The data shows a $1 trillion revenue prediction for Starlink by 2040. The math, however, does not support the narrative. David Friedberg, a potential investor, recently claimed Starlink could generate $400 billion in annual revenue, eventually reaching $1 trillion, and carry 50% of global internet traffic. Elon Musk echoed the sentiment. But as a cold dissector, I see a pattern: hype masking structural flaws. The numbers look attractive until you apply the actuarial lens. Let me walk through the errors.
Context: The Starlink Narrative
Starlink, SpaceX's low-earth orbit (LEO) satellite internet service, currently has around 6 million users and an estimated annual revenue of $60-100 billion. The pitch is simple: as AI and robotics drive bandwidth demand, Starlink becomes the critical infrastructure for global connectivity. The projections assume a linear extension of growth—from 6 million to 300-350 million users, with ARPUs remaining high. But historical data on telecom infrastructure shows a different reality. Bandwidth prices fall, competition intensifies, and capital expenditure never stops. The Friedberg model assumes a 75% free cash flow margin, which is unheard of in the telecom sector. The average telco runs at 10-20%. That single assumption is a red flag.
Core: Systematic Teardown of the Assumptions
Technical Capacity Constraints
Starlink's current satellite constellation consists of approximately 7,000 satellites, each with a capacity of 60-100 Gbps (V2 Mini). To carry 50% of global internet traffic, which is projected to be 396 EB per month by 2027 (peak 1.1 PB/s), you need at least 10,000 to 40,000 satellites depending on capacity upgrades. The plan for 42,000 satellites exists, but the constraints are physical: spectrum availability, orbital debris, and ground station backhaul. The narrative that there are 'no obvious barriers' ignores the physics of radio frequency congestion and the need for thousands of ground stations globally. Code speaks louder than promises. The code of satellite specifications and spectrum allocation tells a different story.
The Unit Economics Trap
Friedberg's $300 billion free cash flow implies a 75% margin. In reality, Starlink's operating expenses include satellite manufacturing, launch costs, ground station maintenance, and customer acquisition. The terminal hardware alone costs $300-600, which is often subsidized. The current ARPU of $100-120 per month means you need 300-350 million subscribers to hit $400 billion in revenue. That's a 50x increase from current 6 million. Even with growth from mobile and enterprise users, the total addressable market for premium satellite connectivity is limited. The number of commercial ships is around 100,000, airplanes 25,000, and off-grid households in developed countries maybe 30 million. The rest are in developing nations where ARPU is lower. Follow the gas, not the narrative. The gas here is the cost of maintaining the constellation. Each satellite has a 5-7 year lifespan. Replacing the entire fleet every 7 years requires a constant launch cadence, which is a massive capital drain. The 75% FCF margin is a mathematical impossibility when you account for this.
The Contrarian Angle: What the Bulls Got Right
Despite the skepticism, Starlink has a genuine technological moat. It is the only LEO constellation with a self-owned launch system, giving it a cost advantage that competitors like Amazon Kuiper cannot match. The military and government contracts are high-margin and sticky. The Direct-to-Device partnership with T-Mobile and others could open a wholesale revenue stream that bypasses hardware costs. The AI demand for low-latency, global connectivity is real, especially for autonomous vehicles, drones, and remote robotic operations. Trust is verified, not given. The bull case is not without merit, but it is a potential $200-300 billion revenue company, not a $1 trillion one. The difference is a factor of 3-5, which is significant in capital allocation.
The Hidden Risks: Governance and Regulation
The reliance on a single entity and a single person for 50% of global traffic is a geopolitical risk. The SEC's regulation-by-enforcement approach has shown that infrastructure assets with concentrated control face scrutiny. But more importantly, the legal structure of Starlink as a private company means that external shareholders have limited recourse. The DAO governance model in crypto has taught me that concentrated control leads to single points of failure. Logic outlives the hype cycle. The hype cycle will fade, but the physical constraints of satellite capacity and the capital requirements will remain.
Takeaway: The Accountability Call
Based on my experience auditing the 0x Protocol v2 smart contracts, I learned that the most dangerous assumptions are the ones that look conservative on the surface. Friedberg's model is not conservative; it's a fantasy built on a 75% margin that defies telecom economics. The real question is not whether Starlink will grow, but whether the market is pricing in the reality of sustained capital expenditure and competitive pressure. When the next bear market comes, the satellites will still need to be replaced. The code does not lie. The ledger does not lie. The physics does not lie. The $1 trillion prediction is a narrative, not a forecast. Keep your eyes on the gas, not the hype.
