The big surprise of the quarter is the company pulled forward the timing of the next generation vehicle by about 9 months.  The shareholder deck announced:

We have updated our future vehicle line-up to accelerate the launch of new models ahead of our previously communicated start of production in the second half of 2025.

Later on in the earnings call Elon commented that he expects production to begin early 2025 or potentially late 2024. There are three key points on the topic.

1. While it’s likely the company meets that aggressive start of production target, it’s worth noting there is a difference between initial deliveries and ramping to scale. For example, the Cybertruck and Semi have both started production but have yet to begin the ramp out. For Cybertruck, the lag to ramp is at 5 months and counting, and the Semi is at 16 months and counting. Our sense is to expect the next-gen vehicle production to ramp early in 2026.
2. The company was vague about if the next gen vehicle is the human-driven Model 2 or the autonomous Robotaxi. They mentioned, “These new vehicles, including more affordable models, will use aspects of the next-generation platform, as well as aspects of our current platforms.” Elon added that we will hear more about the $25k vehicle at the August 8th Robotaxi event. When we put it together, we expect the next-gen vehicle is in fact the Robotaxi. It’s worth noting, investor consensus expects it to be Model 2.
3.  Tesla will use existing manufacturing capacity to build the new model, instead of building Giga Mexico first. This is important because it shows Elon’s willingness to sacrifice long-term margin gain for near-term sales. If the company had stuck to its original plan and first built Giga Mexico, the “unboxed” manufacturing process would likely have been more advanced, efficient, and profitable compared to adjusting the existing Giga Texas lines. That said, by using Austin, they can get a car to market about a year earlier, which is critical when the company looks to compete with more EV competition and change the automotive narrative to autonomy. Separately, we’re encouraged by the change in approach given it’s a sign that Tesla has a start-up-like nose for revenue.

Our take: What matters is when the next-gen vehicle production ramp starts. If it is early in 2026, it would be a win for Tesla investors.

Going into the March results, investors were bracing for the worst when it came to delivery guidance for 2024, and for good reason. Deliveries in March were down 9% y/y based on high interest rates looming, and a weakened EV market, which suggested that the full year would finish down around 5%-10%.

When asked, Elon was quick to downplay lower y/y sales in March by saying he thinks sales will grow in 2024 over 2023. Additionally, he believes the rebound is already on the way, with June shaping up to “be a lot better” than March. Translation: deliveries will still be down in June, likely 4%, which is an improvement from down 9% in March.

In stress testing Musk’s expectations for growth this year, Tesla will need to grow deliveries in the September and December quarters by 8% and 8.4%, respectively, to end the year up 1%. Our take? The EV macro is weak and near double-digit growth in the back half of the year is a stretch. We continue to expect deliveries and sales to be down around 5% in 2024.

Investing in Tesla is going to come down to the success of autonomy. Tesla can build a more profitable EV business than any traditional automaker, but to have meaningful upside to the company’s current $550B market cap, it would require a step function in profitability. The key to making that jump is solving autonomy.  Elon made it clear on the call:

We should be thought of as an AI robotics company. If you value Tesla just like an auto company, … it’s just the wrong framework.

If you ask the wrong question, then the right answer is impossible. So, I mean, if somebody doesn’t believe Tesla’s going to solve autonomy, I think they should not be an investor in the company.

If Tesla cracks the code in the next few years, it will build a highly profitable business on its own (FSD subscriptions, fleets) as well as winning other car makers as customers licensing the tech.

This dream is becoming more of a reality lately. A year ago, Tesla revised its approach to FSD training, and the results have been favorable based on the latest 12.3 software update that came out in March of 2024. As a result, miles driven with FSD surpassed 1.25B miles, with 300m miles being driven on the new FSD v12 Beta alone. That means the growth rate of FSD cumulative miles driven in Mar-24 grew 73%, a sharp acceleration from up 56% in Dec-23, and 64% in Sep-23. Accelerating growth rates off of a higher base is a sign that something is working.

Additionally, Elon said there is good chance Tesla will sign an FSD license agreement with one or two car makers by the end of this year as conversations have started. That said, they believe it will be 2-3 years before one of those cars is on the road given the hardware work the OEMs will need to do.

We see the biggest risk to autonomy to be the regulatory hurdles. Elon brushed away any concerns, even in China, stating that some states have already laid the groundwork, and this can be applied to the rest of the states. Despite Musk’s comments, we see lawmakers as a major factor in when production of the Robotaxi will ramp.

The call had several other takeaways worth mentioning with respect to their future business segments:

Ridehailing: Elon provided more context on the ridehailing service. He compared it to a mix of Uber and Airbnb and the approach is two-fold, with some number of vehicles owned by Tesla and the rest owned by the end user or third-party businesses. The individual user can add and subtract the car to a fleet like an Airbnb and Tesla will have a portion dedicated to driving and computing. Elon is saying there could be 10s of millions of cars in the fleet. Our take: Look out Uber and Lyft.

AWS-like Utilization of Vehicles: Elon also talked for the first time about an AWS-like opportunity with AI distribution to stationary cars. Hardware 4 will be in cars by the end of next year and there is a potential for cars to receive AI-distributed inference. The idea is to efficiently utilize vehicles outside of the ~10 hours of driving time per week.  Elon says the fleet will be used to compute AI. Our Take: This sounds nice, but it won’t move the needle in the next decade.

Optimus: Optimus is performing simple factory tasks and is believed to be able to perform some factory production tasks by the end of the year. Additionally, they think there’s a chance to sell Optimus externally by the end of next year. Increased production efficiencies from a sentiment humanoid robot increases profitability, boosting auto gross margins ex regulatory credits back to “tech company” levels. Our take: Optimus is a long shot, likely to take a decade before its commercialized at scale. That said, if it works, it could create a business that is bigger than Tesla’s EV and autonomy business combined.

Disclaimer