Why Abandon This Ship?
Every past market crash looks like an opportunity, but every future market crash looks like a risk. -- Morgan Housel
Early in the morning of October 3, 1955, the 70-foot merchant vessel MV Joyita set sail from the South Pacific island of Samoa, headed for the island of Tokelau, 300 miles to the north. On board for the two-day trip were 16 crew members and 9 passengers, including a doctor, a government official and two children. It was a routine journey.
At some point during the trip, a corroded pipe began to leak heavily. The bilge pump, a pump located in the lowest point of the hull where water collects inside a ship, was clogged from debris and the crew could not find the source of the leak. As the lower decks began to flood, the crew panicked. To them, it was clear the ship was sinking. The captain gave the order to abandon ship, the life rafts were launched and all passengers and crew evacuated.
What no one on board realized was that the Joyita was practically unsinkable: her hull was lined with 640 cubic feet of cork to insulate refrigerated cargo holds. During that fateful trip, she was carrying nothing but 80 empty fuel drums. These provided so much buoyancy that even if the ship were completely full of water, she would still float.
Five weeks later, the Joyita was discovered drifting 600 miles off course. She was listing heavily and half-submerged, but she was still afloat and perfectly buoyant. Had the passengers stayed on board, they would have been safe and likely rescued within days. Instead, because they panicked and fled to the small, fragile life rafts, not a single one of the 25 people was ever seen again.
Although ultimately tragic, the captain’s reaction is understandable. Who wants to wait and see if the ship will actually float once full of water? By then, it could be too late. In the presence of a fatal threat, the obvious response is to react, to do something.
There are lessons in this tragedy, not just about marine safety, but about human psychology: the way a group collectively reacts to a threat can be as dangerous as the threat itself.
If software was eating the world, what is AI doing now?
You could be forgiven for stating that we are in a software apocalypse. The iShares Expanded Tech-Software Sector ETF has fallen 27%+ since September 2025. Individual names, such as Atlassian (down ~56%), Figma (down ~58%) and Workday (down ~38%) are faring even worse. But what is even more important is the narrative shift. It used to be that software was the best business; low capital intensity, recurring revenue and sticky customers. Software companies commanded crazy multiples and investors piled on, in both private and public markets. But now?
Now AI is eating software. The market appears to be thinking that LLM’s make software companies, and developers, obsolete. Jack Dorsey’s decision to lay off 40% of Block’s workforce makes the threat all the more real. AI *is* coming for developers, even the software companies say so...
Is this reasonable?
The fears have merit: administrative workforces are bloated or wasteful, and a lot of that can be automated or done away with. AI does, in fact, make developers much more productive. Change is coming.
If software companies not only use AI internally, but also deploy it within their applications with autonomous power, that will raise further questions. Does an AI agent require a paid license, or should it share the license assigned to the human that created it? What happens when one human spawns multiple AI agents?
The cost-per-seat model may need to be updated to adapt to the reality, and we still don’t know how it will evolve creating more uncertainty. And yes, some deep-pocketed companies with an internal development team will build their own apps and ditch *some* vendor tools. But that trend has been in place for a while, and although AI will accelerate it, it’s not a new trend.
So why the shift?
The possibilities enabled by AI have been slowly eroding the narrative around software companies for some months now. AI companies have spent millions of dollars hyping their own narrative. Some of them are in the process of raising billion-dollar equity or debt rounds (or both) and some are in the process of going public. The media hype benefits them directly.
The AI Reality Check - Software companies are not dead
But real-world, scalable software is incredibly difficult to do. It’s not just about writing code, it’s about architecture, database design, data security, quality assurance, version control, logging, auditing, software vulnerabilities, debugging and constant fixing and upkeep. And software is never *done*. We are still a long way from AI handling all these tasks competently, transparently, and consistently.
And enterprise software is a different beast. Enterprise buyers will not switch mission-critical applications to vibe-coded slop. For years, large companies have had the ability to hire a professional development team to build and maintain their own applications, and some do, at least for some of their tools. Why don’t they do this for everything? Because it doesn’t make sense to do so. And within the enterprise software space, vertical market software companies are even more insulated.
Vertical market software, or VMS, is specialized, industry-specific software applications designed to meet the unique needs of a particular niche. These are boring but essential tools on which entire industries rely to get things done. They have deeply specific functionality and are compliance-ready, the result of years of embedded industry knowledge and unscalable personal relationships. If they fail, their customers cannot continue to operate.
AI will not replace these companies any time soon. AI, if anything, is a boon to VMS companies by making them more efficient and effective, not obsolete. We should expect changes and a lot of volatility in the software space, but not its complete destruction.
