All week, the headline ran the same way: "Tesla posts record revenue."

$28.2 billion in Q2, up 26% year over year. A record 480,126 deliveries. By any surface measure, one of the best quarters in the company's history.

By Thursday morning, TSLA was down 14.5%, its largest single-day decline in over a year, shedding more than $140 billion in market cap in a single session.

The market was not confused by the headline. It was confused by what was sitting underneath it.

Tesla's Q2 operating income was $398 million on $28.2 billion in revenue. Its operating margin was 1.4%. The company spent $5.79 billion in capital expenditure in Q2 alone to get there.

Today's One Big Idea is about whether $311 is an opportunity or a trap going into this week.

Missed Friday's deep dive on Google, where a $119.8 billion quarter sent the stock lower for a very similar reason?

The bigger point is that the AI trade is starting to broaden beyond Nvidia and the usual chip giants. A recent MarketWatch piece made a similar case, highlighting S&P 500 companies that are using AI deeply inside their operations rather than just talking about it, including examples like Visa with 26,000 employees using AI tools and more than 100 internal AI applications. In plain English, the next AI winners may not only be the companies selling chips, but also the firms quietly using AI to improve margins, productivity, and customer experience.

Caterpillar’s rally is not just about construction equipment anymore. Investors are starting to price in the company’s exposure to data centers, backup power, energy infrastructure, and the heavy machinery needed to support the AI buildout. The real takeaway is that Caterpillar has become a less obvious AI infrastructure play, but after the stock doubled in a year, the risk is that a lot of that good news is already reflected in the price.

AMD and Cerebras are trying a different approach to AI inference by splitting the workload: AMD’s Helios rack-scale systems handle prompt processing, while Cerebras’ Wafer-Scale Engine handles token generation. That matters because inference is becoming one of the biggest bottlenecks in AI, especially as agentic AI needs faster, lower-latency responses. With Microsoft Azure expected to deploy Helios across data centers in the second half of 2026, this could become a serious test of whether specialized hardware partnerships can challenge the one-size-fits-all GPU model.

Every manager faces the same situations before lunch: a message to land, a meeting to run, a hiring call, a report due. The AI Report built 37 Claude prompts for exactly those moments, organised by the situations every manager faces. 

Copy the prompt, fill the brackets, run it in Claude, and get back 10+ hours a week. Oh, and it's free. 

All you have to do is subscribe to The AI Report, a 5-minute daily AI brief read by 400,000+ business leaders at IBM, AWS and Microsoft, and the full prompt pack lands in your welcome email. The newsletter and the prompts, both free. Subscribe and grab both

Subscribe and get 37 prompts free

Tesla (TSLA) posted Q2 results that look like everything bulls have been waiting for. $28.2 billion in revenue, a Q2-record 480,126 vehicle deliveries, and trailing twelve-month revenue above $100 billion for the first time in the company's history.

The obvious read: demand is not the problem.

That reading is correct. What it misses is that demand at 1.4% operating margins is barely worth more than breaking even on.

Start with what is genuinely working, because it is genuinely working.

Revenue beat analyst estimates of $25.71 billion by more than $2.5 billion. Automotive revenue grew 23% to $20.52 billion. Services and Other grew 50% to $4.58 billion — wait, no em-dashes. Let me fix:

Revenue beat analyst estimates of $25.71 billion by more than $2.5 billion. Automotive revenue grew 23% to $20.52 billion. Services and Other grew 50% to $4.58 billion, the line item where robotaxi revenue will eventually show up.

FSD attach rate hit 55% in Q2. More than half of all Tesla buyers are now paying for the autonomous software. Robotaxi pilot operations have expanded into six US cities. Optimus production begins at Fremont this month or next, targeting 1,000 units for 2026.

That is a legitimate bull case. The autonomous and robotics future is arriving faster than it was priced six months ago.

Tesla's Q2 operating income was $398 million, a 57% decline year over year on a record revenue quarter. The operating margin is 1.4%.

Here is the number that stopped the market cold: Tesla's Q2 capex was $5.79 billion, up 142% year over year. The car business generated operating profit that covered less than 7% of what the company spent on its future in the same quarter.

Non-GAAP EPS came in at $0.33, missing the $0.54 consensus by 38%. Free cash flow went negative for the first time in years at negative $1.09 billion. Tesla then raised full-year 2026 capex guidance above $25 billion and said spending will increase for the next two to three years.

