I’m all for CSU, but your WACC seems way too low. The premium over U.S. Treasuries is minimal, and even if the beta is indeed low - say 0.7-0.8 - something in the 7%–9% range probably makes more sense.
That said, I also think using an LTM valuation is less appropriate. We’re already in Q1/26, and LTM still includes Q4/24. It would be more reasonable to think in terms of 2026 FCFA2S. If we back out the IRGA liability (which isn’t really a liability), we’re probably looking at roughly $2.2–$2.3 billion USD in 2026, which implies a multiple of around 16x.
Great breakdown of the VMS moat. Your point in section 7.3 about "institutional memory" is key—CSU’s 30-year database of failures is the ultimate defense against AI disruption, as "vibe-coding" can't replicate decades of niche legacy data.
However, struggling with the Mark Miller transition. If his Volaris model—reinvesting FCF locally rather than sending it to the parent—is now the lead strategy, doesn't his promotion signal the "death" of the classic CSU model? If every unit stops upstreaming cash to grow autonomously, does CSU eventually stop being a unified compounding machine and just become a loose collection of ETFs?
if you continue to grow at high pace - at some point you'll be the size of the entire market.
so regardless of how you love a business for its leadership, ethics, etc - growth will plateau at some point.
assuming a stock price is impacted by past growth, it might mean that high growth is priced in. when plateau starts to materialize - the price will cave, as the stock no longer holds the promise it once had.
in case of CSU: the PE is high, even after recent rout.
this suggests that the market as a whole still thinks they have growth in the cards, so there's room for correction.
the main issue in my opinion is not whether an existing customer starts to vibe-code something that's already working.
the issue is to whether there's enough verticals to buy to keep growth at that pace.
the game used to be:
you open a business to serve some niche.
the business succeeds, and you need some software to help you manage.
you find some software that does exactly what you need, or something close enough + pay for adjustments.
then CSU comes and buys your vendor's business.
I'm seeing many people serving niche customer base with n8n, lovable, base44 etc.
I'm not sure all those businesses will succeed, and i'm not sure all will stay with the hack solutions they built, but at least some would never feed the food chain that ends with CSU.
i think the re-pricing reflects that some investors expect that to have immediate impact on growth and thus demand a smaller premium.
Great writeup! You have a small typo under section 10.1: You wrote "They have increased their revenue from $1.8B in 2025, to $11.1B for the LTM." It should say "They have increased their revenue from $1.8B in 2015, to $11.1B for the LTM."
Great article about $CSU and VMS in general. Almost at the same time, i published an article about Vitec Software, often called the Nordic CSU. Maybe you want to check it out:
i'm a holder of a long position - so i like that someone questions this position.
i'm not sure i agree that this should really be the cause for a pause.
the way i see it their biggest short term liability (by a huge margin, more than 40% of total current liabilities) - is deferred revenue.
depending on the business and how it bills its customers, this might not indicate an issue. e.g. if you're charging your customers for 12 months in advance for a service you already know you can deliver - an increase in deferred revenue might indicate that customers are buying more of the stuff you know how to deliver.
when one looks at the asset side, one sees a small increase in AR - which seem reasonable consider the business is growing. DSO looks stable too.
you're worried about something specific in the current assets?
When CSU received advance payment 3.21B, 3.21B is recorded as Deferred Revenue in Current Liability and 3.21B added in the Cash and Equivalents in Current Assets.
.
Current Assets = X + 3.21B = A
Total Liabilities = Y + 3.21B = B
Current Assets - Total Liabilities
= X + 3.21B - (Y + 3.21B)
= X - Y
= (A-3.21B) - (B-3.21B)
= A - B
.
Summary:
Deferred Revenue will not affect the NCAVPS final result.
I’m all for CSU, but your WACC seems way too low. The premium over U.S. Treasuries is minimal, and even if the beta is indeed low - say 0.7-0.8 - something in the 7%–9% range probably makes more sense.
That said, I also think using an LTM valuation is less appropriate. We’re already in Q1/26, and LTM still includes Q4/24. It would be more reasonable to think in terms of 2026 FCFA2S. If we back out the IRGA liability (which isn’t really a liability), we’re probably looking at roughly $2.2–$2.3 billion USD in 2026, which implies a multiple of around 16x.
