Elon Musk has frequently said that Tesla FSD will make Tesla cars immensely more valuable than their … [+]
This week, during Tesla’s quarterly earnings call, Elon Musk repeated a claim that thanks to Tesla FSD, Teslas will all become worth far more — five times— what people are paying for them, and that the $15,000 FSD package is “actually a low price, not a high price.”
Robotaxi technology is indeed very valuable, and companies around the world are investing tens of billions to build it, but Musk’s math needs some serious adjustment.
Here’s what he said:
“It’s not about getting more share. It’s just that you can think of every car that we sell or produce that has full autonomy capability as actually something that in the future may be worth as much as 5 times what it is today. Because average — vehicle is doing like maybe 10 hours of driving a week. If sort of — if this says 1.5 hours a day on average, that’s 10 hours a week-ish. If you’ve got on autonomous — if the vehicle is able to operate autonomously and use either dedicated autonomous or partially autonomous like Airbnb, like maybe sometimes you allow your car to be used by others. Sometimes you want to use it exclusively just like Airbnb — doing Airbnb with a room in your house. The value is just tremendous.”
We think of private cars as having a lifetime in years, and to some degree in miles. Taxi lifetime, on the other hand, is almost exclusively in miles. In California, the average car lasts about 20 years and drives about 200,000 miles. It’s pretty worn out at the end of that, in part due to design. In New York, taxis last 250,000 miles and around 5 years. They have some life left after that if maintained well.
Once your car’s life is measured in miles, the math is quite simple. Whether it’s on the road all day or half the day affects how many years it will last, but doesn’t change the economics a lot. The car becomes a consumable asset, like a tank of gasoline, rather than a fixed asset that is being wasted if it sits there unused in your garage. A tax of gasoline isn’t more valuable if you use it for 10 hours/week instead of 1.5 hours/week. Real estate is the exact opposite. It ties up capital and if you don’t rent it or live in it, you’re wasting it. Cars and robotaxis are not completely at one end or the other, but as taxis they become much closer to consumable. A fully consumable asset is only worth at best a tiny bit more if you use it up more quickly.
The value of the robotaxi
Of course, a robocar is more valuable than a regular car in every mile it drives. However these values are similar for the private car and robotaxi:
- It ideally drives more safely. That’s worth a lot, but we have a price on that, which is what you pay for liability and collision insurance — around 5 cents/mile.
- You get free time as you ride, not having to drive. That’s huge, depending on how you value that time.
- It offers electric travel with no compromises. That’s cheaper and less hassle than gasoline.
The robotaxi is a great business proposition for the fleet operator. They can now sell customers miles instead of cars. Customers get the value of not having to own and maintain them, not needing a garage, and not paying for parking. You do get more effective utilization of assets. You can allocate “the right car for the trip” so that the 80% of car trips that are solo are done in solo cars at half the cost. You could go on and on about these advantages, and I have done so in these pages. But they don’t add up to five times the value per mile — it’s a better case perhaps for double. Which is nothing to sneeze at, to be sure, but not five-fold.
The best argument for Elon’s case is that electric powertrains are much more durable. Maybe you can make an electric car that can last for a million miles over 20 years. Turns out taxis, even when they work all day in shifts, don’t do much more than 250 miles/day unless they are the rare ones doing a lot of highway. In that case it’s hard to actually wear one out in 20 years. Some parts aren’t going to last a million miles — the seats and various interior elements will need replacement long before butts have sat in them for a million miles. In places with rust and road debris, the vehicle might still drive but nobody will want to pay to sit in it. People expect a relatively new looking vehicle, and they also expect modern features in their ride — after enough years, like a computer, cars will become obsolete and very downmarket before they physically have to go to the junkyard.
The hard reality is that having your car drive a mile will use of up a mile of its life. That includes battery life until we get amazing new batteries. You can use that mile for yourself, as the owner of the car, or you can sell it to a robotaxi passenger, but it doesn’t change the value of the car to you at all unless you can get a lot more miles out of it and still have it be in as good condition when done.
