Constellation Software Stock Analysis (CSU)

Constellation Software (TSX:CSU) Stock Fundamentals, Overview

Constellation Software Inc. (CSU) is a Canadian technology holding company founded by Mark Leonard in 1995 in Toronto. The company employs approximately 50000 people worldwide and owns more than 1000 software companies, typically small, vertical, so-called Vertical Market Software, VMS, companies specializing in markets. Constellation does not develop software in-house, but continuously acquires smaller, profitable, stable customer base software companies and then operates them in a decentralized manner over the long term.

The company's philosophy is the “buy and hold forever” approach: It does not speculate, it does not try to maximize short-term returns, but builds value over decades. Mark Leonard, also known as Canada's Warren Buffett, is a genius who discovered one of the most boring but profitable segments of the software market: small, industry-specific software companies that are essential to their clients' operations, all of which are mission-critical solutions, and therefore maintain an extremely loyal customer base. The Constellation model thus became one of the most pure-running machines of capitalism: constant cash flow, low risk, high returns, all done discreetly, avoiding publicity.

Market capitalization:
Investor relations:
iO Charts stock page: CSU.TO stock page

📒Table of Contents📒

I have created a table of contents to make it easier for you to navigate the longer articles:

Constellation Software (TSX:CSU) Specialties
How does Constellation Software (TSX:CSU) make money and what market advantages does it have?
Constellation Software (TSX:CSU) metrics
Constellation Software (TSX:CSU) acquisitions
Constellation Software (TSX:CSU) management
Competitors: Constellation Software (TSX:CSU) opponents
What risks does Constellation Software (TSX:CSU) face?
Constellation Software (TSX:CSU) valuation
Major news and the last quarter
Other interesting facts about Constellation Software (TSX:CSU)

〽️Market segment analysis〽️

In this section, I examine the dynamics of the market segment, how it operates, who the main players are, and what tailwinds or headwinds the players in the given market have to deal with. I will not analyze companies in depth, but I will touch on the market share of individual companies.

Constellation Software (TSX:CSU) is a serial acquisition company that operates in the VMS market segment. VMS is an abbreviation for vertical market software, but to understand the segment, we first need to clarify the difference between horizontal and vertical software.

🧑🏻‍💻What is horizontal software?

Horizontal software is typically not an application that serves unique needs, but rather a kind of Swiss Army knife solution with an incredibly large market size. We mean general-purpose software under the umbrella name that sIt can be used in any industry, company or private user because it solves a universal problem. Very typical examples are:

💻 Operating systems: Windows, Linux, macOS, Android, iOS, etc.
🧾 Software suitable for word processing and spreadsheet management: Office 365, LibreOffice, etc.
🎨 Image editing apps: GIMP, Photoshop, InDesign, etc.

Each of them is characterized by the following::

🌍 Wide customer base, many industries.
📈 High volume market, many competitors.
🔁 Often standardized, easier to replace.

There are a thousand and one such applications, but the essence of all of them is that they appeal to a very wide user base, can spread virally, very often they spread according to the principle of network effects. They are often tried to be pushed out of the market by disruptive methods, but in return they can grow quickly.

👨🏻‍💻What is vertical software?

In contrast, vertical market software is software that is designed for the specific needs of a specific industry or niche market segment and is deeply integrated into those processes.

Examples:

⚰️ Cemetery management system (for funeral service providers only)
🏥 Hospital patient record software (for healthcare only)
💧 Water utility billing software (for utility companies only)
🚌 Bus company timetable and ticket management system

Features:

🎯 Narrow customer base, but extremely high customer retention.
🧩 Strong specialization, high switching cost (difficult to replace).
💰 Often developed by smaller companies with long-term stable cash flow.
🐍 Market leaders often "eat" their own market, crowd out competitors, generate a lot of cash, but cannot grow at all, and in fact, often decline.

In connection with the above, it is very important to mention some key factors that characterize the VMS market. One is the extremely high customer retention rate, which usually results in two things:

📉 Low attrition rate.
🔄 Large-scale renewable revenue, for which the company doesn't actually have to do anything.

As an IT professional for 12 years, I have come across several of these in the public sector. One of the best examples is the software that manages parking in certain districts of Budapest. Such applications are tailored to the needs of the given districts, contain an extremely large number of special elements that are included in the IT systems due to local specificities, but they also have to comply with countless legal elements. What does this mean in practice? That:

🏗️ although there is a basic software, no two districts are exactly the same,
🧱 almost irreplaceable, because there is no other player on the market that would program the same special software,
💸 its pricing power is brutal, practically the authorities plan it in the budget and pay the fee the same way every year,
🔁 almost zero churn, such software has always generated recurring revenue,
📊 Since the number of districts and the population are not expanding, the market is becoming saturated and essentially stagnant, but revenues are increasing due to strong pricing power.

Based on the above, you can find countless other similar software that works very similarly in a concrete market. Of course, in the US, you can find many orders of magnitude more such companies, whose growth size is limited to a few million USD by the specialized market.

📊 Comparison in brief

Point of view	Horizontal software	Vertical Software (VMS)
Target market	Broad, industry-independent	Narrow, industry-specific
Function	General problems (e.g. office work)	Special industry processes
Clientele	Many, heterogeneous	Few, homogeneous
Race	Strong, global (Microsoft, Google)	Few players, local dominance
Replacement cost	Lower	High (due to complex integration)
Revenue model	Volume License / SaaS	Stable, long-term niche income

What is the market size of Constellation Software (TSX:CSU)?

Although we'll get to the exact numbers later, there's a fund manager, REQ Capital, who has launched two funds in the past, both of which are almost exclusively serial acquirers. One is a global fund, which includes Constellation Software and its spin-offs, and the other is a Nordic fund, which includes serial acquirers from the Scandinavian countries. Luckily, REQ Capital has already done the homework for you and estimated the approximate size of the European market, which you can see in the image below.

REQ Capital, number of European SMB companies in millions — source: REQ Capital, number of European SMBs in millions

In the figure above you can see the number of SMBs, or small and medium-sized enterprises, in the EU, in millions. There are approximately 23.5 million such businesses in total, of which 60000 are acquisition targets, even according to REQ Capital's relatively strict filter, and their number is increasing over time. There are even more SMB companies in the US, according to some sites there were 34.7 million three years ago and the number of companies here is also increasing. Constellation's database alone has about 40-50000 acquisition targets, so the number of opportunities is huge.

Constellation Software (TSX:CSU) has realized that there are many small, niche software markets around the world that have limited growth potential but are highly lucrative and have high cash-generating capabilities. Since these markets are not growing very much, it:

💵they can be acquired cheaply, Constellation Software (TSX: CSU) usually acquires these at 1-1.5x P/S or 3-5 P/E, but I have also read the 5-8x EBITDA number in several places, these can represent essentially similar amounts,
🚀 serial founders often sell their stagnant software company and start another business,
⏳ founders get tired of stagnation, get old or die,
🤝 They cannot achieve further growth in a closed market, so they are looking for partners for further growth or mergers while maintaining their leading position in the company. In fact, they are looking for a permanent home for their company.

In other words, they are ideal acquisition targets if the cash in them can be channeled into a holding company and then reinvested there without harming the operations of the small software company.

🤝What's the problem with acquisitions?

Acquisitions are generally not lucrative deals. If you go back a little to the question of what a company can do with the cash it generates, one answer is: it can acquire other companies. They usually acquire other companies because they are looking for synergies, that is, an additional, untapped opportunity, which looks something like: 1+1>2. That is, the acquiring company hopes to create additional value by absorbing another company, and therefore usually overpays for the acquisition.

The image above is an image from a 2012 McKinsey study examining acquisitions between 1999-2010, with the outperformance after various acquisitions shown on the left. It's worth looking at the large deals row, where you can specifically see underperformance, but the result of various mergers is that acquisitions fail 66% of the time and destroy shareholder value. Here are a few examples of the above:

💣 IBM's gigantic acquisition failures, e.g. RedHat,
💸 HP–Autonomy acquisition, worth 11 billion USD,
📞 Ebay–Skype, worth 2.6 billion USD,
📱 Microsoft–Nokia, for 7.2 billion USD. Where is Nokia today? Nowhere, the phone business is owned by HMD.
🧀 Kraft–Heinz, its $55 billion acquisition burned a lot of capital,
📺 AT&T and Warner merger for 85 billion USD, most of which had to be written off,
🕳️ AOL–Time Warner was one of the most devastating acquisitions of all time from an ownership perspective, valued at $165 billion.

What happens with acquisitions? On the accounting side, goodwill is an intangible asset that is recorded when a company is acquired at a price higher than its book value. These can be completely real items, such as brand power, intellectual property, customer relationships, business reputation and a thousand others, but the part of the overpayment that was unnecessarily paid for with acquisitions also goes here. I wrote quite a bit about goodwill in the Adobe analysis: Adobe Inc. Stock Analysis (NYSE: ADBE).

☝🏻What happens if it turns out after an acquisition that the acquirer overpaid during the acquisition? It is usually written off as a loss in the books, meaning the capital invested is destroyed, which is particularly harmful from the owners' perspective.

Unfortunately, management tends to force acquisitions when they have nothing to show for it, so they want to make something big. A great example of this is General Electric, which under Jack Welch acquired everything from financial services to insurance to healthcare, but they also had media companies, energy, aviation, and manufacturing divisions, so the company grew into a formidable conglomerate. However, in 2000, the company began to decline, and by 2018, it had lost 85% of its value due to its unmanageable, overly broad spectrum.

For those interested in more information on the topic of failed acquisitions, I recommend a few links:

McKinsey study
Paul B. Carroll - Billion Dollar Lessons: Chapter 3 – Deflated Rollups
Scott Management, LLC: Serial Acquirers

🤝Why can VMS or other acquisitions still work?

The above examples show that several problems arise when assessing acquisitions:

💰 high pricing: acquirers overpay for acquisitions
🧩 Lack of integration: they were unable to integrate the acquired company into the company structure
🪓 the management of the company was taken over by the acquiring company, autonomy was lost

As you can see from the above, these three things are critical in acquisitions, yet countless serial acquirers have been in business for decades, with Constellation Software (TSX:CSU) being the king, and the spin-offs that are strongly tied to Constellation. But why is it that some companies are successful in this endeavor while others are unsuccessful? It's because not all acquirers operate in the same way, as shown in the image below:

types of acquiring companies — source: Scott LLP, Types of Acquiring Companies

📌 1. Roll-up companies

📘Definition: The essence of the roll-up model is that many smaller, very similar, commodity-like businesses are acquired and fully integrated into a larger unit.
🎯Main goal: economies of scale, cost reduction, central synergies (that's the word that makes me shiver anyway).
🧠Feature: everyone uses the same system, common brand, common operation, autonomy is lost.
🏢Examples: Waste Management, Berry Global, Johnson Service.

📌 2. Platform

📘Definition: A platform-type acquirer buys companies that are related to each other but not completely identical, and then divides them into sub-platforms, clusters organizes them.
🎯Main goal: sharing product differentiation and corporate methodology programs (best practice, lean, etc. the latter may be familiar to startups).
🧠Feature: not full integration, but strong management control and common systems.
🏢Examples: Danaher (classic), Roper, Interpump, Lifco, Assa Abloy.

📌 3. Accumulator

📘Definition: The accumulator buys many companies but hardly integrates them. The emphasis is on differentiated products and independent operation.
🎯Main goal: companies should remain independent, but there should be a loose umbrella over them (e.g. common capital allocation, financial discipline).
🧠Feature: federation model, each company retains its identity, the center mainly provides capital and benchmarks.
🏢Examples: Constellation Software (classic accumulator example), Addtech, Judges Scientific, Vitec Software.

