Investment Vehicle Vehicle
Autonomous cars are coming. Rideshares aren’t going anywhere. Ergo, rideshare drivers will lose their jobs. What should they do? Retrain, upskill, become data scientists. Yes, yes. But what are their actual alternatives?
I asked Rafael, an Uber driver I met in New York last week. He said, “I’ll buy a driverless car and have it make money for me on Uber.”
I’ve heard this before. He seemed earnest. So I decided to look at the numbers. How viable is a driverless car as an investment vehicle…vehicle?
Here’s the spreadsheet if you don’t want to read my words.
The scenario Rafael envisions is this: In the not too distant future, Uber and it’s ilk no longer engage human drivers. Their fleets consist of crowd-sourced driverless cars. Crowd-sourced because owning, operating, and maintaining millions of cars worldwide isn’t fun. So Rafael, a 4.95 star driver, is now out of work. What does he do? He buys (or leases) a driverless car. He enrolls it in all the rideshares. His car is hailed. The customer is charged. The rideshare takes a cut. And Rafael gets the rest. All the while he’s lounging about on his couch. His cumulative fares are a living wage. Passive income, baby.
So let’s dig into Rafael’s dream.
First we need to make some assumptions. Then we need to look at his costs. Then his revenue. Then his profit or loss.
This is a projection. I am not a soothsayer. So we have to make certain assumptions. Let’s err on the side of conservatism. In other words, when in doubt, let’s assume things aren’t going to get better, cheaper, faster. Here are the assumptions:
- Location is New York City
- The car is parameterized by a Tesla Model 3
- Fares and demand are constant
- No EV tax credits, incentives, and so on
- Cars will opt to Supercharge
- Supercharger network density will mean <10mi to nearest node from any point on an urban map
- Battery capacity doesn’t wane during this window
- Maintenance once a month, downtime of 24hrs
- No change in electricity costs
I think that’s it.
So Rafael’s car of tomorrow is an upgraded Tesla Model 3. Why?
- With a few upgrades, all the hardware needed for automation is ostensibly baked in
- It’s a real car, not a prototype
- There’s a lot of data available about it
Maybe production and competition will ramp up and Rafael’s car will be cheaper and more capable. Maybe car ownership will plummet, the technology will stall, and more expensive, incapable cars will become the norm. The former is more likely. Regardless, this seems like a reasonable starting point.
Let’s look at the Tesla Model 3’s parameters:
Here’s how it’ll work. The car starts fully charged. It heads out to work. It’s average operating speed is 17.6mph (NYC is a congested town). It operates until a remaining range of 5% (15.5mi). At that point it heads back to recharge. Once charged it has completed a single lap. It then begins it’s next lap. It continues in this way ad infinitum. Actually, it continues until it needs maintenance.
Tesla says the Model 3 needs maintenance once a year or every 12,500 miles, whichever comes first. With our usage, 12,500 miles is going to come first. Way first. Like every 12 days. That doesn’t seem practical. Apparently, some drivers with double or triple this mileage were advised by Tesla to nevertheless come in only once a year. We want to be conservative. So let’s bring our car in for maintenance once a month. According to Tesla owners the annual maintenance takes 24hrs. We’ll assume we’re not given a loaner.
- Average Urban Speed (mph) 17.6
- Recharge Trip Buffer 5%
Alright, next we’ll consider the costs. Here are the items I could think of:
- Vehicle down payment
- Vehicle monthly lease payments
- Rideshare Commission
- Sales tax
- Income tax
Again, to be conservative, we’ll ignore incentives. We’ll also assume insurance premiums remain constant. Here are the costs associated with the vehicle.
And of course, taxes and fees (remember we’re in NYC):
Rafael won’t want (or be able) to drop $50k cash on his car. So I borrowed some financing parameters from a Tesla owner’s forum. These are the least favorable terms I could find.
So let’s tally up our monthly costs. Any costs that aren’t monthly (e.g. down payment) we’ll amortize over 12 months.
This is the easy part. Rafael will make whatever his car brings in. The car will bring in money as long as it’s picking up and dropping off passengers. It won’t make money when it’s idle. It’s idle when it’s driving back to re-charge. It’s idle when it’s recharging. It’s idle when it’s in maintenance.
We can avoid worrying about utilization rate, trip duration, surge pricing, and so on by using published data on average fares and trips per hour.
- Average Fare (NYC) $25.91
- Average Trips per Hour 2.00
This, in principle, should bake in those other factors. It also allows us to proceed with our assumption that fares and demand are constant.
A lot hinges on the accuracy of these data. Different sources offer different numbers. It’s often unclear whether the numbers are pre- or post-rideshare fees. I tried to choose se the least favorable.
So now we simply need to figure out the car’s total earning road time.
And the total revenue:
Alright. We’re at the finale.
That’s a lot of money.
I’ve almost certainly made mistakes. Please point them out to me. Here are a few interesting observations from this analysis:
- New York City caps drivers to 10 hours of driving per 24 hours as part of the Fatigued Driving Prevention Rules. Our car can operate at 21.3 earning hours per day. That’s a 113% advantage in earning potential
- A lot hinges on the accuracy of our data. Specifically, Gross Average Fare, Trips per Hour, and Average Urban Speed. Different sources offer different numbers. I tried to choose the least favorable.
- Vehicle warranty (50k mi) would be exhausted by mileage in 48 days. Battery warranty (120k mi) would be exhausted by mileage in 114 days. Tesla recommends maintenance every 12,500 miles, which would translate to every 12 days. More rugged commercial vehicles may need to be developed.
- Profit margins are too big. Competition will narrow margins. Consumers will see significant savings; car owners likely to see greater commissions
- Fares could be reduced 40% without negatively impacting status quo. Fares could be reduced 70% before drivers stop making money
- Autonomous cars will have at least 1 additional free seat; capacity will increase, revenue per hour should increase as well
- Supercharging to 80% takes 40 minutes. The last 20% takes another 40 minutes. In practice, it may be more economic to only partially recharge and make more frequent recharge stops
- As predictable cashflow assets, fleets of autonomous vehicles might be securitized and sold on capital markets
Fork the spreadsheet and play with different scenarios. Let me know what you discover.
Autonomous cars in rideshare services may generate a monthly profit of over $10k. Cost of rideshares will almost certainly plummet.