Byte # 55: My Reasons for Believing in Constellation Software (CNSWF/CSU)

Dear Readers,

Welcome to this week's investing byte on Constellation Software, the company I covered in Byte #28 back in November 2025.

Since then, the stock has fallen more than 50% from its peak before recovering to around $2,100 today. Much of that decline was driven by fears that AI could disrupt vertical market software businesses.

The question for investors is simple: has AI broken Constellation's moat, or has the market simply repriced a great business?

Here are six reasons why I believe the long-term thesis remains intact.

First: The business nobody can replicate

Constellation Software is Canada's second-largest technology company by market cap, behind only Shopify - and arguably the more durable business of the two. It operates as a permanent, decentralized acquirer of vertical market software (VMS) businesses: niche, mission-critical software serving industries like hospital billing, emergency dispatch, transit operations, agricultural logistics, and municipal governance.

The model's beauty lies in its economics. CSU's software is essential but relatively inexpensive in the context of a customer's overall cost base, and switching costs are prohibitive. Customers would absorb retraining costs, integration risk, and operational disruption - often exceeding any savings a competitor could offer. They don't switch.

A large share of revenue is recurring, which is why the business compounds so reliably. Organic revenue grew 6% in Q1 FY 2026, suggesting the core portfolio remains healthy and that the feared AI-driven churn has simply not shown up in the actual results.

Second: AI fears are structurally overblown

The market has sold CSU down ~50% from peak on fears that AI will commoditize VMS software. That thesis is compelling on the surface but collapses under scrutiny for three reasons.

VMS moat isn't code. Customers buy trust, compliance, and liability support - not lines of code. An LLM cannot take legal responsibility for failure in a police dispatch system.

Fragmentation equals protection. AI needs massive datasets and large addressable markets to be economically viable. CSU operates in thousands of micro-niches so specifically they don't register on Big Tech's radar.

AI as accelerant. Lumine's CEO put it plainly at the 2026 AGM: development cycle times went from months to days to hours. AI compresses product delivery, deepens customer integration, and drives more cross-sell - not erosion.

Third: What Q1’ 2026 showed

Constellation's Q1 2026 results were strong. Revenue grew 20% year over year, organic growth came in at 6%, and diluted EPS increased from $6.41 to $17.32. Free cash flow available to shareholders (FCFA2S) rose from $510 million to $733 million - reinforcing that the business continues to generate meaningful cash even while scaling.

Capital deployment also remained robust: $809 million deployed in Q1, including deferred payments, plus over $627 million in pending commitments after quarter-end. That matters because Constellation's growth model depends not only on operating performance, but on the continued ability to find attractive capital allocation opportunities.

The Permanent Engaged Minority Shareholder (PEMS) strategy expands that toolkit by allowing CSU to take meaningful stakes in larger public software companies that wouldn't be realistic full acquisition targets - broadening the addressable deployment universe meaningfully.

There was a modest margin decline in the quarter, but management attributed it to acquisition mix and expects improvement over time. That reads as a near-term integration effect, not structural deterioration.

Fourth: Valuation is compelling on FCFA2S

GAAP metrics distort CSU's true economics. Acquisition-related amortization under IFRS creates significant non-cash charges that obscure the underlying cash engine. FCFA2S - management's preferred metric - is the right lens.

Annualizing Q1 2026 FCFA2S of $733 million implies a ~$2.9 billion run-rate, growing at an 11% CAGR over five years. The valuation compression tells the real story: at peak mid-2025, the market was paying ~48x FCFA2S. Today, on a materially higher and accelerating cash flow base, you are paying approximately 14x - a level not seen in years. At ~$2,100 per share, the gap between price and fundamentals looks sentiment-driven, not structural. Analyst consensus targets point in the same direction, though I treat them as a secondary input rather than the core of the thesis.

Fifth: Culture – The Invisible Moat

Mark Leonard did not just build a company - he built an institution. What makes Constellation unusual is that its culture appears designed to survive leadership transitions because it was never dependent on one person alone.

Alignment is built into the system. Managers buy shares in the open market rather than receiving options, headquarters stays disciplined on capital allocation, and operating groups are given a high degree of autonomy. The result is a structure that minimizes bureaucracy and keeps incentives aligned with long-term value creation.

The permanence of ownership is also a major advantage. Founders selling to Constellation are often choosing continuity, not just liquidity. That creates a deal-flow advantage that private equity and strategic buyers cannot easily replicate.

Mark Miller, Constellation’s new President, is also deeply steeped in this culture. Having founded the first company Constellation ever acquired and spent decades inside the organization, he looks less like a break from the past and more like a continuation of the same playbook. That makes the succession story look more like a strength than a risk.

Sixth: Key risks - and why they are manageable

The biggest risk is AI disruption, but so far the numbers do not support that fear. Organic recurring revenue is still growing, and AI appears to be improving internal productivity rather than weakening the business.

Capital deployment at scale is the second risk. Sustaining historical growth rates becomes harder as the company grows larger. PEMS helps by broadening the opportunity set, but execution will matter.

Leadership transition is worth monitoring, though I view it as manageable given how deeply embedded the operating model is across the organization. The bigger risk here may be sentiment, not operations.

Finally, GAAP earnings will remain noisy due to acquisition amortization. FCFA2S is the number that matters.

Sometimes it's the stock that's broken, not the company - and knowing the difference is one of the most important skills in investing. I'll let time prove the thesis.

As always, this isn't a recommendation to buy or sell - it's about revisiting the process and stress-testing whether the original thesis still holds.

Keep investing smarter 😊

Cheers,

Pooja