📌 4. Holding company (Hold Co)

📘Definition: The holding company's portfolio consists of completely independent, different businesses.
🎯Main goal: capital investment, management through the board of directors, but no operational intervention.
🧠Feature: There is no integration, no corporate methodology programs, the center is just a portfolio manager.
🏢Examples: IAC, Leucadia, Bollore, HAL Holding, Investor AB.

😟I mentioned spin-off companies earlier, they actually strip the acquired company of its management and force it into the corporate structure in which the parent company has operated so far, causing the acquired company to lose most of its identity.

The other extreme is the holding company, where there is no particular connection between the profile and operation of the companies, they operate more like when someone puts unrelated products, such as shampoo, pruning shears, and eggs, into a shopping cart. Without integration, it is not possible to exploit the advantages that these companies could offer each other, because there is no similarity in operation between them, the holding company only exercises ownership rights through management members.

The remaining two categories of interest to us are platform and aggregator companies. A prime example of the latter is Constellation Software, which you can read more about in the next chapter. Platform companies differ from the previous ones in that they group the acquired companies into a set, where they try to exploit their synergies. Such grouping can result in economies of scale, but in the long run cross-selling can also occur. Although the above examples are also excellent, Kering (PPRUY) also fits into the Platform or Holding company type of acquiring companies, mostly a hybrid of the two. I analyzed the company, you can read about it here: Kering Stock Analysis (PPRUY) – Gucci is not selling.

🙋‍♂️Specialties of Constellation Software (TSX:CSU)🙋‍♂️

In this section, I examine what specialties the analyzed company has, what its position is in the market, and whether it does anything differently than its competitors. If so, what and how, and what impact does this have on their operations.

To understand how Constellation Software (TSX:CSU) works, you need to know the name Mark Leonard, an amazing financial genius and the founder of Constellation Software Inc. Even his contemporaries recognize his greatness, and it's no coincidence that he is often referred to as the Canadian Warren Buffett, but because of his appearance, the nicknames Software Santa Claus and Santa Mark have also been raised several times. The reason for this is partly his appearance (Mark is 2 meters tall and weighs about 130 kilos), and partly because, like Warren Buffet, he also wrote investor letters containing incredibly interesting thoughts, although in much smaller numbers and shorter lengths. You can read Mark Leonard's investor letters here: President Letters.

source: Gulf News, Mark Leonard, founder and former CEO of Constellation Software (TSX:CSU)

Mark Leonard founded Constellation Software (TSX:CSU) in 1995, which went public on the Toronto Stock Exchange in 2006. Over the past 19 years, the company has produced an annual CAGR of 31.8%, with a total increase in company value of 21036%. This is incredibly brutal, among other things, it ruined companies like Microsoft and Google in terms of price growth. But what was it that Mark Leonard knew better than anyone? Capital allocation, and how Constellation Software could turn this into the acquisition of more and more mini-companies, the number of which now exceeds 1000.

It's worth knowing where Mark Leonard started, because it will help you understand how Constellation Software works. Mark Leonard started his career at Barclays Bank, where he wasn't particularly successful, but he managed to make the transition to venture capital. While learning about how VCs operate, he noticed that venture capital funds tend to pursue investments, primarily startups, that did not generate much cash but had very high expected growth potential, and in return, such companies cost significantly more. Since I have built a startup portfolio myself, I know that venture capital is basically invested in companies that have a survival rate of 1:10, so the expected return is roughly a minimum of ten times the return on capital, so that the capital fund can also cover the losses of the other 9 companies. I have written quite a lot on the topic, you can access these articles here:

Mark Leonard's problem was that venture capital funds were incredibly bad at picking which high-growth companies would be successful in the future, burning through an awful lot of investor money. All the while, they ignored low-growth, highly cash-generating companies that could generate huge returns on invested capital and were valued at far lower than startups.

The other groundbreaking idea that later led to the founding of Constellation Software (TSX:CSU) was that since highly skilled managers could effectively run almost any company, you first need to find the right people, and then give them the company they could run. That leader was Steve Scotchmer, who had extremely high expectations for the companies and ultimately failed to find an investment target, but Mark Leonard learned from Steve the criteria by which he could discover quality companies.

Constellation Software (TSX:CSU) corporate structure — source: REQ Capital, corporate structure of Constellation Software (TSX:CSU)

After examining the market segments based on the criteria mentioned by Steve, Mark ended up with vertical market software, or VMS, companies that ticked all the important criteria for him, except for one: they could not invest enough capital in them due to their small size. This is where the idea came from, what if they founded a company called Constellation Software (TSX:CSU), withdrawing the capital from the acquired companies, they would reinvest in new ones and keep them forever. This effectively turned Constellation Software (TSX:CSU) into a kind of asset management company, where the wealth was created by the companies themselves.

💡The whole structure should be thought of as a huge machine operating on the principle of compound interest, where the fuel is provided by the cash extracted from the acquired VMS companies. As long as Constellation Software (TSX:CSU) is able to acquire more and more companies, the flywheel never stops, just producing a huge amount of cash, which, when invested again and again, can achieve a huge internal rate of return.

In practice, this means that between roughly 1995 and 2021, Constellation Software (TSX:CSU) acquired around 500+ companies, where they were able to work with an internal rate of return, or IRR, of around 20-30%. Constellation Software (TSX:CSU) calls this the hurdle rate, and it should be noted that it works with relatively small companies, between $1 million and $10 million. This entails a few additional things:

☝🏻no competition: Constellation Software (TSX:CSU) has no real competition in the VMS market due to its low organic growth potential and small size. It is too small for venture capital funds, and large-cap companies do not divide or multiply companies of this size, but they would have to run the risk of integration, which they are not experienced in.
☝🏻adequate number of acquisition targets: As long as there are enough targets, Constellation Software (TSX:CSU) can keep feeding the compound interest machine. If the targets run out, the machine will stop.
☝🏻Effect from size increase: The bigger Constellation Software (TSX:CSU) gets, the more capital it has to deploy, meaning either the number of companies it acquires increases or the size of the companies increases, which in turn reduces its IRR. The same goes for Warren Buffett's Berkshire Hathaway (BRK-B) problem is that the company can no longer effectively allocate the enormous amount of money.
☝🏻Replication of intellectual capital: It was clear that Mark Leonard couldn't make 100 acquisitions a year on his own, so he needed to replicate his mentality with leaders who followed similar principles. It's also important to keep Constellation Software's (TSX:CSU) corporate structure, which has become very complex over time, flexible and manageable.

🧩Responses to Constellation Software (TSX:CSU)'s problems

From the above points, it is not difficult to conclude that the problems arising from the operating principle of Constellation Software (TSX:CSU) had to be solved somehow. One of them was the number and size of acquisition targets. Over the past 30 years, Constellation Software (TSX:CSU) has built up an incredibly rough database, in which:

🗃️According to some sources, there are data on 40, and according to other sources, more than 50000 VMS acquisition targets (while there are millions of tiny VMS companies on the market),
🖇️CSU maintains long-term relationships with acquisition targets, even following companies for years,
🔬They experiment with acquired companies, so they have an awful lot of data on how to integrate them, what works and what doesn't.

I mentioned the hurdle rate earlier, which is an internal rate of return that is most often compared to IRR. If I were to translate this into the language of fundamental analysis, I might use ROCE, ROIC, and similar metrics. Although the IRR for Constellation Software (TSX:CSU) is confidential, various sources estimate it at 20-30%, so you can expect an IRR somewhere around 25%. The problem is that this works well for companies with a market cap of $1-10 million, but not for those larger than that. Since Constellation Software (TSX:CSU) has to spend roughly $1 billion to $1000 billion a year in cash, plus it can use debt, it can no longer do this by acquiring only companies with a market cap of $1500-10 million, because that would mean 100-150 companies with a market cap of $10 million.

"There are three hurdle rates. Less than a million in revenue is 30% IRR; above four million, you can drop to 20%; and 25% is for everything in between. You can go to 20% because those bigger businesses are more competitive, but everything which falls in the middle at 25% is 90% of deals. There is another hurdle rate of 15%, but that is on deals so large they're not in CSI's wheelhouse. Above 50 million, you can go to 15% but they might only do one of those per year." - In Practice article

Therefore, Constellation Software (TSX:CSU) will have to increase the amount of acquisitions in some cases, i.e., it will have to absorb companies with larger capitalizations, which in turn will reduce the IRR but will still keep the compound interest machine in motion. This in turn means that:

📊the Constellation Software (TSX:CSU) model became known, copycat companies appeared (some of which they created themselves), competition emerged,
📊private equity funds are also moving into larger companies, which increases the acquisition price,
📊less opportunities for optimization and cost efficiency in larger companies because they operate more efficiently.

📌In practice, this means that: Small companies were typically bought at a price of P/S=1-2, meaning one to two times their annual revenue, or P/E=3-5, meaning 3-5 times their earnings, which is considered very cheap in today's market. The SP500 is valued at a P/E of 30,97, so you can see how terribly cheap Constellation Software (TSX:CSU) has been in recent years. In the case of larger companies, the acquisition amount may move into the P/S=3-5 range, meaning the IRR will decrease, somewhere in the 12-15% range. To illustrate the above situation with an example, let's assume that:

💰they invest half of their capital in small companies at an IRR of 25%,
💰they invest half of their capital in medium or large companies at an IRR of 15%,
💰the result will be the average of the two, i.e. 20% IRR.

Since the ratio will likely shift toward larger acquisitions in the long run, the internal rate of return will also deteriorate. What can be done in this case? A part of the company should be spun off from the Constellation Software (TSX:CSU) organization, so that both new companies will have smaller market capitalizations, and everything can move forward. The other possibility is to move away from the VMS market and into another, more fragmented segment, where hundreds of thousands of other companies could be potential acquisition targets. What is ideal? Let's do both at the same time❗

🔀Constellation Software (TSX:CSU) spin-off companies

Constellation Software (TSX:CSU) has spun off several companies over the past few years. To understand the logic behind this, you first need to understand how Constellation Software (TSX:CSU) is structured. Constellation Software employs a total of 20 people in its headquarters office, but the structure below shows that:

🧑🏻‍🍼It is divided into 6 operating groups (OG), these are: Volaris, Harris, Jonas, Vela, Perseus and Topicus (which is now a spin-off),
👱🏻each operating group comprises business units, which are assembled from similar companies (business units, BU),
👥There are additional companies in each business group.

Most companies operate completely autonomously, and in countless cases the founders still run their own companies, but they are now owned by Constellation Software (TSX:CSU). Day-to-day decisions are made by the companies, while capital allocation decisions are made by the group leaders who head the operating groups. These leaders decide on acquisitions up to approximately USD 20 million, only requiring CSU approval for larger acquisitions.

Constellation Software (CSU) architecture (the star is Perseus OG) — source: Generalist Substack channel, built by Constellation Software (CSU) (the star is Perseus OG)

The above arrangement shows that for quite some time now, it is not CEO Mark Leonard and Bernie Anzarouth, the head of mergers and acquisitions, who decide on acquisitions, but the six division heads, and in some cases the BUs also have a say in such activities. In other words, Mark Leonard recognized that he also poses a personal risk, so he began to train other leaders under him who act with a similar mentality and stewardship during acquisitions.

💡The aforementioned risk unfortunately materialized, in late September 2025 Mark Leonard probably suffered a stroke, which forced him to resign as CEO, and was replaced by Mark Miller, COO and head of the Volaris group. Mark Leonard will remain chairman of the board.

