Matt Franz of Eagle Point Capital presents an investment case for Constellation Software, a serial acquirer of vertical market software (VMS): hyper-specialized, mission-critical applications serving niche industries such as ship management, public transit routing, country club operations, and grain elevator management. A typical subsidiary commands 50-80% market share, charges customers less than 1% of their sales, and retains 98% of them. Franz traces how the company has completed roughly 1,200 acquisitions across 100 countries, maintaining a pipeline of 40,000 relationships that it contacts three to four times a year, targeting businesses too small for private equity or venture capital and priced at private valuations that have not risen with public markets — it has deployed five times more capital year-to-date than in the prior two years combined. Underpinning this is an extremely decentralized organization of just 14 HQ staff, with M&A itself pushed down to the operating units and divisions capped near 100 people (Dunbar’s number), and a combined ROIC-plus-net-revenue-growth metric designed so neither figure can be gamed alone. Free cash flow has compounded 26% since the 2006 IPO with zero shares issued in twenty years; organic growth runs a modest 1-6% while essentially all real growth comes from acquisitions.
The central opportunity, in Franz’s view, is a roughly 50% decline in the stock — from a peak of 47x free cash flow to about 15x — driven by AI fears he considers unsupported by the data, alongside the health-related retirement of founder Mark Leonard (who was succeeded by Mark Miller, founder of Trapeze, Constellation’s first acquisition in 1995) and doubts that a company this large can keep compounding. His variant perspective is that “the code is not the moat”: mission criticality, regulatory validation, decades-long switching costs, and customer trust are, much as banks still run COBOL on IBM mainframes because ripping out working systems is too painful even when better technology exists. He argues AI may in fact be a tailwind through the Jevons Paradox — more software means more VMS to acquire — with a central AI hub available to all business units and 1,500 independent subsidiaries functioning as parallel experiments that learn faster than centralized rivals. Franz frames forward returns of roughly 25% annualized over five years, built on 20% incremental ROIC and multiple expansion from 15x back toward 19-20x. In the Q&A he identifies organic growth as the canary in the coal mine for AI disruption — currently accelerating to 4-6% rather than declining — notes that about 85% of revenue is stable maintenance and SaaS fees with the license-to-SaaS transition still underway, highlights European spinoff Topicus as a compelling “baby Constellation,” and describes the “pen” strategy of taking minority stakes in public companies such as Sabre when private valuations run too high.
Presentation summary generated by AI