Tesla Posted Its Worst Quarterly Sales Drop in 10 Years

Retail traders have christened a new class of meme stocks. Opendoor, Kohl’s, and Krispy Kreme all surged last week.  

This wave is being driven by hype, not fundamentals.

Take Krispy Kreme. The company’s revenue fell 15% last quarter. It paused its store expansion plans and suspended its dividend. But none of that mattered online. Bullish mentions of $DNUT exploded by 3,500% on Stocktwits, fueled by hashtags like #KrispyMoon. 

Opendoor and GoPro, another new meme stock, are struggling, too. Both posted double-digit declines in revenue and roughly $400 million in net losses last year.  

We’ve seen this movie before, and it didn’t end well. During the 2021 GameStop surge, individual investors lost an estimated $13 billion. GameStop is still down 70% from its peak, and 60% of shares are still held by retail investors. “Sticking it to the man” might be more exciting, but sticking to index funds is almost always more profitable.

The bigger story is that young people have a limited menu of real investment options. The S&P is trading at 25x earnings. When my parents were my age, it was closer to 11. Housing is unaffordable. The median home price just hit a record $435,000. And the companies actually driving innovation, like OpenAI and SpaceX, are staying private. We are locked out of the upside.

So what do we do? We invent our own asset classes. Crypto. Meme stocks. Anything that offers a chance at returns we can’t access through traditional paths.

But this isn’t about taking on Wall Street. It’s not us versus them. It’s us versus us. Every time someone buys in, someone else is looking to cash out. We are not investing in real businesses. We’re playing a game of musical chairs where the only way you can get rich is if you sell at the peak. 

Oracle and OpenAI are scaling up their AI infrastructure, announcing an additional 4.5 gigawatts of new power capacity for their data centers. For context, 1 gigawatt is enough to power roughly 750,000 homes in the U.S. The move builds on the Stargate initiative, the $500 billion project announced back in January.

Oracle’s cloud infrastructure unit now makes up 43% of its total revenue. It grew more than 50% last quarter, and that growth is expected to accelerate to 70% this year. Investors are paying attention. The stock is up 46% year to date, and Oracle now trades at 56x earnings, roughly equivalent to Nvidia.

But Oracle is only one piece of a much larger trend.

To understand what’s happening, start with where AI actually runs. Not on your phone or laptop, but inside massive data centers, packed with GPUs, cooled by industrial systems, and powered by electricity-intensive servers. These are the steel mills of the 21st century.

Big Tech is pouring money into them. Microsoft, Amazon, Meta, and Google combined are on pace to spend roughly $340 billion on capex this year — more than the GDP of Finland. 

The AI boom has become an infrastructure boom. Case in point: Capital expenditures on AI data centers are already more than the peak in telecom spending during the dot-com bubble and the buildout of 5G networks as a percentage of GDP. 

Larry Ellison took Oracle, a mature company that had been returning capital to shareholders, and made a bold shift. He reduced profits in the short term, ramped up capital spending, and bet big on AI infrastructure. That kind of move requires conviction. 

Then you have Sam Altman, who is playing a different game. He’s keeping OpenAI laser-focused on the front end of the technology — the user experience, the branding, the product. His goal is to make OpenAI the default name in AI, the way Kleenex is for tissues or Xerox was for copies. To make that possible, he’s offloading the infrastructure to partners like Oracle. He knows what OpenAI should specialize in, and he’s not letting anything distract from that.

OpenAI is now having real problems with Microsoft. They’re starting to realize that giving up so much control may have been a mistake. 

A recent example makes this clear: OpenAI tried to acquire a company called Windsurf, but the deal fell apart when Microsoft stepped in and demanded full rights to the IP. That didn’t make sense, especially since Windsurf was supposed to compete with Copilot. It was a wake-up call. OpenAI had effectively partnered with a company that is, or soon will be, a direct competitor.

That’s why they’re turning to Oracle. They need a partner, not a gatekeeper. Someone who can support them without competing for the same prize. Oracle fits that need.

And this is exactly what I said last year. At the time, Oracle was undervalued. Now it’s clear to everyone that it’s a serious player in AI. I’ll say it plainly — I was right.

____________sponsored content ____________

Stop wasting time on repetitive explanations. Guidde’s AI creates stunning video guides in seconds—11x faster.

How it works: Click capture on the browser extension, and Guidde auto-generates step-by-step video guides with visuals, voiceover, and a call to action.

Best part? It’s 100% free.

👉 Download the extension now

____________sponsored content ____________

Second quarter earnings from GM and Tesla rolled in last week, and both were grim. GM took a $1.1 billion hit from tariffs, and Tesla posted its steepest drop in quarterly revenue in over a decade. 

GM had already warned about tariffs, lowering its full-year guidance in May to account for a $4–$5 billion hit. Unlike many companies, GM hasn’t raised prices to offset the cost — a decision that preserves volume but led to a 35% drop in net income, meaning the tariff burden falls squarely on shareholders. GM has delivered just a 2.6% annualized return over the past 15 years, compared with 11.8% for the S&P 500.

Despite the headwinds, GM’s EV business is showing signs of life. Electric vehicle sales more than doubled in Q2, even as GM’s overall U.S. sales fell 2%. The company now holds 14.9% of the U.S. EV market, up from just 0.3% three years ago.

Tesla had a rough quarter across the board. Revenue dropped 12%, its steepest decline in over a decade, and net income slid 16%. Free cash flow collapsed to just $146 million, an 89% drop year over year and far below Wall Street’s $760 million estimate. 

