137 Ventures' Alex Jacobson on why SpaceX-style private liquidity beats public markets for employees

Summary

Alex Jacobson of 137 Ventures frames the firm's portfolio strategy around what he calls a 'private Mag Seven' — a concentrated set of late-stage private companies that combine dominant market position with large, still-expanding total addressable markets. The thesis borrows directly from the public Mag Seven's compounding logic but requires different execution in private markets.

The SpaceX Liquidity Model as a Talent Weapon

SpaceX is the clearest proof of concept. The company issues equity at hire and runs regular tender offers, giving employees a predictable path to liquidity. Jacobson's key point is that the tender price sits materially below what the open public market would pay — and that discount is a feature, not a flaw. Because SpaceX controls its own pricing and the stock moves in one direction, employees joining early can underwrite their own appreciation with high conviction. Public company equity doesn't offer that; once a stock is perfectly priced by the market, upside is uncertain and volatility driven by macro noise — tariffs, rate moves — affects compensation regardless of company performance.

The structural advantage compounds on the hiring side. Issuing stock at the 409A valuation while controlling subsequent pricing creates a deterministic compensation outcome that public company RSUs cannot replicate. Jacobson argues this is systematically underestimated as a competitive advantage in talent markets.

Retail Access to Private Shares: Skepticism with Nuance

On efforts to tokenize private company equity or create on-chain SPVs for retail access, Jacobson is dismissive of the mechanics but acknowledges the fairness concern underneath them. The core problem is that circumventing private market controls — shareholder selection, price setting — strips the very properties that make these companies valuable to hold. Forcing a company into de facto public status without its consent is not a service to the company or to retail buyers who would receive no disclosure, no insider trading protections, and no meaningful governance rights.

He speculates that if Stripe shares were listed on Solana, the initial pop could push the implied valuation to $400 billion or more, illustrating how retail demand, untethered from fundamentals, would distort pricing immediately. The legitimate version of broad access already exists indirectly: pension funds — teachers, firefighters — invest as LPs in funds like 137 Ventures and capture private market returns. But Jacobson concedes that leaves out a large population with no pension exposure and no mechanism to direct their own allocation toward privates.

The Take-Private Question

On whether public companies would ever go private to recapture the advantages Jacobson describes, he points to the scale of today's private capital markets as the key structural shift. The old argument that public markets offered access to more capital no longer holds — private markets can now absorb large-scale investment in companies like SpaceX, Anduril, and Gusto. The real appeal of going private would be escaping quarterly reporting pressure and regulatory overhead, not capital access.

Dell's take-private and subsequent re-IPO at a significantly higher price is the canonical example, though Jacobson acknowledges such outcomes are rare. A company like Snap, down 17% on earnings the day of the conversation versus Apple up 5%, Meta up 1%, and Amazon up 4%, is an illustrative case — at a $13 billion market cap the capital exists to take it private, but private buyers who watched the public performance would be a hard sell.

Armada: Stranded Energy Meets Stranded Data

Jacobson traces 137 Ventures' involvement in Armada back to a simple infrastructure insight: traditional data centers are optimized around fixed points of power and fiber connectivity. Starlink dissolves the connectivity constraint by making broadband available everywhere, even if per-location bandwidth is lower than fiber. That unlocks a rethink of where compute can physically sit.

The original model Jacobson sketched — a shipping container with unfolding solar panels powering a compute rack — produced roughly a 3% IRR, unexciting on its own but sufficient to justify testing. The real commercial breakthrough came when Dan (Armada's CEO) called contacts in the energy sector and landed a conversation with the CTO of Ramco, an oil and gas software company. Oil platforms generate enormous volumes of operational data that currently go unprocessed — 'stranded data' sitting alongside stranded energy. On-platform compute is typically a closet server managed by a rig worker running Halliburton software. Armada's proposition is to deploy an edge cloud at exactly that intersection of stranded energy, stranded data, and Starlink connectivity — capturing value that the existing architecture leaves entirely on the floor.