From the above, it is clear that there are many excellent capital allocators in the company, who were raised on the creed of Mark Leonard, which is why it is relatively easy to spin off the original structure and create a similar organization to the one that Constellation has. These are also publicly traded companies and operate with a very similar logic to Constellation Software (TSX:CSU):

🏢Topic (TOI): was the first spin-off in 2021 and focuses on European VMS companies. It is effectively a European CSU with a smaller market cap, created from the merger of Topicus BV and TSS group. Through TSS, Constellation Software (TSX:CSU) owns roughly 40% of the ownership, but also more than 50% of the voting rights.
🏢Lumine Group (TSX:LMN): The second spin-off was in 2023, CSU spun off Lumine from Volaris OG and merged it with a company called Wide Orbit, which they acquired at that time. The two companies were registered as Lumine Group, a Canadian company. Not VMS, but an acquirer with a telecommunications and media focus. CSU also owns 33% of this, but here too they have more than 50% of the voting rights.
🛢️Computer Modeling Group (TSX:CMG): is not owned by Constellation Software (TSX:CSU), but CEO Mark Miller is involved in the company's operations as a board member. Andrew Pastor, who is a board member of CSU, owns 26% of CMG through his EdgePoint fund. They travel in VMS-like, gas reservoir simulation software. The CMG is particularly interesting because here you can observe how they are building the company almost from scratch by filling leadership positions with former Constellation Software (TSX:CSU) members, such as Birgit Troy, Vipin Kulhar, Mohhamad Khalaf, and establishing the classic Constellation structure. I also analyzed the company, you can find the article here: Computer Modelling Group Stock Analysis (TSX:CMG)
🧑🏻‍💻Sygnity SA: Sygnity is a Polish IT services and software company, TSS, a European operating group of Constellation Software (TSX:CSU), acquired majority ownership, roughly 73%, in 2021. VMS acquires companies.
🇵🇱Asseco Poland: also a Polish company, in which Topicus has a roughly 25% stake.
🏛️Togetherwork: is not a publicly traded company, but a private sector company, similar to Constellation Software (TSX:CSU). John Billowits, the chairman of Constellation Software, is also a board member of Togetherwork, which is the connection between the two companies.
🇦🇺Kelly Partners Group (KPG): Australian financial advisor and accountant network. Lawrence Cunningham, vice chairman of Constellation Software, is also on the KPG board. They also believe in the model of eternal growth plus decentralization, but they operate in the accounting industry, not VMS.

I'm almost certain I couldn't untangle the entire Constellation Software (TSX:CSU) network of contacts, but if anyone knows of any other names, please let us know.

🧩Benefits of operating at Constellation Software (TSX:CSU)

Although the above may not be fully understood, as a serial acquirer grows, not only does the amount of capital to be deployed increase, making the deployment increasingly difficult, but the company structure must be designed to achieve a lower and lower level of decentralization. This is very important because it is not the amount of capital available that will hinder growth in the long term, but the number of expert capital allocators the company has. That is why Constellation Software (TSX:CSU) has always strived to develop as many people as possible in its corporate structure who are capable of conducting an acquisition with the appropriate quality and to ensure that acquisitions can be carried out as low as possible, even at the BU level.

source: Exploring Context blog, the number of stores acquired by serial acquirers

The chart above shows how many acquisitions each serial acquirer was able to make in a year. Constellation Software (TSX:CSU) stands out from the crowd, and its spin-offs are not included in the picture. It is clear from this that in the case of most serial acquirers, a few senior executives decide on acquisitions, which is why even the largest ones are not able to make more than 20-30 acquisitions per year, with the exception of Constellation, which you can see on the Y axis. This results in the longer term that they are forced to increase the size of acquisitions, which in turn drives down the previously mentioned IRR, which is now expressed as the 3-year average EBITA value, which you can see on the X axis. The above is of course true not only for CSU, but also for its spin-offs.

☝🏻So one of the unrivaled advantages of Constellation Software (TSX:CSU) is the organizational structure and the team of professionals that allow them to increase the number of acquisitions much more than their competitors. This is a very difficult attribute to copy, because the underlying logic of the companies would have to be changed in order to compete with CSU, which I think is almost impossible.

source: Exploring Context blog, the internal rate of return of serial acquirers

The other idea that comes into play here is that the internal rate of return and thus the performance of serial acquirers fundamentally depends on two things:

📊 from the aforementioned internal rate of return (in the picture this is ROIC)
🌱 and the organic growth they can achieve in acquired companies
🔁 IRR + organic growth = one of the benchmarks for serial acquirers (e.g. Bergman & Beving uses a different one)

The first has already been discussed, so it is worth saying a few words about the second. Many serial acquirers acquire another company because they expect operational efficiency improvements and organic growth due to synergies. In fact, the internal rate of return and the sum of organic growth are often considered benchmarks for serial acquisition companies, and Mark Leonard also uses this. So, if the acquirer acquires a company with a high internal rate of return, which necessarily depends on the acquisition price, then the organic growth can be zero or negative! The acquirer can still achieve a high return on it. This is shown in the image above, where Constellation Software (TSX:CSU) once again stands out from the crowd. By the way, the image only takes into account half of the organic growth due to conservative scenarios.

🌊Resources to understand how Constellation Software (TSX:CSU) works

Since I think it's very important that you not only read the in-depth analyses, but also that my articles have some educational nature, I've gathered some resources for those who would like to do further research on Constellation Software (TSX:CSU), or any other serial acquirer, for that matter. On the one hand, I recommend the materials of REQ Capital, who operate two funds that mainly hold serial acquirers, and on the other hand, they have created several very solid studies on the operation of serial acquirers. What's more, they have also written a book called The Compounders, which you should also read if you can:

🔗REQ Capital: A Deep Dive into Shareholder Value Creation by Acquisiton-Driven Compounders (2023)
🔗REQ Capital: A Deep Dive into Shareholder Value Creation by Acquisiton-Driven Compounders (2025)
🔗REQ Capital: Lessons from Acquisition-driven Compounders
🔗REQ Capital investor letters
🔗The Compounders (book)
🔗Studying serial buyers (Substack article)

The above roughly 1000 pages of material can be researched and not only gained a deeper understanding of Constellation Software (TSX:CSU), but also analyzed dozens of other serial acquisition companies.

💰How does Constellation Software (TSX:CSU) make money and what market advantages does it have?💰

In this section, we examine what exactly the company does to generate revenue, what products and services it has, how indispensable they are. Does it have any competitive advantage (economic moat), how defensible is it, and whether the trend is decreasing or increasing, and what is likely to happen in the long term.

As I mentioned, Constellation Software (TSX:CSU) acquires small, vertical software companies and reinvests the free cash flow from those acquisitions into new acquisitions. It's a complicated process, but here are a few metrics to help you understand what Constellation Software (TSX:CSU) is doing:

📈 number of companies acquired: It primarily provides information on whether or not more acquisitions have been made compared to previous years,
🏢 Size of acquired companies: larger companies allow for more capital to be invested, but reduce the internal rate of return,
💰 valuation of the acquisition: how much they paid for the acquisitions. The less, the higher the return you can expect,
💵 percentage of cash placed: Because the company generates a lot of cash, there are years when they cannot deploy 100% of it, but there are also years when they deploy previously generated cash.

One more thought about cash on hand: Constellation Software (TSX:CSU) has to put away not only the cash it generated in a given period, but also the amount it didn't put away from previous years, so this number can be higher than 100%. However, it can't put away all of its cash, because then it wouldn't be able to pay other expenses and the company wouldn't be able to operate. Therefore, we consider it 100% when the cash actually generated in the previous year was invested. This is the cash from investing line, and within it, a cash acquisition section, which shows how much capital has flowed into acquisitions.

Of course, CSU also tries to improve the efficiency of the acquired companies, so organic growth, although it is usually not very high, is also worth following. The company reports its revenues in the following segments (as of December 2024):

📜license revenues: ~3.8%
💼professional services: ~19.2%
🖥️hardware and other revenues: ~3%
🛠️maintenance and other recurring revenues: ~73.8% (recurring revenue, which is why they say that more than 70% of CSU is recurring revenue)

When it comes to territorial distribution, most people think of the Canadian and US markets when it comes to Constellation Software (TSX:CSU). However, this is not entirely true, as the revenues are as follows:

🇨🇦Canada: 9.2%, $929 million
🇺🇸USA: 44.7%, 4504 million USD
🇬🇧🇪🇺UK/Europe: 32.7%, $3293 million
🌍Other: 13.1%, USD 1341 million

As you can see, the company has significant exposure in Europe, although the traditional name for European Constellation Software is Topicus (TOI), which is a spin-off of CSU. Of course, since Constellation Software (TSX:CSU) has acquired over 1000 small companies, which are typically monopolies or oligopolies in their own markets, diversification is out of the question. It's even possible that some companies will decline or go bankrupt over time, which is a natural process, as Constellation Software has been acquiring companies for 30 years. But, each of these has an impact of less than 1 thousandth of the company's total.

One more thought about micro-business growth: analysts usually expect 2-3%, but there have been all sorts of growth rates between 0 and 6% in the past few years. Obviously, this is not the driving force behind Constellation Software (TSX:CSU), but a little extra growth is always welcome.

source: fiscal.ai Constellation Software (TSX:CSU) stock price drops

Constellation is a typically countercyclical company. Due to the recurring nature of their revenues and the stickiness of their services, these are business-critical software that is difficult to replace. A very good example of this is that since 2006, i.e. the IPO, there has not been a drop in share prices of more than 30%, regardless of whether there was a crisis or not. In other words, the current economic cycles have a relatively small impact on business, which is also reflected in the price.

💡Typically, these are very sticky, super-critical software for a very low amount, typically 1% for an amount of around are provided to partners, so they are relatively cheap compared to their importance, and therefore they are very rarely replaced for cost-cutting reasons.

This also means that Constellation Software (TSX:CSU) has a partner retention rate of roughly 90%, although this number is no longer disclosed as it used to be.

🏰Economic moat🏰

In this segment, I examined whether the company has any economic competitive advantage, which Warren Buffett referred to as the “economic moat,” which deters competitors from besieging the company’s fortress, i.e. its business, and taking its market. In this case, these could be the following:

🫸Cost/scale advantage: none in the traditional sense. These tiny VMS companies are less likely to grow, but they are monopolies in their own markets. However, because Constellation Software (TSX:CSU) groups similar companies, cross-selling can occur.
🫸Switching cost: very high. The products of VMS companies are very sticky, the rate of recurring revenue is extremely high, and in these niche markets, the acquired companies are often in a monopoly position.
🫸Network effect: none.
🫸Intangible assets, know-how, trademark: yes, very strong. Constellation Software (TSX:CSU) has a culture and a team of professionals that are virtually unrivaled. The more skilled people there are who are skilled at finding, selecting, and successfully executing acquisitions for VMS companies, the stronger an acquiring company is. CSU also has a wealth of data, with tens of thousands of companies in its database, and 30 years of acquisition experience aggregated within the company.
🫸Barriers to entry: very high. Although it seems that all you need to acquire a company is capital, which is why many private equity and venture capital funds are taking this path, think of the SPAC institution for example, yet most of them typically fail. First, the acquiring holding company must have the capabilities mentioned above, i.e. the team of experts, the data, the selection methodology and the relationships built over the years. This certainly cannot be bought from the market, it takes decades to build such capabilities. Despite this, there are many acquiring companies on the market, you can read about them in the Competitors section, but their efficiency does not even come close to the performance of Constellation Software (TSX:CSU).

Constellation Software (TSX:CSU) is a very special, extremely diversified, non-cyclical company that has been able to repeat its successes for 30 years, virtually unchanged. It is surrounded by one of the strongest economic moats in terms of corporate culture that I have ever seen. However, I must also mention that in some places there is a bit too little data available to adequately assess how strong this moat is, a good example of this is the discontinuation of the disclosure of the above partner retention rate, or the disclosure of data on all acquired companies in quarterly reports. However, I think this is primarily because countless copycat companies have emerged over the years, against which Constellation Software's (TSX:CSU) business operations must somehow be protected, but this may also mask any decline in profitability indicators.

