur in the 2020 time frame. Big auto will need to substitute something for its lost revenue and will likely need to provide some level of TNC-like services themselves to do so.
The market for mobility isn’t shrinking. To the contrary, the mobility market will grow if we do indeed see 66% cost deflation. The market for cars themselves, on the other hand, will shrink dramatically. ARK Invest has modeled out this scenario and believes annual volume in developed markets (America, Europe) will drop by half over the next decade:
AVs pose an existential crisis not just for Uber, but for the entire automobile industry. Personal transportation is a commodity. If you aren’t the low cost provider, what are you in business for? If the cost per mile of AVs shrinks beneath the cost of owning a vehicle, it’s game over for the current business model of auto sales. That’s going to cause a lot of hurt, and create a lot of opportunity.
Authors
