PodBrowser
Impact Theory

Why 3 Private Companies Are Worth 45 Years of Public Tech — And You're Locked Out

Tuesday, 29 September 2026 · 4 min read · Listen to the episode ↗

SpaceX, Anthropic and OpenAI are valued together above the combined value of U.S. companies that went public over the preceding 45 years, the speakers say, leaving public investors with less access to their growth. They warn that delayed IPOs can make public buyers “exit liquidity.” With Anthropic valued at about 45 times revenue and AI-sector debt growing faster than revenue, they also question whether the opportunity comes with excessive risk.

SpaceX, Anthropic and OpenAI are together valued above the combined value of all U.S. companies that went public during the preceding 45 years, the speakers say. They argue that firms staying private for 10 or 15 years let private investors capture more of their growth before public investors get access. SpaceX is not publicly buyable, while Anthropic and OpenAI are planning to list, according to the discussion. Delayed IPOs can leave public buyers paying too much or serving as “exit liquidity,” one speaker warns, though they would rather have access than none. Microsoft’s IPO valuation, they note, was below $100 million.

Public investors also face a concentrated market. About a quarter of U.S. household net worth is tied up in stocks, but the top 10% of households own 93% of stock assets, and five companies account for 30% of the S&P 500. The speakers support index funds while cautioning that diversification can disguise heavy exposure to large technology companies. They say the CAPE ratio, historically around 16 times, is now above 40; one speaker claims every earlier move above 40 preceded catastrophe and expects a bubble to break eventually, though not necessarily soon. The dot-com market took 15 years to regain its prior level, and crowded ownership could speed a decline if buyers disappear.

The speakers describe mounting pressure from U.S. debt, which they put above $40 trillion, with interest payments already the largest federal spending line item. One cites the 10-year Treasury rate above 5% and the 30-year rate higher still, arguing that rising borrowing costs threaten the budget. They criticize buying government debt with newly created money as a way of artificially suppressing rates, and warn that a country becoming the main buyer of its own debt risks stagnation. One speaker favors austerity and a balanced budget as the most moral response, but expects policymakers may instead cut rates while inflation remains high. The speakers also disagree about Japan: one calls its low-cost borrowing approach successful, while another points to decades of stagnation, zombie companies and a lower quality of life for workers.

The Strategic Petroleum Reserve has fallen for 26 consecutive weeks to its lowest level since 1982, the speakers say, even though it is almost three times larger than about a decade ago. The president released another 400,000 barrels the previous week and had said 22 days earlier that the reserve would be refilled with Venezuelan oil. One speaker argues that Venezuela’s heavy crude does not meet reserve specifications and that restoring neglected oil fields and refineries would take time. The drawdown can temporarily support supply and restrain oil prices and inflation, but leaves the U.S. more exposed to disruption. An oil shock could raise diesel and freight costs and then food and building-material prices; the speakers qualify the risk by noting that oil flows were roughly at pre-war levels at the time of recording and limited disruption could have a much smaller effect.

AI could deliver unprecedented productivity gains if adopted broadly, but the speakers warn that the sector’s financing and valuations are risky. They say Anthropic is generating revenue unusually quickly and in extraordinary quantities, while AI-sector debt is growing faster than revenue; one valuation cited is about 45 times revenue. One speaker describes AI companies as “default dead” while that debt imbalance persists and calls the concentration of capital in them frightening. At the same time, the speakers argue that technology is winning and that today’s firms can have substantial revenue and cash flow, unlike many dot-com-era companies. They say growth on the scale needed to ease the debt challenge would be difficult without a massive contribution from AI, but caution against concentrating investments in the sector.

As alternatives, one speaker favors cash-generative businesses that can prove resilient, including credit-card networks whose percentage-based fees rise with transaction prices, waste-management companies needed in both strong and weak economies, and inflation-linked toll businesses. Reported trades by Donald Trump included purchases of Berkshire Hathaway, Visa, Mastercard, Home Depot, Tractor Supply and Republic Services, and sales of Meta, Palantir and Netflix. The speaker reads these as a shift toward established businesses and “toll booths” and away from crowded technology stocks, but says the trades are a pattern to notice, not a recommendation to copy. They urge investors to consider what is working now, what has worked historically and where companies have been vulnerable, while recognizing that suitability depends on risk tolerance and time horizon.

This summary was generated from the episode transcript and can contain mistakes.