Great breakdown of the VMS moat. Your point in section 7.3 about "institutional memory" is key—CSU’s 30-year database of failures is the ultimate defense against AI disruption, as "vibe-coding" can't replicate decades of niche legacy data.
However, struggling with the Mark Miller transition. If his Volaris model—reinvesting FCF locally rather than sending it to the parent—is now the lead strategy, doesn't his promotion signal the "death" of the classic CSU model? If every unit stops upstreaming cash to grow autonomously, does CSU eventually stop being a unified compounding machine and just become a loose collection of ETFs?
Miller is not going to apply the Volaris model to the whole CSU constellation.
I merely pointed out he had a different strategy.
He said multiple the business will continue ‘‘as is’’.
great post.
my main concern is the runway.
the theory:
if you continue to grow at high pace - at some point you'll be the size of the entire market.
so regardless of how you love a business for its leadership, ethics, etc - growth will plateau at some point.
assuming a stock price is impacted by past growth, it might mean that high growth is priced in. when plateau starts to materialize - the price will cave, as the stock no longer holds the promise it once had.
in case of CSU: the PE is high, even after recent rout.
this suggests that the market as a whole still thinks they have growth in the cards, so there's room for correction.
the main issue in my opinion is not whether an existing customer starts to vibe-code something that's already working.
the issue is to whether there's enough verticals to buy to keep growth at that pace.
the game used to be:
you open a business to serve some niche.
the business succeeds, and you need some software to help you manage.
you find some software that does exactly what you need, or something close enough + pay for adjustments.
then CSU comes and buys your vendor's business.
I'm seeing many people serving niche customer base with n8n, lovable, base44 etc.
I'm not sure all those businesses will succeed, and i'm not sure all will stay with the hack solutions they built, but at least some would never feed the food chain that ends with CSU.
i think the re-pricing reflects that some investors expect that to have immediate impact on growth and thus demand a smaller premium.
Great writeup! You have a small typo under section 10.1: You wrote "They have increased their revenue from $1.8B in 2025, to $11.1B for the LTM." It should say "They have increased their revenue from $1.8B in 2015, to $11.1B for the LTM."
Well spotted, I'll adjust it later! Thank you.
I believe there are a couple more mistakes in there, good luck finding them all :D
Great article about $CSU and VMS in general. Almost at the same time, i published an article about Vitec Software, often called the Nordic CSU. Maybe you want to check it out:
https://manuelwalz.substack.com/p/vitec-software-is-claude-code-the
Thanks for sharing my deep dive Manuel. Saved your post on Vitek to check out later!
Amazing work mate, we're in the same boat with CSU and I believe it'll be a market beating one :)
Earnings Quality ttm
= FCFPS ttm÷EPS ttm
= 3,460÷930.33 (data from investingdotcom)
= 3.7191104232
.
NCAVPS
= Negative
.
With such an extremely high Earnings Quality, the NCAVPS however still maintains NEGATIVE, I don't feel comfortable with that.
i'm a holder of a long position - so i like that someone questions this position.
i'm not sure i agree that this should really be the cause for a pause.
the way i see it their biggest short term liability (by a huge margin, more than 40% of total current liabilities) - is deferred revenue.
depending on the business and how it bills its customers, this might not indicate an issue. e.g. if you're charging your customers for 12 months in advance for a service you already know you can deliver - an increase in deferred revenue might indicate that customers are buying more of the stuff you know how to deliver.
when one looks at the asset side, one sees a small increase in AR - which seem reasonable consider the business is growing. DSO looks stable too.
you're worried about something specific in the current assets?
When CSU received advance payment 3.21B, 3.21B is recorded as Deferred Revenue in Current Liability and 3.21B added in the Cash and Equivalents in Current Assets.
.
Current Assets = X + 3.21B = A
Total Liabilities = Y + 3.21B = B
Current Assets - Total Liabilities
= X + 3.21B - (Y + 3.21B)
= X - Y
= (A-3.21B) - (B-3.21B)
= A - B
.
Summary:
Deferred Revenue will not affect the NCAVPS final result.
Great deep dive, Tacticz!!! Congratulations on the work.
It clearly shows the amount of thought and work you put into it.