Dear Ark Invest team,
Thanks for taking the time to run a detailed Monte Carlo analysis of Tesla’s 2025 fortunes. It’s easy to criticize, and not so easy to do the work. I appreciate your efforts.
I apologize in advance for my skepticism. On the bull case, I have my doubts, which I will detail below. I hope my thoughts serve to fine tune your model. There are three major areas I will focus on: Tesla’s cash flow, vehicle sales, and robotaxi dominance. I will be using your “Estimated Values in Bull and Bear Example Cases” as my basis.
|2020||Example Bear Case 2025||Example Bull Case 2025|
|Cars Sold (millions)||0.5||5||10|
|Average Selling Price (ASP)||$50,000||$45,000||$36,000|
|Electric Vehicle Revenue (billions)||$26||$234||$367|
|Insurance Revenue (billions)||Not Disclosed||$23||$6|
|Human-Driven Ride-Hail Revenue (net, billions)||$0||$42||$0|
|Autonomous Ride-Hail Revenue (net, billions)||$0||$0||$327|
|Electric Vehicle Gross Margin (ex-credits)||21%||40%||25%|
|Total Gross Margin||21%||43%||50%|
|Total EBITDA Margin*||14%||31%||30%|
|Market Cap (billions)||673||$1,500||$4,000|
|Free Cash Flow Yield||0.4%||5%||4.2%|
Take bull case cash flow and market cap at face value
Let’s assume Tesla makes it to 10 million vehicles by 2025. From the above table, the model estimates Tesla will be worth $4 trillion and have a Free Cash Flow Yield of 4.2%.
If we multiply $4 trillion market cap by 4.2%, we get Free Cash Flow of $168 billion a year.
Apple and Microsoft had the most cashflow of all US companies in 2020. This took them more than 40 years to achieve. It’s far less than Ark’s bull case, which has Tesla doubling Apple’s cashflow, 15 years from the time Tesla went public. For Tesla to achieve $168 billion in Free Cash Flow would be shocking and unprecedented. I don’t think it’s that easy.
|Market cap (billions)||$628.58||$2,014.40||$1,737.35|
|2020 Free Cash Flow (billions)||$2.79||$73.37||$45.23|
|Present Free Cash Flow yield||0.4%||3.6%||2.6%|
All data from macrotrends.net as of March 23, 2021
- Tesla’s technoking (Elon) and master of coin (Zachary) would spend that $168 billion on hiring people, factories, body shops, service centers, galleries, Boring tunnels, Superchargers, and R&D to speed up the renewable energy timeline. The Free Cash Flow wouldn’t remain free for long. Tesla’s focus is not on making money for itself. If that was true, it would have stopped at the Model S and Model X.
- Competitors will enter the market at negative margins to gain some of this cash flow and market share. This will depress future Free Cash Flow.
- Regulators will take a keen eye if Tesla needs antitrust action or regulatory guidance on robotaxi fares. We see concerns rising around Big Tech. At that level of cash flow, scrutiny would be very high.
- If Tesla was valued similarly to Apple and Microsoft’s Free Cash Flow yield, at 3.1% Free Cash Flow yield, it would be worth $90 billion. No one is claiming Tesla will go down that far, but it’s food for thought.
10 million vehicles by 2025
In the below chart, let’s break down a simple way Tesla gets to 10,000,000 vehicles a year. In 2021, Tesla produces 2 million vehicles, and every year afterwards, Tesla produces 2 million more vehicles. The third row shows Tesla’ cumulative vehicle production, which would be close to 32 million in 2025. In the last row, we estimate how many doubles kick in on the Cumulative Vehicle Production. If Tesla was to achieve 10 million vehicles produced by 2025, Wright’s Law would kick in three times. Each time, the average selling price would be expected to decline 10%.
|Ark’s estimated vehicle growth for Tesla, as an example|
|Vehicle Production Per Year||500,000||2 million||4 million||6 million||8 million||10 million|
|Cumulative Vehicle Production||2 million||4 million||8 million||14 million||22 million||32 million|
|Wright’s Law Production Double / 10% ASP Cost Reduction||First Double||Second Double||Third Double|
Tesla’s 2020 Average Selling Price was $50,000. Reducing that by 10%, 3 times, gives us an Average Selling Price of $36,450, very close to Ark’s 2025 Average Selling Price of $36,000.
Big assumption: let’s take Tesla’s presumed historical growth rate of 50% year over year growth and apply Wright’s Law. In Ark’s footnote #3, they allude to the fact that Elon said he believed Tesla could achieve a growth rate “meaningfully above 50%” for 2021. For Tesla to achieve 10 million vehicles a year from a base of 500 thousand, Tesla would need to grow vehicle production by 82% every year for the next five years.
|Tesla’s estimated vehicle growth is 50% year over year growth|
|Vehicle Production Per Year||500,000||750,000||1,125,000||1,687,500||2,531,250||3,796,875|
|Cumulative Vehicle Production||2,000,000||2,750,000||3,875,000||5,562,500||8,093,750||11,890,625|
|Wright’s Law Production Double / 10% ASP Cost Reduction||First Double||Second Double|
With a 50% presumed growth rate, Wright’s Law kicks in twice, reducing Average Selling Prices to $40,500. Vehicle production of 3.8 million is close to but below Ark’s bear case, and the ASP is below Ark’s ASP of $45,000. This is conservative, and even if Tesla only got to 3.8 million vehicles at an ASP of $40,500 in 2025, it would be a smashing success. If Tesla achieved 50% growth, we could expect $153 billion in 2025 vehicle revenue (vehicles produced times Average Selling Price). Although Tesla has grown at a mad clip, can it sustain 82% growth for the next five years? In my mind, it’s possible, but unlikely.