Not because Tesla is losing the EV race. Because Tesla is running the EV business at near-zero margins to fund autonomous vehicles and humanoid robots that have not yet demonstrated commercial returns.

TSLA was trading near $360 heading into the July 22 earnings release. The stock fell roughly 14.5% on July 23, and spent the two sessions after still drifting lower, settling near $311.40 heading into the weekend. That is less than $14 above the 52-week low of $297.82.

The drift is the detail worth watching more carefully than the initial drop. When a stock falls 14.5% on earnings and still cannot find buyers in the two sessions that follow, the market is not overreacting. It is still doing the math.

The level that matters most going into this week is $297.82. If it holds, a recovery toward $340 to $350 is the base case for bulls who believe margin deterioration is temporary and autonomous revenue is two to three quarters away. If $297.82 gives way, the conversation shifts from "buying the dip" to "what multiple does a car company with 1.4% operating margins and negative free cash flow deserve at a $1.18 trillion valuation," and that is not a conversation that resolves in the bulls' favor quickly.

Tesla's gross margin fell to 16.8% in Q2, well below analyst estimates near 19.4%. Watch for whether Q3 shows any recovery. A move back toward 19% changes the profitability conversation immediately and gives investors a foundation to hold through the capex cycle. If gross margin stays flat or falls further, the bull case requires robotaxi and Optimus revenue to close a gap the car business cannot close on its own.

Cybercab is running in six US cities. Watch for whether any material robotaxi revenue shows up in the Q3 Services and Other line item. Tesla is spending $25 billion per year on autonomous infrastructure on the assumption that monetization will eventually justify the cost. The moment actual revenue becomes reportable, the thesis shifts from speculative to testable. That is the single most important data point for this stock over the next two quarters.

Optimus production targets 1,000 units for 2026. Watch for any Q3 guidance on pricing, customers, and initial deployments. One credible Optimus revenue number does more for the TSLA thesis than any number of record delivery reports at 1.4% margins. The question is not whether humanoid robots will eventually be valuable. It is whether Tesla can get them into paying customers' hands fast enough to matter to the current balance sheet.

The 52-week low of $297.82 is the most actionable near-term level for TSLA heading into Monday. Watch whether the stock opens the week above that level without a specific positive catalyst. A clean hold signals the post-earnings repricing is complete and value buyers are stepping in. A break below $297.82 opens a conversation about what the correct entry level is for a company whose base business cannot yet fund its own ambitions, and that conversation does not start at current prices.

Let me be direct about the Tesla story this week.

The deliveries are real. The revenue is real. The 55% FSD attach rate is real, and the robotaxi expansion into six cities is moving faster than most analysts predicted six months ago.

The problem is what the business generating all of that revenue is actually earning.

$398 million in operating income on $28.2 billion in revenue. A 1.4% margin. Operating profit that covered less than 7% of what Tesla spent on its future in the same quarter.

Not because Tesla is losing. Because Tesla is running a near-zero-margin car business to fund autonomous vehicles and humanoid robots that have not yet shown up on an income statement.

Markets stay patient when the base business is solid. At 1.4% margins and negative free cash flow, this base business is not solid enough to carry a $1.18 trillion valuation on its own. The autonomous revenue needs to become real in the next two quarters. At $311, $13 above a 52-week low, the market is sending a clear signal about how much patience it has left.

Neither will I.

You don’t have to buy OpenAI or Anthropic to invest in AI. Each large language model requires specialized technology to keep running and growing.

MarketBeat’s The Infrastructure’s Backbone: 10 Stocks Powering the AI Buildout report reveals 10 companies supplying the memory, storage, connectivity, fabrication, power, and cooling behind AI’s next phase.

Send My Free Report

The Tesla story this week is really about what markets ask of businesses they believe in.

Record revenue. Record deliveries. A 14.5% selloff that separated the growth story from the profitability story.

Growth without profit is a plan. Profit is what makes a plan self-funding.

Tesla needs the autonomous revenue to arrive before the margins compress the balance sheet further.

Know which part of the business you are paying for. If it is still being built, get the timeline from management first.

Know someone watching Tesla from the sidelines right now? Forward them this edition before the week gets moving.