Growth for growth’s sake vs disciplined growth. Management was explicit that some competitors are chasing volume, whereas NLB targets risk-adjusted, properly priced growth matched by deposit growth.

Pricing in Slovenia runs c 30bps above competitors, attributed to underwriting discipline and to NLB’s market-leading position, with the competition following.

Croatia is off the table for now - following the failed Addiko bid management indicated the issue now has to be resolved at the political level.

Overall, I believe the earnings call was very informative and worth a watch - particularly the analyst question at 44:45-49:45, and the answer to my own question at 54:15.

Link to webcast here.

In addition, while during H1’25, a EUR 2.5m release of impairments and provisions was made, a charge of EUR 32m was made in H1’26. To this end, the profit before impairments and provisions actually grew YoY - EUR 332m for H1’26 vs EUR 322m in H1’25. This was driven by 3% net operating income growth, partially offset by opex growth of also 3%. However, the 3% cost growth is also the good news: uncontrolled opex was one of the key items to watch out for in my initial deep dive - and it seems management has finally managed to start controlling it, which is a good sign; hopefully it lasts. NPLs also remained flat at 2%. Total assets grew 8% YoY but fell -1.3% or EUR 417m QoQ - this seems to be driven by the early repayment of the EUR 500m senior preferred note (at 7% interest, meaning EUR 35m of annual interest cost are now saved) and the EUR 138m June dividend payment, partially offset by the EUR 252m of net income earned in Q2.

On the income side, the growth was driven by continued strong loan book growth at 12% YoY, partially offset by still marginally declining NIM (from 3.25% in Q4’25 to 3.21% in Q2’26). While NIM seems to have not fully fully normalised (as per assumed in my deep dive - though 2 quarters are still left), the decrease is a rounding error. Thus I think group level NIM should still fully bottom out this year. And while Slovenia has already fully bottomed out (even starting to tick-up), the SEE NIM as a whole continued to fall into H1, bringing the group average down. This was namely driven by a particularly sharp drops in Serbia and Bosnia, which seem to be driven by a shift in the deposit mix towards more term deposits. On the positive side, NLB has gained market share YoY in all markets, including Serbia and Bosnia, which are under pricing pressure - so the pricing reset might be at least in part a deliberate move to take market share (though market shares thus far are reported for Q1 only).

As per above, at least for Serbia, the steep decrease in NIM seems to be driven by the increased share of term deposits in the country:

Second rather bad news - though not connected with the H1 earnings per say - is that the Addiko acquisition fell through. Raiffeisen Bank International secured 56.2% of the shares and won, so Croatian market entry by the backdoor was again denied and remains a pipe dream at scale.

Finally, the asset management arm is continuing its strong performance, with EUR 150m of net inflows in H1’26, AUM now at EUR 4.4bn, and market share up from 42% last year to 43%. NLB also announced they are planning to launch a new listed real estate fund, which, as per my previous post is a terrible deal for incoming investors, but which - since it probably still will go through - of course means it is a fantastic deal for NLB shareholders.

Conclusion

Overall, despite the margin pressure in Serbia and Bosnia, and as per my original deep dive, I believe NIM in now likely fully bottoming out. Starting in 2027 we should see an inflection in profitability - especially if opex remains under control, as it was in H1’26.

Back of the envelope, assuming NIM will indeed flatten out in 2026 while the underlying loan book will continue to grow at 10% into 2027, and holding opex increase at H1 level of 3%, we should see c. EUR 130m of incremental net operating income in 2027. Less c. EUR 20m of incremental opex, and after tax and provisions, this would come out at c. EUR 90m of incremental net income. As such, we are eyeing EUR 590m of net income for 2027, up from c. EUR 500m likely for this year. Each subsequent year should then be subject to similar incremental unit economics, though the NOL is set to expire in the next years. But even considering a full 22% tax rate, the above perhaps slightly optimistic scenario should still comfortably clear EUR 1bn of pre-tax income by 2030, or c. EUR 800m in net income terms. That lands slightly below where management’s own organic roadmap points but overall we are well on track for the Base Case outlined in my original deep dive, so the target price ranges still hold.

Disclosure: I currently hold a beneficial long position in NLBR through stock ownership. I have no business relationship with the company mentioned. This post reflects my personal opinions and is not a recommendation to buy or sell any security. My views are subject to change without notice as new information becomes available.

i) The briefcase pays out 10k per year, in perpetuity

ii) You can never open it and access the actual money

iii) If you ever want to sell it, the most anyone can pay you is 200k

How much is that briefcase worth? Probably not 1m. The money inside is real, you just can’t get to it. What you have is a 10k annuity plus a lottery ticket that someone, someday, changes the rules.

Below is a deep dive into such a briefcase trading on the Zagreb Stock Exchange (Croatia).

The pitch that sounds too good to be true

ACI d.d. (ZSE: ACI) operates 22 marinas with over 6,000 berths along the entire Croatian coast, representing the largest marina chain in the Mediterranean. These are irreplaceable assets with high pricing power: since EU accession, permitting a new marina on the Croatian Adriatic is close to impossible, and ACI holds some of the best locations, assembled decades ago when nobody else was competing. Its 23rd marina, currently under construction in Rijeka under a 50:50 partnership with Lürssen (admittedly a bigger, superyacht-grade project than the existing portfolio), carries an investment value of c. EUR 50m, before cost overruns – and that’s for a single marina with 230 berths.

The financials, at first glance: c. EUR 40m of revenue, c. EUR 10m of annual operational cash generation (EBITDA), market cap of EUR 89m, and EUR 45m of cash and term deposits sitting on the balance sheet against minimal debt. Strip out the cash and you are paying an enterprise value of under EUR 50m - roughly 5x EBITDA and around EUR 8k per berth - for a monopoly asset class that private equity is falling over itself to buy.

And they really are. Literally this week, CVC sold D-Marin - the closest regional comparable, with 28 marinas and c. 14,300 berths across the Mediterranean - to InfraVia for over EUR 1bn, roughly 15x EBITDA and c. EUR 70k per berth. ACI trades at close to a 90% discount to that on a per-berth basis (granted, D-Marin’s marina’s are more upscale, but still a stark discount).

So: unbelievably undervalued. BUT. Essentially all of ACI’s concessions expire in 2030. The Croatian state owns c. 79% of the company and runs it accordingly. The free float is under EUR 20m and the stock barely trades. After a lot of thought, I decided not to take a position. Here is why.

The asset

ACI was founded in 1983 and built out its network when the Adriatic coast was still open for development. Today that network cannot be replicated. The Rijeka datapoint implies c. EUR 200k per berth for a new premium build; even a conservative blended EUR 40-80k per berth on ACI’s c. 6,000-berth portfolio gets you to EUR 250-500m of replacement value. And replacement is mostly theoretical anyway - environmental and maritime-domain permitting means very few new marinas will ever be built in Croatia. Scarcity value on top of replacement cost.

Below the locations of ACI’s marinas:

The concession cliff

When it comes to concession topic, there is currently a structural problem in Croatia. ACI’s marinas (as well as all other marinas) sit on “pomorsko dobro” - Croatian maritime domain, which cannot be privately owned and can only be leased out via 20-50y concessions. In addition, at concession expiry, everything built on it (piers, breakwaters, buildings) is legally part of the public domain and reverts back to the state, without any compensation. One of ACI’s anchorage concession expires in 2026, one marina in 2027, and everything else during 2030.

Due to this reason, ACI’s capex has collapsed to bare minimum in recent years - EUR 1.4m invested in all of 2025 (vs EUR 9m of annual D&A) - because management openly states that investing into concessions expiring in four years is uneconomic. A EUR 13.5m capex plan for 2026 exists on paper, likely contingent on the extension.

ACI has formally requested extensions based on its historical over-investment relative to the concession contracts, and a government decision has been expected imminently for months. The complication: under EU rules concessions generally must be tendered, and Croatian law only allows extension without tender within limits - so this is genuinely uncertain, not theatre.

That said, I think the concessions are, in practice, rather safe. First, the state owns the company - the government tendering its own marinas away to third parties would be self-harm. Second, this is not an ACI-specific issue: the whole Croatian marina sector operates on expiring-in-2030 maritime-domain concessions, and the entire industry is holding back capex because of it. A model where nobody invests in the country’s most valuable tourism infrastructure is obviously broken, and the state will have to fix it - for everyone. Third, the D-Marin sale is instructive: InfraVia just paid over EUR 1bn for a portfolio including Croatian marinas carrying the same concession-renewal risk. Sophisticated infrastructure capital clearly believes Med marina concessions will get rolled.

Indeed, assuming this would be a for-profit ran company, the worst case would have been well covered – EUR 45m of cash on hand + 5 more years of EUR 10m of annual cash generation would result in cash balance of EUR 90m by end of 2030, equal to market cap. So in theory, even if all concession were to be lost, one would still be made whole as an investor (save for the tax lost on dividend payout and severence payments).

The financials - and the embarrassing comparisons

FY2025: revenue EUR 40.3m (+3%), EBITDA EUR 9.5m, net profit EUR 0.4m. EUR 0.4m of net profit on EUR 40m of revenue and a 3% revenue growth from one of the hottest premium yachting destinations in the world, coming from a monopoly is very unusual. Cash and term deposits of c. EUR 45m against modest remaining debt. Dividends are relatively low and not consistent: EUR 4.82 gross per share for 2024, a EUR 0.5m total payout - a c. 0.6% yield on a company sitting on half its market cap in cash.

Part of the terrible net margin is optical - the aggressive amortization to 2030 described above. But compare EBITDA margins and it stays ugly. Marina Punat, a family-owned operator of essentially one large marina on Krk, generated EUR 11.2m of revenue in 2024 with EUR 4.7m of EBITDA and EUR 3.5m of net profit - a 42% EBITDA margin and 31% net margin, on 98 employees. ACI: a 23% EBITDA margin and 1% net margin, on c. 350 employees. Horrendous.

So one privately-run marina earns roughly 10x ACI’s net profit and 50% of ACI’s EBITDA on a quarter of its revenue. The gap is structural: ACI’s staff costs of EUR 16.4m represent 41% of revenue and are growing double digits - partially a consequence of ACI operating multiple small marinas, which still need to be staffed, but partially a sign of inefficiently ran organisation. Moreover, ACI outsources the high-margin ancillary businesses (service, restaurants, charter) to third parties and keeps only the landlord fee - a suboptimal business model for a marine operator.

Run ACI’s EUR 40m of revenue at Punat’s 41% EBITDA margin and you get c. EUR 16-17m of EBITDA - roughly EUR 7m per year, every year, left on the table. And this is in addition to indirect loss of potential revenues due to unoptimized pricing (and investment) structure. Rijeka will eventually add a high-end revenue stream on top (via the 50% JV stake), but it does not fix the core problem: this business does not seem to be ran to maximize profit. Instead, it seems to be ran as an employment and regional-policy vehicle that happens to have shareholders. Which is funny, considering most of the paying customers are international, hence maximizing revenue would truly maximize the value also to the country as a whole. But anyway, state company is a state company, motivations are what they are.

Ownership: nobody will ever sell the shore

The Republic of Croatia holds c. 79%, and there is no live privatisation plan. There have been attempts to monetise ACI over the decades; all died. I can understand why: privatising the marinas would likely be a political suicide, seen as “selling the Croatian coast” by the public. Expect indefinite state ownership as the base case.

Valuation: three scenarios

Bear case - concessions not extended. In 2030 the marinas revert to the state and ACI is left with its cash pile, its 50% of the Rijeka JV (secured to 2053), and whatever it wins in new tenders. The cash alone is c. EUR 90m by then against an EUR 89m market cap, plus 50% ownership stake in Rijeka - call it EUR 20m of equity value (majority is to be financed with debt). Downside from here is very limited and not catastrophic - but this is in theory: it assumes state winds down the company under this scenario and pays out the excess cash, which is not a given.

Base case - concessions extended, state keeps running it. The most likely outcome. You own a perpetual EUR 10m EBITDA stream managed for employment rather than shareholders, paying out a negligible dividend yield, with the cash pile likely earmarked for capex rather than distribution (expect a massive, multi-year capex ramp-up if concessions are renewed ; profitability of such investments, given the state ownership, is also not guaranteed). At 5x EBITDA the market is pricing exactly this: a monopoly whose owner is the problem. Nothing re-rates. Technically you need to value it on cash-flow, which is the dividends, which means the current valuation is still not attractive. This is the briefcase.