Multiple factories need to be in flight or planned for immediately after Giga Austin and Giga Berlin are complete. Giga Shanghai needs to continue to scale.
How long is the ramp up time to maximum capacity for each factory?
What is the maximum capacity for each factory?
How much can Tesla really grow year over year?
The most amazing figure of all is the $327 billion in modeled Autonomous Ride-Hail Revenue, if Tesla launches a robotaxi service. Let’s assume 70% of cumulative vehicle production is enrolled in the Tesla Robotaxi service, the upper end and most generous value in Ark’s model. Additionally, Tesla gets 30% of each mile, and the Robotaxi charges $1 per passenger mile. For simplicity sake, no fees are charged for traffic.
|Autonomous Ride-Hail Revenue (A)||2025 Cumulative Vehicle Production (B)||70% Participate (C = B x 70%)||Robotaxi Annual miles (D = A / C)||Tesla + Owner’s Cut: Annual miles (E = D / 30%)||Passenger Miles Per Day (F = E / 365)|
Similar to Apple’s App Store, let’s say Tesla gets 30% of the cut for each passenger mile, and to keep the math easy, let’s assume Tesla charges $1 per passenger mile.
There will be 22.4 million vehicles that participate, based on a cumulative vehicle production of 32 million and 70% participate as a robotaxi. If we divide 327 billion in Ride-Hail Revenue by 22.4 million vehicles, each Robotaxi would only need to have 14,598 passenger miles per year. But, Tesla has to pay the owner.
We divide 14,598 passenger miles by 30%, which shows us each robotaxi has to drive 48,670 passenger miles per year, with 70% of the miles to the owner, and 30% of the miles to Tesla. Taking 48,670 passenger miles per year and dividing by 365 days, in Ark’s model, each robotaxi would need 133 passenger miles per day to reach their revenue figure. That seems high. As a personal example, I live 20 miles from work. Due to traffic, it takes one hour. Rush hour and lunch time are the most productive times of the day for a robotaxi, when people are out and about. The rest of the time, the robotaxi is milling around or charging, picking up small trips here and there.
Tesla can charge fees for traffic. It can double up on the passengers to compensate. Or it can increase the base rate charged per mile. Even at $0.50 per minute in traffic, my cost would be $40 one way. Round trip, my cost doubles to $80. Over 200 days, that would be $16,000. That’s much higher than the cost of my Tesla Model 3, insurance, and maintenance. Ark’s bull numbers heavily rely upon Tesla reaching 10 million vehicles to drive robotaxi penetration.
What happens if Tesla only grows 50% per year?
|Autonomous Ride-Hail Revenue (A = C X D)||2025 Cumulative Vehicle Production (B)||70% Participate (C = B x 70%)||Robotaxi Annual miles (D from above)||Tesla’s 30% cut — Annual miles (E from above)||Productive Miles Per Day (F from above)|
An Autonomous Ride-Hail Revenue of 121 billion is a big number. Huge, honestly. It’s about 1/3 of Ark’s Bull Case, and significantly higher than Ark’s Bear Case. How much revenue Tesla can pull in depends on how Tesla prices the service and when the service is ready.
What happens to black cabs, Uber, Lyft, Didi, Ola, and all the rest?
Under what scenarios does a robotaxi not make sense?
Will robotaxis co-exist with regular ride-hailing services?
In our exercise, a large portion of every Tesla produced in history will need to have FSD purchased and enabled.
FSD cost is currently not a major part of Average Selling Price.
Elon has mentioned FSD will go up significantly in price once FSD is complete. Based on the above numbers, it makes lots of sense to put your vehicle in the robotaxi pool. This will drive up the price for old and new Teslas, offsetting the vehicle cost reduction from Wright’s Law.
Will private sale of Teslas to individuals stop or be reduced? If Tesla does that, it realizes all of the benefit of the robotaxi, and it needs a far lower number of passenger miles per day to be viable, which increases the chance of success. It would be a gutsy move.
Ark, there is a path for Tesla to get to 10 million vehicles per year by 2025 and $327 billion in robotaxi revenue in Ark’s bull case. To do that, though, many things would have to go right, including production growth, FSD becoming feature complete, many people purchasing Tesla vehicles with FSD and adding it to the robotaxi fleet, and high passenger miles per day from the robotaxi fleet.
I can’t rule it out as happening in this crazy world, but given all of the above, it seems unlikely and low probability. The case I laid out would not result in Tesla being worth $4 trillion a year, but it would be more valuable than the $628 billion it is worth today. That’s the reason many people are hanging on. Elon is late on his promises, but he delivers eventually. Ark’s bear case seems more likely to be achieved by Tesla, though. I would assign it a much larger probability than Ark’s bull case.
Note: The author owns 3 shares of Tesla as of today’s analysis. The author has no plan to add or buy Tesla shares for the next 72 hours.