The Agentic Stub: Why Sabre Is One of the Most Asymmetric AI Setups

A patient, engaged shareholder base, including Constellation Software, combined with estimated equity upside of 3-7x

This piece follows my recent article, “Who Gets Paid When AI Agents Subscribe to Data?”, in which I argued companies whose data is sensor-produced, regulator-cited, or embedded in a two-sided network stand to benefit from the growing adoption of agentic AI. I called out Amadeus (Spain: AMS) as “perhaps the best risk/reward large-cap on the list.”

Today I discuss Amadeus’s smaller and, in my view, more interesting comparable: Sabre Corporation (US: SABR). An important caveat up front: Sabre is highly leveraged relative to its equity cap, so if your investment process eschews debt-laden companies, this article will not appeal to you.

Sabre is the world’s #2 global distribution system (GDS), a centralized network that facilitates automated, real-time transactions between travel service providers (airlines, hotels, car rental companies) and travel agencies. It sits on top of the same agentic data flywheel as Amadeus and, in several places, is further along on agentic infrastructure. The enterprise, however, trades at a fraction of Amadeus’s valuation, as Sabre carries roughly $4.2 billion of debt against a market cap under $750 million. This is a genuine equity stub, as defined in Chapter 10 of my book, The Manual of Ideas.

Equity stubs are uncomfortable, and the bear case for Sabre is zero. However, a combination of strong cash generation, deep proprietary data, aggressive deleveraging, and active engagement by 10+% shareholder Constellation Software (Canada: CSU) tilts the risk-reward in a way I find difficult to ignore. Buffett follower Discerene Capital, with close to 10% ownership, appears to hold a similar view.

Let’s understand why the uncomfortable optics of Sabre equity may be deceiving, and why investors may capture 3-7x upside as the company executes its agentic AI strategy and delevers.