In the rapidly evolving landscape of automotive technology and artificial intelligence, the race for fully autonomous driving remains the industry's most coveted prize. While numerous tech giants and legacy automakers are pouring billions into research and development, a new analytical note from Morgan Stanley suggests that the competition may not be as fierce as it appears on the surface. Andrew Percoco, taking the reins of Tesla coverage for the financial powerhouse, has issued a bold declaration regarding the electric vehicle manufacturer's position in the sector: when it comes to self-driving dominance, it is "not even close."
This assessment comes at a pivotal moment for Tesla, as it navigates a transition from a pure hardware manufacturer to a company increasingly defined by its software and artificial intelligence capabilities. Percoco's note provides a fresh perspective on the market dynamics, particularly in light of emerging competition from semiconductor behemoth Nvidia. By analyzing the fundamental requirements for achieving Level 5 autonomy—specifically the necessity of real-world training data—Morgan Stanley has reinforced the narrative that Tesla's years of fleet deployment have created a protective moat that competitors may find nearly impossible to breach.
For investors and industry observers, Percoco’s analysis offers a stabilizing view amidst a week of high-profile announcements from competitors. While new tools and platforms are being launched to democratize autonomous vehicle development, the Morgan Stanley note argues that the raw infrastructure required to train the necessary neural networks—millions of miles of diverse, real-world driving data—remains a resource where Tesla holds a near-monopolistic advantage. As the coverage of Tesla at Morgan Stanley shifts hands, the firm's long-term conviction in the automaker's technological superiority remains unshaken, setting the stage for a deeper look into the mechanics of this perceived dominance.
The Changing of the Guard at Morgan Stanley
The context of this new note is significant not just for its content, but for its author. For years, Adam Jonas has been the face of Morgan Stanley's automotive research, particularly known for his bullish and often visionary takes on Tesla's potential to disrupt mobility. However, the landscape of the market is shifting, and so too is the coverage within major financial institutions. Andrew Percoco has recently overtaken coverage of Tesla stock from Jonas, a move that signals a broader realignment of how these companies are categorized.
Adam Jonas has transitioned to covering "embodied AI" stocks, moving away from the traditional automotive sector. This shift itself is a tacit admission by the financial sector that the lines between carmakers and robotics firms are blurring. Percoco, now stepping into the spotlight, has wasted no time in establishing his analytical framework. His first significant move after assuming coverage was to adjust the price target for Tesla, raising it from $410 to $425. Concurrently, he adjusted the rating from 'Overweight' to 'Equal Weight.'
While a rating change to 'Equal Weight' might typically suggest a neutral stance, the context of the price target increase and the commentary within the note suggests a nuanced view. It acknowledges the massive valuation Tesla has already achieved while pointing toward the specific technological pillars that support its long-term viability. Percoco’s analysis suggests that while the stock price reflects much of the current optimism, the fundamental technological gap between Tesla and its peers is widening, not shrinking.
Data is King: The Mathematics of Dominance
At the heart of Percoco’s bullish thesis on Tesla’s technology is a simple yet profound truth in the world of machine learning: data is king. The note explicitly highlights that Tesla’s advantage is derived from its ability to collect vast amounts of training data through its massive fleet. Unlike competitors that rely on small fleets of test vehicles in geofenced areas, Tesla has millions of cars driving throughout the world, gathering millions of miles of vehicle behavior on the road every single day.
Percoco writes:
“It’s not even close. Tesla continues to lead in autonomous driving, even as Nvidia rolls out new technology aimed at helping other automakers build driverless systems.”
This statement underscores the difference between theoretical capability and practical application. To train a neural network to drive safely in all conditions, it must be exposed to "corner cases"—rare, unpredictable, and complex scenarios that occur infrequently. A fleet of a few hundred test cars might take decades to encounter the same number of corner cases that Tesla’s fleet encounters in a single afternoon. This data feedback loop allows Tesla to refine its Full Self-Driving (FSD) software at a pace that competitors cannot mathematically match without a similar volume of vehicles on the road.
The "shadow mode" capability of Tesla vehicles, where the computer runs in the background comparing its decisions to the human driver's actions without actually intervening, provides an infinite stream of error-correction data. Percoco identifies this infrastructure as the primary reason why Tesla’s lead is secure. While other companies can buy faster chips or better sensors, they cannot easily purchase the billions of miles of real-world driving context that Tesla has already banked.
Nvidia's Alpamayo: A Different Approach
The timing of Percoco’s note was likely influenced by recent headlines surrounding Nvidia, the ruling champion of AI hardware. Nvidia recently launched its own self-driving suite, dubbed "Alpamayo." As an open-source AI program, Alpamayo represents a significant development for the broader automotive industry, offering a lifeline to legacy automakers who lack the in-house software expertise to build autonomous systems from scratch.
