Level setting on Austin. The robotaxi pilot is small, expanding, and most importantly, has had no safety issues. After its mid-June debut with about 15 vehicles, it has expanded to about 35 vehicles. The geofenced zone has about doubled and now covers 42 square miles, which is about half of Waymo’s Austin service area. In May, Musk outlined plans targeting 1,000+ robotaxi vehicles in a few months (and in multiple cities) and indicated the Bay Area (San Francisco) is likely the next launch by the fall.

The key metric and commentary to listen for on the call is the pace of expansion, not the absolute fleet size. While Tesla has only a few dozen cars in one market, versus Waymo’s 1,500+ vehicle fleet operating in several cities, the rate of expansion has been faster than what I was expecting. If Tesla can get a couple hundred robotaxis on the road by the end of September, it would be viewed as a positive development by most investors. Less than 100 would be a disappointment. In short, the market will likely forgive small numbers early on; demonstrating an ability to accelerate deployment is the real win to watch for.

A second focus area on the call will be updated comments about the long-awaited mass market model. Most Tesla watchers, present company included, believe it will be a slimmed down Model Y and could account for 25% of deliveries by full year 2027 and half long-term. In other words, it will be as important as the Model 3 and Y. Earlier this year, management suggested production would begin by mid-2025, but recent reports suggest the project will start production by the end of the year. My expectation is comments on the call will move the target to the first half of 2026, which means the ramp will likely be late in 2026. If Giga Mexico will be the manufacturing hub, that window would be pushed into early 2027 given the factory broke ground in 2023 with site prep and infrastructure, but full construction has been on hold since mid-2024. Austin could serve as an initial manufacturing line given it will likely run on the Model Y framework. Once the company has established product-market fit, they can ramp production at Giga Mexico.

The impact of an affordable Tesla is key to the long-term autonomy story, given it would expand Tesla’s addressable market by opening up a lower price tier. More vehicles on the road means more shots on goal for the autonomy offerings. Musk has indicated a price point under $30,000 for the next-gen vehicle, which would squarely target the mass market segment dominated by the likes of VW, Toyota, and BYD. This move would also counter the narrative of Tesla losing share due to a lack of cheaper models. My take is the company will continue to build investor excitement for the product and stop short of giving any hard details on this earnings call.

When production begins for Cybercab will be woven into the conversation around the more affordable model. In the end, it doesn’t matter if it’s a slimmed-down Model Y or Cybercab, the company needs a more affordable model to be produced at scale by the end of 2026.

Deliveries over the next two quarters will give us little insight into the underlying auto growth due to the tax credit expiration.

The high level is roughly 35% of Tesla’s deliveries occur in the U.S., and an estimated 60% of those currently qualify for the $7,500 credit, which means the policy change will impact about 20% of global deliveries. The expiration at the end of September will pull forward demand into September at the expense of December and March of 2026. In June, deliveries were down 14% YoY. While this was as bad as the 13% decline in March, the result was better than the most bearish whispers expecting down 18%. For September, the Street is expecting deliveries to be down 6% YoY. I expect it will be closer to flat. The math behind that prediction is sales for buyers who qualify for the credit in the U.S. will be up 20% YoY in September.

For December, the Street is looking for flat YoY. I believe it will be down 5%. By March of next year, the Street is expecting a rebound to 10% YoY. I am currently at down 3%, which assumes no affordable model in the March quarter.

Despite the focus on autonomy, profitability will remain a key topic. In March, automotive gross margin excluding credits was about 12.5%. The Street is expecting 13.6% in June. There are several puts and takes on margins: continued pricing pressure (especially in China and on older Models S/X) likely weighed on margins, but lower raw material costs and factory scale efficiencies provided some relief. Tesla also benefited from a higher mix of Model Y (its best-margin vehicle) and some one-time FSD software revenue recognition that should be a positive for automotive gross profit.

Going forward, autonomy is the wildcard for margins. If Tesla can break even on selling cars (65% of revenue today) but then make money on FSD and robotaxi at scale, profitability will improve dramatically. I’m going to hold off on putting an exact margin target out there until after earnings. Management has often emphasized that the lifetime value of an FSD-enabled Tesla (through software subscriptions or robotaxi usage) matters more than the manufacturing margin.

A new focus area among Tesla watchers: “hardware anxiety.” As Tesla pushes toward full self-driving, the capability of its in car computers has become a talking point, more important than battery range. The reason its becoming a bigger topic is there are round 1.4 million Teslas in the U.S that run on the HW3 computer, that won’t be able to run full autonomy. As for HW4, there are about 1.3m on the road in the US today, which is good enough to run the company’s spruced up Austin robotaxi version, but in my view won’t be enough to run more advanced models coming out next year. Musk has said the next-gen HW5 chip will likely launch in early 2026 and be 10× more powerful than HW4, which he believes is what’s needed to achieve scalable full.

Why this is important: As more Tesla buyers focus on hardware, more will wait for HW5 which further complicates the delivery outlook over the next year.

Disclaimer