☝🏻It is also difficult to see what unique competitive advantage each of the 1000+ companies has, and I think it is essentially unanalyzable on an individual basis. Since Constellation Software's (TSX:CSU) economic moat is an average of these, and there are both weaker and stronger companies, most sources attribute only a narrow moat to the company.

However, it should not be forgotten that due to the high number of acquired companies, the risk is also very dispersed, The decline of one of these has very little impact on the company. Since the VMS market is not cyclical, in many cases CSI is more protected against market impacts than companies with a different model. The longer-term problem for Constellation Software (TSX:CSU) may be that as it grows, like all similar companies, think Berkshire, it finds it increasingly difficult to deploy its capital, so it has to acquire larger targets over time, which reduces efficiency. This could reduce internal rate of return, which in the long run could force Constellation Software (TSX:CSU) to venture beyond the VMS markets into uncharted territory. However, I think this will last at least 10+ years, so there's no need to worry about it for now.

🎢Metrics of Constellation Software (TSX:CSU)🎢

In this section, I examined what metrics characterize the company, how it stands on the revenue side, what margins it operates with, whether it has debt, what the balance sheet shows. I look for items that are extreme – too high debt, high goodwill, etc. - what return on capital the company works with, what its cost of capital is, how the revenue and cost sides are structured. I also examine trends, owner value creation, and how the company uses the cash generated.

📈What is the S&P 500 yield?📉

Compared to previous analyses, I have introduced a new section to compare the metrics below. Since many people use the US stock market index as a benchmark and also buy S&P 500 ETFs, it is worth looking at what companies are doing in aggregate (obviously, you should be happy if the company you are analyzing outperforms these values).

S&P 500 2024 data:

SP&500 revenue growth: +7%
SP&500 profit growth: +10%
SP&500 gross margin: 45%
SP&500 net margin: 13%
SP&500 ROE: 15%
S&P 500 ROIC: 12%
S&P 500 ROCE: 11%

Constellation Software (TSX:CSU)'s metrics are sometimes quite difficult to decipher. There are no colorful prospectuses like other companies, and CSI doesn't even issue quarterly forecasts, which is quite right. Since acquisitions are ongoing, it can be a bit difficult to tease out exactly what the company spent money on, and the holding structure can also mask certain things. So I recommend always reading the exhaustive and not very well-structured quarterly and annual reports, because they have everything in them, you just have to dig a lot.

Let's start with the usual, revenue, which here means the aggregate revenue of all companies. You don't have to look at too much in the picture, the three-year revenue growth average is 23.3%, while the five-year is 23.8%, which is shockingly rough. Of course, this is not organic growth, only Constellation Software (TSX:CSU) shows revenue growth from acquisitions as a whole. Don't let the last LTM column fool you, this is the average of the last 3 quarters, so the company can probably see a similar percentage growth this year.

Constellation Software (TSX:CSU) revenue — source: fiscal.ai, revenue of Constellation Software (TSX:CSU)

Returning to organic growth, in the image below you can see exactly this, the important ones are the purple and green graphs, the first one is from maintenance, and the second one is the total growth. The image shows that over the past five years, organic growth has varied between -3 and 5%, with an average of ~2.5%, so apparently this is not where Constellation Software (TSX:CSU) gets its additional revenue, but it also shows that revenue from software maintenance and updates is growing by an average of 5-6%. Why is this important? Because this accounts for 74% of total revenue, and since 90% of customers are returning customers, it is clear how difficult it is to replace Constellation Software (TSX:CSU) programs.

Constellation Software (TSX:CSU) organic growth — source: fiscal.ai, organic growth of Constellation Software (TSX:CSU)

Here are the margins of Constellation Software (TSX:CSU), which are very interesting compared to the analyses you may be used to, you can read these here. Almost all values are low, on paper, if you just look at it, you might think that this is not a good company. Let's quickly summarize why, for example, the net profit margin might be low. It's because it's an accounting issue, and acquisitions are recorded on the expense side, so in this metric, the accounting costs of acquisitions eat into the book profit. This is not a corporate weakness, but an operational necessity.

Constellation Software's (TSX:CSU) margins — source: fiscal.ai, Constellation Software (TSX:CSU) margins

That's why I've included Constellation Software's (TSX:CSU) FCF/net profit ratio in the image below, which shows how much cash is generated compared to the profit that the company can reinvest. For those who aren't familiar with this ratio, Google, Microsoft, and Nivida also have FCF/net profit ratios below 100% in most years. How can it be so high for Constellation Software (TSX:CSU)? So:

📉 Mature VMS companies are characterized by low growth but very high profitability,
🏦 Many monopoly VMS companies need very little money to maintain their business, but they generate a lot of cash,
💸 VMS companies often receive money from customers in advance, so they finance the development of functions, which creates negative working capital,
📊 goodwill amortization, which is not included in the free cash flow, i.e. FCF indicator, you need to look at the descriptions for this.

The above perfectly supports what Constellation Software (TSX:CSU) does: it acquires cash cow companies that are very difficult to replace.

Constellation Software's (TSX:CSU) Margins 2 — source: fiscal.ai, Constellation Software (TSX:CSU) margin 2

To avoid getting bogged down in the topic of goodwill and write-offs, I looked at how much of the goodwill generated by the capital raised by Constellation Software (TSX:CSU), that is, intangible assets such as business value, brand strength and other intangible assets, had to be written off in the books due to overpayments. Remember what I said earlier? The high internal rate of return, or IRR, of Constellation Software (TSX:CSU) is supported by the very low acquisition price, as can be seen in the images below. They almost never pay more for anything than it is worth, so there is no need to write off the surplus value on the books. These are excellent numbers and indicate excellent use of capital.

goodwill of Constellation Software (TSX:CSU). — source: fiscal.ai, goodwill of Constellation Software (TSX:CSU)

Here's the usual debt ratio, which is also quite special in the case of Constellation Software (TSX:CSU). I compiled the data based on the status as of December 2024:

💰income: 10742 million USD
🤑profit: 641 million USD
🫰🏼cash: 2575 million USD
💸net debt: $2582 million (~24% of revenue)
💶net debt/EBITDA: ~ 1.3
👛interest coverage, EBIT/interest: 6.5

The above numbers would be scary for any other company, but here several things greatly cloud the picture. On the one hand, the debt of the acquired companies is consolidated by Constellation Software (TSX:CSU), and on the other hand, they issue corporate bonds, such as a $500 million senior note at an interest rate of 5.158%. Why? There are two reasons:

🏦 acquisitions are not always cash-only, but are often secured with credit
⚙️ the loan acts as leverage, as they achieve a much higher return on it than the interest on the loan

So there's no need to worry about this, debt actually acts as a catalyst in the compound interest mechanism. Of course, you have to pay attention to the amount, but the current one is not terrible at all. I have read in several places that part of the negotiated purchase price will only be paid later, under other conditions, if the company meets certain indicators, so it is actually a deferred payment. The longer the time horizon, the more time Constellation Software (TSX:CSU) has to generate cash.

Constellation Software (TSX:CSU) debt — source: fiscal.ai, debt of Constellation Software (TSX:CSU)

🧮What do ROIC and ROCE metrics show?🧮

ROIC – Return on Invested Capital – shows how efficiently the company uses its total invested capital to generate profit. Read more here.

It shows the company's fundamental value creation capability.
It filters out the impact of the financing structure.
If ROIC exceeds the cost of capital (WACC), the company is creating value.

ROCE – Return on Capital Employed – shows how efficiently the company uses its long-term financing sources. Read more here.

It measures the profitability of business activities.
It does not take into account tax effects.
A good basis for comparison between different industry players.

Indicator	What does it measure?	When is it considered good?
ROCE	Total return on capital	If higher than the industry average
ROIC	Return on invested capital	If higher than WACC
ROE	Return on equity	If stable and sustainably high

Constellation Software (TSX:CSU) Ownership Value Creation

On the owner value side, I usually look at what the company uses the free cash generated. Basically, a company can do the following things with cash:

reinvests in the business (almost zero growth)
reduced debt (debt acts as leverage here)
pays dividends (very rarely, in the form of a single dividend)
buys back shares (never bought back)
buys up other companies (base case)

Constellation Software (TSX:CSU) is a textbook case of capital allocation. As long as a company continues to grow its business, either through what we call organic growth or through acquisitions, it can generate very high internal rates of return. Until then, there is no point in paying it to investors, as the surplus will be priced into the share price in the long term, so it can be taxed as a capital gain. This is significantly more efficient than taxing dividends, but the company still pays a very small dividend of 0.1%. Sometimes they also pay special dividends, the reason for this is that if there is not enough target, they will pay the money to investors at a better price, but this is quite rare.

ROIC, ROE, ROCE metrics for Constellation Software (TSX:CSU). — source: fiscal.ai, ROIC, ROE, ROCE metrics of Constellation Software (TSX:CSU)

As for share buybacks, Mark Leonard often calls their biggest mistake the IPO, when they issued shares to let their previous investors exit the company. This has cost the company a terrible amount of money over the past decades, as this stake would be worth billions of dollars today. I don't completely agree with the fact that they don't buy back shares, I think this is a particularly lucrative thing for shareholders if the stock is trading below its fair value. This is almost never true for CSU, at least in the last 10 years it has been very difficult to find a time when the company was truly undervalued, so there was no reason to allocate capital for it. Constellation Software (TSX:CSU) has had a constant share count since its IPO.

source: fiscal.ai, ROIC, ROE, ROCE metrics of Constellation Software (TSX:CSU)

Finally, a chart that illustrates very well why Constellation Software (TSX:CSU) does not look like a good company at first glance. The shareholder yield is low, because what does it consist of? Debt reduction, dividend payment, share buyback. The first of these is used by Constellation Software (TSX:CSU) as leverage, the latter two are negligible or non-existent, so they cannot appear in the metric. What is not visible in this case? The high internal rate of return on capital employed. And this also brings me to why it is difficult to debug certain indicators in the case of Constellation Software (TSX:CSU).

☝🏻Constellation Software (TSX:CSU) metrics are difficult to interpret

I'll go back a bit to the first picture showing ROE, ROIC, ROCE values, because you can see that the numbers are constantly deteriorating. Before anyone panics, this is because acquisitions really distort metrics like ROIC:

🤔Distortion of goodwill and intangible assets: Constellation Software (TSX:CSU) is constantly acquiring smaller software companies. With each acquisition, a portion of the purchase price is recorded as goodwill and intangible assets such as software licenses, customer relationships, brand name, etc., as I mentioned earlier. These do not generate direct cash flow, yet they are included in the “capital invested”, thus diluting the denominator of the formula, therefore the ROIC will be artificially low.
🤔The capital of acquired companies doubles when Constellation Software (TSX:CSU) buys a company, its book value "restarts" at the purchase price. Due to accounting revaluation, the total invested capital appears higher than the actual working capital used. The result: an apparently low return, even though the internal rate of return relative to cash flow is actually high.
🤔Deferred revenue and negative working capital: Software companies bill upfront, e.g. annual subscriptions, so they often work with negative working capital. The classic ROIC calculation does not always handle this well, thus underestimating the return on investment.

In contrast, in the case of ROCE, goodwill and intangible assets are often excluded, or only capital committed to actual operations is taken into account. Since Constellation operates the acquired companies extremely efficiently, ROCE gives a much more realistic picture of capital efficiency, and therefore moves at much higher values. If you are interested in the formula, you can search for it online, but it is a very serious accounting distortion that can be easily misinterpreted at first glance.

