For two decades, Constellation Software (TSX: CSU) has been the closest thing public markets had to a perpetual compounding machine.

Under the stewardship of Mark Leonard — a reclusive former venture capitalist turned operator — the company executed one of the most disciplined and successful vertical market software (VMS) serial acquisition engines in modern history.

From a modest Canadian software consolidator, CSU evolved into a sprawling network of over 1,500 downstream companies, each operating in niche verticals with high switching costs and mission-critical workflows.

The formula was like clockwork:

• Acquire small VMS companies at low multiples
• Decentralize operations and preserve entrepreneurial autonomy
• Harvest predictable free cash flow
• Deploy capital into the next acquisition
• Continually repeat…

This wasn’t just M&A — it was a well oiled capital allocation engine in hyper drive.

But in 2025 – 2026, something started to change…

The tone from long-time investors wasn’t what it used to be. Quietly at first, then more openly; the market was beginning to question whether the flywheel at Constellation Software could still spin as effortlessly as it once did for the past 20 years.

The Bear Case: Why Conviction Is Cracking

1. End of the Mark Leonard Era

Leonard’s transition away from day-to-day leadership — and more importantly, from the board — represents a structural break.

This is not a typical CEO succession story.

Leonard wasn’t just setting strategy; he defined the culture, underwriting discipline, and capital allocation philosophy. His letters became required reading across the investment community, often compared to those of Warren Buffett.

Now, the wizard is gone.

As highlighted in the latest ROE Newsletter by Jason Donville, President DKAM – a long time investor in CSU who recently exited his position entirely in 2025 between $3,250 – $3,760 – the concern isn’t just about Leonard’s abrupt exit but compounded by the fact that Constellation Software is becoming increasingly more challenging to model. And the very real threat of AI disruption is not helping the outlook.

2. The Flywheel Is Losing Torque

Donville goes on to write in his newsletter that historically, CSU thrived on a powerful structural advantage:

• Acquire VMS businesses at ~0.8x revenue
• CSU trades at 4–6x revenue (a handsome multiple)
• Reinvest internally generated cash at similar spreads

This created embedded multiple arbitrage, supercharging returns.

But that engine is evolving.

The Shift: Minority Investments

CSU has increasingly allocated capital into minority stakes in public and private companies.

The Problem:

• No operational control
• No guaranteed access to cash flows
• No ability to redeploy capital internally

In other words, less flywheel, more portfolio-investing.

This is a fundamental shift from operator to allocator.

And markets are rational to question whether the same premium multiple applies.

3. The AI Conundrum: A Portfolio Too Big to Valuate

Finally, Donville laments that boasting over 1,500 businesses across dozens of industries; CSU has become too vast and complex to model at the edge.

That was once a strength.

Now it’s a risk.

AI introduces asymmetric disruption:

• Some VMS businesses may become significantly more valuable (data + workflow ownership)
• Others may be quietly obsoleted by AI-native challengers

The issue isn’t direction — it’s uncertainty.

If even 10–20% of the portfolio faces structural erosion, the compounding math changes materially.

And unlike a focused SaaS company, CSU cannot easily communicate this exposure.

It’s a black box at scale.

The Bull Case: Why the Compounding Machine May Still Work

1. Decentralization Was the Point All Along

Bulls argue the market is over-indexing on Leonard.

Constellation is not a centralized empire — it’s a federation of capital allocators.

Its six major operating groups (including Topicus, Lumine, Harris, Jonas, etc.) already function semi-independently.

Over decades, CSU has effectively trained hundreds of operators to think like disciplined investors

The argument:
Leonard didn’t just build a company — he built a system.

2. The New CEO – Mark Miller – is an Insider

With Mark Miller stepping into the CEO role, the market is now recalibrating around a different kind of leader.

Miller is not an outsider parachuted in to “professionalize” the business. He is a Constellation insider, having spent years inside the organization, most notably as a senior executive within operating groups like Volaris.

In many ways, he represents the purest continuation of the CSU playbook — a capital allocator trained inside the system rather than a strategist imposing a new one.

What Bulls Like

• Cultural continuity: Miller is steeped in CSU’s disciplined capital allocation and decentralized philosophy
• Operator DNA: Unlike traditional CEOs, he has direct experience overseeing acquisitions at the ground level
• No reinvention risk: There’s no indication he intends to “fix” what isn’t broken

What Bears Question

• Is he Leonard — or just Leonard-adjacent?
• Will underwriting discipline hold at scale without the founder’s final say?
• Does he have the authority to say “no” as often as Leonard did?

The real issue isn’t competence — it’s comparative mythology.

Leonard became larger than the company itself.

Miller inherits not just the role, but the expectation of near-perfect capital allocation.

