When there’s lasting quality, timing becomes incidental
I’ve been watching Constellation Software for what feels like ages, admiring it for years, but couldn’t get on terms with the valuation. It was never unreasonably expensive like Palantir, just a bit too pricey for me. Until last week, when I finally opened a position.
For months, I kept telling myself I’d wait for the “perfect” entry around 3,500 CAD. That was my mental anchor. One, honestly, set quite arbitrarily in hindsight. So I realized that the level has always been attractive after the stock’s 25% drop from all-time highs. And when it sat at 3,700 CAD, I had to admit the obvious: I was letting precision get in the way of sense. For a company like Constellation — a world-class compounder, cash-rich, and still expanding at 15%+ revenue growth — the difference between 3,500 and 3,700 CAD is just noise over a long-term horizon. At some point, hesitation can become more expensive than action. Something I had to learn painfully, when I couldn’t fully convince myself to open a position I great companies that then subsequently ran away. For me, the risk of not getting into the Constellation because the stock never hits my 5%-below target is far greater than the one of not owning the business at all.
What led me to even consider the company for my portfolio is its unique approach to growth. Constellation isn’t a typical software roll-up chasing scale for scale’s sake. It’s an ecosystem of hundreds of niche vertical-market software businesses, each running independently, each optimized for cash flow and longevity. The decentralized model isn’t just a fancy buzzword, but rather a wide moat. One that hasn’t really been replicated anywhere. Culture is a big asset for this compounder, which is why Mark Leonard’s departure isn’t the end of the world, because the groundwork has already been laid over decades. The brilliance of Leonard’s system is that it’s built to outlast him. Every acquisition adds another small, durable income stream, and together they compound quietly in the background.
And on the AI fear? Still overblown. The kind of clients Constellation serves (municipalities, niche service providers, specialized industrials) aren’t hiring developers to build their own software from scratch with ChatGPT, let alone start vibe coding in the back of an industrial company to find groundbreaking solutions. They want reliability, integration, and support, not DIY code.
So yes, I paid 5% above my “ideal” entry. But I finally own one of the most durable compounding machines in global markets, and it’s not about trading the stock — you just let it do what it’s always done: grow, compound, and ignore the noise.