Best case - privatisation. Apply anything remotely like D-Marin or Marina Punat economics - a proper EBITDA margin, a private-market multiple, optimised revenues, EUR 70k per berth - and you are into the EUR 300-500m range, i.e. a multi-bagger. But this scenario requires a Croatian government to voluntarily commit political suicide. I attach a very low probability to it, and there is no catalyst on any visible horizon. A minority stake in an eternal state vehicle is an option with no expiry date but also no discernible path to exercise.

Conclusion

Everything about ACI screams value: irreplaceable assets, fortress balance sheet, a sector where private capital just paid EUR 1bn for the neighbour’s version of the same concession risk. And yet.

The cash is in the briefcase, and the briefcase does not open. Even thought the downside case is - in theory - well covered, the upside case requires a miracle. All that while also the base case doesn’t call the stock to re-rate. The state controls the payout (0.6% yield), the state controls the operating model (41% of revenue to wages), the state controls the exit (there is none), and the free float is small enough that even changing your mind is expensive. What is left for a minority shareholder is a EUR 10k annuity on a EUR 1m note - worth something, but nowhere near what is inside.

Would you pay EUR 1m for the briefcase? Neither would I. But I also wouldn’t pay EUR 89m - sorry, EUR 50m net of cash you’ll never see - for ACI, until the day a Croatian politician decides the shore is for sale. I’ll keep watching for that day. I don’t expect it soon.

Not investment advice. No position in ACI. My opinion is subject to change without notice.

Subscribe now

1. Executive Summary

This piece is really two articles in one. The first is my macro view on Slovenian residential real estate - the default wealth-building asset of every Slovenian household, and one I believe is heading into a structurally weaker decade(s). The second is a deep dive on Equinox Nepremičnine (”Equinox” or “EQNX”), the currently still only listed way to own Slovenian real estate, and the natural test case for whether the “just buy an apartment” playbook still deserves its default status.

On the macro: the extraordinary run of the past decade - Ljubljana prices roughly doubling from from 2015 to today - is not the new normal. It was the product of several tailwinds arriving at once: a population still growing, continued urbanisation into the capital, shrinking household sizes, a multi-decade collapse in interest rates, and the mortgage tenors / credit expansion. As I lay out in Section 2, each of these drivers is now either plateauing, or outright reversing: the population peaks around 2026, tenor extensions from here add little borrowing capacity, the rates supercycle cannot repeat, and Ljubljana is now less affordable relative to salaries than most other EU cities. My conclusion: expect mid-single-digit nominal returns from Slovenian residential real estate going forward in the (realatively optimistic) base case, not the 8-10% of the past decade. History is full of housing markets that “always went up” - until they didn’t.

However, if you still want to hold real estate exposure, the question becomes what is the best way to do it - and whether the standard Slovenian answer (buy an apartment, maybe take a mortgage, maybe rent it out) still beats the alternatives. Which brings us to Equinox.

Equinox is Slovenia’s first - and at the moment still only - listed real estate company, spun off from Union Hoteli d.d. in 2021 and listed on the LJSE in January 2022. It owns 8 properties in central Ljubljana - 4 hotels and 4 commercial assets - with a total area of c 68,000 m² and Q1 2026 appraised fair value of EUR 139.9m. Three of the four hotels are covered by a 20-year master lease with Spanish operator Grupo Hotusa, providing contractual rent of over EUR 110m through 2042. The company runs with near-zero financial leverage (LTV of c 8%) and distributes c 50% of FFO (funds from operations, aka cash flow) as dividends.

The investment thesis rests on two key pillars:

• Equinox is a passive proxy for prime Ljubljana real estate: At the current share price of c EUR 63.5, the stock trades at 0.81x NAV (NAV of EUR 78/share), a 19% discount. Combined with a starting dividend yield of c 3.1%, a fortress balance sheet, a 20-year contracted anchor lease and zero operational management burden, Equinox offers a meaningfully better risk- and effort-adjusted return profile than the typical Slovenian retail alternative of buying one’s own apartment to rent out.

• The unlevered balance sheet is both a feature and a limitation: The very conservatism that gives the stock its “safe haven” character also caps its upside. A direct real estate purchase in Slovenia can be financed at up to 70% LTV, potentially amplifying returns (or losses). However, despite EQNX leaving “money on the table” by running de-facto fully unlevered – in a business where leverage is a very standard way to increase returns – as shown in the later chapters, even for a levered apartment to start meaningfully outperforming EQNX, an assumption of annual real estate appreciation of 5%+ is required. That is precisely the assumption my Chapter 2 argues against. Not to mention that a single apartment also comes with far higher concentration risk, illiquidity and personal time cost.

Based on my valuation, assuming real estate prices continue increasing only with inflation at 2% p.a. - i.e. my macro view playing out, but without a crash / drop in value - I see a total expected return of c 8% p.a. over a long-term investment horizon, made up of c 2.4% net-dividend yield, 2.5% effective NAV growth, and an additional c 3.2% growth coming from FFO reinvestment. Notably, this is a return profile that holds up without needing the real estate bull case – which is exactly what one should demand from a real estate allocation today. But even more importantly, it is also a return profile assuming no mean reversion or real bear case - which I personally see as more likely than the bull case (over the long term).

As such, I see Equinox at 8% p.a. as too-low yielding of an asset to be considered a Buy for an active investor. Nevertheless, for very conservative investors, whose main priority is aversion of loss of principal, and whose alternative is buying an apartment as an investment outright, I believe Equinox offers a relatively more attractive and hassle-free proposition. The diversification across 8 central Ljubljana assets, the Hotusa anchor contract through 2042, and the embedded 19% discount to NAV together make it a reasonable “hands-off real estate alternative” – but not the core wealth-building equity vehicle that an investor seeking to maximise their portfolio returns should make.

2. The drivers of real estate returns - my take

While below will focus on Slovenian real estate market, since that’s where Equinox’s portfolio is located, it should by and large apply for the entirety of Europe.

The standard (Slovenian/European) retail view on real estate is, broadly, that “real estate always goes up”, which is based on “rear view mirror analysis”. To this end, if one has spare EUR 10k of savings, they keep them on the bank (great for NLB). But once you get to EUR 100+k, you will go ahead and buy an apartment, as it’s still “just as safe” + you will get some principal appreciation, maybe even some rent yield - maybe even 10% per year! You also ignore for the ammout of work you had to put in, and the results look(ed) great. This is because, by and large, investing the savings in stocks is still a complete no go in the region. It’s either cash or real estate.

Over the past 10 years, Ljubljana apartment prices are up c 2x (the median price went from c EUR 2,400/m² in 2015 to c EUR 4,900/m² in 2025), far outpacing inflation (officially up only 30% for the same period) – especially if one bought a property with a large mortgage. For an entire generation of Slovenian savers who watched their parents’ apartments appreciate through communism-to-capitalism, EU entry, eurozone accession and the post-GFC recovery, “buy an apartment” is a no-brainer - risk free and highly profitable.

However, a closer look reveals that the near-universally positive historical experience was likely driven by the convergence of several specific secular tailwinds, some of which are now either neutral or turning into headwinds.

2.1 Population, urbanisation and household size

Population: Slovenia’s population grew from below 1.5m in 1960 to 2.13m by 2024. Both Eurostat and United Nations project that as of 2026, the population has already peaked and is expected to start declining going forward. EUROPOP2023 projection – adopted by the Statistical Office of Slovenia as the official forecast – has the population peaking in 2026 and slowly declining to c 1.95m by 2100 (-7%). The UN’s 2024 revision is considerably more pessimistic, implying a population decline of -6% already by 2050 - see below:

To this end, one of the key drivers for increased country-wide real estate demand - and in turn price appreciation, is now expected to go into reverse. But let’s have a look at it in a bit more detail, by considering also the urbanisation trends:

Urbanisation. Slovenia has been on a long, slow urbanisation trajectory. Share of population living in Cities (Ljubljana) has been on a steady upward trajectory - both in terms of absolute numbers (in part helped by the overall Population growth), as well in relative terms. Most importantly for Equinox, the City based population has grown from c 100k in 1950s to c 300k today - but again, according to UN, it has now fully peaked and is expected to remain flat / slowly decline going forward.

As per above, the historically fast growing population of Cities of 20% between 1975-2000 and 10% between 2000-2025 is now expected to turn marginally negative at c -4% for the 2025-2050 period.

Considering this, the picture for Ljubljana located real estate is much less pessimistic than looking at total population for Slovenia, as most of the decline in population will be in rural areas. Nevertheless, historically strong tailwind - from the years when the city was actually still growing - will subside. Further price appreciation assumption should take this reality into account.

Household size / household formation. Slovenia has been on the same household-fragmentation trajectory as the rest of Europe – fewer multi-generational households, more single-occupant and two-person households. The average Slovenian household had 2.48 members per the 2011 census, and 2.41 in 2021 (SURS register-based census data). A seemingly small change: but at a national level, moving from 2.48 to 2.41 average members, on a roughly stable population, implies c 3% more households needed – and therefore 3% more units demanded – for the same total population. Over 10 years, that is roughly 0.3 p.p. per year of structural housing demand tailwind just from household splitting. Unfortunately I couldn’t find any reliable projections of this trend, however, with the increasing share of single people, one could plausibly expect this trend to continue. With more housing demand due to smaller households, even if population stagnates, there could still be some positive demand side pressure.

All in all, on the population side, there used to be strong tailwinds from overall population and urbanisation side, both of which are now expected to actually reverse. Even considering a continuation of demand from household size / occupancy side, de-celeration (or even decrease) of price increases can reasonably be expected from this front.

2.2 Relative price of real estate

One of the most important factors in gauging the affordability and thus sustainability of real estate pricing is the rent-to-income ratio. As per below comparison table, you can see Ljubljana ranks among the top 10 worst EU capitals, with rent to income ratio at a staggering 61.2%. The standard rule for financial discipline is to spend no more than 30% of your net income on rent - for a single person, Ljubljana is already at double that. This leaves only very limited room for further price appreciation, as housing will literally become unobtainable - and at the same time leaves a plausible mean reversion (i.e. downside) scenario.

2.3 Financialisation of real estate

Housing, like energy, utilities etc. is a natural monopoly business. There is only so much land available, and, for a well functioning (long-term) society, it is in country’s best interest to at least partially regulate such industries, in order to prevent price gouging. However, the governments are currently not agreeing with this, at least with regards to real estate. By and large, there is little to no restrictions on real estate or rent pricing, or ownership. To this end, the financialisation of housing, especially since 2008 financial crisis, has been accelerating (throughout the world). Financialization is the process of treating housing as a financial asset and / or investment, rather than a social good. People are no longer buying apartments to live in them, but instead as an investment. This, especially combined with this being de-facto only investment class for an average Slovenian, creates upward pricing pressure, as people with actual housing need have to compete on pricing not only with other people searching for a place to live, but also with “rich” investors, who may even leave the flat idle / without renting it out. And the worst part is that a large portion of those competing “rich” investors see real estate as the only possible investment vehicle for their money, hence are prepared to pay literally any per-m² price asked.

I see a plausible scenario where an increase in the financial literacy (which I already opined upon in my NLB and ZVTG deep dives), which is becoming an imporant agenda item for all EU countries, will (hopefully) result in more people pursuing productive (equity) assets, i.e. less money will be left over for real estate. This could tame the demand and continued price pressure via financialisation channel.

2.4 Zoning & political aspect

Slovenian construction permitting remains slow and conservative by Western European standards. The OECD’s 2024 review explicitly flags that urban density and building heights in Slovenian cities are “far below the levels seen in other countries” and that there is “room for densifying already urbanised areas without necessarily increasing the share of artificialised land.”

The key beneficiaries of this undersupply are existing property owners, and in Slovenia the median existing property owner is a boomer who has already paid off their home. Boomers are also the largest voting demographic. The political incentive to meaningfully loosen zoning and accelerate construction is therefore structurally weak – any government that did so would be transferring wealth from the largest voting bloc to a much smaller group of renters and first-time buyers. This is not unique to Slovenia - it is true in most developed housing markets. What I would thus expect is not a sudden loosening (which would be required for a genuine housing price “crash”) but rather a very gradual deceleration of the undersupply premium as some modest amount of new building does come online, secondary-city price competition increases, and affordability pressure eventually forces marginal policy action. All this will also be aided by the falling / no-longer-increasing population.