From this perspective, the ROE value is the best, but even that can be a bit misleading. The point of the analysis is that if the decline in ROE persists for several years, while cash flow and acquisition volume also slow down, then it may indicate that:

❗it is more difficult to find targets that yield 20–30% IRR,
❗the market is saturated with easy-to-buy vertical software companies,
❗or management is forced to resort to larger, lower-yielding acquisitions.

Then ROE truly reflects that The efficiency of capital deployment is decreasing, as Mark Leonard mentioned. This phenomenon is being referred to in many places as CSU 2.0, which means that sooner or later Constellation Software (TSX:CSU) will be forced to make much larger acquisitions or venture beyond the market stack of VMS companies.

💵Constellation Software (TSX:CSU) Acquisitions💵

In this section, I examine how acquisitive the nature of the company is and what impact each acquisition had on the life of the company, if any.

Constellation Software (TSX:CSU) acquires about 100 smaller companies and a few larger ones each year, so unlike other companies, there’s no point in listing the 1000+ companies. What matters is how much return CSU can generate on them. Actually, we’ve already covered this in Chapter 2, but to quote from the previous In Practise article:

There are three hurdle rates. Less than a million in revenue is 30% IRR; above four million, you can drop to 20%; and 25% for everything in between. You can go to 20% because those bigger businesses are more competitive, but everything that falls in the middle at 25% is 90% of deals. There is another hurdle rate of 15%, but that is on deals so large they're not in CSI's wheelhouse. Above 50 million, you can go to 15% but they might only do one of those per year.”

So the point is that:

🪙For companies with less than 1 million USD in revenue, the internal rate of return is 30%,
💰 Between 1-4 million USD 25%,
🤑while above 4 million USD the IRR drops to 20%.

90% of acquisitions are businesses with an IRR of around 25%, which means that if they were to invest the cash generated in a year in companies of this size, they would have to acquire hundreds of million-dollar companies, compared to roughly 100 acquisitions in the busiest year.

This means that a second internal indicator had to be created, the aforementioned IRR, which applies to companies with revenues greater than 50 million USD, but we do not know what the upper limit of this is. In the list below you can see the largest acquisitions, and these are not acquisitions of 50 million USD, but much larger ones. Even 2 of the large acquisitions below are enough to deploy the annual cash, although not every year there is such a significant acquisition, as the quote above indicates. Let's see how the larger acquisitions turned out:

source: Andvadi Associates, Q4 2023 Newsletter, Constellation Software (TSX:CSU) major acquisitions

The picture shows that valuations have clearly increased with larger acquisitions, but in many cases, revenue data cannot be gleaned from a reliable source. However, since Constellation Software (TSX:CSU)'s internal rate of return depends on it, it is very important to monitor them. Based on how CSU has grown in recent years, this will certainly be a challenging period.

source: fiscal.ai, Constellation Software (TSX:CSU) revenue growth over 3/5 years

If a company maintains its current revenue growth, it will have to spend more and more money on acquisitions, so the balance sheet will inevitably shift towards higher-value acquisitions. Of course, it is impossible to calculate exactly how much money Constellation Software (TSX:CSU) will spend, because it depends on:

❓the amount of cash used,
❓from loans taken out, as purchases can be made not only with cash,
❓and also the number of suitable targets.

Constellation Software (TSX:CSU) debt and cash-based acquisitions — source: fiscal.ai, Constellation Software (TSX:CSU) debt and cash acquisitions

I don't think anyone can estimate exactly how much capital will be raised by Constellation Software (TSX:CSU), as the identities of the targets are not public, but the image above clearly shows the spread in the loans taken out and the cash spent on the acquired companies.

🤵Constellation Software (TSX:CSU) management🤵

In this section, I examine who runs the company and how. What is the bonus system, how much risk – skin in the game – do the managers take on while running the company? Is there a family connection, or perhaps a special “heritage” factor?

Constellation Software (TSX:CSU) is led by an outstanding management team, even though Mark Leonard has been in the CFO's shoes since his departure in September 2025. In fact, the management team is excellent primarily because they have carried on Mark Leonard's teachings, so it is essentially the corporate structure that makes the leaders truly strong. Since Mark Leonard also posed a kind of personal risk, having previously been responsible for executing acquisitions, he designed CSU's corporate structure so that the subunits could make acquisition decisions at the lowest level possible.

The situation is a bit like Berkshire Hathaway, where Warren Buffett and, while he was alive, Charlie Munger were the intellectual leaders, but as they got older, Greg Abel was entrusted with operational tasks and capital allocation. Todd Combs and Ted Weschler have previously received similar assignments, but it is a completely normal process that as a company grows, the organization develops exceptionally good managers. This happened in the case of Constellation Software (TSX:CSU), but here the network of connections between the board and management must also be taken into account. It is also worth connecting countless companies that share the Constellation creed but are not directly part of Constellation Software to the members of the management.

⚙️ Mark Miller – Chief Executive Officer and Chief Operating Officer (CEO, COO)

👔 In 2025, he took over Mark Leonard's role in managing day-to-day operations and capital allocation, effectively becoming the CEO of Constellation Software (TSX: CSU)
🏛️ Longtime CSU veteran who carries on Leonard's philosophy while also being responsible for major acquisitions and organic growth
🔗 Also a member of the board of directors, thus a direct link between management and the board of directors

📌In practice: Mark Miller previously worked at Trapeze Group, a company that became a transportation and transit-focused software company. He was one of the owners of Trapeze and a key figure in its product development. When Mark Leonard founded Constellation Software in 1995, his first acquisition was Trapeze. This made Miller's company the first acquisition in Constellation's history. Thus, Miller and Leonard had been working under the same roof, at least indirectly, since 1995. Miller later gradually became involved in Constellation's central management: in 2001 he became the company's COO, and has since played a decisive role in the operational and acquisition strategy. Since he is already the CEO, COO, board member and head of the Volaris group, this also suggests that he may be giving up some of his duties.

It is also important to note that Mark Miller is also a board member of Computer Modelling Group (TSX:CMG), so the two companies have a relatively close relationship. I also analyzed the company: Computer Modelling Group Stock Analysis (TSX:CMG).

💰 Jamal Nizam Baksh – Chief Financial Officer (CFO)

💼 Coordinates the finances and bond financing of more than a thousand subsidiaries.
💰 Responsible for consolidated financial stability and cash flow management.
🏦 while he is in charge of the company's loan portfolio and debt management. Since capital allocation is very important here, he is also a very skilled leader. He has been with Constellation Software (TSX: CSU) since 2003, is a qualified accountant, but also has a degree in mathematics.
➡️ Overlap: also a board member.

📊 Bernard Anzarouth – Chief Investment Officer (CIO)

💼Professionally responsible for the company's acquisition and investment decisions.
💰He leads the valuation models, target company selection, and financial due diligence.
🏦He will share capital allocation responsibilities with Mark Miller, continuing Leonard's previous role. He is the company's professional responsible for acquisition and investment decisions, having previously held senior roles at IBM.

🏗️ Farley Noble – Deputy Managing Director, Large Acquisition Group

💼Constellation's big hunter. He is the one who identifies, prepares and carries out acquisitions worth hundreds of millions of USD.
💰His role has grown in recent years as CSU has gradually opened up to larger transactions. He has been with the company since 1999, previously serving as CFO and head of the Volaris group.

📌In practice: If you recall the organizational chart of Constellation Software (TSX:CSU), you would have seen 6 operating groups. One of them was the Volaris group, led by Farley Noble. Lumine was spun off from this group when it merged with a company called Wide Orbit in 2023 and was listed as a separate company under the name Lumine and the ticker TSX:LMN.

You can read more about the roles of management members here: CSI management team.

🏛️ Constellation Software (TSX:CSU) Board of Directors

I omitted Jamal Baksh, who is also a member of management, from the list, and I also did not include Mark Leonard as a former CEO due to his resignation in September 2025.

👤 Mark Leonard – Founder, former Chairman and CEO

🐐Founded Constellation Software in 1995, he is the intellectual father of the company and the creator of its culture. He stepped down as CEO in September 2025 and handed over the baton to Mark Miller.
✨He built the legendary buy and hold forever model, which was used to acquire over a thousand vertical software companies. He is also responsible for the spin-offs of Topicus (TOI) and Lumine Group (TSX:LMN).
💫Although he has now stepped back from the role, he remains a member of the board, providing strategic direction while delegating day-to-day management to others. He is also credited with shaping the creed of the other leaders, as the Constellation Software (TSX:CSU) model functions as a kind of religion.
Mark Leonard hasn't taken ANY pay from Constellation Software (TSX:CSU) in 20 years. Fun fact: he's so humble that when he has to fly, he chooses economy class.
➡️ Overlap: He was previously the CEO, currently a board member and founder.

📌In practice: It's hard to explain what a talented investor Mark Leonard is, the easiest way is to read his investor letters, which will give you a deeper understanding of his creed: CSU investor letters.

📚 Lawrence Cunningham – Vice Chairman of the Board of Directors

💼Experienced financial leader, previously CEO of Home Capital Group and member of the board of directors of the non-listed company Togetherwork.
💰A professor at George Washington University and renowned investment writer, known for his study of Warren Buffett and Berkshire Hathaway, which was also published in book form: The Essays of Warren Buffett.
🏦Cunningham is a representative of long-term value creation, shareholder thinking, and ethical capital allocation on Constellation's board, a kind of philosophical assurance for the Buffett-style direction.

📌In practice: He is also a board member of not only Constellation Software (TSX:CSU), but also Kelly Partners Group (KPG) and Markel Group, the former of which is also a serial acquirer in Australia, acquiring accounting firms.

👔 John Billowits – Chairman of the Board of Directors

☝🏻An experienced financial executive, he was previously CEO of Home Capital Group and a member of the board of directors of the non-listed company Togetherwork, as is John Billowitz.
🎮Its task is to maintain strategic control of management and balance of governance.

📊 Andrew Pastor – Independent Director

🤔Partner at EdgePoint Investment Group, previously worked as an analyst and portfolio manager.
💡 At Constellation, he brings an investor mindset and a "rational use of capital" approach to the board.
📈 He specifically shares Mark Leonard's philosophy of building compounder-type companies.

📌In practice: Andres Pastor's role is actually interesting because EdgePoint Investment owns 26% of Computer Modeling Group (TSX:CMG) through the investment fund.

💼 Claire Kennedy – Independent Director

⚖️Senior Advisor at Bennett Jones LLP, formerly an attorney and partner, with extensive experience in legal, regulatory, and corporate governance issues.
⚖️ Strengthens ethical, compliance and legal controls on Constellation's board, especially during international expansions.
🌍 One of the most respected women in the Canadian business community.

💸 Donna Parr – Independent Director

👨🏻‍⚖️Experienced investment professional who has held leadership roles at several Canadian private equity and institutional investment funds.
💼 Previously, he was a manager at Northwater Capital, with extensive experience in software and service industry transactions.
📊 Capital market and acquisition experience on the board Because of this, he plays a key role and is often involved in the audit of M&A decisions.

💻 Laurie Schultz – Independent Director

⚙️Tech industry veteran with over 25 years of leadership experience in the software industry (Sage, Intuit, Galvanize).
🚀 As the former CEO of Galvanize, he sold his company to Diligent, so he has first-hand knowledge of software M&A processes.
🧩 Strengthens the operational, product and customer-focused approach in the board of directors, counterbalancing the financial orientation.

🧠 Robert Kittel – Independent Director

🫰🏻Financial and strategic advisor, previously a portfolio manager and corporate executive.
🧠He has experience in both corporate restructuring and long-term investment decisions.
💬 One of the voices of the discipline of risk analysis and capital utilization on the board.

You can read more about the roles of the board members here: CSI board of directors.