Take the opportunities that are handed to you
One ship that has been abandoned as if destined to fail and sink is Topicus.com Inc. It’s the Joyita story all over again: the market is abandoning ship as if catastrophe had already struck. But Topicus is not only seaworthy, she’s also in terrific shape and poised to navigate what’s ahead. Sure, choppy and stormy seas are approaching, but is it really time to jump into a dinghy life raft?
No.
Topicus.com Inc. is Constellation Software’s European subsidiary. While the parent company has become a $57 billion behemoth, Topicus is executing the identical strategy across Europe’s fragmented, regulation-heavy markets. Same model, same team, earlier innings.
To understand why Topicus matters, it’s important to understand the machine that built it.
The Mothership
Constellation Software was founded by Mark Leonard in 1995, following a career in venture capital. He realized that small, niche software products had staying power and could become quiet compounders.
At Constellation, Mark acquired small VMS companies, implemented best practices, and let the acquired management teams run their companies independently, with the flexibility, speed, and energy of small teams, all while supported by a head office that ensures they are well-capitalized and managed. Constellation’s core model could be summarized as “buy, improve, never sell”.
Mark and his team have been very disciplined acquirers and proven capital allocators. Over time, they developed an edge in quickly finding, analyzing and closing on acquisition targets. They have utilized that discipline to their advantage, making dozens, sometimes hundreds, of acquisitions each year.
Initially focused on Canada, they expanded into the US and other markets, including Europe. The company went public in 2006 and has consistently been extremely cautious when issuing new shares, which is almost unheard of among serial acquirers.
The Next Frontier
The model has worked beautifully. With Constellation’s core model clearly defined and tested, the next question is how far this success can scale beyond the US and Canada.
Enter Europe.
Europe is a large, rich and highly fragmented market. Each country, by itself, is relatively small compared to the US, yet has its own tax code, labor laws and regulatory framework, as well as its own language and cultural idiosyncrasies.
The European VMS market represents a unique opportunity, and Topicus is well-positioned to take on the challenge and deal with the complexities. To most competitors, this is too daunting a challenge.
Topicus is not a sequel or a do-over. Topicus is a subsidiary of Constellation Software, and the expert team of capital allocators, acquirers, and operators applies its experience and knowledge directly. The question isn’t whether the model works. The question is whether Europe’s software market is as rich with hidden gems as Canada and the US have proved to be.
Topicus.com Inc. is a Toronto-based holding company spun off as a dividend-in-kind in December 2020 and began trading on the Toronto Stock Exchange in early 2021. Constellation holds over 31% of its shares, has a super-voting share that gives it 50.1% of the votes and has the right to name six out of the ten board members.
Robin van Poelje, the chairman and CEO of Topicus, has been acquiring European VMS companies since 2006. And he’s not messing around. From 2014-2020, TSS (Topicus precursor) executed over 60 deals, averaging €6M each. Since 2021, the year they were spun off, activity has accelerated, with €1.06B deployed across 135 deals, with a median ticket size of €3.9M.
At Topicus, bonuses are based on the return on invested capital minus a hurdle rate set by the board. Between 25% and 75% of bonuses must be used to acquire shares in the open market, subject to holding restrictions of three to ten years. This results in executives owning a considerable portion of their net worth in shares they *bought*, not shares they were granted. No accounting tricks, no stock dilution, no executives making millions at the expense of shareholders. Insiders own close to 30% of the shares.
What kind of targets are van Poelje and crew after? The boring kind. They acquired Cipal Schaubroeck NV, a local-government software provider based in Belgium with €110M in 2024 revenues and over 590 employees. Cipal offers solutions such as Citizen Administration (population management, registrar services), Financial Tools (local-government tax collection, debtor management, payroll/HR), Social Welfare, Land Planning, and smart city apps, among other services. Once a local government makes the switch, it’s more likely to buy more functionalities and stick with Cipal for a very long time. This is but one example.
Full Steam Ahead
Constellation’s track record provides the benchmark. Operating cash flow has averaged 22.74% of revenue over the last ten years — the lifeblood of a serial acquirer. Using the company’s own ROIC methodology, returns have averaged 31.75% annually over fifteen years.
Topicus is in the early stages of its European conquest. Some see the selloff and narrative shift as a reason to abandon ship. The market panic will also affect private-market valuations, making future acquisition targets cheaper and still more accretive. Topicus is built to deploy capital; if it can execute, the current environment is an opportunity to pick what others tossed: more sellers, fewer buyers, lower prices, better results.
It’s not the time to panic. The negative narrative is likely a tailwind for Topicus. Shares of Topicus are trading 43.9% below their all-time high, handing long-term investors an attractive entry point.
