Sections e, f & g are for paid subscribers. They include detailed data and commentary on guidance, valuation, the shareholder letter and the conference call, as well as my take on the quarter.

If you'd like to read those sections and full reviews on 40+ companies this season (and so much more), you can access 25% annual discounts and upgrade below.

Upgrade

In case you missed it:

South Korean growth remains robust, as that market is now bigger than every other country besides the USA and China. Deliveries in that nation, Taiwan, Japan and Singapore set new records. Across Europe, Model Y has been the top-ranked vehicle in Norway, Switzerland and Iceland so far in 2025. And momentum appears to be healthy, as this car became the top seller in Finland this quarter and the top seller in the Netherlands and Denmark last month.

Lower regulatory credits Y/Y and lapping higher full self-driving (FSD) revenue recognition were both revenue headwinds. The FSD factor is related to last year’s Cybertruck FSD release and popular feature debuts like “Actually Smart Summon.” As Tesla reaches product milestones, it recognizes deferred revenue from previous FSD payments as revenue. It reached an abnormal number of milestones last year, leading to the tough comp.

The EBIT and EPS misses were related to operating expenses (OpEx) coming in 17% higher than Bloomberg consensus estimates. R&D and SG&A were 6% and 13% higher than anticipated, respectively. That was heavily tied to AI, while legal fees added to the SG&A number. There was also $238M in restructuring and other expenses (within the chip design team), which was larger than a few of the buyside analyst estimates I have access to. Without that item, EPS would have been closer to in line with expectations and the EBIT miss would have been below 1%.

“Our employee-related spend is increasing, especially in R&D, as we have recently granted various performance-based equity awards to employees working on AI initiatives. Such spend will continue to increase going forward.” – CFO Viabhav Taneja

Tesla reiterated its $9B annual CapEx guide. It added that CapEx will meaningfully grow in 2026.

Tesla trades for a whopping 225x EPS. EPS is expected to fall by 30% this year and compound at a 33% clip over the next two years.

Auto Product Line-up:

Earlier this month, Tesla unveiled a new version of the Model 3, which starts at $36,990 and is more than $5000 cheaper than the previous iteration. It also launched a new Model Y “Standard” vehicle, which starts at $39,990 and is also about $5000 cheaper than its predecessor. In both cases, the battery life is still over 300 miles and the cars come with many of the popular features that its more expensive vehicles do.

These products come as Tesla pushes to remain affordable following the September 30th expiration of the $7500 EV tax credit. It’s important to note that while this expiration will hurt demand for all EV makers, Tesla’s relative margin advantage vs. other EV programs, its EV scale and its ability to release these cheaper cars all put them in good shape to weather the blow. You can therefore argue that this will hurt its competition more than Tesla and will be a net positive for the company long-term. 

It will be interesting to see how the combination of this tax note and the cheaper models nets out in Q4 results. For now, it’s important to note how this expiration helped demand during the quarter. Many consumers rushed to buy cars before the overall cost of that purchase effectively rose by $7500. This helped overall auto revenue and its market share in North America. While it’s great to see that market share rise Q/Q, this is still imperative context. It creates risk of this rise being short-lived. We shall see. 

In China, it launched the Model YL. This is a 6-seat, 3-row car, which gives it a much better product for the family car category there. Some Chinese news outlets recently reported that this is already the 4th-ranked SUV in that important market.

Energy Business:

This segment remains the shining star of Tesla’s financials. It’s booming. The Megapack 3 (commercial battery storage system) is set for production in Houston late in 2026 and will be able to deliver 50 gigawatt-hours (GWh) annually. The new Megapack will be equipped with a “Megablock” that can connect 4 Megapack 3s, switchgear and a transformer into one unit. This is expected to simplify construction, improve energy density and help cut hardware connections by nearly 80%. That goes a long way in giving Megapack 3 more than 90% round-trip efficiency (energy output / energy input), which is roughly in-line with best-in-class competitors. As Musk told investors, the USA has one terawatt of power continuously available, yet only ends up using about half of that. Simply upgrading the grid with better storage systems like these can nearly double energy availability with the exact same base of power plants. It’s a massive opportunity and should be a lucrative one for Tesla and the field.

AI, Full Self Driving (FSD) and Robotaxi:

Version 14 of supervised FSD debuted this month. There are some reports on the software being a bit rough around the edges, but Musk reminded us that’s always the case with new versions. They’re confident in their ability to “smooth the rough edges” with 14.2. 

This new iteration of FSD makes a lot of the Robotaxi AI model work available for consumers updating to the latest software. Per the team, this update also brings better handling of fringe and complex use cases – including road debris maneuvering.