A few thoughts on the above: In May 2025, a number of former board members were not re-elected, such as Jeff Bender, Robin van Poelje, and Susan Gayner, one of the former two being a director of the Harris Group and the other of Topicus. From what we have seen so far, the network of connections between members and their companies is incredibly thick, and there are incredibly experienced professionals among the members. There are board and management members who have been working in this industry for 20-30 years, as well as many senior executives of spin-offs spun off from Constellation Software (TSX:CSU), the parent company, or who have interests in similar serial acquisition companies. Personally, I think this is the best management I've seen so far.

💰Constellation Software (TSX:CSU) Compensation Plan

Constellation Software (TSX:CSU)'s compensation is a prime example of how the long-term interests of the company and management can align. There is virtually no compensation structure beyond salary, short- and long-term incentives, or stock options like other large companies. Instead, Constellation Software (TSX:CSU) pays out relatively large cash bonuses, but 75% of that bonus is paid to employees through the purchase of Constellation Software (TSX:CSU) shares. The number of shares has remained unchanged since the IPO in 2006, and the company has never issued additional shares.

This is brilliant because in the short term, employees don't really benefit from bonuses, but in the long term, they benefit greatly from stock appreciation. This has resulted in the following over the past 30 years:

💼 founder Mark Leonard owns 1.86% of the company, worth $1.5 billion,
🏗️ former COO and Perseus OG CEO Dexter Salna owns 1.2% of the company, worth approximately $976 million,
👔 CEO Mark Miller owns 1.06% of the company, worth $872.5 million,
📊 CIO Bernard Anzarouth owns 0.7% of the company worth $575 million,
🧩 Former board member and Jonas OG CEO Barry Alan Symons owns 0.62% of the company, worth approximately $507 million,
👥 Approximately 3000 Constellation Software (TSX:CSU) employees own shares in the company.

It's hard to imagine a greater motivation for management and employees to do their best for the success of Constellation Software (TSX:CSU) than this.

💭Interesting information about the management of Constellation Software (TSX:CSU)

There are few companies that don't provide forecasts regarding their annual revenue and other expectations, but Constellation is one of them. The reason for this is that the company's core belief is that it focuses exclusively on long-term value creation, so it doesn't want to give short-term tips, which I find particularly welcome. It's completely irrelevant what effects the company will have in a quarter or two, what's important is how it is managed over many years.

They don't hold quarterly conferences either. On the other hand, if there is an impact that threatens the business, the company is willing to dedicate a separate investor seminar specifically to this. Such was the conference call held by Mark Leonard on September 22, in which he spoke about the effects of artificial intelligence, you can read more about this in the risks section. You can access it here: impact of Artificial Intelligence technologies (“AI”) on software businesses.

🆚Competitors: Constellation Software (TSX:CSU) opponents🆚

In this section, I examine who the competitors of the analyzed company are, what is their market position, whether they are in a subordinate, secondary or superior role. What is their market share and what is their specialty? Are they losing or gaining market share to their competitors?

The strange thing about Constellation Software (TSX:CSU) is that while there have been many copycat companies over the years, very few have been able to replicate the success that Constellation Software (TSX:CSU) was able to achieve. In effect, its biggest competitor has become itself, due to its own growth in size. That's why they started spin-offs, which is how Topicus (TOI) and Lumine Group (TSX:LMN) were created, which are essentially small Constellation Softwares, the former in the European market and the latter in the media company market. Since serial acquisition companies are virtually unanalyzable using traditional metrics, it is quite difficult to understand how effectively they operate, and this is also true for competitors.

Constellation Software (TSX:CSU) opponents — source: Interactive Brokers, Morningstar, opponents of Constellation Software (TSX:CSU)

However, there are quite a few opponents, the question is whether we take serial acquirers as a basis or companies operating in the VMS software segment. Both approaches can be correct. I vote for the former, while Morningstar mixed the two approaches a bit, as you can see in the picture above. Cisco Systems and Oracle are both horizontal software manufacturers, which I think is quite difficult to classify as either VMS or serial acquirers. Roper Technologies (ROP share subpage) is a serial acquirer, but it makes much larger, more transformative acquisitions:

🧩 Roper Technologies (ROP) Key Acquisitions

Vertafore (insurance software, 2020): $5.3 billion
Deltek (project management, 2016): 2.8 billion USD
Frontline Education (education administration software, 2022): $3.7 billion
CIVIC Technologies, Aderant, Data Innovations, CliniSys: all VMS-type software companies.

👉 These transactions are much larger than what Constellation usually does, but the philosophy is the same: Acquiring niche, sticky, high-margin software and keeping it forever. However, there is not really a common cross-section due to the different acquisition targets.

However, I think the key here is the company structure, so you need to find companies that make acquisitions in series according to a similar logic, and Roper Technologies actually fits into this. There are dozens of companies on the list of serial acquirers, but I would like to show you a few pictures, which are not all active in the VMS segment, but they are telling.

Statistics from the book The Compounders — source: REQ Capital, statistics from the book The Compounders

In the image above you can see large acquiring companies, along with their annual and total market cap growth and when they entered the market. However, here you run into the same problem I mentioned earlier, they are not competing in the VMS market:

🧱Bergman & Beving: wholesale and distribution of industrial products and tools, its model is very similar to that of Constellation Software (TSX:CSU). B&B's spin-offs are Addtech, Addlife, Lagercrantz, Indutrade and Lifco! too. Swedish company.
✈️Heico: aerospace and defense supplier, also a serial purchaser.
🔬 Judges Scientific: scientific and laboratory equipment. A British holding company that buys up small, high-profit lab and measurement equipment manufacturers and then holds them for the long term. Its philosophy is almost 1:1 to Constellation's model, only in the industrial hardware sector.

As you can see, the list above actually includes a total of three companies, as Bergman & Beving has also spun off a large number of companies. Are these the rivals of Constellation Software (TSX:CSU)? Hardly, but once they leave their own market segment, they could even enter the VMS market with the same philosophy as CSU. The image below is of the Exploring Context Substack blog You can see a list of serial acquisition companies:

source: Exploring Context, list of serial acquirers

Many of the names are familiar, but there are a few new ones. The biggest of these are Danaher Corporation and Ametek, but Heico and Rollins are also quite large. But if you look behind the scenes, you will again find that Rollins (ROL), which has been acquiring pest control companies, will probably never be a real competitor to Constellation Software (TSX:CSU), which has interests in VMS software companies.

The following list was compiled by Scott Management LLC., who ranked companies based on total shareholder return. The source is a bit old, but it gets the point across:

Scott Management LLC, serial acquirer list — source: Scott Management LLC, list of serial acquirers

Almost all of the names were on previous lists, but there are some new ones. Vitec Software is very similar to Constellation Software (TSX:CSU), acquiring VMS companies, but in the Scandinavian market. SP Group is in the plastics industry, while Rotork is active in the pumps and valves market, so they are not really its competitors either. Although not in the picture is the Australian company Kelly Partners Group, which was mentioned earlier, but they also follow a very similar Constellation Software (TSX:CSU)-like model, only in the accounting and financial services market.

🧾 Summary – CSU DNA Map

Category	Examples	CSU similarity
🧠 Pure decentralized serial acquirers	Indutrade, Lifco, Addtech, Addlife, Lagercrantz. Judges Scientific, Vitec Software	80-100%
💼 Centralized capital allocator holdings	Roper, Danaher, Ametek, Heico, Transdigm	60-75%
🧱 Smaller decentralized roll-ups	Diploma, Halma, Teqnion, Tyler Tech, Rentokil, SP Group, Rotork	40-60%
⚙️ Integrated growth companies, partly from M&A	OpenText, SiteOne, Waste Connections, TopBuild	30-50%

⚡What are the risks of Constellation Software (TSX:CSU)?⚡

In this section, I examine all the risks that could affect the company's long-term future. Currency, regulatory, market disruption, and so on.

Constellation Software (TSX:CSU)'s risks stem mostly from its operating model, but it also provides protection against a number of common risks. So, unlike what I've seen in my previous stock analyses, these are not mentioned here in many cases, which I'll discuss later. But let's take a look at the problems the company faces:

👨🏻‍⚖️ Mark Leonard and the Cult Problem

💡 The philosophy of the CSU is decentralized to an almost religious level.
🏗️ The entire system was invented by Leonard, including capital allocation principles, return discipline, and trust in a model built on decentralization.
⚠️ If Leonard or his direct disciples, such as Anzarouth or Miller, withdraw, cultural erosion may occur: the company may slowly transition to a Danaher/Roper-like, more centralized system.
👉 An analogy could be: Berkshire Hathaway after Buffett. The financial engine keeps turning, but the philosophy is gradually disappearing.

📌In practice: I've actually covered the above issue before, and I think that in an organization this large, Mark Leonard has never been the only one responsible for capital allocation. It's no coincidence that they created the 6 operating groups, the OGs, and under them the business units, the BUs. They've been responsible for capital allocation for years. However, large acquisitions are decided at a higher level, so if Mark Miller gets involved in a major, multi-million dollar bad deal, it could erode returns. However, I think that if there is a company that is capable of cultivating good capital allocators at the organizational level, it is Constellation Software (TSX:CSU).

📉 IRR pressure due to larger acquisitions

📈 As Constellation grows in size (~60 billion CAD market cap, 10+ billion in revenue), it simply can't close enough micro deals to meaningfully impact the company's growth.
💰 That's why larger transactions come (Topicus spin-off, Altera, Optimal Blue, etc.), where a realistic target is 15–20% IRR, not 30+.
📉 The aggregate IRR is therefore likely to gradually fall below 20% as the big fish effect takes hold.

📌In practice: I think this is the biggest risk for Constellation Software (TSX:CSU), which while not fatal, will inevitably lead to a deterioration in the return profile over the long term. We have already seen this in the case of Berkshire, on the one hand, the return has decreased proportionally to the increase in size, and on the other hand, it is sitting on USD 344 billion in cash or equivalent securities because it cannot place it.

The big question is how long this will take. Based on revenue growth of around 20% per year, if Constellation Software's (TSX:CSU) IRR drops to around 20%, it will still perform much better than the SP500 average, so I think it will still have many years to run. And if that's not enough, there is also the possibility of spin-offs as a lifeline.

🔎 VMS companies are running out

🌍 There are probably hundreds of thousands of vertical software companies operating in the world, CSU only owns a little over a thousand.
⚙️ However, most niches that produce high-quality, stable cash flow have already been acquired, so the quality of the stores may also deteriorate.
🚀 This is offset by new spin-out operator groups (Volaris, Jonas, Perseus, etc.), which hunt for new companies locally and autonomously.
💡 So it's not the market that's running out, but the very cheap but high-quality stores. Over time, CSU will probably normalize to a 15-20% IRR.

📌In practice: I don't think this is a terrible risk at all in the case of Constellation Software (TSX:CSU). The situation is a bit reminiscent of the case of Dino Polska (WSE:DNP), where I tried to estimate the saturation of the market and came to the conclusion that it will definitely not happen there for another 10 years or so. You can read his analysis here: Dino Polska SA Stock Analysis (WSE: DNP; ADR: DNOPY).

Here, the situation is even better, as according to some rumors, there are 40-50000 companies in the CSU database, and if 1000 of these have been acquired in three decades, and this pace is even accelerating, we are still talking about decades.

🏗️ Organizational complexity

🏢 Over 1000 subsidiaries, 100000+ customers, dozens of operational groups.
⚠️ The biggest meta-risk is that decentralization itself becomes unmanageable.
🔄 The flow of information may slow down, incentives may be distorted, or hidden bureaucracy may develop.
💬 This is the paradox of Constellation: the more successful decentralization, the harder it is to maintain.