We’re already seeing some signs of potential change, with the historically strong NIMBY (not in my backyard) mentality now being challenged by a competing YIMBY (yes in my backyard) movement. Some people are starting to realize that without a collective shift toward a bit less individualism - since everyone wants the problem fixed, but only if it comes at zero detriment to themselves - nothing can actually change. If YIMBY becomes “fashionable,” and in turn politically popular, things could turn quickly. Potentially bad news for housing prices.

2.5 Falling interest rates

The 10-year German bond went from c 9% in 1990 to below 0% in 2020, and real estate everywhere in Europe re-rated in lockstep - via a compression in required rental yields (and cheaper mortgages). This decreasing and zero interest rate cycle now seems firmly over. The ECB’s deposit rate at c 2% in 2026 is far from the 2021 lows, and a return to the zero lower bound would require a material economic shock. More importantly, even if the ECB cuts another 50-100 bps from here, the structural tailwind from a multi-decade downward drift in discount rates is simply not replicable – you cannot take bond yields from +9% to 0% twice in one investor’s lifetime. The most an investor today can realistically hope for is that rates stay where they are, which means the real estate repricing machine that has been running for 30+ years is now idling. Another tailwind gone.

2.6 Credit expansion – mortgage tenors and what’s left to give

Slovenia’s household debt / GDP ratio went from near-zero in the early 1990s to c 25-30% today, while Western European averages sit at 50-70%. The continued convergence is part of my NLB (and ZVTG) thesis. Mortgage lending was essentially non-existent under Yugoslav socialism, and has steadily expanded as banks became more comfortable extending long-duration unsecured credit and households more comfortable taking it on. Each incremental euro of mortgage credit extended is, in effect, an additional euro of demand for housing, and this has provided a secular bid for real estate prices over the past 25 years.

A cleaner way to see how much of this tailwind has already been captured vs how much is left is to look at what a typical household can afford to buy with a mortgage, as the mortgage tenors extend. The math is mechanical: at a given monthly debt-service budget and interest rate, longer tenors let the same household service a bigger principal, which in a supply-restricted market translates almost directly into a higher house price.

Working with a representative EUR 1,000/month debt-service budget at 4% fixed rate, the maximum mortgage principal each tenor supports is:

Two observations matter here. First, the tenor extensions we have already been through – roughly 10→15 year norms in the early 2000s, to 20-25 years by the mid-2010s, to 30 years today, have already unlocked the biggest gains. Moving from a 10-year norm to a 30-year norm roughly doubled household borrowing capacity at a given monthly budget (EUR 99k → EUR 210k, +112%). That is an enormous increase in purchasing power - in turn, due to limited supply, driving up the prices for essentially the exact same amount.

Second, further tenor extensions from here yield progressively less. Going from 30 to 40 years adds only 14% of extra borrowing capacity; from 40 to 50 years only 8%. At 6% interest rates these increments are even smaller. In practice, this means that even if Slovenian banks start offering 40-year mortgages as standard over the next decade – which is plausible but not certain – the residual price tailwind from “more tenor” is likely capped at 10-20% of cumulative house price uplift, spread over many years. That is not nothing, but it is not another housing supercycle either.

On the positive note, Slovenia still sits meaningfully below Western European household debt levels, so theoretically if people start taking more mortgages the supply of money chasing same housing stock would be larger – but the runway is still shorter than it was 20 years ago.

2.6 Tourism

The final driver, especially relevant to the hotel-heavy nature of the Equinox portfolio, is Slovenian tourism - and here the data is genuinely positive. Slovenia had a record year in 2024 (16.9m overnight stays, +5% YoY, +7% vs 2019 pre-COVID) and then surpassed that again in 2025 with nearly 17.8m overnight stays (+6% YoY), per SURS data. The Slovenian tourist board’s 2022-2028 strategy targets continued moderate growth with a sustainability/premium orientation rather than volume push.

This is a direct and continuing tailwind for the Equinox portfolio, because the Hotusa master lease has a variable rent component tied to hotel operating performance - so incremental tourism growth translates into incremental rent to Equinox, on top of the inflation-indexed minimum (and the real estate value appreciation). It does not, however, translate directly to the apartment rental market (which is primarily local Slovenian rental demand) and therefore does not neccesarily bail out the residential thesis if the other drivers fade.

2.7 Conclusions

In all honesty, I have not the slightest idea how to quantify how much each of the above factors will contribute to NAV appreciation (de-celleration) going forward. Neither is my thesis a prediction or call that real estate can be expected to crash or fall. Instead, the point I am trying to make is that real estate does not always go up. Moreover, I am convinced about is that taking historical real estate price appreciation rates, which had ALL of the above factors as a strong tailwind, and extrapolating them into the (long-term) future, is an idiotic assumption.

With that, my NAV growth story is tamed, and what you’re left with with EQNX is mainly a stable, yet (relatively) low FFO / dividend yield investment. Of course, it may continue to grow above the inflation rate, especially on the short term, but that is equivalent to multiple expansion / discount rate compression investing. But at least saying Slovenain stocks should re-rate from 10x PE to 20x PE seen in other markets has some ground - whereas with real estate, Ljubljana already starts with above-average rent to income ratio of 61%, so if anything the logical trend would be downward.

3. Valuation: Equinox vs buying an apartment

The natural comparison for Equinox is not some other listed stock – an unlevered listed “REIT”, unless trading at very significantly below NAV and with a clear convergence path (which EQNX, at least on short to medium term, in my view does not have), is not that comparison. Instead, Equinox’s comparable is the typical Slovenian retail investment: buy an apartment in Ljubljana, rent it out and collect the income (and maybe sell it later). This is the path walked by generations of Slovenian savers. An investment in Equinox has to compete first against this alternative.

Below a simple apples-to-apples comparison using a EUR 300k investment sum - roughly the cost of a reasonable 60 m² apartment in Bežigrad, Šiška or Vič – across three scenarios: unlevered direct real estate, Equinox, and levered direct real estate. And while comparing residential to hotel and commercial real estate isn’t an exact apples-to-apples match, the deeply compressed yields of individual residential units in Ljubljana, in my opinion, simply do not justify such a massive risk-reward discount - especially when compared to the highly secure, fully contracted 20-year cash flows backed by Equinox’s master lease agreement with de-facto zero tenant black-swan risk.

I run the comparison of buying an apartmnet outright with cash only (Option A), buying an apartment with a 70% morgage (Option B) and simply buying Equinox stock instead (Option C). For all 3 options I show final return assuming the underlying real estate assets continue appreciating at 0%, 2% and 5% going forward.

Explanatory note: the different figures in row 5 (Capital appreciation) arise because leverage and the NAV discount change what the 2% NAV increase is measured against. For Option A (unlevered apartment), a 2% NAV increase is a direct 2% return on the invested amount. For Option B (70% LTV apartment), the 2% NAV increase applies to the full EUR 296k flat value, but is measured against only EUR 100k of equity - turning it into a 6% return. For Option C (Equinox), the 2% NAV growth applies to EUR 368k, and since you’re buying in at a 19% discount to NAV, this translates into a 3% return on your invested amount.

All-in-all, assuming real estate prices continue to rise in-line with inflation at 2%, Equinox should continue to deliver c 8% return at the current price levels. Assuming the growth in real estate prices fully stalls to 0%, the return would still stand at tolerable 5.5%. However, it’s the upside case that I find very underwhelming - even assumping a 5% NAV growth in perpetuity (an impossibility over the long term, btw), Equinox would still deliver “only” 11.7% return. In my opinion, this compares unfavorably to other alternatives, currently available on the market, and is not attractive given the (very real) risk of a bear case.

On an unlevered, like-for-like basis, Equinox (Option C) is a significantly better investment than a direct unlevered real estate purchase (Option A) on all dimensions – higher expected total return in all scenarios, much higher diversification, lower management burden, and lower transaction costs. Equinox outperforms an unlevered apartment in all 3 scenarios (0%, 2% and 5% annual NAV appreciation).

With the use of leverage, a 70% LTV apartment purchase (Option B) naturally amplifies the upside of both the 2% and 5% annual NAV appreciation scenarios, which an unlevered EQNX portfolio cannot match. Nevertheless, in 0% further NAV growth scenario, Option B with a return of 0.7% starts to underperform EQNX, which would still return 5.5%. In addition, a levered apartment purchase at least in theory also carries a plausible 100% principal loss scenario (e.g. if real estate prices drop more than 30%), whereas it would be almost impossible to lose the entire investment in EQNX (due to its low leverage).

NAV discount (currently 18.5%), remains another, albeit mostly theoretical upside to EQNX. But, since this discount is consistent with – even slightly tighter than – the 20-30% discount typical for European hotel REITs, I view it as reasonable given the illiquidity premium Slovenia commands broadly and the small-cap nature, but not so wide as to scream “deep value”. More importantly, due to low float and trading volumes, and Axor’s selling pressure, I do not see a realistic path in this spread shrinking over the short to medium term.

4. Company background and portfolio

Equinox was incorporated in May 2021 as the legal successor of Union Hoteli d.d., which previously both owned and managed the real estate / hotels. As the market was assigning a significant discount to such a bundled portfolio - it was essentially valuing the company mainly based on the value of its (meager) cash flows from the hotel management business, at hotel management business discount rates - instead of recognising also the value of the actual real estate it owned. Thus, the management had decided to - following also the footsteps of other Western European and US peers – split the asset management and the real estate divisions into separate entities, in order to close the NAV vs valuation gap. The old operating entity retained the hotel operational business and continued to trade on the LJSE as “Union Hotels Collection” briefly, until being liquidated (under a ticker GHUG), while Equinox was spun out with the underlying real estate portfolio and listed on the LJSE in January 2022 - making it Slovenia’s first listed real estate joint-stock company.

As mentioned, the rationale behind the split was straightforward: asset-light hotel operations and asset-heavy real estate ownership are fundamentally different businesses with different risk profiles, different capital structures and different optimal shareholder bases. Separating them allowed Equinox to pursue a clean “real estate investment company” strategy - acquiring properties, leasing them to operators on long-term contracts, and distributing rental cash flow to shareholders - while letting the operational hotel business (now under Hotusa’s management) be valued on its own merits. In turn, Equinox was able to obtain a fairer and much higher valuation. To this end, market cap of previous pre-split Union Hoteli stood at EUR 29m in 2020, compared to EUR 112m market cap of Equinox today.

It should be noted that while EQNX aimed to create a true REIT structure - where the company’s own profits are not taxed, but only the actual distributions to shareholders are - it did not manage to convince Slovenian regulators to approve such a structure. Hence, it is running under a standard corporate entity structure (i.e. CIT at the company level + tax on dividends), which is not ideal for a de-facto REIT.

4.1 The portfolio

The portfolio as of year-end 2025 consists of the following assets:

The geographic concentration is high, but in a relatively good way. The average distance from Prešernov trg – Ljubljana’s central square – is under 300m across the entire portfolio. As management likes to point out, this is the most centrally located hotel real estate portfolio of any EU capital city, a claim that is more than just marketing. In the long run, location drives most of real estate’s absolute return, and Equinox’s locations are effectively irreplaceable – there is no new land to be created 300m from Prešernov trg. See below EQNX’s portfolio locations (excl. Nama Kočevje, which is the only asset not located in Ljubljana):

The 20-year master lease with Grupo Hotusa, covering Grand Hotel Union, uHotel and Hotel Lev, is by far the company’s most important contract. The cumulative contractual rent of over EUR 110m is combined with annual indexation and a variable component tied to hotel operating performance, meaning Equinox partially participates in the upside if Ljubljana’s hotel market outperforms. Hotusa itself is a large European hotel group operating 200+ properties globally, which provides substantial counterparty comfort - though it is obviously not guaranteed. The 20-year tenor is long enough that, from an equity perspective, the bulk of hotel cashflow is de-risked through 2042, removing hotel operational volatility from the equity story entirely.

