Despite posting record revenue figures, Tesla's stock price fell 4% in after-hours trading.
On July 22nd, after the U.S. market closed, Tesla released its FY2026 Q2 results. Revenue reached $28.24 billion, a 26% year-over-year increase, marking the first time its trailing 12-month revenue surpassed $100 billion. Adjusted EPS was $0.33, missing market expectations of $0.51 by 35%. Operating profit was $398 million, with the operating margin declining from 4.1% to 1.4%. While revenue grew by 26%, profit plummeted by 57%.
The record revenue is real, but the pace of decline in automotive profits is currently outpacing the growth in AI-related revenue. This interim gap represents Tesla's most pressing challenge.
Examining the Gross Margin
The Q1 gross margin was 21.1%, which plunged to 16.8% in Q2. This 4.3 percentage point "collapse" alarmed the market.
However, the elevated Q1 figure was inflated. It included a $230 million warranty reversal and over $200 million in tariff benefits. These one-time factors artificially propped up the margin. Excluding them, the true Q1 gross margin was approximately 17.5%, compared to Q2's 16.8%, a difference of only 0.7 percentage points.
While 0.7 points is not a massive drop, the overall trend continues to be a slow decline. The automotive gross margin, excluding regulatory credits, was 16.3%, down from 17.2% a year ago, slipping a little each quarter.
The real warning sign is in the operating margin. The drop from 4.1% to 1.4% is largely attributable to a 47% surge in operating expenses—driven by increased AI R&D, new product ramp-up costs, and depreciation/amortization for computing power. Each of these is a result of deliberate management investment.
Revenue from regulatory credits also fell from $380 million in Q1 to $146 million. This represents a structural decline, not a temporary fluctuation. As other automakers advance their electrification efforts, supply and demand in the credit market are reversing, making this "free profit coupon" increasingly difficult for Tesla to cash in.
The FSD Strategy: A Car with the Software
The take rate for Full Self-Driving (FSD) on new vehicles in North America reached 55%, with 1.48 million global paying users, a 56% year-over-year increase. CEO Elon Musk made a notable comment on the earnings call: "Many people come to the store saying they want to buy FSD, and they'll take a car with it."
The FSD subscription fee is $99 per month, carrying an extremely high pure profit margin. With 1.48 million users paying $99 monthly, the annualized revenue is approximately $1.76 billion. This already represents revenue on the scale of Tesla's second-largest business. A take rate exceeding 50% signifies that FSD is transitioning from an "optional feature" to a "default configuration." When over half of new car buyers choose FSD, Tesla is no longer just selling vehicles.
A contradiction arises here: higher FSD penetration and a greater share of software revenue should theoretically lift gross margins, yet they are declining. The cost growth from AI R&D and computing power depreciation is far outpacing the revenue growth from FSD. Software revenue is not yet substantial enough to cover the investments.
Tesla is navigating a transitional phase: automotive profits are declining while AI revenue grows, but the latter cannot yet fill the void left by the former.
Significant Quarterly Capital Expenditure
Q2 capital expenditures (CapEx) reached $5.789 billion, a 142% year-over-year increase and a doubling from the previous quarter. The full-year guidance exceeds $25 billion, with expectations for continued growth over the next 2-3 years.
Free cash flow turned negative at -$1.09 billion. However, operating cash flow was $4.697 billion, up 85% year-over-year—indicating strong cash generation from operations, which is being consumed by CapEx. The $43.5 billion cash reserve is sufficient to fund operations for a considerable time.
Musk described this as "the fastest industrial-scale expansion in the U.S. since World War II." Projects include Cybercab production lines, Optimus production lines, the Dojo supercomputing cluster, next-generation AI hardware R&D, and the TeraFab in-house chip factory—with a site announcement imminent. The CFO explicitly stated that free cash flow is not expected to turn positive until 2029.
In summary: from now until 2029, Tesla will be in a sustained period of significant investment.
Robotaxi Progress and the Long Road for Optimus
The Cybercab is already in production, with Robotaxi service covering seven metropolitan areas. The system has logged 380,000 miles of unsupervised driving with zero major accidents. Version 15 of FSD is being tested in the fleet, with weekly miles driven growing at a double-digit rate.
The safety record is a positive. The data of 380,000 miles with zero major accidents is more impressive than Waymo's figures for a comparable period.
Progress on the Optimus humanoid robot is more measured. Musk admitted, "Every component of the robot is new; there is no supply chain." The first production line is being installed at the Fremont factory—replacing the former Model S/X line. The forecast is for production of 100,000 units per month in five years, but the early part of the S-curve is described as "long and flat." The production ramp will be a lengthy process.
The connection with SpaceX is deepening. Starlink will be integrated into Robotaxi to ensure network connectivity, and the TeraFab wafer fab location will be announced "soon." Tesla's equity stake in SpaceX contributed $1 billion in unrealized gains. Questions about a potential merger were sidestepped: "The earnings call is not the place to discuss that."
Valuation Metrics and Divergent Analyst Views
In Morgan Stanley's $417 price target, the automotive business accounts for only $47. The remaining $370 is entirely based on the AI narrative—Robotaxi, Optimus, and FSD.
The forward P/E ratio stands at 467, with a Return on Invested Capital (ROIC) of 3.8%. With quarterly profits at just $400 million against a market cap exceeding $800 billion, capital efficiency at the current valuation is arguably "destroying value."
Analyst frameworks have split into two camps. The automotive fundamentals-focused group is concerned with gross margin and EPS, and the earnings miss has made them more cautious. The AI narrative-focused group is tracking FSD take rates and Robotaxi progress, with the 55% take rate and zero-accident data bolstering their confidence.
The divide is widening: near-term profit deterioration and long-term AI investment are occurring simultaneously. A 467x P/E ratio implies the market is already pricing in a story for 2029, creating real valuation pressure during a projected three-year period of negative free cash flow.
Key Metrics to Monitor
For Q3, focus on three signals:
Whether the FSD take rate can rise further from 55%—a core indicator of Tesla's shift from an auto company to an AI company.
Whether CapEx continues to accelerate—and if the $25 billion annual guidance is raised toward $30 billion.
Whether the gross margin stabilizes in the 16.3-16.8% range. Continued declines would undermine the narrative of "trading profits for future growth."
Tesla is using its automotive profits to fund its AI ambitions. While revenue hit a record $28.2 billion, profits dwindled to just $400 million. Free cash flow has turned negative, with significant investment expected through 2029. The $43.5 billion cash reserve provides a long runway. The critical question is whether, after this period of heavy spending, the AI investments will deliver on their promise.
Comments