📌In practice: Many multinationals face a decline in efficiency because what works on a small scale no longer works on a large scale. It is impossible to control all capital allocators in an organization of this size, so management must be very good at conveying the philosophy to the lower levels of the organization.

🧠The impact of artificial intelligence on VMS software

The CSU portfolio includes over 1000 subsidiaries, most of which are vertical market software (VMS): medical billing, parking, train scheduling, accounting software, legal management systems, etc. Their characteristics include:

⚙️ business critical,
🧲 difficult to replace,
⚖️ they operate in a regulated environment,
🏛️ not markets that are quickly adapting to AI: e.g. municipal, education, utilities
🧠Constellation Software (TSX:CSU) businesses are currently experimenting with 15000 unique AI models, according to letters from Akre Management, which is affiliated with Chuck Akre, so the company has already started integrating them at the enterprise level.
➡️ So in the short term, AI is not a threat, but can marginally improve development efficiency (e.g. code generation, customer service automation).
CSU's philosophy is to own companies forever anyway, so they don't rush after new technologies; the stability of 20-30 year old customers is their main value.

📌In practice: As for how seriously Constellation Software (TSX:CSU) took the above issue, Mark Leonard spoke for about an hour in a webinar a few days before his illness about the impact of artificial intelligence on VMS software. He also emphasized that no one knows exactly what will happen in this market in the future, but:

☝🏻Most VMS software is unique, tailored to the needs of customers, so it is not really possible to generate a uniform code for them,
☝🏻It is much more likely that AI tools will be integrated by software provider Constellation Software (TSX:CSU) than that customers will replace Constellation's applications,
☝🏻In many cases, the market is surrounded by regulations, for which VMS software contains unique solutions. Such organizations find it extremely difficult to replace their software and are especially reluctant to implement changes that cause downtime.

So, the point is that artificial intelligence tools, which I have been using for years, can be used excellently for general things, a great tool for increasing efficiency. It can also generate general codes very well, just yesterday I had ChatGPT write table codes for the WordPress engine, but it is much more difficult to apply them to special tasks. There is also another effect that applies to the VMS software market: Software that is difficult to replace is necessary for basic operation, which is why many places still use ancient, legacy software that no one likes to mess with. This is especially true for the state and local government sectors.

💡Moreover, it is often not worth developing the code with AI, because the underlying operating logic definitely requires an expert due to the specialties of the small market, but the number of available users is low. So the development will never actually pay off, it is much more likely that Constellation Software (TSX:CSU), which has already developed the software, will integrate AI tools into its own operations.

In contrast, for horizontally integrated software companies, where a single software can reach a huge market, the application of artificial intelligence can be a much more lucrative business. A good example of this is the market for Adobe's image editing software, which we also wrote about in our analysis of the company: Adobe Inc. Stock Analysis (NYSE: ADBE).

🔐Risks not affecting Constellation Software (TSX:CSU)

Constellation Software (TSX:CSU)'s operating model means that they are simply not affected by many of the risks discussed in my other company reviews. I think this greatly distorts the picture, and I'll list a few in a telegraphic style:

🌍Customs, geopolitics: There are no physical goods that need to be customs cleared, and there is no reason to ban their software because they are too small for that.
💱Foreign exchange risk: It has minimal impact on the company, and most of their costs and revenues are generated in USD.
📉Non-cyclical company, recession-proof: The VMS software industry is not cyclical, as is evident from Constellation Software's (TSX:CSU) share price declines, which have never been greater than 30% in the 20 years since its IPO, but their revenues are also consistent. Since software is very difficult to replace, attrition is low and the rate of recurring revenue is high.
🚢Supply chain problems: they don't sell a physical product, they are not affected by this problem.
⚔️Direct competitors: We discussed it above, but there are hardly any companies of this size operating in the VMS industry, even serial acquirers. Vitec, Valsoft, Roper Technologies and so on, but they are also much smaller.
⚖️Regulatory risks, regulation: Since they only make smaller, non-transformative acquisitions, the authorities that regulate the market and protect competition do not often block merger deals.
🔥Inflation: VMS software is a business-critical, hard-to-replace system that accounts for only a very small portion of the costs of user organizations, so Constellation Software (TSX:CSU) can pass on price increases caused by inflation to its customers in the vast majority of cases.

I made a self-check list that confirms the thesis about the company:

low or zero debt: YES/PARTLY/NO
significant economic advantage that can be protected in the long term: YES/PARTLY/NO
excellent management: YES/PARTLY/NO
excellent indicators, significant owner value creation: YES/PARTLY/NO
The majority of the total return comes from reinvesting the cash generated, not from dividends: YES/PARTLY/NO
appropriate company valuation: YES/PART/NOT

I think that we can safely say that Constellation Software's (TSX:CSU) management, culture, operating model, and capital utilization are at a world-class level. This essentially eliminates the main risks of a company. Beyond its internal problems, however, the way the company operates brings with it one more thing: it usually trades at a very high apparent valuation. In other words: not only is it expensive, but in the vast majority of cases it is also difficult to analyze because traditional valuation metrics do not tell you much about the company, as you will see in the next chapter.

👛Constellation Software (TSX:CSU) Valuation👛

In this section, I will examine the company's current valuation compared to historical values and consensus fair values.

Rating metrics

In the two rows below you can see valuation metrics. The first row shows the current valuation, the second row shows the historical valuation. While I don't think these metrics are particularly good - they hide a lot - they can be used as a benchmark, in general. However, in the case of Constellation Software (TSX:CSU), they are very misleading, see the explanation below:

Share price (2025-10-11) 2815 USD; P/E: 95.82; EV/EBITDA: 24.37; P/FCF: 24.97(Based on Finchat.io)
~~Historical median valuation (10-year average): P/E: 62.32; EV/EBITDA: 22.45; P/FCF: 26.17(Based on Gurufocus)~~

Why don't you see a DCF model in this segment? Because each input data produces a huge variance in the output, and most of the data is an estimated value. Therefore, the valuation will never actually be a single exact number, but rather a range can be defined where the current valuation falls.

You should apply a margin of safety to this price range, according to your risk appetite.

So don't expect an exact price, no one can say this for a stock. However, there are fair value prediction services, almost every major stock screening site has one, I've aggregated them below. However, if you want a good stock support service, subscribe to The Falcon Method (The Falcon Method), entry prices are given for the stocks analyzed there.

Rating (TSX:CSU is the Canadian ticker, CNSWF is the ADR)

Wall street estimates: 2474-3869 = 3172 USD (I took into account the Alphaspread, the average of the two extreme values:)
Peter Lynch Median P/E: $2050
Morningstar: $3731 (4 stars)
Gurufocus: $3600
AlphaSpread: 2959 USD (5% undervaluation compared to the base case)
SimplyWallst: $3568

Average (based on 6 reviews): $3180 (11% underrated)

How to interpret the numbers? The above “margin of safety” rule should be applied according to your convictions, so if you really believe in the company, you can buy it at fair value, but if you proceed in 10% increments (whose convictions are strong), the math would look like this:

10% margin of safety: 3085*0.9=2777 USD
20% margin of safety: 3085*0.8=2468 USD
30% margin of safety: 3085*0.7=2160 USD
40% margin of safety: 3085*0.6=1851USD
50% margin of safety: 3085*0.5=1543 USD

Of course, the list could be continued indefinitely, but the point is that the right purchase price for you is determined by the level of your conviction. However, in the case of Constellation Software (TSX:CSU), it doesn't really make sense to wait for a very big drop in price, since it hasn't fallen more than 30% since it's been on the market. It typically did this from an overvalued level, roughly to its fair value. So you should be very happy even if the stock is trading at its fair value.

source: fiscal.ai, Constellation Software (TSX:CSU) price appreciation and cash flow

I left out the NOPAT yield because in the case of Constellation Software (TSX:CSU), acquisitions distort NOPAT. And that brings me to the problem that CSU is not a traditional business, but a capital allocator, so for example, the P/E will be terribly distorted because the net income is massively below the real cash generation. There are similar problems with EV/EBITDA and P/FCF ratios, which is why I pulled them up. I mentioned earlier that Mark Leonard and Constellation Software (TSX:CSU) both use IRR, but you can't evaluate the company based on it because it's not a public number. So, for lack of a better word, I looked at how much the increase in free cash flow per share correlates with the increase in the share price. The answer is that it does, it's worth looking at the CAGR numbers here, the share price is growing a little faster, due to the premium pricing.

In addition to the above, there is also the issue of how to factor in the ownership stakes held by Constellation Software (TSX:CXU) in spin-offs into the overall picture. The difficulty of truly valuing such a complex serial acquirer using traditional models is well illustrated by the discussion on page 61 of the REQ Capital summary, part of which I have translated into Hungarian to make the point clear:

"The chart below shows Constellation Software's stock price over the past 10 years, excluding small dividends and a 2019 special dividend. The stock price has risen 41% annually over that period, supported by more than 30% annualized earnings growth. The gray line shows the price at any given point in time that would have earned a 500% annualized return.

source: REQ Capital, Constellation Software (TSX:CSU) price increase

This clearly shows that CSI has always been significantly undervalued and, as we will show below, it still is today. If we had paid CAD 862 per share ten years ago, when the stock was trading at CAD 69, we would have achieved a 10% annual return by now. And if we had paid CAD 1657 three years ago, when the stock was trading at CAD 866, we would still be achieving a 10% annual return today.

The question today is what kind of earnings growth is expected over the next ten years. If CSI can maintain its impressive 30% compound annual growth rate (CAGR) over the next decade, the stock is currently significantly undervalued. In this historical scenario, it would be reasonable to pay as much as CAD 13776 per share and still achieve a 10% annual return over the next ten years. The company grew its earnings by 31% in 2020. Even at a much slower growth rate of 15%, a valuation of CAD 3624 would be justified, which is well above the current market price.

We believe that the right factors, reinvestment opportunities, scalability and optionality (new verticals or even completely new businesses beyond the software industry), are all in place to ensure the company continues to grow strongly and profitably over the next ten years. This is why CSI is a core investment in our portfolio."- REQ Capital: Lessons from Acquisition-driven Compounders.

I found a great saying online, which I think is very true for CSU: Constellation is not to be appreciated, but to be understood. But also a little bit of faith that it will continue the 17224%, or 30.4% annual price increase it has achieved in the past two decades. I have been following this company for five years, when the stock was trading around 1000-1100 USD, so everyone should decide whether to believe in it or not (despite the fact that the past is not a good indicator for the future).

🌗Significant news and the last quarter🌗

In this section, I will examine what happened in the last quarter, whether there were any significant news/events. If the company reports semi-annually, we examined this period.

Below is a summary of Constellation Software's (TSX:CSU) 2025Q2 earnings. As usual, there are no forecasts or targets, except that they stick to the same creed as before:

💰 Constellation Software (TSX:CSU) Q2 Financials

Indicator	2025 Q2	Change
Income	2.844 million USD	+ 15 %
Organic growth	5% (excluding FX 4%)	—
Net profit	56 million USD	-68%
EPS (diluted)	2.66 USD	-68% (2024Q2: $8.35)
Operating cash flow (CFO)	433 million USD	+ 63 %
FCFA2S	220 million USD	+ 20 %
Quarterly dividend	$1/share	unchanged

💼 Constellation Software 2025Q2 other events

Constellation Software Inc. further strengthened its position in the second quarter of 2025, mainly through growth achieved through acquisitions.
The company's total revenue grew by 15% to USD 2844 million, of which 5% was organic growth (4% adjusted for currency).

Although revenue increased significantly, net income attributable to common shareholders fell 68% to $56 million. This decline was primarily driven by the company's aggressive acquisition activity, with new acquisitions totaling $469 million during the quarter, of which $380 million were cash payments.