On a per-m² basis, the Q1 2026 EQNX fair value implies c EUR 2,055/m² (including only appraised gross value of real estate). This is well below headline Ljubljana residential prices of c EUR 4,900/m² (2025 median), but what is more important is the cap rate, which stands at c 5.5% (2025 EBITDA / investment property fair value of EUR 132m). This seems meaningfully lower than publicly cited Ljubljana office prime yields of c 6.5-8%. The c 200 bps spread, is at least partly justified by the 20-year Hotusa anchor lease (long-dated contracted income from an international operator should trade tighter than generic vacant office) and partly by the trophy nature of EQNX’s locations.

5. Financials

The operational track record of Equinox since spin-off is one of a steady revenue growth (which is likely to slow down), margin expansion and a disciplined, predictable capital return profile (50% FFO dividend payout). The table below summarises the key financial trajectory:

Over the 2022-2025 period, revenues have grown at a c 8.5% CAGR – driven by three factors: i) annual rent indexation on existing leases; ii) completion of the 2023-2024 renovation cycle on commercial properties (particularly Delo Tower and parts of uHotel), which brought those assets back to full occupancy at higher rents; iii) higher variable rent component from the hotel portfolio as Ljubljana’s tourism market has kept delivering record results. The EBITDA margin jumped from 74% to 78% in 2025 on completion of the renovations.

On a per-share basis, FFO of EUR 4.12/share against a EUR 63.5 share price gives an FFO yield of 6.5% - a useful framing for thinking about the stock as a real estate cash-flow play rather than a growth business.

The balance sheet is essentially (too) unlevered. Net debt of c EUR 11m against a EUR 140m portfolio gives an LTV of under 8%, compared to 30-50% typical of European REITs. Management has explicitly stated a preference for low leverage. While I understand the prudence, unless one’s key goal is never to lose money, running real estate fully unlevered may not the optimal capital allocation decision for one’s portfolio.

Another point to mention is that over the past years, management has also built up an investment portfolio of c 16.5m, constituting of:

• c EUR 12m is a newly established stake in Fondul Proprietatea, a listed Romanian investment trust

• EUR 3.3m shares in other real estate companies

• EUR 1.2m in other shares & bonds

6. Management, ownership and dividend policy

Equinox is controlled by Axor Holding d.o.o., a privately-held Slovenian investment vehicle, with 75.5% of the share capital. State-owned pension fund manager Kapitalska družba (KAD) holds 18.75%. This leaves a free float of approximately 5-6%.

It shall be noted that since the spin-off, the idea of the majority shareholder Axor has been to downsize their position, thus they continue to sell their shares today & can be expected to do so for the foreseeable future.

The two-person executive leadership is Matej Rigelnik (Executive Director) and Peter Krivc (Chairman of the Board). Both have long tenures with Union Hoteli / Equinox dating back to 2017-2018 and were the architects of the 2021 spin-off and 2022 IPO. Presence of KAD as a sizeable minority shareholder technically provides a useful governance counterweight – but unlike in Cetis I don’t see a real need for it in Equinox.

The execution track record has been, in my view, good on the “hard” metrics: successful restructuring of Union Hoteli, completion of the 20-year Hotusa master lease on favourable terms, consistent delivery on the stated dividend policy, steady rent growth, disciplined capex on renovations. Communication with shareholders is exceptional for Slovenian listed company - each quarter there is a management call where anyone is free to ask questions that the management answers (online). The limitation though is that most materials are in Slovenian only and I am also not very pleased with the granularity of information provided to some key topics, which would be needed for valuation (e.g. more details on Hotusa lease agreement, Fondul Proprietatea etc.).

Dividend policy is clearly articulated: c 50% of FFO paid out annually, subject to the company’s capital needs and board discretion. In line with FFO growth, the dividend per share has grown from EUR 1.2/share in 2022 to EUR 2/share for FY 2025.

Similar to Brodnjak (NLB CEO), Rigelnik also often appears on podcasts and his appearances are also worth a listen, as he also has a very grounded and well-rounded views on both real estate, as well as general investing topics.

6.1 Free float, trading liquidity and the buyback programme

Equinox shares are very illiquid. As per Equinox’s Q1 2026 disclosure, the stock trade volume of around 25 shares (c EUR 1.5k) a day in Q1 2026. Nevertheless, for larger blocks a direct sale agreement with the majority owner (Axor) is probably possible.

The buyback programme has been running since February 2023. Cumulative shares repurchased through 31.3.2026 is 9,558 shares – c 0.5% of total shares, or c 10% of the free float over a roughly three-year period. Given the 19% discount to NAV, every share bought below NAV is immediately accretive to remaining per-share NAV. I would love to see the buyback scaled up meaningfully, hower due to low liquidity and free float that’s not doable.

7. Key Risks

• Concentration in Ljubljana tourism / hotels. majority of NAV is hotel real estate, all leased to a single operator (Grupo Hotusa). If Hotusa defaults or requires major lease renegotiation in a downturn – a 2020 COVID-style scenario is the obvious template – the near-term income base takes a direct hit. While the 20-year tenor and Hotusa’s 200+ hotel platform provide substantial comfort, they are not guaranteed.

• Cap rate expansion. If long-term interest rates re-rate higher or tourism sentiment turns, NAV itself could fall. A 100 bps cap rate increase would reduce gross property value by c 19%, erasing the current discount entirely.

• Concentrated ownership. Axor controls the company. Decisions made by Axor for its own portfolio interests (rather than Equinox’s) are difficult to contest given the tiny free float.

• Capital allocation drift. The Fondul Proprietatea stake and the growing portfolio of other listed REIT shares, while good in theory, are data points requiring monitoring vis-a-vis the pure REIT play strategy.

• Liquidity. c 5-6% free float, c EUR 1.5k of daily exchange turnover means Equinox is a buy and forget investment; anyone expecting daily trading opportunities should look elsewhere. Position sizing must reflect a multi-year lock-up assumption.

• Real estate macro. All the concerns outlined in dedicated section apply – Equinox is, at the end of the day, a bet on Ljubljana real estate prices over the long term. If the asset class under-delivers, so will Equinox.

8. Conclusion

For the typical Slovenian saver weighing a real estate allocation, Equinox Nepremičnine at current prices is a compelling alternative to the “buy an apartment to rent out”. It offers diversification across 8 central Ljubljana properties, a 19% discount to NAV, a 20-year anchor lease with a credible international hotel operator, a 3.2% and growing gross dividend yield, tax equivalence to direct rental income, and zero operational hassle.

However, Equinox, with an expected return of c 8% (assuming NAV growth in line with inflation of 2%) is in my opinion not a substitute for a “standard-company” stock purchase. As such, I do not consider EQNX a Buy at the current price levels. However, in EQNX’s defense, the lower return is indeed largely compensated for with relative safety of principal, and as such the stock might be well tailored for a very specific purchaser who seeks real estate exposure.

10. Appendix - further breakdown of Options A, B and C

10.1 Option A: Direct real estate, unlevered

Scenario: buy a 60 m² apartment in a decent Ljubljana neighbourhood (Bežigrad, Šiška, Vič) at c EUR 4,900/m² = EUR 294k, plus c 4% round-trip transaction costs (2% real estate transfer tax, agent and legal/notary/energy certificate). Total upfront cost c EUR 306k.

• Gross rental yield: c 3.8% p.a. in Ljubljana outside the city centre (per Numbeo Q1 2026 data and Global Property Guide). Gross rent of c EUR 11,200 p.a. on the EUR 294k asset price.

• Direct operating costs: c 25% of gross rent altogether. c 10% of rent for maintenance / renovation reserves; c 8% for property management (if outsourced, which most working professionals do – if you do it on your own you should count at least the same amount for your time, else you are cheating yourself); c 3-5% vacancy allowance; 3% building-level common charges, municipal property tax.

• Net rental income before tax: c EUR 8,400 p.a.

• Tax: Slovenian rental income is taxed at a 25% flat “cedular” rate on net rental income (with a 10% standard deduction, or actual documented costs).

• Net after-tax cash yield: c EUR 6,300 p.a. = c 2.1% on the EUR 300k invested.

• Capital appreciation: 0-2-5% p.a. nominal – though, as per previous sections, I am cautious on the higher end of this range holding over the next decade

• Total expected return: c 2.1-4.1-7.1% p.a (under 0-2-5% nominal NAV appreciation, respectively)

Additional considerations: 100% of the invested capital is exposed to a single property, a single tenant, and a single district. The black-swan event of a non-paying tenant you are unable to kick out is very real (Slovenian tenant eviction processes are slow and favour the tenant). Exit is measured in 3-12 months, with a transaction-heavy, legally fraught process. Management hassle is real and constant – tenant search, lease negotiation, maintenance calls, tax filings, the occasional eviction. For a working professional, the implicit opportunity cost of time spent on property management is meaningful and rarely properly accounted for.

10.2 Option B: Direct real estate with 70% LTV mortgage

Scenario: same EUR 294k apartment, but financed with a 70% LTV mortgage. The investor puts down EUR 90k + c EUR 10k transaction costs ≈ EUR 100k of equity, and borrows EUR 210k at c 4.0% p.a. (Slovenian fixed-rate mortgage in early 2026, 25-30 year amortisation). All assumptions same as for Option A, with the following additions:

• Annual mortgage interest (year 1): c EUR 8,400, declining over time as principal amortises.

• Net rental income after op costs: EUR 8,400 (same as unlevered base).

• Pre-tax net cash flow: c break-even, i.e. rent roughly covers mortgage interest and op costs, with principal amortisation being a “forced savings” mechanism.

• Capital appreciation: 0-2-5% p.a. on the full EUR 294k asset value = EUR 0-6-15k p.a.

• Total expected return: Since rental yield is used to pay the mortgage, the NAV appreciation is the most relevant factor. On a EUR 100k equity base, the EUR 0-6-15k of annual price appreciation translates into 0-6-15% annual equity return, before accounting for mortgage amortisation (which adds c 0.7% p.a. of “forced savings” equity build-up in early years). Expected equity IRR over a 7-10 year hold: c 0.7-6.7-15.7% in a base case (under 0-2-5% nominal NAV appreciation, respectively); materially higher in a bull scenario; negative in a potential downturn.

This is the option that Equinox, by its very structure, cannot replicate. A listed real estate company with 6% LTV cannot be levered further by the individual investor in any practical sense (margin loans exist but are expensive and risky).

10.3 Option C: Equinox EUR 300k equity position

Scenario: buy 4,724 shares of EQNX at EUR 63.5 = EUR 300k, with round-trip commission of c 0.2% on IBKR or similar broker.

• Implied share of NAV: 4,724 × EUR 77.9 = EUR 368k. I.e. EUR 300k of cash buys EUR 368k of net real estate value – an implicit 18.5% discount, aka “immediate EUR 68k NAV gain”. This is the “structural entry bargain” Equinox offers vs buying direct real estate at market prices.

• 2026E dividend: c EUR 2/share (50% of 2025 FFO of EUR 4.12/share). Total dividend of c EUR 9,700 = 3.2% starting gross dividend yield.

• Dividend withholding tax: 25% for Slovenian tax residents (exact same cedular rate as rental income – so the tax treatment is completely equivalent).

• Net after-tax dividend yield: c 2.4% p.a.

• Retained FFO (the other 50%): c EUR 2.05/share × 4,724 = c EUR 9,700 p.a. reinvested into the portfolio via capex, new acquisitions, or (ideally) buybacks. At 0.81x NAV, every euro used for buybacks immediately accretes 23% to remaining NAV per share – a mathematical feature that management is exploiting, but due to illiquidity the progress is very slow. That’s 3.2% of additional yield (not paid out hence no dividend tax applies)

• Expected NAV growth: c 0-2.5-6.1% p.a. (same numbers as per single-apartment example, but 0.5-1% higher to reflect that 18.5% NAV discount amplifies growth)

• Discount compression: could add 0-2% p.a. if the market eventually assigns a tighter discount. This is a free option that does not need to work for the base case to be acceptable –and is not assumed for valuation purposes

• Total expected return: 5.5-8.0-11.7% p.a.

