Tesla’s core business is in reverse. Its stock is driving autonomously

At Tesla, the wheels are slowing, but the hype engine is humming.

On Wednesday, Elon Musk’s EV empire will report its second-quarter delivery figures, a once-predictive metric for the stock’s movement that has been seemingly reduced to background noise. Now, Tesla’s stock isn’t driving itself off of fundamentals anymore — it’s running on the idea of autonomy.

Call it the great EV bait-and-switch. Just as its car sales stall in key global markets, Tesla has managed to reroute investor attention toward a shinier, driverless future. The robotaxi is (sort of) here — at least in Austin, Texas, with some carefully curated demo rides. Whether that detour can carry the company’s trillion-dollar narrative through another quarter of underwhelming results is the question facing Tesla and its increasingly speculative shareholders.

Tesla’s second-quarter delivery numbers, once expected to clear 440,000, are now forecasted by a growing chorus of analysts to fall somewhere between 355,000 and 377,000 units (while Bloomberg’s consensus sits at 392,000). That’d be a 15–20% year-over-year drop. Deutsche Bank trimmed its projection to 355,000 — a nearly 20% year-over‑year decline, versus last year’s high of roughly 440,000 — citing sharp year-over-year erosion in Europe and a stubborn sales slump in China as BYD pulls further ahead and Xiaomi racked up 200,000 preorders for its new EV in under five minutes.

JPMorgan, largely in the bear camp, has warned of an “accelerating decline” with its estimate landing at 360,000 units. The bank now expects Tesla’s yearly unit sales to fall to 1.575 million, down from the 1.7 million consensus. Analysts pointed out that Tesla would need a far stronger second half to hit any rebound.

Meanwhile, Tesla’s details are grim. The company reported a dismal first quarter. Vehicles registrations in the EU cratered by almost 30% in May compared with a year ago; numbers in China remain slightly below the first quarter despite deep discounts; and even the U.S. market showed cracks earlier this quarter, before a late-stage Model Y refresh (Juniper) helped paper over the gap. Tesla’s global brand has been “damaged the most” in Europe “where competition is intensifying,” according to Deutsche Bank, while JPMorgan sees “no end in sight” to what it calls a multiyear expectations reset.

“We see material risk to the outlook for full year deliveries also, given that consensus requires a sharp pivot from underperformance to outperformance of expected seasonal pattern despite the likely significant near-term curtailment of EV subsidies,” JPMorgan analysts wrote in a Monday note. Tesla’s declining numbers have a structural feel to them, and even the bullish buy-side has tempered expectations. This is more than a stuttering recovery; it feels like a business sliding into reverse.

But here’s the kicker: None of that data seems to matter. Despite falling stock prices (down 6.64% over the last month and 15.84% year to date), investors stay bullish. Tesla’s P/E ratio is 176.34 — meanwhile Nvidia, the most valuable company in the world, has a P/E ratio of 50.85. A mere whisper of a robotaxi pilot with Model Ys in Austin sparked a 9.2% single-day spike, which added roughly $95.7 billion in market cap, more than the annual GDP of Croatia.

Autonomy is the detour of the decade

Musk’s robotaxi dream has become the backbone of Tesla’s valuation.

William Blair, in a recent note, pegged the value of Tesla’s robotaxi business at $299 per share, nearly 10 times the estimated value of the core automotive segment ($28.09). That math makes a lot of things make sense — such as why a company with declining sales and shrinking margins still trades at over $1 trillion. Wedbush analysts have echoed that sentiment, arguing that Tesla’s robotaxi unit cost could be just a third of Waymo’s, cementing Tesla’s position as a disruptive force in autonomous mobility.

“Our analysis reveals that Tesla’s valuation is increasingly dependent on the robotaxi business,” William Blair analysts wrote. “We fundamentally believe in Tesla’s long-term solution of neural nets and vision only, but acknowledge it opens up attack vectors for the inevitable hiccups to come. We are encouraging investors to use bumps along the road tactically. The counterbalance to the robotaxi opportunity is a challenging environment for the core businesses today.”

According to the firm’s 2040 forecast, Tesla will own 35% of a $1.4 trillion autonomous ride-hailing market, hauling in nearly $250 billion in robotaxi revenue. With projected EBITDA margins nearing 60% — thanks to Tesla’s hardware stack — it’s no wonder some investors are happy to discount today’s softness for tomorrow’s robot army.

But “tomorrow” is doing a lot of heavy lifting.

Despite the heady valuation, there are cracks even in Tesla’s self-driving narrative.

Videos from the robotaxi pilot show unsettling signs — phantom braking, erratic lane behavior, sudden stops — attracting fresh scrutiny from the National Highway Traffic Safety Administration. And then there’s the competitive heat. Waymo just launched service in Atlanta, adding to an already robust presence in San Francisco, Los Angeles, and Phoenix, Arizona. Even as Tesla pivots to AI, the competition isn’t exactly stuck in neutral.

Plus, another big bet from Tesla seems to be slipping. Deutsche Bank noted that the anticipated Model Q budget car, originally pegged for a June unveiling, appears delayed. Without it, Tesla may not regain meaningful sales momentum until the fourth quarter — assuming the launch even materializes.

Wednesday could show the fork in the road

When the second-quarter delivery numbers hit, Tesla won’t just be reporting how many cars it sold. It will be reporting how far its narrative can stretch before it snaps. If deliveries hover near 355,000, Tesla may still escape punishment — buy-side expectations have been tamped down, and the market has largely decoupled the stock from its sales trajectory. But a deeper miss (say, under 350,000 units) could raise questions: Is Tesla just a regular car company with a well-funded science project on the side, or is it really driving toward a new economic model?

On the other hand, a modest beat — maybe closer to 375,000 units — might give bulls just enough fuel to keep the stock aloft. But even then, the company will need to offer more than AI optimism. Is the Model Q still coming this year? Can Tesla stabilize its China business? Are the Austin robotaxis more than glorified demos? Tesla’s magic has always been its ability to sell both cars and dreams. But as the cars struggle and the dreams get more expensive, the company’s margin for error narrows.

Wedbush analyst Scott Devitt wrote in a Monday note, “While we think the near-term financial impact to established ridesharing platforms is limited, over time, we expect AVs will disrupt the current status quo. … Tesla’s robotaxi approach circumvents affordability as a limiting factor on scale.” 

But affordability cuts both ways — and so do expectations. If the robotaxis don’t scale fast enough or if they meet resistance from regulators or the public, investors may find themselves holding a lot of promise and not nearly enough product.