That’s an almost impossible benchmark.

3. Vertical Market Software Still Wins

The core thesis behind VMS remains intact:

• High switching costs
• Deep workflow integration
• Low relative cost to customers
• Often mission-critical

Examples include software for:

• Municipal governments
• Healthcare providers
• Utilities
• Education systems
• Dealerships

These are not easily replaced.

Even in an AI-driven world, incumbents often win because:

• They own the data
• They control the workflow
• They have distribution

Firms like Morningstar have long argued that VMS companies may benefit from AI via upselling and automation.

4. Downturn = Opportunity

Constellation has historically thrived in fragmented, inefficient markets.

If AI or macro tightening creates distress among smaller software vendors, CSU could:

• Acquire at lower multiples than historical norms
• Re-accelerate organic and inorganic growth
• Reassert its dominance

In this framing, today’s uncertainty is tomorrow’s hunting ground.

The New Reality: A Valuation Reset

The real shift isn’t operational — it’s psychological.

For years, CSU enjoyed:

• Premium multiples
• Near-unquestioned trust
• “Buy at any price” status among quality investors

That era may be ending.

The market is now asking:

• What is the sustainable reinvestment rate?
• How repeatable is the acquisition model at scale?
• What is the true AI exposure?
• Who replaces Leonard as the philosophical anchor?

This is not a collapse.

It’s a repricing of certainty.

For another long-time investor in Constellation Software, though, the current climate of uncertainty isn’t changing his course: “I have been personally adding to CSU and TOI (for clients we are holding)”, says Jason Del Vicario, Portfolio Manager at Hillside Wealth, a division of iA Private Wealth.

In a conversation over email, Jason Del Vicario elaborated:

”The true moat at Constellation Software lies in their deep customer relationships and a culture of governance and incentives that is globally second to none. While some may view recent moves into public equities like SABR as a departure, CSU (through Topicus) has a proven track record of finding value where others don’t. They are built to excel at deploying capital effectively, potentially even beyond the VMS space in the long term. We remain long-term believers. Rather than taking a ‘victory lap’ on short-term price movements, the real test of a thesis is where the price sits in five to ten years. For those of us focused on high-quality businesses and perfect incentive alignment, we’re happy to step in when others decide to exit.”

For investors reassessing Constellation Software, which companies are you looking to deploy capital into going forward? For investors buying more or holding, what’s the rationale?

Regards,
Robin Speziale

Disclosure: own CSU

As I write this issue of the newsletter, I’m high up in the air and on my way to Las Vegas. Probably the worst destination to attempt any sort of wealth creation — but the sunny 30+ degree weather and escape from Toronto is drawing me in 🙂 I’m sure most of you know by now that I much prefer the proven long term way to build wealth: The stock market. And I’m not alone…

There are nearly 20 million Tax-Free Savings Accounts open in Canada as of March 2026.

Naturally, most are ordinary:

• Holding cash
• A few ETFs
• Bank stocks (that I call “Canada’s Magnificent 6”)

In fact:

• ~95% of TFSAs are under $100,000
• Only a tiny fraction exceed $200,000
• Just a few hundred have crossed $1 million
• Fewer than 30 Canadians have built TFSAs worth over $5 million

It’s those $1 million plus TFSAs that are the focus of The Globe and Mail’s “TFSA Trouncers”, a recurring series that uncovers a parallel universe where ambitious and bold DIY investors reveal how they parlayed up to ~$100K of lifetime TFSA contributions into seven-figure portfolios (note: the profiled investors must submit their brokerage statements to The G&M to verify their balances).

Let’s be honest: $1M+ TFSAs are not typical outcomes. They are statistical outliers.

But the stories behind these TFSA Trouncers interestingly all follow a similar script.

Read on to learn how these extraordinary investors turned $100,000 into millions — tax free.

1. The Shopify Engineer — ~$2M–$3M TFSA (20x–40x Return)

Core holding:

• Shopify

An engineer made one of the most consequential retail investing decisions and more than 20x’d his TFSA: he went all-in—early—on into Shopify.

And then:

• Ignored volatility
• Avoided trimming
• Let it run

As Shopify compounded, his TFSA:

• Crossed $1M
• Then climbed toward $2M–$3M

At a time when 95% of Canadians hadn’t even reached $100K, one stock alone – Shopify – created generational wealth.

2. The U.S. Tech Visionary — ~$1.5M–$2.5M TFSA (15x–30x Return)

Core holdings:

• Tesla
• NVIDIA

This investor bet on the most disruptive and innovative U.S. leaders.

The payoff:

• Tesla’s explosive run
• NVIDIA’s AI-driven surge

Result:

• A multi-million-dollar TFSA

In a country where only a few hundred accounts exceed $1M, this portfolio got there by concentrating in only two stocks.