Additional considerations: the EUR 300k is spread across 8 properties in central Ljubljana, with the bulk of income backed by a 20-year contract with a large international operator. Majority commercial real estate, i.e. risk of stranded non-paying tenent black swan event is effectively null. Exit at a “screen price” is measured in days to a few weeks at this size, but the realised exit price can deviate meaningfully from screen – see the illiquidity discussion in section below. Management hassle is zero. Time cost is zero beyond the quarterly report read.

Important note: due to high illiquidity, building a EUR 300k equity position in Equinox could take months (unless Axor is open to direct sale - which I believe they are).

To put the numbers into perspective - Addiko earned €44M in 2025 on net banking income of €317M. Optically, NLB is thus offering a 14.7x P/E -rich for a CEE-based bank. However, Addiko standalone is too small a bank to run an efficient modern multi-jurisdiction bank (remember that the key reason for NLB’s 50-2-1 strategy is precisely to get sufficient scale to remain competitive in terms of size, i.e. to be able to amortise overhead and development costs across a large enough client and asset base). To this end, Addiko runs a horrendous cost-income ratio (CIR) of 61.7%, while NLB runs a CIR of 47%. Once integrated into the NLB Group, Addiko’s CIR should be able to converge closer to NLB Group’s level (single app and technology development, rationalisation of staff and admin functions across a much larger platform, etc.). Best case, Addiko’s CIR would fully converge with NLB’s. So, if we apply NLB’s CIR to Addiko’s revenue base, we “immediately” get ~€37M of additional after-tax earnings for a fully synergised Addiko. While at face value 14.7x entry earnings is not cheap, on synergised earnings NLB is paying a relatively attractive 8.0x - below its own market multiple. In addition, NLB gets access to Croatia.

Pro forma NLB after Addiko acquisition:

However, as outlined in the original deep dive, the problem with Cetis results is that they are bundled up with the impact of subsidiaries, with further noise created by the ongoing construction of the New Factory - making results difficult to interpret at face value. To remind, my investment thesis is centred around the standalone mother entity Cetis d.d., operating in security printing and identity solutions. To this end, a number of adjustments need to be made to exclude the impact of subsidiaries - most notably I remove Financial income, as it mainly relates to dividends and interest received from subsidiaries, which are not part of the valuation, coming out to c. EUR 5m annually. To note, Financial income also includes minor FX gains for both years, which I did not include in my original 2024 deep dive but do add here. For a fair like-for-like comparison to 2024, two additional adjustments are needed:

• Depreciation: increased by EUR 1.1m, likely driven by equipment and assets related to the New Factory programme already placed into use ahead of full opening. De-facto the equipemnt orders from the current capex cycle (New Factory) deteriorate financial results in the short term - despite not yet producing any income. Depreciation can be expected to further spike in H2, once the New Factory is commissioned. And while deducting depreciation from net income is fine in normal times, during the capex surge / factory construction phase this distorts results. We should therefore adjust net income for it until the factory is opened and ramped up.

• EGP write-down: due to poor performance and outlook, the value of the investment in subsidiary EGP d.o.o. was written down by EUR 1.8m, which resulted in a negative P&L charge. As we are valuing only the mother company, we must exclude this one-off charge and adjust net income accordingly.

Adjusting for the above, we get to a “normalised” 2025 net income of EUR 11.0m, 6% below the 2024 comparable of EUR 11.7m.

With the numerator (profits) being squeezed, while denominators continue increasing (asset base; New Factory construction), most metrics have expectedly deteriorated, i.e. debt, assets and equity are up, while cash and profits are down. This trend can be expected to again turn only after the ramp-up of the New Factory:

Subsidiaries:

The annual report also reveals net income for the subsidiaries, with significant deterioration noted at Cetis Flex and EGP. Most notably, Cetis Flex net income decreased from EUR 2.4m in 2024 to EUR 0.8m in 2025. At EGP, the loss widened from -EUR 0.4m to -EUR 1.3m - though due to negative earnings it was already counted at zero in the SOTP Subsidiaries valuation. The only real positive was at Zarja Elektronika, which continued its rapid growth with net income jumping from EUR 1.3m to EUR 1.6m in 2025.

I would note that while the plastic and paper packaging industries (Cetis Flex and EGP) had a tough year overall and are facing increased competition, namely from Asian players, annual reports for none of the subsidiaries are public yet, so not much can be concluded at the moment (revenue is also not yet known - only the net income item disclosed as part of Cetis mother company reporting). However, since Subsidiaries are not the core of the investment thesis and are essentially a “free option”, we can again do a quick and dirty valuation impact - the cumulative value of subsidiaries has decreased from EUR 33.2m at the time of my original post to EUR 17.5m today. I believe the simple 10x PE valuation methodology I used already in my original deep dive, is still conservative.

Receivables:

Share of international sales (i.e. outside Slovenia) increased from 68% in 2024 to 75% in 2025, mainly driven by growth in LATAM and Africa, which cumulatively increased from 51% to 58%, or EUR 7m. This overall is a good sign that international expansion is going well, but, importantly, governments in these regions are slow payers. Yet they are still governments, so eventual collection remains very likely - just with a delay. Cetis accordingly reported a large increase in receivables of EUR 4.8m, and at first sight concerning EUR 3.4m increase in overdue receivables. For the time being I don’t see a reason for concern given the geographical sales mix shift, but the low provisions made by management on these receivables need to be monitored.

Loan to MSIN:

The annual report reveals that Cetis now has EUR 2.8m of loan to MSIN (controlling shareholder) outstanding. I have seen my fair share of questionable de-facto zero interest rate “personal” loans at Datalab - hopefully this is not the same & definitely I do not approve that no explaination whatsoever is offered in the annual report. Unfortunately, such loans are “part and parcel” in Slovenian small caps, but at EUR 2.8m this is a sizeable related party exposure and at least a footnote in the report would be due - hopefully Kapitalska druzba, SDH and KAD take notice, as in theory this should be one of the benefits of having them in the shareholding structure.

Conclusion:

The report confirms the New Factory is expected to open by mid-this-year. Thus, for 2026, I would expect to see continued stagnation - and probably slight bottom line deterioration - as revenues remain capped at maxed-out capacity while opex and salaries continue rising. For H2, it will be interesting to see whether meaningful ramp-up benefits already kick in - I doubt it, and more likely additional fixed costs will match any incremental revenues. In short, despite this year being an election year in Slovenia, which should be a slight tailwind for domestic revenues, one should not expect 2026 results to be above 2025. Cetis is a 2-3 year play on the mother entity’s New Factory, not a short-term one.

For this year, the dividend will likely be significantly below last year’s - cash is lower following EUR 24m in capex throughout 2025 and further EUR 9.6m paid out in dividends. In my original deep dive I had correctly predicted a c EUR 10m cash shortfall for planned capex, which I assumed would be covered by EUR 10m in new debt. In reality, Cetis collected EUR 5m in dividends from subsidiaries, needing only EUR 5m in new debt. Nevertheless, with poor subsidiary performance - which seems to have also been “fully cashed out” last year - and with only EUR 6m of cash on the balance sheet as of YE’25, a materially lower dividend (or none) this year would not surprise me. Let’s see.

With that said, the investment thesis remains intact. No updates to any numbers - other than the Subsidiaries valuation - or investment thesis is needed at the moment. Cetis remains a buy for a patient long-term investor, with a clear mid-term catalyst ahead.

Nova Ljubljanska banka (“NLB”) is the largest banking group headquartered in the ex Yugloslav region, with total assets of EUR 31bn and a with a presence across seven markets. Following a dramatic near-death experience following the 2008 financial crisis - with NPLs peaking at 28% and the bank requiring a EUR 1.9bn state bailout - NLB has been methodically transformed into one of Europe’s most profitable regional banking franchises. Under CEO Blaž Brodnjak’s leadership since 2016, the bank has evolved from a state-controlled zombie into a lean, acquisition-hungry platform that now generates EUR 500m+ in annual profits and trades on the London and Ljubljana stock exchanges.

The investment thesis is built on the following pillars:

• Dominant regional franchise: NLB holds top-3 positions in five out of seven markets where it operates, with a leading 33% market share in Slovenia. This provides a defensive, cash-generative core business with significant pricing power and deep customer relationships built over decades

• High profitability: Following the brutal 2008-2013 period, NLB has emerged with record profitability. The bank delivered EUR 515m and EUR 503m in net profit for 2024 and 2025, respectively, demonstrating the normalized earning power of the franchise. ROE for 2025 stood at 14%, with a cost-to-income ratio of 47% that management targets to compress below 45% by 2030

• The SEE growth & compounding runway: Similar to outlined in my Triglav Insurance company’s deep dive, the Balkans represent Europe’s last frontier of financial services underpenetration. Banking loan-to-GDP ratio is still catching up, and NLB operates in markets with 5% average nominal GDP growth, providing a clear runway for 14% annual mid-to-long term compounding rate at today’s market capitalisation

• Inflation protection and pricing power: Unlike Western European peers, NLB’s deposit base exhibits remarkably low beta (12% vs 40-50% in Western Europe) - deposits in Slovenia, Serbia, and other SEE markets have historically repriced much slower than in developed markets. This creates significant NIM expansion during rate hiking cycles, providing for a strong inflation hedge

• The 50-2-1 strategy: Management’s ambitious 2030 plan targets EUR 50bn in assets (from EUR 31bn currently), EUR 2bn in revenue (from EUR 1.3bn), and EUR 1bn in profit (from EUR ~500m). This represents a 10% revenue CAGR and a 15% profit CAGR from 2025 levels, achievable through both organic growth and strategic M&A. The bank has proven acquisition capabilities, having successfully integrated Komercijalna Banka (Serbia, 2020) and N Banka (former Sberbank Slovenia, 2022)

• Croatia re-entry catalyst: After 30 years of absence from Croatia (due to legacy depositor issues from Yugoslavia’s breakup), NLB has re-entered via its September 2024 acquisition of Summit Leasing Slovenia and its Croatian subsidiary Mobil Leasing. While the first Addiko Bank acquisition attempt failed, NLB has just made a re-bid for the bank at a considerably improved price. The SLS acquisition has already given NLB its first foothold in Croatia, but a successful Addiko bid would transform that into a meaningful presence - completing the final missing piece in NLB's SEE footprint

With a current market cap of EUR 4.6bn (EUR 230/share), NLB trades at 1.2x book and 9.1x 2025A earnings. Relative to CEE banking peers like OTP Bank (2.0x P/B), Erste Group (1.2x P/B), or even Western European giants like UniCredit (1.5x P/B) and Intesa Sanpaolo (1.5x P/B), at the current valuation NLB still presents a solid entry point for investors willing to tolerate CEE/SEE exposure. With a clear organic growth tailwind, the stock can be expected to provide investor a long-term compounded annual return of 13.8% (12.4% net of tax on dividends). However, despite much less the case vs the Slovenian companies I analysed previously, investors must still be prepared for relative illiquidity - with only EUR 1-2m in daily volume on the LJSE, this is a position that requires patience and a long-term mindset for larger positions.

2. NLB’s journey from Yugoslav to regional banking champion

Given NLB’s complicated past - Yugoslav succession disputes, the Croatian depositor issue, and a state bailout following the Great Recession - I cover the history in more detail than usual. Feel free to skip ahead if you’re already familiar.

2.1 The Yugoslav era (1919-1991)

The roots of NLB stretch back to 1919, when its predecessor Ljubljanska banka (i.e. before it became New Ljubljanska banka) was founded in the newly formed Kingdom of Serbs, Croats, and Slovenes following World War I. Slovenia, as the most industrialized and commercially connected region of the new Yugoslav state, naturally developed a sophisticated banking sector. Ljubljanska banka quickly became the region’s dominant financial institution by the interwar period’s end.

Following World War II, communist Yugoslavia nationalized the entire banking sector in 1945-1946, liquidating or absorbing private institutions. Ljubljanska banka was restructured and formally re-established in 1955 under Yugoslav socialist banking laws, operating as a state-controlled entity within Slovenia. Unlike the centrally planned Soviet model, Yugoslav “market socialism” allowed for some degree of banking autonomy, and Ljubljanska banka developed relationships with Western banks that would prove crucial later. By the 1980s, Slovenia accounted for roughly 20% of Yugoslav GDP despite being only 8% of the population - a disparity that created political tensions and ultimately contributed to Yugoslavia’s dissolution.

