In Friday's e-mail, I analyzed Tesla's (TSLA) latest earnings report, which had triggered a 14.5% sell-off in the stock on Thursday. And I concluded:
I would never bet against [CEO Elon] Musk and his team. They've created enormous value by achieving things that boggle my mind. As I've said before, humanity owes them a debt of gratitude.
But with most of Tesla's fundamentals trending downward and the stock still trading at an extreme valuation, I wouldn't touch it with a 10-foot pole...
That said, one of the most important skills an investor can have is being open to hearing the other side of the story. So I'm always seeking out facts and opinions that might cause me to change my mind about a stock.
In the case of Tesla, I not only listen to the bulls – I have one on my payroll!
My longtime analyst, Kevin DeCamp, bought a Tesla Model S more than a decade ago and loved it. This prompted him to buy the stock, and he still holds it to this day. As you can imagine, he has made a fortune on his investment. Good for him!
I asked him to share his bull case for Tesla. So I'll turn over the rest of this e-mail to him...
Another Tesla earnings report, more gut-wrenching volatility.
At first glance, it appears that the stock's big drop was justified by declining margins and a big earnings miss, right?
I disagree entirely, and here's why...
Tesla doesn't trade at a nosebleed valuation for its car business, but because it's the leader in real-world AI and robotics. It's barely in the first inning of this next growth phase, and the company revealed some positive news on this front during the earnings conference call.
Even Elon Musk implied on the call that Tesla has a long way to go before its Optimus humanoid robot will be genuinely useful and make a meaningful contribution to the business, but almost all of the new Tesla cars you see on the road today are robots with the capability to fully self-drive with superhuman abilities and reaction times.
Of course, not every Tesla driver has purchased or subscribed to its Full Self-Driving ("FSD") Supervised software, but the majority of owners can choose to do this simply by selecting the option on the touch screen for $99 a month.
I've been using FSD for more than five years and have seen how much better it has become. The version 14 ("V14") update last year and its subsequent improvements have been a complete game changer for my daily driving – even my wife loves it now! My experience with it led me to conclude that FSD would be the next demand driver for Tesla cars – and that is exactly what this last quarter showed.
An incredible 55% of deliveries in North America included an FSD subscription at the time of delivery. On the earnings call, Musk said, "I think for a lot of people, they're actually buying Tesla Full Self-Driving with a car attached, as opposed to a car with FSD... Clearly this is a significant demand driver and as we get approval for FSD in different countries, I think we'll see a similar uptick in demand."
The word is getting out – FSD is the product – and the inflection point is finally here.
Tesla finished the quarter with 1.48 million FSD subscribers – up around 200,000 over last quarter and 56% year over year. The adoption rate is accelerating.
This subscriber count is a worldwide number, but even if we assumed all these FSD users are in the U.S., it's a tiny percentage of the nearly 300 million registered cars and trucks that are on U.S. roads.
With the step-change improvements that the V14 software has brought to Tesla cars, all the other cars on the road are essentially "dumb" ones, with Tesla's being the only "smart" cars on the road, in my view. I believe we are finally at the true "iPhone moment" for Tesla – and the company has barely scratched the surface.
As large as this market is, it's a drop in the bucket compared to the enormous potential size of the robotaxi Transportation as a Service ("TaaS") market that Tesla's unparalleled fleet and data flywheel enables.
Although I've been a bit disappointed at the pace of Tesla's robotaxi rollout, the management team shared important new details on the call that make me more bullish on its near-term prospects...
Ashok Elluswamy, Tesla's AI lead, shifted the focus to the unsupervised portion of the robotaxi fleet, which is currently operating in six metro areas: Austin, Dallas, Houston, Miami, Orlando, and Tampa.
Although Tesla started the robotaxi rollout just over a year ago in Austin, all vehicles had safety monitors in the passenger seats until the end of last year when they started rolling out fully driverless, unsupervised cars in Austin on a very small scale.
Since then, this unsupervised fleet has driven 380,000 miles with an "impeccable" safety record with zero at-fault incidents and has been compounding miles driven at "double digit" rates every week. They are confident that this exponential growth rate will continue through the rest of this year.
Elluswamy took this opportunity to call out the skeptics for Tesla's vision-only approach, saying:
Historically, the so-called experts have always claimed that you need lidars, radars, HD maps, and the entire kitchen sink to drive safely. Here we show that such is not true. You can have safe, comfortable, and affordable autonomy with just cameras. This record should be a huge validation of Tesla's entire AI approach.
Elluswamy said that the time to launch to a new city will continue to trend toward zero, "an end where we operate in entire states as a whole instead of going city by city."
His update made me more confident than ever that Tesla will rapidly scale its robotaxi business, creating enormous value.