3. The Elite Serial Compounder — ~$1M–$1.5M TFSA (10x–20x Return)

Core holding:

• Constellation Software

No moonshot.

No hype.

Just:

• Elite capital allocation
• Serial acquisition engine
• Long-term compounding
• Believed in the magic of Mark Leonard

This investor quietly crossed $1M+ placing them in the top fraction of 1% of all TFSA holders in Canada.

4. The Canadian Tech Stack Builder — ~$800K–$1.5M TFSA (8x–15x Return)

Core holdings:

• Lightspeed Commerce
• Kinaxis
• Enghouse Systems

This portfolio combined:

• Multiple Canadian tech growth names
• Occasional exposure to Shopify (for that added oomph)

Outcome:

• High six figures to low seven figures Canadian tech concentration (imagine that; no exposure to the Mag 7!)

Still an outlier — given that the vast majority of accounts never approach even half that size.

5. The Oil Cycle Insider — ~$700K–$1.2M TFSA (5x–10x Return)

Core holdings:

• Suncor Energy
• Cenovus Energy
• Canadian Natural Resources

This investor understood oil cycles.

Buying when oil was out of favor, then riding recovery:

• Doubles
• Triples
• 5-baggers

Outcome:

• Approaching or exceeding $1M

Again, placing them in elite TFSA territory — shared by only a few hundred Canadians nationwide.

6. The Junior Mining Jackpot — ~$1M+ TFSA (10x–50x Spike)

Core holdings:

• Undisclosed junior miners

Outsized TFSA gain typically driven by one major discovery.

One breakout. One life-changing position.

Multiple investors fit this category:

• Took small speculative bets
• Hit exponential upside in 1-2 junior miners
• Benefitted not just from new discoveries but commodity prices for precious metals (Gold, Silver, etc.) surging to the moon

Often:

• A single junior mining stock turned tens of thousands into hundreds of thousands — or more

In a system where millions of accounts stagnate, this is pure asymmetry – and what’s still possible in Canadian junior mining.

7. The High-Volatility MicroCap Investor — ~$500K–$1M TFSA

Core exposure:

• Microcaps
• Crypto-adjacent equities

This approach:

• Created massive swings
• Occasionally massive gains

But also highlights something critical: for every successful high-risk TFSA, there are many that don’t survive — and those blow-ups are rarely ever talked about.

8. The Late Bloomer (Retiree) — ~$1M+ TFSA (10x–20x Return)

Core holdings:

• Shopify
• NVIDIA

Instead of de-risking later in life, this retiree:

• Leaned into growth – with a cross border mix of technology superstar stocks

Outcome:

• million-dollar TFSA

Highlighting a striking contrast in a country where most retirees:

• Hold conservative, balanced portfolios – with bonds or GICs outweighing stocks almost always
• And never approach these outcomes

9. The Massage Therapist — ~$700K–$1M TFSA (10x–20x Return)

Core holdings:

• Shopify
• Tesla

No formal education in finance.

No network on Bay Street.

Just:

• Conviction
• Patience
• Discipline

Her TFSA climbed toward seven figures — placing her among the top 1% of all Canadian investors.

What we can observe through these nine case studies of $1M+ TFSAs is that each of the “TFSA Trouncers” profiled in the Globe & Mail followed a very similar playbook for outsized returns:

1. Typically one big bet drove the asymmetric outcome within the tech, junior mining, oil & gas, and micro-cap space

• Tesla, NVIDIA, and Shopify were the most common stocks

2. Concentration — not diversification — created wealth in a relatively short period of time

• Position size really, really matters

3. Holding — not trading — captured the gains

4. Canada and the US both presented opportunities to invest in market-beating companies (Go Canada!)

When the TFSA was first launched on January 1, 2009, it wasn’t designed to create millionaires.

And statistically — it doesn’t.

But for a tiny minority:

• One stock
• One decision
• One act of patience

Was enough to separate them from millions of other Canadians.

In a country where most TFSAs quietly compound — or sit idle — a few became something entirely different:

Tax-free engines of asymmetric wealth.

Whether you call these extraordinary stories “gambling” or not is up to your interpretation. For me, they’re nothing short of inspiring. That being said, I’m off to play the slot machines in Vegas — hope I get lucky!

Regards,
Robin Speziale

Disclosure: own SHOP, CSU, NVDA, TSLA

A $65M Canadian company is quietly building what could become the Dollarama of Mexico…

And almost no one is paying attention.

When most investors look at Pesorama they see:

– A small-cap TSX Venture stock
– A company still reporting net losses
– A retailer in an unfamiliar market

And they move on.

But that’s not where the real story is.