The robotaxi service is growing nicely in Austin and debuted in San Francisco this past quarter as well. There are still safety drivers in both markets, but Tesla hopes to remove them in Austin before 2026. This isn’t a regulatory issue. This is Tesla being extra cautious with rollouts, knowing the high-profile nature of their brand and founder means any autonomous blunder will be placed under a microscope and endlessly scrutinized. Fair? Maybe not. But reality. Leadership thinks the vertically integrated and extensible model will allow them to expand to many more cities over time and nearly 10 metro areas by the end of 2025; the plan is for phase one of those launches to all entail safety drivers. 

Finally, Musk told investors that Tesla’s growing confidence in FSD’s viability is leading him to prioritize maximizing production. They think the cars will sell themselves whenever they finally are equipped with unsupervised FSD. They also now think they have enough clarity on that and regulations. The company is willing to risk some inventory bloating to make sure they have the supply to meet the future demand they see ahead.

“I think it's important to emphasize that Tesla really is the leader in real-world AI. No one can do what we can do with real-world AI… I think that Tesla has the highest intelligence density of any AI out there in the car, and that is only going to get better.” – an always charismatic Elon Musk

Samsung:

There was some investor confusion surrounding the $16.5B Samsung chip deal to manufacture Tesla’s planned AI5 chip. This is the successor to its AI4 chip that debuted in 2023 and was manufactured by Samsung. The AI5 chip will be manufactured by Samsung in Texas as well as Taiwan Semi in Arizona (initially Taiwan). The company is happy to overbuild supply for this product, as they’ll be able to reallocate that compute to data centers if Optimus and their cars don’t need it anymore.

Musk has been personally spending a lot of time with the AI5 design team and is excited about the up to 40x performance improvements it will deliver vs. AI4. This certainly will not replace their reliance on Nvidia, but it should diminish it. Tesla has the luxury of building chips that granularly match their specific needs, rather than the needs of several giant customers. Their chips can be far more granular than Nvidia’s general-purpose GPUs. This, Musk thinks, will mean these chips have best-in-class performance per watt for Tesla’s needs. We shall see.

“The net effect is that I think AI5 will have the best performance per watt, maybe by a factor of 2-3x. I think it will also have the best performance per dollar for AI, maybe by a factor of 10x.” – Elon Musk

Software Updates:

Tesla integrated Grok in North American cars, introduced Low Power Mode and added other tools like in-car Tesla Diner ordering. This really gets into what makes Tesla special. Its cars are big pieces of world-class technology. They’re highly capable computers on 4 wheels, which is more true for them than their non-Chinese legacy competitors. This means they can seamlessly push product updates to consumers without needing to sell them a new model. Not only does that make its product experience better, but it also creates so many opportunities to upsell customers more services and to fatten up overall margin. As they’ve talked about in the past, they’re willing to accept lower hardware GPM (by doing things like cutting new Model 3 prices by $5K) because they’re confident in “harvesting” more software-based profit down the road.

Optimus:

Tesla plans to unveil Optimus version 3 during Q1 of 2026. It’s unclear if this will be the version ready for scaled production, as they’ve talked about in the past. They still also have goals to finish a million-unit production line and begin usage by the start of 2027. 

There is no existing supply chain for any piece of Optimus, so Tesla is responsible for advancing every single facet of it. That will take time, but an always optimistic Musk remains exceedingly bullish on this product, and thinks nobody else matches the data, real-world AI and engineering capabilities Tesla has.

Final Notes:

Another uninspiring quarter from Tesla. Not terrible. But I expect great results from this great company and it has been a few years since we’ve gotten them. I realize tariffs are hurting the sector, but interest rates have also fallen in the last few quarters and Tesla comps have gotten much easier. Unemployment is still very low and economic growth is strong. This should be a reasonably good environment for them. Still, auto revenue and deliveries both were in the mid-single-digit growth range, while Ford outgrew Tesla deliveries this quarter. This is despite the expiration of the EV tax credit that accelerated Tesla demand more meaningfully than for Ford. Simply put, in my opinion, the core auto business does not look amazing. The energy business, where tariffs are arguably hitting harder, does look amazing. Auto doesn’t. 

I know there’s reason to be excited about robotaxis and Optimus, but those products are not ready to carry this financial engine and likely won’t be for at least a couple years. They are promising yet speculative. And in the meantime, this trades for an earnings multiple north of 200x with a multi-year EPS CAGR of 33% expected. It feels priced as though investors think Optimus has already worked and Tesla’s robotaxi business has already scaled. Both of those things remain entirely uncertain.

That leads me to the following 2 conclusions. First, Tesla is a company to be deeply respected and studied. It’s a great American growth story that employs a ton of people. It stormed onto the scene and took a massive piece of the EV pie. Musk is a historically successful (and entertaining) founder. But second? This is not a compelling investment in my mind. Bulls are paying a giant price tag for a somewhat challenged core business based on promised future products that have no guarantee of panning out. I prefer to focus elsewhere and continue rooting for Tesla from the sidelines.