More troubling: Tesla’s profits are increasingly decoupled from its core business. The company posted $923 million in operating income this quarter, but only about a third of that came from selling cars. The rest came from selling regulatory credits, earning interest on cash, and gains from its stash of bitcoin.

One bright spot: Tesla’s “Services and Other Revenue,” which jumped 17%. That growth came primarily from its Supercharging network, which added more than 2,900 new stalls — an 18% year-over-year increase.

The difference between net income and free cash flow? 

Net income is profit based on accounting rules, called GAAP in the U.S. It includes noncash items like depreciation: For example, if a company buys a $1 million building, it might count only $50,000 per year as an expense, even though the full $1 million was spent up front.

Free cash flow shows the actual cash left after paying for things like that building. So in that same year, free cash flow would drop by $1 million, even though net income shows only a $50,000 expense.

Net income allows companies across industries to be compared consistently, while free cash flow is key to understanding what a company can actually spend or reinvest. 

According to Goldman Sachs, about 70% of tariff costs will eventually get passed through to consumers — but it’s going to be on different timelines. If you’re selling cars, which are big-ticket items, you’ll likely delay price hikes. If you’re selling lower-priced goods, you’ll raise prices faster. And that’s exactly what’s happening: Walmart has already said it’s raising prices due to tariffs, and so has Best Buy. 

So the sequence is: Earnings take the hit first, pricing follows. GM’s in phase one. Phase two is coming.

Elon Musk is known for ambitious promises that rarely arrive on time — full self-driving by 2016, trips around the moon by 2018, humans on Mars by 2026. But last week, he finally made good on one. In 2018, he floated the idea of a restaurant and drive-in theater at a Tesla Supercharger in LA. Seven years later, it’s real.

Tesla opened a 24-hour diner in Hollywood with 80 charging stations and two 66-foot movie screens. It serves American comfort food in Tesla-branded packaging, and if it does well, Musk says he’ll bring more to “major cities around the world.” 

Ironically, the menu is … “woke.” Substacker Rebecca Thimmesch wrote (emphasis ours):

“At face value, it’s quintessential conservative: burgers, hot dogs, grilled cheese. … Upon closer inspection, it’s actually giving quite woke, quite DEI. The menu is dripping with “organic,” “free-range,” and “wild-caught” labels. …The tortillas for the breakfast tacos come from the Tehachapi Grain Project, which sounds like something the official DOGE account would call gay on Twitter and then take away their USDA grant.”

Tesla’s foray into food might sound like a gimmick, but there’s precedent. Ralph Lauren, Ikea, and RH (formerly Restoration Hardware) have all turned restaurants into high-performing brand extensions.

What do these success stories have in common? They all create an immersive brand experience. The Polo Bar, RH’s rooftop cafés, and Ikea’s food courts all reflect the core identity of each company: luxury and exclusivity, dramatic design, and Swedish hospitality. 

These venues are also visually striking, designed to encourage photos, social sharing, and word of mouth. As people share the experience across social channels, the brand benefits from organic buzz and increased clout. 

They’re also integrated with the core business. Ikea’s café extends dwell time and encourages shopping; Polo Bar and RH both blur the line between showrooms and leisure. 

The Tesla diner hits many of the same notes. It's visually striking, social media ready, and plugged into the product experience, turning idle charging time into a memorable, positive brand experience. 

Tesla is, in my opinion, the most inflated bubble in the world right now. Auto sales are down 16%, and yet the company trades at 180x earnings. Those two things can’t coexist.

What you have with Tesla is a CEO who’s a genius at the intersection of capital markets and narrative. Musk understands that a massive part of shareholder value is perception. That 180x P/E isn’t about fundamentals; it’s a function of his ability to convince the market that Tesla is still a disruptor, a market leader, an AI company.

So he says: “Look over here — we’re building AI, we have a diner, a flamethrower, a tequila brand. We’re not just a car company.” They’re doing everything they can to signal innovation, to suggest there’s always something else coming.

To be clear, this is brilliant marketing. It’s getting a ton of press — and honestly, I want to go to one. 

But, the reality? It doesn’t have any core competency that suggests they’ll succeed at what In-N-Out has spent more than 70 years perfecting.

Tesla’s diner solves two problems. First, it helps restore the brand’s halo, which dimmed after Musk became deeply involved in politics. Second, it tackles one of the biggest friction points for EV buyers: waiting to charge. Tesla’s diners make that downtime feel less like a hassle and more like a perk.

Tesla used to attract coastal environmentalists and tech optimists. Now it’s the ultimate boys’ club. What unites liberal Californians and libertarian truckers? Food. The diner is somehow both nostalgic and futuristic — and, crucially, apolitical. It channels the best of Tesla while sidestepping its most polarizing association.

It’s a smart way to rebuild a more unifying Tesla brand. Honestly? It will probably work, unless Musk sabotages the whole project with more drugs, broken promises, and political distractions.

Alphabet is going to outperform the market and Big Tech over the rest of the year. 

Supposedly AI is an existential threat, but search was up 12% and YouTube was up 13% this quarter, and Cloud is growing like crazy. Still, it trades at a P/E multiple of 20 versus 25 for the broader market. Is Alphabet a below-average company? Would you rather own Dow Inc. or P&G? The answer is no. 

In last week’s Office Hours, Scott shares how he chooses who to mentor, and makes the argument for never getting married or having children. Listen here.