Pesorama is attempting to build a scaled discount retail chain in Mexico — a market with ~130 million people and an estimated 13,700 potential dollar-store locations (est. total addressable market).

Today, Pesorama operates 35 JOi Dollar Plus stores (see all of their listings on Google Maps here)

That’s less than 0.3% market penetration.

Is Pesorama in the early innings of becoming a potential retail compounder? The stock was up +180% (~3x) in 2025, and is up +36% YTD 2026. Let’s explore…

1. Store Growth: The Foundation of Everything

The entire thesis starts here.

Retail winners scale locations first, then optimize everything else.

Pesorama store count trajectory:

Year        Stores
2019        1 (founding year)
2020        ~10
2021        18
2022        21
2023        22
2024        25
2025        31
2026E       ~40-45 (35 as of Mar '26)

What this shows:

• Store acceleration is happening now, not later
• The jump from now → 50 stores is where scale starts to matter
• Nearing the phase where retail models can typically inflect

2. Revenue Growth: Proof of Demand

Growth isn’t theoretical — it’s already showing up.

Revenue (CAD)
2023   $14.5M
2024   $20.5M  (+41%)
2025   $23.5M  (+14%)
2026E  ~$25-$30M+  

Sales increased 14% in 2025 to ~$23.45M, continuing strong momentum.

What matters alongside the revenue growth rate:

👉 Growth is paired with margin expansion
👉 Same-store sales are still positive (~5–6%)
👉 Ticket sizes are increasing (~15–20%)

That combination is what separates early-stage retailers doing something right versus “story stocks.”

3. Margin Expansion: The Hidden Engine

Gross margins are continually improving.

Gross Margin
2023   ~40–41%
2024   ~42.4%
2025   ~44.5%
2026E   ~46%+

Margins increased from ~42.4% to 44.5% in 2025, with further expansion estimated to exceed ~46%+ in 2026.

Why this matters:

This is textbook retail scaling:

• Higher volumes → better procurement
• Better procurement → higher margins
• Higher margins → more capital to reinvest in growth

This is the aspirational flywheel.

4. Unit Economics: The Early Tell

Zoom in further.

Per-store performance:

Metric                     Value
--------------------------------------
Revenue/store             ~$750K–$1M
Store-level profit        ~$100K+
Same-store sales          +5–6%
Ticket growth             +15–20%

Total store profits increased +60% in 2025

Interpretation:

Each store is not just:

→ A growth unit
→ But a cash-generating asset

That’s what eventually enables scale without constant dilution (in theory). That said; Pesorama is still funding store growth via debt/share issuance.

5. Valuation vs Peers: Where It Gets Interesting

Let’s compare.

Company         EV / Sales     Growth Profile
----------------------------------------------------
Pesorama        ~3x            Early / high growth
Dollar General  ~1x            Mature / competitive
Dollarama       ~8x            Premium / scaled

Dollarama operates at ~45% gross margins and premium valuation.

What this implies:

Pesorama is currently priced:

• Too high to be “distressed retail”
• Too low to be “proven compounder”

That middle ground is where mispricing can happen.

6. The Flywheel (Visualized)

This is the entire thesis in one aspirational diagram:

New Stores Open
       ↓
Higher Sales Volume
       ↓
Better Procurement
       ↓
Higher Margins
       ↓
More Cash Flow
       ↓
Fund More Stores
       ↓
(REPEAT)

The key question:

👉 Has the flywheel started?

Based on the data, it appears to be just beginning – but needs to prove-out with an inflection point in scale.

7. What Would This Look Like at Scale?

Let’s project directionally (all estimates, not actuals):

Scenario         Stores    Revenue(E)     Outcome
----------------------------------------------------------------
Today            ~35       ~$25M+        Early stage
Near-term        50+       ~$50M         Scale emerging
Mid-term         100+      ~$100M+       Institutional interest
Long-term        300+      ~$300M+       Category player
...

This is not a prediction.

Rather, it’s a framework for thinking about upside.

8. Final Takeaway

If you zoom out:

• Store growth is accelerating
• Revenue is compounding
• Margins are expanding
• Unit economics are proving out

And yet:

• Market cap: ~$65M
• Coverage: minimal
• Institutional ownership: low

The Big Idea

Most investors wait for this:

✔ Profitability
✔ Scale
✔ Institutional validation

But by then, the biggest gains are mostly gone…

The greatest opportunity is usually when there’s broad uncertainty.

Closing Thoughts

Pesorama is not a guaranteed winner. The risks are real.

But it is:

• Early
• Executing
• Underfollowed
• Showing real signals beneath the surface

Disclosure: $PESO.v (own)

Watch a recent interview (2026) with Pesorama’s Founder & CEO: