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.

$CSU (+3,16 %)

Fake Blood in the Streets or Justified Caution?

I just spent the last few hours digging into Constellation Software, trying to lift the mystery around how much the recent headlines actually matter. The stock has dropped more than 25% — the biggest sell-off since its IPO. That’s not nothing. The trigger? First, as for many software companies, AI worries entered the room. And then the bombshell of founder and CEO Mark Leonard stepping down for health reasons did the rest. He is undoubtedly a legend in investing circles — the Gandalf lookalike who built Constellation into one of the best compounders of the last two decades. The Constellation he imagined wasn’t like PE funds. Leonard’s brainchild didn’t buy to make a quick buck or fire half the staff to make a turnaround — they invested in niche disruptors for perpetuity.

Besides, the numbers themselves still look attractive. Forward EV/revenue of around 5. Projected revenue growth above 15% for the coming years. As good as no leverage. Strong margins. FCF yield approaching 5%. By almost any metric, this looks like one of those setups I like: proven execution, cheap valuation, and a market reaction that looks more like panic than rational repricing. I’ve said it before: I like to buy when there’s blood in the streets. And right now, Constellation is bleeding heavily.

But here’s the tension: a lot of what made Constellation so exceptional came from Leonard himself. He wasn’t just any president. He was an all-time great capital allocator, one of those rare leaders who built not just a company but a culture of decentralized excellence. His stepping down is significant — you don’t just replace someone like that. Everyone was hoping that he’d stay at the helm for longer in a Warren Buffett style. Though, unfortunately, Gandalf’s beard doesn’t grant Leonard immortality.

On the other hand, maybe that’s exactly the point: the culture is so strong and the model so decentralized that Leonard’s departure might not be as damaging as it first looks. Let’s see what the new president, Mark Miller, has to offer. After all, it’s fair to assume Leonard would choose someone capable to build on his legacy. Let’s not forget, Gandalf choosing Bilbo worked out for the quest of reclaiming the Lonely Mountain. Maybe people are underestimating Miller as much as the dwarfs did Bilbo. Enough of the Lord of the Rings references for now. Constellation is built around hundreds of smaller, niche software businesses, each run with autonomy. The machine keeps running, even without constant input from the top. Leonard’s fingerprints are everywhere in that system, and that gives me some confidence it can continue without him.

And then there’s AI. Could it disrupt Constellation? In theory, yes. But I think the risk is overblown — as with many other formerly loved names, most prominently Salesforce and Adobe. Constellation’s customers are often niche operators without the resources or scale to build software in-house, even with AI making developers more efficient. AI can reduce the time it takes to write code, but it doesn’t replace the expertise, maintenance, or distribution that Constellation provides. Building your own software is still expensive, time-consuming, and for most organizations — risky. Many will find it cheaper and far more reliable to keep buying from Constellation rather than attempt DIY projects that may end up buggy or unusable. Especially for non-tech-native businesses that don’t have the capabilities needed to develop complex programs. That’s not something the plumber’s IT administrator “vibe-codes” in an afternoon session.

So where does that leave me? Honestly, weighing it. On one hand: flawless execution, cheap valuation, and what looks like an overreaction to legitimate but manageable risks. On the other hand: the founder stepping back creates uncertainty that you can’t simply model away. For me, this is the classic setup: buy now and bet that Constellation’s culture carries it forward — or wait and see if this transition shakes the foundation more than expected. However, I am worried that I might miss this opportunity if I don’t buy soon, since the market is pretty quick at recovering nowadays.

One thing is clear: this sell-off won’t go down as just another blip. Either this becomes one of those rare “blood in the streets” buying opportunities (more likely in my opinion), or a reminder that even legends don’t live forever.

$CSU (+3,16 %)

Harris Chairman on $CSU (+3,16 %) capital deployment

"We obviously spend a lot of our time thinking about capital deployments and adding new businesses into our own.

But I think making sure that we have leaders who wake up every single morning,

focused on doing what they can to drive our existing businesses forward is really important and something that we can never sort of lose track of."

-Jeff Bender, current Harris Chairman, 2025 AGM