Despite the decline in profitability, cash flow generation remained strong:

💵 operating cash flow increased by 63% to USD 433 million,
📈 Free cash flow available to shareholders (FCFA2S) increased by 20% to USD 220 million.

No conclusion in the world can be drawn from the above, as Mark Miller said in his webinar on 2025-10-01:

"Over time, capital deployment has gone up and it's gone down. I cannot attribute it to any specific reason. I feel very comfortable with what we're doing as far as covering our potential opportunities to deploy capital and just continue to try to improve upon that. I have no major insightful answer to that question, Richard." –Mark Miller, CEO

More important than the above are the two events that occurred this quarter:

🧠 Mark Leonard held a webinar on 2025-09-22 about the impact of artificial intelligence on the company.
🏥 Mark Leonard resigned on 09-25 due to health reasons and will be replaced by CFO Mark Miller. Mark Leonard will remain with the company as a member of the board of directors.
📢 Mark Miller gave a briefing on 2025-10-01 regarding what will change after Mark Leonard steps down. In very simple terms: nothing, everything continues as before.

"I think one of the things that, as Constellation has grown and as our operating groups have grown, we tend to spend more time, and this would be the operating group managers and Mark, studying the larger investments we're making and trying to support each other and thinking through how to do those and how to do those better. I think that won't change. I will spend time with any of our larger capital deployment opportunities and really focus on very much the same things Mark was looking at, compensation, making sure that we're optimizing our compensation to our team across the board, as well as looking at continuing to share best practices among the operating groups as we're seeing evolution in technology, for example, with artificial intelligence, more capital deployment, or just different things that we see as an opportunity to improve our businesses.” –Mark Miller, CEO
"I actually had the opportunity, of course, speaking to all of the leadership team after the announcement was made. I have to admit my comfort level increased as I spoke to the team. I think we've got a great team around the table at Constellation and inside of Lumen and Topicus. I want to continue to just work with them. There are no plans to change any of the management team at all." –Mark Miller, CEO
"Well, we're not considering buying back shares. I don't think there's any, you can extrapolate into that other than, you know, both John, the Chairman, and myself acquired some shares. We really just felt it was sending a signal to people that we were very comfortable with the company and its long-term future. That's all that's to be read into that.” –Mark Miller, CEO

To reinforce the above consistency, there has been a lot of insider buying after Mark Leonard's illness, to emphasize that Constellation Software (TSX:CSU) executives: nothing has changed compared to before.

The following report: 2025.11.06.

✨Other interesting facts about Constellation Software (TSX:CSU)✨

Everything that was left out of the previous ones, or if there is any special KPI - key performance indicator - or concept that needs to be explained, is included here.

Akre Capital Management: Constellation Software (TSX:CSU) and its executives are considered legends not only by me, but also by other big-name investors. Akre Capital Management, led by the extremely well-known investor Chuck Akre, owns 16% of its total holdings in Constellation Software (TSX:CSU) and Topicus (TOI), which represents a brutally high commitment to the group of companies. They share their thoughts in their letter to investors dated 2025-09-30, which you can read here: Akre Focus Fund.

Other funds whose newsletters mentioned Constellation:

(I.e.Andvari Associates, 2023 Q4
(I.e.Sequoia Fund, 2024 year-end

Chris W. Mayer and the 100 Baggers: One of my favorite books about investing: Chris W. Mayer:100 baggersChris Mayer does a great job of describing what makes a company stand out. Chris keeps a focused portfolio, usually 8-10 names in his base, which is: Woodlock House Family CapitalHow does this relate to Constellation Software? Chris's portfolio includes CSU, Topicus, and Lumine, but he also held positions in Teqnion and Heico, which are all serial acquisition companies.

Constellation Software (TSX:CSU) real value creation: Although I didn't elaborate above, CSU actually owns a large stake in its spin-offs, which are also serial acquirers. What better way to invest capital than in a company that does the same thing as the parent company, except they also let the compound interest work? Mark Leonard would probably call this meta-compounding, or compound interest compound interest, if he were to say it, because this is how the internal returns add up. What is very important is that CSU either has a large stake in these companies or has majority voting rights, but this is not priced into the stock in most cases.

Since the market typically evaluates consolidated financial results, I believe these are not included in the company's fair value because they are underrepresented in accounting. For Constellation Software (TSX:CSU) a self-replicating capital management network, but the market prices it as if it were just a software holding company.

🔑Key Performance Indicators (KPIs)🔑

Hurdle rate and IRR: In the case of Constellation Software (TSX:CSU), these are the two metrics you really need to watch. I talked about it a lot above, and I can say the same thing here: deploying as much capital as possible, with the highest possible return, into new acquisitions, that's the point.

Summary of Constellation Software (TSX:CSU)

Summary of the analysis, drawing lessons.

Constellation Software (TSX:CSU) may not seem like an interesting company at first glance, but the more you delve into it, the more their genius will come through. In fact, CSU does nothing more than skillfully reinvest the cash it generates through acquired companies. And this is precisely the problem for the company, that due to the uniqueness of their model, they will eventually outgrow their own market and be forced to make larger and larger acquisitions, which will worsen their returns.

It is also a big question how Mark Miller, who will take over after founder Mark Leonard resigned for health reasons, will replace him, but there are also AI fears that could affect the vertical software market. Because of this, the exchange rate fell by about 30% at the end of September 2025, which can be a good entry point if you think everything will continue as it is.

I think understanding the company, and basically the serial acquirers, is a tough nut to crack, which is why investors often avoid these opportunities. The same is true for valuing companies, the usual metrics don't really work, it's hard to gather all the spin-offs and other interests and understand the full web of relationships. It's quite a deep dive into investing if you bet on serial acquisition companies, but in the past two decades, you could have been rewarded with a 17500% price increase, or ~30% per year, in the case of Constellation Software (TSX:CSU). The million-dollar question is whether this will continue or whether a decline will begin.

Frequently Asked Questions (FAQ)

💼 What should you know about Constellation Software (TSX:CSU)?

Constellation Software Inc. is a Toronto-based, Canadian technology holding company that owns more than 900 software companies specializing in vertical markets worldwide. Founded in 1995 by Mark Leonard, Constellation Software Inc. has become one of the world's most successful and largest serial acquirers. The company acquires stable, mission-critical software and holds it for the long term, based on the principles of decentralized operations and disciplined capital allocation.

🧭 What is Constellation Software's (TSX:CSU) mission statement?

Constellation's philosophy is that capital should always be placed where it will yield the highest return with the least risk. Mark Leonard and his team do not strive for short-term profit maximization, but for capital to continue to grow over decades. The company's motto is practically this: Buy, hold, and never sell.

🔁 What spin-offs did Constellation Software (TSX:CSU) have?

CSU has so far created two major spin-offs: Topicus (2021) and Lumine Group (2023). Both operate as serial acquirers with the same decentralized philosophy as the parent company. Topicus mainly acquires European VMS companies, while Lumine specializes in telecom and media technology software. Constellation has retained significant ownership in both companies, so it earns double interest on its capital through them.

👨‍💼 Who is Mark Leonard?

Mark Leonard is the founder and longtime CEO of Constellation Software. He was a former venture capitalist who founded the company in 1995 with the goal of assembling small, stable software companies focused on vertical markets. Leonard rarely appears in public, rarely giving interviews, but is known for his philosophical, long-term investor letters that echo the style of Berkshire Hathaway and Warren Buffett.

🧠 Who is Mark Miller?

Mark Miller is the current CEO of Constellation Software. He has been with the company for more than two decades, having previously earned his reputation as the head of the Jonas Software division, one of CSU's largest operating units. Miller grew up directly under Mark Leonard and continues Leonard's capital allocation philosophy: decentralized management, low debt, high-return acquisitions.

⚔️ What competitors are worth mentioning in relation to Constellation Software (TSX:CSU)?

CSU has few direct rivals, as few others have adopted the roll-up model based on vertical markets so extensively. The most similar companies are Roper Technologies, Vitec Software Group (Sweden), Valsoft Corporation (Canada), and to a lesser extent Lifco AB and Teqnion AB (both Swedish industrial holding companies). These are all Constellation-like capital raising companies, but none of them reach the level of CSU in terms of size, IRR, and philosophical depth.

📈 What has been the annual price growth of Constellation Software (TSX:CSU) so far?

The company has posted a compound annualized price growth rate (CAGR) of over 30% since its IPO in 2006 (TSX:CSU). The stock price started at around CAD$70 and has now risen to over CAD$4000. This represents a nearly 60-fold increase in less than two decades, making it one of the strongest long-term performers in the software sector worldwide.

🤖 How is artificial intelligence (AI) impacting Constellation Software's (TSX:CSU) business?

AI is not a direct threat to Constellation at this time, as the majority of the company’s portfolio consists of business-critical, hard-to-replace software (e.g. healthcare, utilities, education). However, CSU’s businesses are already experimenting with over 15,000 different language models and have access to the same LLMs as AI-native companies. AI is more of an efficiency tool for them than a competitive risk, the real challenge is how to integrate it into the decentralized structure.

💼 What well-known investors own shares of Constellation Software (TSX:CSU) or its subsidiaries?

Constellation and its spin-offs (Topicus, Lumine) are favorites of several well-known compound interest investors. Chuck Akre, founder of Akre Capital Management, is one of its most well-known shareholders, who recognized CSU's capital allocation power early on. Chris W. Mayer, author of 100 Baggers, has also written several times about Constellation as the ideal compound interest machine. In addition, several Canadian and American funds, such as EdgePoint, Baillie Gifford, and members of the In Practise analysis community, also follow and own shares in the company.

Which broker should I choose to buy shares?

There are several aspects to consider when choosing a broker - we will write a complete article about this - but I would like to highlight a few that are worth considering:

size, reliability: The bigger a broker, the safer it is. Those with a banking background – Erste, K&H, Charles Schwab, etc. – are even better, and well-known brokers are typically more reliable.
expenditures: Brokers operate with various costs, such as the account management fee, the portfolio fee - which is the worst cost -, the purchase/sale fee and the currency exchange cost (if USD is not deposited in the brokerage account)
Availability of instruments: It doesn't matter which broker has which market available, or whether they add the given instrument upon request and how quickly.
account type: cash or margin account, the latter can only be used for options. For Hungarian tax residents, having a TBSZ account is important, but citizens of other countries also have special options – such as the American 401K retirement savings account – which are either supported by the broker or not.
surface: is one of the most underrated aspects, and it can be a real pain. Anyone who had an account with Random Capital, a now-defunct Hungarian broker, knows what it's like to work on a platform left over from the 90s. Erste's system is lousy slow, Interactive Brokers requires a flight test, and LightYear believes in simple but modern solutions.

Based on the above, I recommend the Interactive Brokers account because:

the world's largest broker with a strong background
a few million instruments are available on it, and shares listed on multiple markets – e.g. both the original and the ADR – of a single share are often available
Interactive Brokers a discount broker, they have the lowest prices on the market
you can link your Wise account to them, from which you can quickly transfer money
Morningstar's analyses are available for free under the fundamental explorer (good for analysis)
EVA framework data is available under fundamental explorer (useful for analysis)
they have both cash and margin accounts, Hungarian citizens can open a TBSZ
you can use three types of interfaces: there is a web and PC client and a phone application

Legal and liability statement (aka. disclaimer): My articles contain personal opinions, I write them solely for my own entertainment and that of my readers. The articles published here do NOT in any way exhaust the scope of investment advice. I have never intended, do not intend, and am unlikely to provide such in the future. What is written here is for informational purposes only and should NOT be construed as an offer. The expression of opinion is NOT in any way considered a guarantee to sell or buy financial instruments. You are SOLELY responsible for the decisions you make, and no one else, including me, assumes the risk.

Constellation Software Stock Analysis (CSU) – In Series