The economics on the other side haven't changed. Tesla makes roughly $7,500 in gross profit per vehicle today. Each Cybercab – Tesla's customized vehicle for its robotaxi service – costs about $25,000 to produce at scale and should generate $25,000 to $50,000 per year in revenue at a relatively high gross margin.
Once Tesla clears a safety bar it deems sufficient, it can literally manufacture the entire current Waymo fleet in one afternoon, at a far lower cost. Waymo spent five years getting a head start that Tesla can erase in a single production shift.
Waymo currently completes more than 380,000 driverless miles per day, but Tesla's robotaxi rollout is actually expanding much quicker than Waymo did after it debuted its first commercial operations in Phoenix six years ago. In fact, it took Waymo about three years to expand to a second city.
Back to the recent earnings report... The quarter looked lousy on the surface, but the underlying AI and robotics businesses are rock solid, with zero real competition. I believe they are at an inflection point and will soon produce an enormous amount of value.
Tesla's gross and operating margins were disappointing, but I think the market is smart enough to realize that margins were dragged down by pre-revenue projects like the Cybercab, Tesla semi-trucks, and Optimus robots.
I think most of the negative market reaction is from Tesla reiterating that it expects to spend at least $25 billion in capital expenditures ("capex") for this year. Capex was only $8.3 billion in the first half of the year, which means Tesla plans to spend more than $17 billion in the second half of this year alone – with guidance for a continued increase over the next two to three years.
These are big numbers for Tesla, but peanuts compared with the hundreds of billions of dollars of capex being spent to build AI capacity by the hyperscalers like Waymo's parent company Alphabet (GOOGL) and Meta Platforms (META).
Alphabet announced earnings on the same day as Tesla, and its stock sold off for the same reason – its capex was $44.9 billion during the quarter, causing its free cash flow ("FCF") to turn negative for the first time in the company's history. That's right, Google's capex last quarter was double Tesla's hugely increased capex guidance for this entire year.
It's not too much of a stretch to speculate that, although Elon Musk is the wealthiest person in the world, he might just be suffering from "capex envy." Even the huge, record-breaking SpaceX (SPCX) IPO "only" raised $85 billion – less than one-tenth of what the hyperscalers will spend on the AI arms race just this year.
Turning to the rumored Tesla/SpaceX merger... I've had mixed feelings about this, but the fact that Musk answered questions about it during the earnings call means, in my opinion, that he's 90% likely to pursue it. (Bettors on Kalshi think it's 68% likely to happen by the end of 2027.)
The increasing overlap is undeniable... Grok is now integrated into Tesla cars and is able to make phone calls, search and play music, adjust climate, and set and adjust navigation on the fly. Meanwhile, Cybercabs have Starlink terminals built in, the Terafab (the massive AI chip factory Musk is building) was announced as a joint project, and the data centers in space he has promised need solar panels, to name a few.
Tesla's FSD and robotaxi program alone reminds me of Warren Buffett's $100 billion question: "If you gave me $100 billion and said, 'Take away the soft-drink leadership of Coca-Cola in the world,' I'd give it back to you and say it can't be done."
When you consider how dominant a company that a combined Tesla and SpaceX would be, this becomes the new "trillion-dollar question."
Besides the synergies, the combined company would likely have an easier time raising money for the AI arms race, and its outright dominance would be another reason Musk wants to get this done sooner than later (before a potential Democratic administration gets into power again to block it). But that's a discussion for another day...
Until Tesla proves the robotaxi fleet can scale and produce massive Magnificent-Seven-like cash flow for the big increase in capex plans, this remains a "show me" story.
I'm confident Tesla will get there, but the most important thing to watch will be the growth of the unsupervised robotaxi fleet. Given Tesla's valuation – 141 times next year's estimates, by far the highest among the Magnificent Seven – buckle up. Volatility comes with the territory.
Thank you, Kevin, for sharing your analysis. You capture many of the reasons why I've long thought Tesla was a bad short – there are many open-ended opportunities.
But I'm not persuaded that I should remove it from my "Filthy Five" list of stocks to avoid...
Best regards,
Whitney
P.S. I welcome your feedback – send me an e-mail by clicking here.
P.P.S. My family and I spent two days in Vienna and drove to Salzburg yesterday. On Friday, we spent most of the day at Schönbrunn Palace and saw a concert there in the evening:
On Saturday, we toured all around the historic district, seeing the Belvedere Palace, the Rathaus (city hall), the Parliament building, the Hofburg Palace (including the Imperial Treasury of crowns, robes, etc.), ate Vienna's famed Sachertorte cake, and capped the day with another magnificent concert and McFlurries:
