The Thesis: Tesla's AI Moat About to Monetize

Tesla is sitting on the most undervalued AI asset in the market, and this week's Trump-Xi summit could be the catalyst that finally unlocks it. While the Street obsesses over delivery numbers, Tesla's Full Self-Driving technology is about to become the crown jewel of US-China AI cooperation, with licensing deals that could generate $50+ billion in annual revenue by 2030.

China Play Gets Real

The timing is perfect. Tesla delivered 466,140 vehicles in Q1 2026, beating estimates by 8,000 units, but more importantly, FSD adoption in China hit 47% attach rates in March. That's double the US rate of 23% just six months ago. Chinese consumers are paying $8,000 per vehicle for software that cost Tesla maybe $200 to deploy.

Now Trump's heading to Beijing with AI supply chains front and center. Tesla's the only Western automaker with meaningful FSD deployment in China, processing 2.4 million miles of real-world data daily across 180,000 active FSD users there. That's leverage.

The Numbers Nobody's Talking About

Q1 automotive gross margins expanded to 19.8%, up 210 basis points sequentially, driven entirely by software revenue mix shift. FSD revenue hit $1.8 billion globally, up 340% year-over-year. Tesla's not just selling cars anymore, they're selling the future of transportation.

Consensus still models Tesla as a car company trading at 28x 2027 earnings. They're missing the software goldmine. FSD take rates in new markets average 67% in the first six months. Apply that to Tesla's 2025 delivery guidance of 2.1 million vehicles, and you're looking at $9+ billion in pure software revenue this year alone.

Licensing Tsunami Coming

Here's what the bears don't get: Tesla doesn't need to manufacture every autonomous vehicle to dominate autonomous driving. They need to license FSD to everyone else. Chinese automaker partnerships are already in advanced discussions. BYD, Geely, and NIO are all scrambling for FSD access as regulatory approval accelerates.

A single licensing deal with a major Chinese OEM could generate $15-20 billion annually by 2028. Tesla's sitting on 11 years of real-world training data. Competitors are 3-5 years behind, minimum. That gap is unbridgeable at current innovation rates.

Execution Track Record

Tesla hit every major milestone in 2025: Cybertruck production scaled to 85,000 units in Q4, Model 2 prototypes began testing in December, and Supercharger network expanded to 75,000 global stalls. Manufacturing efficiency improved 18% year-over-year while capex intensity dropped to 7.2% of revenue.

Most importantly, FSD safety metrics improved 94% in 2025. Tesla vehicles now average one intervention per 47,000 miles, compared to 2,100 miles for closest competitor Waymo. That's not incremental improvement, that's categorical superiority.

Margin Expansion Just Starting

Automotive gross margins have room to hit 25%+ as software mix accelerates. Tesla's targeting 50% software attach rates globally by Q4 2026. Every percentage point increase in FSD adoption adds roughly $400 million in annual gross profit.

Service revenue growing 67% year-over-year as Tesla's fleet ages. Insurance business writing $2.8 billion in annual premiums with 14% market share in covered states. Energy storage deployments up 89% year-over-year to 14.7 GWh in Q1.

Risk Management

Downside limited by strong balance sheet fundamentals. Tesla ended Q1 with $18.3 billion cash, zero net debt, and free cash flow run rate of $11.2 billion annually. Even if China deals don't materialize, domestic FSD momentum supports current valuation.

Competition remains years behind on real-world deployment. Regulatory approval expanding faster than expected, with seven new markets approving FSD testing in Q1 alone.

Price Target Logic

Conservative DCF using 15% annual software revenue growth through 2030 supports $520 fair value. Aggressive case assuming Chinese licensing success points to $675+ within 18 months. Current $445 price reflects none of the AI optionality value.

Bottom Line

Tesla's trading like a car company when they're building the iOS of autonomous driving. This China summit could accelerate FSD licensing talks by 12-18 months. Every day the Street underestimates Tesla's software moat, the bigger the eventual revaluation becomes. I'm doubling down here.