2.2 Independence and the depositor crisis (1991-1994)

Slovenia declared independence on June 25, 1991, becoming the first Yugoslav republic to break away. The 10-Day War that followed was brief and relatively bloodless compared to the conflicts that would engulf Croatia, Bosnia, and Kosovo. However, the banking implications were immediately catastrophic. Ljubljanska banka became separated from the central bank of the former Socialist Federal Republic of Yugoslavia, continuing with all its previous assets - but not all its liabilities.

The most contentious issue was the treatment of foreign currency deposits held by Croatian citizens at Ljubljanska banka’s branches in Croatia. When Slovenia declared independence, it effectively ‘locked’ these deposits, arguing that the Yugoslav successor states should negotiate a comprehensive settlement of assets and liabilities. Croatia, facing its own war of independence and hyperinflation, passed legislation in 1991-1993 requiring Croatian banks to honor Ljubljanska banka deposits for Croatian citizens, then sued Ljubljanska banka (and later NLB) to recover these sums plus decades of accumulated interest. This created a legal nightmare that persists even today, with Croatian courts awarding billions of euros in judgments against NLB - claims that Slovenia contests as violations of international succession law. It shall be noted that NLB itself is de-risked - any potential penalty is de-facto backstopped by the Slovenian government as per the law passed in 2018, when the resolution to this issue was required to facilitate the IPO. As such, this is not a live contingent liability on NLB’s balance sheet in any material way.

In addition to the depositor issue, Slovenia faced the challenge of transforming a socialist banking sector into a market-based one. From 1991 to 1994, the broader Slovenian banking system underwent restructuring, as the heritage of past communist political interference left banks overexposed to unprofitable state-owned companies. The government issued reconstruction bonds and provided direct capital injections estimated at around 20% of GDP to cover bad debts, enabling the separation of domestic banking assets from loss-making Yugoslav-era operations.

In July 1994, the Slovenian Parliament adopted the Constitutional Law that formally established Nova Ljubljanska banka (literally translated to ‘New Ljubljana Bank’) (“NLB”). The new entity took over a large portion of Ljubljanska banka’s assets and liabilities, but with a clean legal structure focused on Slovenian operations. The residual entity, still called (old) Ljubljanska banka, was left to wind down legacy Yugoslav positions. This restructuring effectively created the modern NLB, owned 100% by the Republic of Slovenia.

2.3 Growth and partial privatization (1994-2008)

With a clean slate and state backing, NLB focused on consolidating its dominant position in Slovenia’s banking market. By 2001, the government adopted a privatization plan, seeking to reduce its stake while maintaining strategic control. In 2002, Belgium’s KBC Bank acquired a 34% stake for EUR 435m, with EBRD taking another 5%. This marked NLB’s first step toward becoming a commercial, profit-oriented institution rather than a policy tool. However, while KBC had representatives on the management board and certain veto rights under a shareholder agreement, the Slovenian government effectively held the steering wheel. By 2006, because the state refused to let KBC become the majority owner, KBC officially withdrew from “active management” and reclassified its investment as a passive financial stake. From that point until the 2013 bailout, operational control reverted almost entirely to state-appointed management.

During this period, NLB also began expanding across Southeast Europe, targeting former Yugoslav markets where it had historical relationships and local knowledge. The bank entered Macedonia, Montenegro, Bosnia & Herzegovina, Kosovo, and Serbia - everywhere except Croatia, which remained off-limits due to the ongoing legal disputes. By 2008, NLB had established a network of subsidiary banks across the region, leveraging its “home market” advantage: shared language, cultural understanding, and regulatory familiarity that foreign Western banks could not easily replicate in the Balkans.

Slovenia’s economy boomed during this period. Having joined the EU in 2004 and adopted the euro in 2007, the country experienced rapid credit growth and rising asset prices. Banks, including NLB, still under de-facto state and political control, extended loans aggressively to corporates and real estate developers. The NPL ratio in 2008 stood at an optically healthy 3.8%. However, there were serious concentration risks - much of NLB’s corporate loan book was tied to leveraged buyouts and construction projects that assumed perpetual growth.

2.4 The crisis & bailout (2008-2013)

The 2008 global financial crisis hit Slovenia hard, with GDP having contracted 7.6% in 2009 - one of the worst performances in the EU. The cheap credit that had fueled management buyouts and construction booms vanished overnight, leading to widespread bankruptcies. Slovenia’s export-oriented economy, heavily dependent on trade with Italy (itself in crisis), faced collapsing demand. For NLB, this resulted in its non-performing loans to explode from 3.8% in 2008 to 28.2% by 2012. The bank’s entire capital base was wiped out by loan losses.

In 2012, KBC - facing its own capital pressures from the European debt crisis - sold its 34% stake back to the Slovenian state for a nominal sum, re-nationalizing NLB. By 2013, Slovenia’s entire banking system was on the brink of collapse, threatening to push the country toward an EU/IMF bailout like those imposed on Greece, Portugal, and Ireland. The ECB and European Commission conducted an Asset Quality Review and stress test, identifying a EUR 1.7bn capital shortfall at NLB in the adverse scenario.

To avoid an international bailout, in December 2013, Slovenian government implemented a “bail-in” of NLB’s subordinated debt and hybrid capital instruments, wiping out EUR 250m of junior creditors. It then injected EUR 1.9bn of fresh capital, fully nationalizing the bank. The bailout was approved by the European Commission under strict conditions: NLB was required to restructure, divest non-core assets, and eventually be sold to private investors.

Consequently, in 2013 NLB posted a record loss, reflecting massive write-downs. The bank’s market reputation was shattered - it was perceived as a zombie institution, kept alive only by state life support. The restructuring mandate forced NLB to transfer EUR billions in bad loans to Slovenia’s “bad bank”, cleanse its balance sheet, and radically shrink its operations. Employee headcount was slashed, unprofitable branches closed, and management replaced.

2.5 The Brodnjak Era (2012-Present): Resurrection, transformation and acquisitions

Blaž Brodnjak joined NLB’s Management Board in December 2012, in the midst of the crisis. By the time, he already carried 25+ years of prior experience across various international banking and insurance groups (Bank Austria Creditanstalt, Raiffeisen, BAWAG P.S.K., Hypo Alpe Adria and Zavarovalnica Triglav). In July 2016, he was promoted to CEO - a position he has held since.

Brodnjak’s task was to transform a state-owned zombie bank into a commercially viable institution capable of attracting private capital. His strategy centered on four pillars: i) Complete the balance sheet cleanup and drive NPLs toward Western European norms; ii) Rebuild profitability through cost discipline and selective loan growth; iii) Strengthen the regional franchise through strategic acquisitions; iv) Prepare the bank for privatization by demonstrating consistent, sustainable earnings.

The results speak for themselves. By 2017, NLB posted a record EUR 225m net profit. NPLs fell from 28% in 2012 to under 7% by 2018, then to 2% by 2025 - lower than many Western European peers. Cost-to-income ratio compressed from over 60% in the crisis years to below 47% today. Management maintained strict underwriting discipline, refusing to chase market share through aggressive pricing.

In November 2018, NLB completed its long-delayed IPO, listing on both the Ljubljana Stock Exchange and London Stock Exchange (via GDRs). The Republic of Slovenia reduced its stake to 25% plus one share, retaining strategic control while bringing in a diversified international investor base including EBRD and Brandes Investment Partners. The IPO price valued NLB at approximately EUR 1.2bn - still depressed relative to normalized earnings (EUR 204m and 194m for 2018 and 2019, respectively), reflecting lingering skepticism.

With the acquisition of Serbia’s Komercijalna Banka in December 2020, NLB emerged as a true regional champion. For EUR 395m, NLB acquired an 83% stake in the third-largest Serbian bank with EUR 4bn in assets, at 0.77x its book value and 6x 2019 net income. It allowed NLB to instantly become a top-3 player in the Balkans’ largest economy. The deal was executed in the midst of COVID-19 pandemic panic, a testament to NLB’s execution capacities. Full squeeze out and 100% ownership was reached in 2022, and the integration was completed successfully by 2023, with Komercijalna merged into NLB’s existing Serbian subsidiary to create NLB Komercijalna Banka, which made EUR 153m in net income in 2025 alone.

In March 2022, NLB pulled off another successful merger. When Russia invaded Ukraine and the EU imposed sanctions on Sberbank, the Single Resolution Board transferred Sberbank’s Slovenian subsidiary (1,200 employees, EUR 1bn in assets) to NLB overnight for token consideration. Later renamed N Banka, it was fully integrated into NLB by September 2023. This opportunistic acquisition - essentially ‘free’ in economic terms - added 100,000+ customers and further consolidated NLB’s Slovenian dominance.

NLB is again also bidding for Addiko - NLB made a EUR 29/share all-cash offer in April 2026, valuing the bank at ~EUR 566M, issued one day after Raiffeisen announced its own indicative offer at EUR 23/share.

For some context - Addiko is a Vienna-listed CSEE bank with ~EUR 6.4bn in total assets, ~EUR 900M in equity, and EUR 44M net profit in 2025. It’s running branches across Croatia, Slovenia, Bosnia & Herzegovina, Serbia and Montenegro, predominantly in consumer and SME lending. Importantly, Croatia is Addiko’s core asset: with ~EUR 2.3bn in assets it’s its single largest subsidiary, accounting for roughly a third of its group balance sheet - and this is what’s especially attractive to NLB when it comes to Addiko. Since a direct, organic Croatian market entry for NLB, due to the historical disputes, would be very difficult, an indirect entry via an acquisition is the only real option.

NLB already attempted to acquire Addiko in 2024 - at EUR 22/share, but got only 36% shareholder acceptance against a 75% threshold. The new bid is at a meaningfully higher valuation. Whether it lands is genuinely uncertain, as the acceptance bar remains high. To this end, this acquisition is not assumed in my valuation. Standalone, EUR 29/share is not a bargain, implying 12x 2025A PE, albeit an attractive PB of only 0.6x. However, the strategic case for NLB specifically is solid: Addiko unlocks Croatia at scale and the cost synergy math is straightforward when you’re integrating a (too) small bank into a large, existing regional platform.

That said, Addiko deserves a word of caution - Addiko was carved out of HETA Asset Resolution - the wind-down vehicle of Hypo Group Alpe Adria, one of the largest bank failures in Austrian history, whose aggressive Balkan expansion in the 2000s ended in a EUR 19bn state bailout by Austria and years of legal disputes across the region. Addiko has been substantially cleaned up since, but the legacy of that franchise is not the same as acquiring a clean-sheet regional bank. Croatian retail banking is also notably more competitive and margin-compressed than NLB’s core markets, with Erste, Raiffeisen, and ZABA (UniCredit) all deeply entrenched. Integration complexity here is meaningfully higher than it was for Komercijalna or N Banka. At ~12% of NLB’s current market cap, a misstep would not be existential, but it would not be trivial either.

Nevertheless, management has earned the right to be trusted on M&A execution. Komercijalna in 2020 and Sberbank Slovenia in 2022 both worked out - the former during a global pandemic, the latter acquired literally overnight under geopolitical emergency conditions. Croatia is the final missing piece in NLB’s regional footprint, and given the structural barrier to organic entry, this may be the only realistic window to get there. I back management’s judgement on whether to pull the trigger.

For those looking to learn more about Blaž Brodnjak, NLB, or the regional economy, I’d strongly recommend listening to his podcast appearances. Episode 4 of Banking on Success podcast, for example, is available in English. Brodnjak is a banking chief with first hand feel for the regional economies he operates in, and is able to articulate their strengths and vulnerabilities with great clarity.

3. NLB’s markets

NLB operates across seven markets covering 16 million people. Relative to Western Europe, these are underbanked, fast-growing markets converging toward Western European financial norms over a multi-decade horizon, providing for a significant organic growth tailwind - similar to what I’ve already described in my deep dive of Zavarovalnica Triglav. As of 2025, NLB’s market shares by total assets were as follows:

Importantly, these markets are also characterised by relatively low loan penetration as a % of GDP - hovering between 35-60% vs 80+% Eurozone average. This provides a “buffer” against defaults in case of a downturn, as both individuals and corporates are relatively little levered. In addition, this structural under-penetration provides multi-year upside for relative loan growth as incomes converge, investment activity accelerates and financial deepening continues.