However, Percoco points out that Nvidia’s approach differs from Tesla’s in a significant fashion, particularly regarding hardware. The Nvidia strategy plans to utilize a sensor fusion approach, combining:
- LiDAR: Light Detection and Ranging, which uses lasers to create a 3D map of the environment.
- Radar: Radio waves used to detect the speed and distance of objects.
- Vision: Cameras that interpret the visual world similar to human eyes.
In contrast, Tesla has famously pivoted to a "Vision-only" approach, removing radar and ultrasonic sensors from its vehicles and relying entirely on cameras and neural networks. Elon Musk has argued that since roads are designed for human eyes, a biological neural net, an electronic neural net with cameras should be sufficient for driving.
Percoco notes that while Nvidia’s announcement is impressive and beneficial for the industry at large, it does not impact Morgan Stanley’s long-term opinions on Tesla. The divergence in strategy highlights Tesla's vertical integration. Tesla designs its own inference chips, writes its own software, and manufactures the cars that collect the data. Nvidia, while powerful, is providing a platform for others to build upon, which introduces integration challenges and data fragmentation that Tesla does not face.
Validation from the Competition
Perhaps the most compelling evidence supporting Percoco’s "not even close" assertion comes from the competition itself. In the world of high-stakes corporate rivalry, it is rare for a CEO to praise a competitor's product openly. yet, Nvidia CEO Jensen Huang did exactly that following the launch of Alpamayo. His comments provide a layer of technical validation to the financial analysis provided by Morgan Stanley.
Huang, widely regarded as a visionary in the AI space, commended Elon Musk and Tesla for their early conviction in end-to-end deep learning. Huang stated:
“I think the Tesla stack is the most advanced autonomous vehicle stack in the world. I’m fairly certain they were already using end-to-end AI. Whether their AI did reasoning or not is somewhat secondary to that first part.”
This quote is critical for understanding the current state of the technology. "End-to-end AI" refers to a system where raw image data from cameras is fed into a neural network, and steering/braking controls are outputted directly, without traditional code-based rules in the middle. Tesla’s move to Version 12 (v12) of its FSD software marked a transition to this end-to-end architecture, replacing hundreds of thousands of lines of C++ heuristic code with neural nets trained on video.
By acknowledging Tesla's stack as the "most advanced in the world," the CEO of the company making the chips for everyone else effectively conceded that Tesla’s head start is genuine. This aligns perfectly with Percoco’s assessment. If the supplier of the industry's best AI hardware believes Tesla is ahead, the financial markets are right to take note.
The Financial Implications of Autonomy
The reiteration of the $425 price target by Percoco is inextricably linked to the success of these autonomous initiatives. For Tesla to justify its market capitalization—which frequently exceeds that of the next several automakers combined—it cannot simply be a car company. It must be a robotics and AI company. The "Equal Weight" rating suggests a balanced risk-reward profile at current levels, but the price target implies significant upside if the autonomy thesis plays out as expected.
If Percoco is correct and the gap between Tesla and competitors is insurmountable, the implications for the Robotaxi market are profound. The company that reaches Level 5 autonomy first, with a scalable manufacturing base, stands to capture the lion's share of a multi-trillion-dollar transportation-as-a-service market. Morgan Stanley’s analysis suggests that while other automakers are still figuring out the tools (via partnerships with Nvidia or Mobileye), Tesla is already refining the finished product.
This dominance also has implications for software margins. If Tesla’s FSD becomes the gold standard, the licensing potential—or simply the high-margin revenue from software subscriptions on its own fleet—could radically alter the company's profitability profile compared to traditional manufacturing margins.
Conclusion: A Lead Built on Miles
As Andrew Percoco settles into his role covering Tesla for Morgan Stanley, his inaugural major note serves as a reminder of the unique position the company occupies. The transition from Adam Jonas to Percoco has not diluted the firm's view on Tesla's technological supremacy; rather, it has refocused the lens on the specific data advantages that sustain it.
The assertion that the race is "not even close" is a bold one, especially in an industry characterized by rapid innovation. However, the logic underpinning the claim is robust. In the era of artificial intelligence, data is the fuel that powers progress. With millions of vehicles acting as data-gathering nodes in a global network, Tesla has constructed a data engine that competitors, regardless of their silicon prowess or sensor suites, struggle to replicate.
While Nvidia’s Alpamayo and other emerging technologies will undoubtedly raise the floor for the entire automotive industry, Morgan Stanley sees Tesla’s ceiling as significantly higher. As the industry moves toward a driverless future, the analyst consensus suggests that while many will compete, Tesla’s years of groundwork have ensured it remains the one to beat.