The company reported earnings on February 25. Revenue grew by 20% to 438.67 million EUR, net income increased 41% to 79.4 million EUR and operating cash flow increased 35%. Yet, it missed revenue expectations and the market shrugged it all off. It’s like somebody found a small puddle of water in the lower deck and called for the evacuation of the ship.
Even after the selloff, Topicus is not cheap, by any metric. But investors waiting for Constellation Software to become cheap have missed out: shares have *multiplied* 5.16x over the last ten years, a compound annual growth rate of 17.83%. Quality is very rarely *dirt* cheap.
2025 reported diluted EPS: 0.50 EUR
March 9 stock price: 111.06 CAD
EUR/CAD exchange rate: 1.36 CAD
EPS in CAD: 0.50 EUR x 1.36 = 0.80 CAD
P/E Ratio: 111.06 / 0.80 = 138.83x
However, serial acquirers have large non-cash amortization charges from previous acquisitions, which seriously understate their true earnings power. A more accurate measure is to use P/FCF.
Topicus generated €403.1 million in FCF in 2025. This adds to their available cash for new acquisitions and implies a P/FCF multiple of 26.37x. Rich, but not outlandish.
This is not for the faint of heart. Multiples can still contract further, but over a long enough time frame, Topicus is poised to outperform the rest of the market. Investors seeking to hold for anything less than five or seven years should look elsewhere.
Not Without Risks
One of the biggest risks materialized on September 25, 2025, when Mark Leonard announced his sudden resignation as Constellation’s president, citing health reasons. Mark Miller, Constellation’s COO and part of Leonard’s team for over thirty years (founder of Trapeze Group, Constellation’s first-ever acquisition), took on the role. Leonard remains on the board.
This matters to Topicus because Constellation controls it. Constellation’s culture, discipline, and capital allocation expertise flow directly to Topicus, which means any deterioration at the parent level could affect the subsidiary.
That said, Leonard built a team and a system, not a one-man show, yet Topicus shares have fallen by more than 35% since the news. Miller has been implementing this playbook for three decades and is a very capable captain. However, it is still early to tell how this will impact Topicus, if it ever does. In any case, directly executing the company’s strategy falls to van Poelje, and execution risk is present in any investment.
A bigger concern is multiple contraction, which is largely already underway. For whatever reason, the market may decide to pay even less for Topicus’s earnings. But patience and a long-term outlook will likely be rewarded.
Steady as she goes
The crew of the Joyita didn’t know the ship was unsinkable. They saw water rising and made the only logical decision available to them, and it cost them everything. Patient investors have something Joyita’s crew didn’t: time to think.
Topicus sells software to Belgian municipalities, Dutch hospitals and German regional banks. These customers are risk-averse and unlikely to switch to an app built with a chatbot; if they ever do, it won’t happen quickly. In the meantime, van Poelje and his team will keep buying boring European software businesses no one else wants to touch, at prices the current panic is making cheaper by the day.
AI will change software; it won’t kill it. VMS will be the last category to fall, if it ever does. The risks are real: Leonard’s absence, rich multiples, and execution uncertainty across dozens of jurisdictions. Yet none of them are reasons to abandon ship.
Constellation Software has never been cheap. It still isn’t. But investors who bought at rich multiples and stayed the course reaped the rewards.
Disclaimer and disclosure
The author holds a long position in some or all of the securities mentioned in this publication, either directly or through entities he controls. He may buy more, sell, or change his mind at any time and without notice. Do not assume his position has remained unchanged since the date of publication.
Nothing in this publication constitutes a recommendation to buy, sell, or hold any security or financial instrument. This is not financial, investment, legal, or tax advice of any kind, and no professional or advisory relationship is created between the author and any reader. The views expressed are solely those of the author, are based on information available at the time of writing, and may change without notice.
The author has no obligation to update, correct, or supplement any information contained herein. Some information is sourced from third parties believed to be reliable; however, no representation or warranty is made as to its accuracy, completeness, or timeliness. Readers should independently verify any information before relying on it.
Certain statements in this publication are forward-looking in nature and involve assumptions, risks, and uncertainties. Actual results, outcomes, or developments may differ materially from those described or implied. Past performance is not indicative of future results.
All investments carry risk, including the possible loss of some or all of the capital invested. The securities discussed in this publication may be volatile, illiquid, or otherwise unsuitable for certain investors. You should consider your own financial situation, risk tolerance, and investment objectives before making any investment decision.
This publication is for educational and informational purposes only. Conduct your own independent research and consult a qualified financial professional before making any investment decision. The author is not responsible for any financial loss or damage resulting from reliance on the content of this publication.