I've seen this counter argument to a stock buy back many times but it's always been a flawed counter argument and doesn't really hold any weight to me
The reality is that a Robotaxi service is at least 1.5-2 years away. I think we can all agree on that. With each Gigafactory, we have an idea of just how much one can add to FCF once it's at volume production. We have two new factories going through that process right now. And Shanghai has a further expansion ahead of it. By the time a Robotaxi fleet needs to be built, and remember a lot of 3/Y's will be coming off lease by then that will be inducted into the fleet, Tesla will likely have 5-6 Gigafactories at volume production. You're talking 5-6 billion in FCF every quarter.
A robotaxi fleet isn't going to need 500,000 vehicles right at the start. It will grow over time. Let's say Tesla wants 150,000 Model 3's to act as a Robotaxi for the first year in the US. That's roughly 5.2 billion for Tesla to build those 150,000 Model's ($50k ASP minus gross margin - 35K cost to Tesla to build).
So just one quarters FCF will cover the start of the Robotaxi service.
But wait........you're forgetting another important aspect. If a Robotaxi fleet needs to be made, well then that means FSD has been solved and FSD is already in wide release, unlocking billions in additional profits and cash. I don't know why people don't get this. If Robotaxi is a thing.....then FSD was completed, unlocking huge levels of profits and cash. You can't have one without the other.
And lastly, it completely ignores the point of the post that I made. In that, it's better for Tesla to use the extra 2 billion in FCF per quarter for stock buy backs, especially when the stock has been cut in half by 50%.
Say Tesla succeeds with FSD, and then needs capital for Robotaxi, then they can do a offering......ya know I think Tesla's market cap just might be a bit above where it is today in a scenario where FSD is achieved
Tesla has 2022 already mostly sold out. The only trims that aren't are the high end 3/Y's. Tesla's order log is about recession proof as possible. Tesla has no debt and 18 billion in cash. Tesla is already prepared for a recession. In fact, I'd say they're overprepared. But that's never a bad thing.
It is bad to just have cash sitting there while you generate 2 billion more in cash each quarter in high inflation environment.
Elon’s frequent comments about the fleet awakening with a bit flip seem to indicate that there is no plan for a long, drawn out Tesla Network rollout if we take them at face value which is usually a safe bet with him since he talks literally. Tesla is all about speed.
We are also living in an unceasingly tragic road safety emergency that Tesla wants to solve ASAP, and besides it’s directly advantageous to the sustainable energy mission to deploy Tesla network as fast as possible to displace ICE miles quicker. If you listen to the passion in Elon’s voice in the Financial Times interview when was talking about FSD safety and he adamantly said “if we believe it will reduce the probability of injury…we will do the right thing and get sued”, you can see how he views the moral importance of FSD’s role in society. If I remember correctly, the FSD program was started because a Tesla driver passed out at the wheel and fatally struck someone riding a bicycle and they decided they should develop active crash avoidance software. Every day of delay is more funerals and more hospital visits that could’ve been avoided.
So the main question is, will the demand be there so quickly? I would predict that a fleet of 500k Robotaxis (3/Y/2024 one) sitting in inventory could almost instantly be valuable and deployed.
The US alone has 326 cities/incorporated areas with at least 100k population.
The average American household is spending $10k on transportation out of average after-tax income of $70k for 14% of their income. For the lower class this percentage is more like 20%-40% because in the vast majority of the USA there is little to no decent mass transit options such that people are left with no good option except to drive cars they can barely afford, leaving them one reliability failure from missing a shift and potentially losing their job. The Tesla Robotaxi is being heralded as the lowest-cost transportation option ever. I doubt many people would be willing to pass up on the savings if it takes the edge off of paying bills and getting frustrated driving through traffic. Once a few people start using the service they’ll rave so hard about how great it is that FOMO will kick in and demand will rise rapidly. As long as the service is safe, convenient, cheap and comfortable I think word will spread fast locally.
Millions more in the middle and upper classes are busy and stressed by the logistical difficulties of shuttling their kids around town between school and activities and would welcome some relief. And teenagers under driving age will love to be able to visit friends or go out without having to be chauffeured by an adult. Most teens are somewhat embarrassed being seen around their parents, can barely afford their own car if not being gifted one by family, and they generally care more about looking cool than worrying about safety so I doubt they’ll be concerned about perceived risk.
We also have the adult party demographic who want to get drunk and not have to drive themselves home. Robotaxi will catch on quickly with them too. Rides will be dirt cheap late at night when overall demand is lower.
If all else fails, delivery companies like Doordash, Instacart, Uber Eats, Grubhub and Leafly could easily soak up the services of several hundred thousand robotaxis probably in a matter of weeks. Doordash alone has an army of 2 million couriers. They could just make a new model where restaurants drop off the food in the car and it heads off to the customer.
At Tesla’s estimated $30k per year profit per car, a Robotaxi would need around 1 year to pay for its upfront expenses from fare profits. So, scaling at a rate higher than 100% would require working capital just for paying for CoGS on each vehicle. But then they’ll also be spending many billions up front on actively expanding factories, installing charger capacity, and developing mines and refining facilities. So the combined effect is that the sheer rate of production growth at 70-80% per year (“or probably faster than that in the future”) will swamp the influx of profits and require a lot of cash, probably at $1-3B per quarter.
Then there’s virtual power plants. Tesla is quietly maturing Autobidder, partnering with major real estate developers, and registering as a utility in Texas and other jurisdictions. What if they make 10 GWh per quarter (Lathrop production) of Megapacks for themselves at $200/kWh? Another $2B per quarter CapEx. Now suppose they 10x solar deployments from today by ramping production and buying up supply from other manufacturers. 1 GW of solar per quarter at $1/W is $1B per quarter. Extreme scale for VPP development needs $3B per quarter at this rate. On top of this, maybe they’ll start buying existing solar and wind assets for in order to squeeze more value out of the with Autobidder and profit off the arbitrage while accelerating the phase-out of fossil fuel plants.
An upcoming Indonesian battery Terafactory could be another $1-2B per quarter too.
Investment opportunities will outpace cash generation for the foreseeable future.
Pierre Ferragu asked on the call about how Tesla will spend $500B cash by 2030, and the response was:
Elon Musk: (47:05)
…That seems like a lot of cash. I don’t know. We’ll try to do something useful with it. [inaudible 00:47:31] that’s for sure.
Zach Kirkhorn: (47:36)
I think we have to take this one step at a time. We have investments that are happening right now to get Austin and Berlin up and running. And then as Elon mentioned, installing capacity for robotaxi production. And there’s some decisions that, as Elon alluded to, just to share in the future about and what the economic model looks like for robotaxing. And so the way Elon and I have discussed this is-
Elon Musk: (48:13)
Sorry. Maybe just, yeah … Yeah, everyone just mute if you’re …
Zach Kirkhorn: (48:19)
Yeah. So our focus is to get to the point where robotaxis are on the road, Optimus is in use, get the economic model for that dialed in, and then evaluate the size of cashflows at that point and make decisions then as to what’s next.
So it doesn’t sound like share buybacks are on the table any time soon.