Consequently, NLB’s loan-to-deposits ratio is also relatively low at 76% vs 94% average for Eurozone, providing i) liquidity cushion and ii) pricing power (i.e. low deposit beta) vis-a-vis the deposit holders. More of this in next chapters.

Below also a snapshot of NLB’s activities & performance throughout the region:

4. Financials

4.1 Historical performance

NLB’s financial history since the 2013 restructuring follows a clear narrative of steady recovery, a step-change from the Komercijalna acquisition in 2020-2022, and an earnings explosion from 2022-2024 driven by the interest rate environment.

Over the past 9 years, net interest income has expanded from EUR 317m in 2016 to EUR 947m in 2025, for an annual CAGR of 13%. Of course, much of this can be attributable to the interest rates (as well as Komercijalna acquisition), with the most notable step-jump seen in 2023, when, following ECB’s rate hiking cycle, NLB’s interest margin peaked at 3.6% in 2024. However, despite the interest rate’s normalisation, 2025 net interest income continued to grow, mainly thanks to overall balance sheet expansion.

Similarly, net fee income has also shown a steady growth, with a CAGR of 10% over the past 10y period. Nevertheless, the increase in opex, especially during the 2023-2025 period has mostly negated the growth in the net operating income over the past 3 years - and the (de)celeration of growth in opex will be one of the key items to look at over the coming years. While the profits have remained rather steady over the past 3 years, the 31% increase in total equity position during the 2023-25 period has resulted in a compression of ROE from its peak of 20.5% in 2023, down to 14.1% as of 2025. Management has been intentionally running over-capitalised structure, mainly to allow for potential larger M&A. In case the 2nd Addiko attempt also fails, an increase in dividend payout ratio (or hopefully even a buyback - though very unlikely) should be at the top of management’s agenda, in order to prevent further ROE deterioration.

4.2 Revenue mix

A common question about European banks is whether their profitability is sustainable as interest rates normalize. For NLB, Net interest income accounted for approximately 73% of Group operating income in 2025 - clearly the dominant driver. This is higher than some of its largest regional competitors, such as Unicredit (65%) and Raiffeisen (70%).

However, the net fee income trajectory is also encouraging. NLB’s fee revenue grew from EUR 146m in 2016 to EUR 343m in 2025 - a 2.3x increase over nine years. Importantly, fees income is continuing its growth also following the interest rates stabilisation after 2023. This reflects a structural shift in banking behavior across NLB’s markets: more card payments, more digital transactions, more investment products, more insurance cross-sell, and more sophisticated corporate treasury services. Fee income is more stable than NII across rate cycles, and management has explicitly targeted growing fee income as a proportion of total revenue. NLB Skladi’s rapid AuM growth is part of this strategy.

The critical question for 2025-2027 is NIM trajectory. Management noted that margin stabilization began in Slovenia toward year-end and is expected to extend across the SEE region. This can be seen also in the chart below - in Q4 2025, due to a +7bps uptick in Slovenian NIM, which was enough to fully offset a still -7bps reduction in SEE banks NIM, the actual NLB group’s NIM also increased from 3.24% to 3.25% during Q4.

Moreover, despite the still falling trend of NIM in 2025, net interest income increased 2% YoY, driven primarily by robust loan growth. To this end, we will take 2025 as a base “normal” year for our valuation purposes, since, especially with the most recent uptick in Euribor during Q1’26, group NIM can be assumed to have fully plateaued.

4.3 The deposit beta

The deposit beta - how much of the central bank rate change flows through to customer deposit rates - is the single most important variable for bank NIM analysis in a rate cycle. In Western Europe, mature banking markets with sophisticated retail depositors and intense competition saw deposit betas of 40-70% during the 2022-2023 rate hiking cycle. German savings banks, Dutch retail banks, and UK building societies all saw significant deposit repricing pressure as customers sought better returns.

NLB’s SEE markets are structurally different. The cumulative deposit beta at Group level through the entire 2022-2024 rate hiking cycle was only approximately 12% - meaning that for every 100bps of ECB rate increase, NLB’s average deposit rate only rose ~12bps. This is also helped by NLB’s stable and high 90% share of sight deposits in Slovenia and 70% for its subsidiaries in other countries.

Many retail depositors across Slovenia and SEE are less financially sophisticated than their Western European counterparts. They do not actively shop for the best deposit rate; they park savings in current accounts and rarely switch banks. They do not have a brokerage account with access to public market alternatives either.

This extraordinary outperformance relative to Western European peers and means that, while NLB’s NIM, considering 2% long term Euribor rates, has now most likely already stabilized, NLB is very well positioned to reap extraordinary profits shall the inflationary period return. This exemplifies that NLB enjoys a strong pricing power moat, which is rather unusual for a “commoditised” industry. Below you can see illustration of just how little the yield on deposits (red line) have moved following a sharp increase in Euribor, yield on loans and interest margin during the 2022-2024 period:

To the contrary, it shall be noted that in the (unlikely) event that Euribor would again drop well below 2%, NLB would suffer more than other Western banks, as low deposit beta becomes a negative in a declining interest rate enviroment. Additionally, the low beta dynamic may shift over time. As loan-to-deposit ratios gradually approach Western levels, banks will need deposits more - and when that happens, the fix is simple: raise deposit rates. Besides established local competition, Revolut is also gaining ground offering interest on sight deposits - though this does not present an imminent threat to older (and corporate) clientele with the largest cash balances. But for now, this gives regional banks something rare: true pricing power on both sides of the balance sheet.

5. The loan book

5.1 Loan portfolio breakdown

Over the past 8 years, NLB’s gross loan portfolio has shown a steady growth at 12% CAGR, which was fully followed by equal 12% CAGR in deposits. 52% of total loan book is to Individuals (of which 52% for housing), with the remainder to corporates (o/w 61% SMEs) and state.

Corporate and SME lending covers a broad range of industries. Key sector exposures include manufacturing at 24% (Slovenia’s export-oriented industrial base), construction & real estate at 16%, and trade at 19%. Importantly, exposure to automotive industry, currently under most stress stands at only 4% of corporate portfolio. NLB has also consciously avoided the kind of highly leveraged buyout financing that nearly destroyed it in 2008-2010. The credit culture today is notably more conservative, with strict sector concentration limits and explicit underwriting standards.

5.2 NPL Ratio: The Long Road from 28% to 2%

NLB’s NPL ratio evolution is one of European banking’s great rehabilitation stories, having dropped from 28% in 2012 to 2% today, which is a result of 12 years of systematic workout, write-off, and disciplined new underwriting.

It is worth noting that the Komercijalna Banka acquisition brought Serbia’s historically higher NPL ratios into the Group. Serbian banking system NPLs stood at around 3-4% when NLB acquired Komercijalna (vs. European average of 2-3%). The successful integration and systematic workout has brought NLB Komercijalna Banka’s NPL ratio in line with Group norms.

The slight uptick in NPLs to 2% in 2025, while not disclosed in reports, I believe is due to the on-going Unior (manufacturing of metal end products) and SIJ (steel manufacturing) restructurings. Both companies have since initiated restructuring, so we might see a downward revision in Q1-Q2. Nevertheless, the era of ultra-low sub-2% NPL ratios of late- post GFC era is likely coming to the an end generally, not just for NLB. But, to note - a non-performing loan, so long as the actual borrower does not default and / or the bank was sufficiently secured and manages to collect the full principal, can actually be earnings upside for a well capitalized bank, such as is NLB.

6. The Asset Management business

Asset management is NLB’s fastest-growing non-banking business line. It generates fee income with minimal capital requirements, provides a recurring revenue stream uncorrelated with the credit cycle, and deepens customer relationships by expanding NLB from a transaction bank into a comprehensive financial partner.

NLB Skladi (NLB Funds) is Slovenia’s largest mutual fund management company and the country’s largest asset manager by AuM, holding a 42% domestic market share - an almost monopoly-like position that reflects both the company’s early-mover advantage and NLB Group’s captive distribution through its existing clients.

As of YE’25, NLB Skladi had AUM of EUR 3.6bn, recorded EUR 245m of net inflows throughout 2025 and served over 100k customers. This reflects both the broader secular trend of Slovenians slowly shifting savings from low-yield deposits into equity and bond funds, and NLB’s success in cross-selling investment products through its branch and digital channels. As financial awareness grows and deposit rates decline (making funds more competitive), AuM should continue expanding. Furthermore, the newly introduced INRV accounts in Slovenia (tax advantaged savings plans similar to 401k, that were just introduced) should also provide for a tailwind.

By keeping a high market share in funds management, NLB should also be able to at least partially mitigate the impact of its deposit base becoming less sticky & more sophisticated over the long term - with potential decrease in deposits growth not having a significant impact on the group’s performance anyway due to a ceiling on the loan growth.

With combined net profit of EUR 15m as of 2025, the asset management business is still far from a major profit contributor in absolute terms - but with the Slovenian unit’s net income tripling over the past 5 years, it’s on a very fast growth trajectory, and is carrying extremely high incremental margins. Importantly, management’s target of EUR 2bn+ in revenues by 2030 explicitly includes a material contribution from non-banking financial services, of which asset management is the most scalable. We shall see - generally, NLB should be well positioned to take advantage of this secular trend of increased financial literacy & penetration in the region, due to its current market presence and brand perception. Furthermore, gradual rotation from US asset companies and asset managers should also provide a slight tailwind to the local players. Nevertheless, the competition for this lucrative market - both from established (US) players, as well as local incumbents - is relatively high.

7. Management & the 2030 strategy

7.1 The team

The management board beyond Brodnjak, which I already introduced in the historical section, is a compact team of six, with two names worth flagging in particular. Archibald Kremser has been CFO since 2013 and is now also Deputy CEO - an Austrian, who came to NLB via a decade in consulting (EY, Bain) followed by senior finance roles restructuring Kommunalkredit Austria’s troubled balance sheet after the GFC. He’s been the face of NLB’s capital markets activity. Second is Andreas Burkhardt, also on board since 2013, who serves as a CRO - a German national who spent his pre-NLB career in the Volksbank network across the Balkans and CEE, running risk and finance for subsidiaries in BiH and Romania. The most recent addition is Reinhard Höll, who joined as Chief Transformation Officer in mid-2025. He’s a former McKinsey partner who started his career at Lehman and Standard Bank before spending over a decade advising banks on digital transformation; his appointment is a direct signal of how seriously NLB is taking the technology execution risk embedded in the 2030 plan. The remaining board members - Andrej Lasič (CIB), Hedvika Usenik (retail), and Antonio Argir (governance and payments) - are all long-tenure internal promotions, which speaks to a degree of institutional continuity that, for a bank with NLB’s history, is probably a feature rather than a bug.

The management team is exceptionally international by Slovenian standards - a factor I consider a significant competitive advantage. Relying solely on a limited local talent pool often restricts a company’s potential (as well as can be sign of political influence), whereas sourcing global / regional expertise ensures a higher caliber of leadership. I expect, and hope, to see this trend of internationalization spread to other Slovenian firms, mirroring the established practice of major organizations in the Western markets.

Nevertheless, the management team of NLB has a large flaw - collectively, the entire management team only owns 6,564 of NLB shares (worth EUR 1.5m), which is less than half their 2025 net take-home pay. This means that, beyond bonus & salary, their actual personal fortunes are not tied to the company’s long term performance at all. While not unusual for Slovenia, it does not reflect well on the team.

7.2 Strategy 2030: The 50-2-1 ambition

NLB unveiled its five-year strategy in 2024. The headline target is what management calls ‘50-2-1’: EUR 50bn in assets, EUR 2bn in revenues, and EUR 1bn in profit by 2030. From the 2025 baseline (EUR 31.5bn assets, EUR ~1.3bn revenues, EUR 503m profit), this implies:

• Asset growth: 10% CAGR over 5 years - achievable through a combination of 7-8% organic loan growth and selective M&A

• Revenue growth: 9% CAGR - in line with the asset growth, aided also by fee income

• Profit growth: 15% CAGR - operating leverage, reflected in management’s aim for <45% CIR from the current 47%

According to the management, the ambitious EUR 50bn balance sheet strategy was set out of a necessity, noting that this is what they deem to be the minimum critical mass for sufficient economies of scale to kick in, considering the regulatory (compliance) requirements. One of the core pillars of the strategy is also strong development of digital and mobile first retail banking platform, which will allow a significant increase in efficiency, in turn also allowing both the expansion of relatively more lucrative retail business, as well to do so without much overhead. Also here a critical mass is important, as competition with Revolut - as well as other regional players such as RBI, Unicredit and Erste - who can amortise their software costs across pan-European base is only possible with significant scale.