The robotaxi market will also be competitive, especially if the taxis are living for a million miles. That competition will drive down prices. It’s not sustainable for a product to cost you $50,000 and make you $250,000. Margins will be profitable, but not remotely that profitable. Not on a commodity which a billion people own. The very logic that would make your extra miles incrementally cheap makes them cheap for everybody, and triggers a race to the bottom on prices.
Worse, your personal car, which (if modern sales are any clue) is probably a 5 seat SUV, is a poor competitor in the market for the bulk of the robotaxi market,which is solo rides across town. We buy SUVs as all-around cars to do all the trips in our life, from skiing to errands. They won’t compete with cars that live their life for the sweet spot. (That’s one reason I, as a matter of disclosure, have invested in the development of cars for solo urban rides.)
It should be noted that Musk’s mistake is very common. It’s almost normal for some people to focus on the ability for a car (rather than a ride) to be “shared” as one of the premium values of robotaxi (or all taxi services.) The statistic that cars sit around parked all day but this can change if they are used by multiple people is frequently cited. As such, Musk is probably unaware of his error. Put simply, once cars wear out by the mile, then the number of cars needed equals the number of miles people travel divided by the number of miles a car lasts — how many people share the vehicle at different times isn’t even a factor in that equation.
Fleet owned or subcontracted?
If you stay in AirBNBs, you probably have noticed that while they began as being people’s real apartments and houses taking in guests, most of them today are full time AirBNBs, often owned by somebody who owns several such properties. This has parallels to the idea of hiring out your own car as a robotaxi in Tesla’s network, or Uber’s. There are economies of various types in having a fleet of vehicles dedicated to the task. I helped my students create a revolutionary company called Getaround that rents out privately owned cars, and it works, but it’s still a tiny part of the rent-a-car market, and many of the cars available within it are actually full-time rental.
The likely best use of private cars that hire out is to handle the peaks of demand. Robotaxi companies will build company-owned fleets big enough to serve almost all demand. But there will be unusual peaks and surges where more cars are needed to maintain service levels. Then they will put out the call, “we need more cars” and private cars will hire themselves out if the price offered is acceptable and they aren’t being used by their owner. But these surges won’t drive your car 80,000 miles/year, they will just add some bonus money. If you’re willing to hire your car out at a bargain rate, you may sell more miles — but at a fairly modest margin.
This might vary from location to location. One attraction in some places would be the option to own a car that you otherwise could not afford, with the hire-out income making that possible. It will come with costs — you must keep your car fully clean and ready at all times, and often you will find it’s not available for you if you didn’t put a hold on it and you need to use some other robotaxi. Your car will get old faster, though you might enjoy replacing it more frequently.
I
Tesla concept with no steering wheel.
t may become common that cars spend the first 3 years with a private owner who is willing to pay that premium to have a new, late-model car. Then from age 3 to 7 they could enter the robotaxi market. Nobody insists their taxi is brand-new. This was actually the plan Tesla announced some years ago — they hope to lease new Teslas to customers, then convert them to robotaxi when they come off-lease. It’s a plan only Tesla can do because they design the car for just that — the spartan interior needs modest modifications. Other car designs are hard to retrofit as a robotaxi. Even the Tesla, though, won’t look like a Zoox or Cruise Origin with face to face social seating and sliding doors. Zoox has bet that people will want to ride only in a custom vehicle that was never meant to be a private car.
People could own cars designed to be robotaxis, with no steering or pedals. But they could also just subscribe to a robotaxi service that sends one any time they need it, one tailored to their trip. There are attractions to both.
In the short term, though, there is the problem that making a self-driving car for a local taxi fleet is much easier than making a car that’s a good private purchase.
So if Tesla FSD ever works — and that’s probably a long way off — it could add $15,000 of value to your car, or at least to the years of life left in it. There are people who bought it who have since sold their car, getting no value from it and finding they got less than $4,000 for it on resale. (Temporarily, this quarter, Tesla is allowing customers who buy a new Tesla to transfer the FSD software licence to it, something owners have been demanding for some time.)