The 2030 targets assume no large M&A - i.e. it’s based purely on organic growth plus bolt-on acquisitions. A larger acquisition, such as a Croatian bank, would be on top of these targets - I believe potential Addiko acquisition is also outside these targets. This framing is important: it means the 2030 plan does not depend on NLB finding the perfect acquisition target; the business grows organically regardless.

8. Ownership structure & dividend policy

8.1 Ownership structure

NLB has an unusually clean and transparent ownership structure for a post-communist bank, reflecting the successful IPO and subsequent management of the shareholder register.

Following the IPO, the Republic of Slovenia’s stake has been reduced to 25%+1, meaning the government is no longer an activist shareholder - it generally supports management’s commercial strategy - but it does retain veto power over extraordinary decisions (major M&A above certain thresholds requires government consent, given the state’s golden share provisions). For practical purposes, the government has been a supportive and financially disciplined shareholder: it wanted dividends (which NLB provides) and it wanted privatization (which was successfully executed). The relationship is cooperative, not adversarial.

The EBRD has also been consistently supportive of NLB’s management and has generally voted in line with Board recommendations at general meetings.

From a trading liquidity perspective, NLB has approximately EUR 1-2m of daily trading volume on the Ljubljana Stock Exchange, with additional volume on the LSE via GDRs. For institutional investors managing large positions, this is limited liquidity and requires patience when building or exiting. As per outlined in my other deep dives, this illiquidity is a structural feature of the LJSE and is not specific to NLB - it is a premium the patient investor collects in exchange for accepting that any exit may take time.

8.2 Dividend policy

Management has committed to a 50% payout ratio from 2024 profits onward, and has so far been delivering on this commitment. At the current run-rate profit rate of EUR 500+m, this implies a minimum EUR 257m p.a. (which was the distribution in 2025), or a 5.6% dividend yield based on the current market cap.

The state as 25% shareholder “guarantees” the dividends - the government explicitly budgets NLB dividend income as a revenue line in the Slovenian state budget. This creates a natural political constituency for maintaining the dividend policy, even in years where the commercial logic might suggest retaining more capital for M&A. Management navigates this by structuring dividends in two tranches (one approved at the spring General Meeting, one in autumn) to preserve flexibility if a major acquisition emerges mid-year.

9. Funding and the Balance sheet

9.1 Funding

NLB’s balance sheet is predominantly deposit-funded. Total deposits stand at EUR 24.5bn, representing approximately 77% of total liabilities. This mostly retail deposit base provides the most stable, cost-effective funding available to any bank - and in NLB’s case, at very low deposit beta (c 12%) as discussed in chapters above.

Beyond deposits, NLB has a diversified wholesale funding stack to meet MREL requirements and support the growth plan. Key outstanding issuances as of early 2026 include:

Importantly, the instruments with rates on the upper end are the ones with earlier issuance dates (2022-23), being callable already mostly during 2026-27. As they were issued in a period of elevated market rates, NLB should be able to re-issue them at currently prevailing (lower) rates, in turn lowering total annual interest expenses from the current run-rate of c EUR 150m down to an estimated c EUR 109m by 2030. That’s an 8% increase in earnings yield just from lowering funding costs.

9.2 Capital Ratios

NLB’s capital position is strong and provides capacity for continued growth, dividends, and selective M&A.

With a 15.5% CET1 ratio, NLB holds c EUR 800m in excess capital, providing the necessary headroom to sustain a 50% dividend payout while maintaining capacity for strategic bolt-on acquisitions. Following S&P’s upgrade to BBB+ (positive outlook) in June 2025 - a reflection of improved credit quality and lower NPL ratios - management is now targeting an A- rating by 2027-2028. Moodys also upgraded their credit rating in March this year, from A3 to A2. This trajectory is expected to further compress wholesale funding costs and solidify NLB’s competitive standing in the capital markets.

10. Valuation

10.1 Relative valuation

NLB trades below the median on mostly all metrics, especially so on P/B basis where it trades at 1.2x vs median average of 1.5x. It also offers among the highest forward dividend yields at 6.1% vs median average of 4.9%. Nevertheless, where NLB lacks is in ROE, which is below the median 16.8% of its peer group, at 14.1%. However, when compared to more Eurozone centric comparable banks, such as Unicredit, Intesa, Erste and RBI, which together average ROE of 12.1%, NLB’s ROE compares more favourably. The other banks with significantly higher ROE, such as OTP (HUF), BT (RON) and PKO (PLN) all carry with them significantly higher local currency exposure, in turn driving the excessively high absolute ROE.

Similarly, while NLB ranks above the pure-play CEE and EE players (OTP, BT and PKO; at 61-77%) in terms of loan-to-deposit ratio, it still has some headroom to reach its Eurozone + SEE peers (Unicredit, Intesa, RBI, Erste; 71-92%).

10.2 My valuation

As already detailed in my Zavarovalnica Triglav deep dive, I believe the continued relative under-financialisation of the Adria and Balkans region, combined with its above-Eurozone growth, will remain a strong structural tailwind for all financial businesses in the region.

Approach A: Back-of-the-envelope compounding

For a stock with a stable long-term growth trajectory, the expected buy-and-hold return can be expressed simply as:

Expected return = Dividend yield + Expected earnings growth rate (+ Multiple Expansion)

NLB’s target markets can be expected to grow at ~5% nominal GDP (~3% real), and with loan-to-GDP ratios of 35-60% across those markets versus 80%+ in Western Europe, an additional ~1.5x multiplier from continued financial deepening is reasonable for the foreseeable future - at this pace, full convergence with Western loan-to-GDP levels would take roughly 25-30 years. This yields a ~7.5% organic annual earnings growth runway on the interest margin side. In addition, and given it’s heavy focus by the management, we can reasonably assume net fees income to continue growing at the historical 10%. With the net interest income currently represending 73% of total revenues, and the net fee income the remaining 27%, the weighted annual growth rate can thus be estimated at 8.2% (7.5% x 73% + 10% x 27%). Plugging into the formula: 5.6% dividend yield + 8.2% earnings growth = 13.8% gross annual return, or 12.4% net of the 25% dividend withholding tax (5.6% x (1-25%) + 8.2%). Notably, this remains materially more conservative than management’s own 2030 plan implying a 15% net profit CAGR over the next 5 year period. Furthermore, it assumes no value appreciation via multiple expansion.

Approach B: dividend discount model

For the DDM, I model out the value on assumptions largely in line with those described under Approach A above, most notably:

• Loan & deposit CAGR of 7.5%, as per explained above

• Net fee income growth of 10%, in line with historical average

• Loan yield and deposit yields remaining flat at Q4’25 levels of 5.2% and 0.5%, respectively

• Opex continuing to grow at 8%, in line with avg. revenue growth (notably, this assumes CIR remains mostly flat vs management’s 2030 target of 2% decrease in CIR from current levels)

• Provisions run-rate of 40bps ( above recent years averages, but more in line with what can be expected long-term & a midpoint of management’s target)

• MREL funding costs gradually declining from the current EUR 148m to EUR 108m by 2030

• 50% annual dividend payout ratio

With the above assumptions, I arrive at 2030 total assets, revenues, and net income of EUR 44bn, EUR 2.0bn, and EUR 742m under the Base Case, respectively - still materially below management’s 2030 50-2-1 target, leaving ample headroom. I then sensitise the 2030 exit across various P/B exit multiples and 2030 net profit scenarios (sensitised purely through net interest and net fee income growth rates), yielding the following outputs (relative to the current EUR 4.6bn market cap):

Under the Base Case, where net income grows to EUR 742m by 2030 per the assumptions outlined above and P/B holds at the current 1.2x, the IRR comes out at 14.0% (before dividend taxes) - marginally above the 13.8% under simpler Approach A.

Under the Upside Case, which assumes management’s target of EUR 1bn in net profit is met by 2030 - an outcome that would likely also prompt a re-rating to 1.5x P/B — the IRR rises to 21.3%. My 2030 revenues under this case stand at EUR 2.3bn and total assets EUR 50bn.

Under the Downside Case, which effectively assumes the company flatlines at the current run-rate profit of EUR 500m and the P/B re-rates down to 1.0x as a consequence, the IRR still comes in at 7.8% - a decent return for such a conservative scenario.

Conclusion

As outlined above, despite the strong run-up in price over the past 3-5 years - driven in part by absurdly depressed starting valuations (recall that at the 2020 trough, the entire company traded at a market cap of just around EUR 700m) - I continue to see NLB as a strong long-term Buy. A 13.8% annual compounder with a clear organic growth runway that also offers a meaningful inflation hedge remains a good addition to a portfolio. While the stock likely offers little further multiple expansion - that re-rating has already largely played out over the past two years - the underlying business’s clear organic growth path still makes it a compelling long-term compounding play. It remains a Buy at anything below a EUR ~5bn market cap.

11. Key Risks

11.1 The future of banking

My main long term concern is about how people, as financial literacy matures globally and locally, will handle their money. Over the next years and decades, more people will have access to investment accounts and will look for better deals at digital banks to protect their savings from inflation. NLB could match those higher rates to keep customers, but that would directly cut into its NIM and NII. This is a long-term risk rather than an immediate one, because the older generation is less likely to switch, but it’s a trend that will definitely present a challenge as younger customers become the majority.

Even if neobanks never get to the level of being able to offer mortgages and personal loans at scale (which is also possible, albeit it shall be noted Revolut is already rolling-out a mortgage product), if majority of savings switch to them, deposits will be fully commoditized - and the low deposit beta moat of NLB would be gone.

11.2 Interest Rate Sensitivity

NLB’s NIM is structurally exposed to ECB rate cuts. Management confirmed that NIM has already stabilised in Slovenia, and they expect stabilisation also in other markets shortly. As of Q4’25, NIM stood at 3.25% in 2025 (vs. 3.24% in Q3 2025 - i.e. an uptick / stabilisation). Nevertheless, a more aggressive ECB cutting cycle (e.g., rates to 1.5% by end-2026) could push NIM below 3.0%, which at current asset volumes would translate to a c EUR 50m annual net profit reduction. This is manageable given the current profit base but bears monitoring - but, especially given the most recent dynamics, appears as highly unlikely.

It shall also be noted that, if the ECB cuts rates below 2%, the low deposit beta (currently a strenght) could become a headwind. In a declining rate environment, Western banks often see NII hold up better because their deposits reprice down faster. NLB’s “sticky” low-cost deposits might not have much room to move lower since they never moved high in the first place.

11.3 Serbian Political and FX Risk

Serbia contributes ~30% of Group profits but carries the most concentrated political and FX risk. The 2024-2025 street protests following the Novi Sad station collapse showed that Serbia’s political stability is not guaranteed. A government crisis, snap elections, or policy shifts - particularly around the banking sector tax (0.2% balance sheet levy) - could dent Serbian profitability. The RSD/EUR rate has been stable, but a depreciation scenario is not inconceivable, particularly in a regional risk-off environment. A 10% RSD depreciation would reduce Group earnings by c 5%.

It is also worth noting that, although NLB's non-Slovenian markets have experienced relatively elevated inflation, I have not applied any adjustment for the FX losses, given that the relevant local currencies have historically maintained stability against the Euro. Should this dynamic change, however, the 8% bottom-line growth assumption may need to be revisited to account for potential FX depreciation.

11.4 EU accession stagnation

The investment thesis implicitly assumes SEE economies continue converging toward Western European norms. However, Serbia’s EU accession is currently effectively frozen, and Bosnia remains deeply problematic politically. A stalled accession timeline could hinder the general development of the region, in turn the the financial deepening tailwind could weaken materially.

11.5 Acquisition Execution Risk

NLB has proven it can execute acquisitions - Komercijalna Banka (Serbia), N Banka (Slovenia), Summit Leasing. But each acquisition carries integration risk. A failed integration or an acquisition at a bad price would damage management’s credibility and capital efficiency. Importantly, on standalone basis, the latest Addiko bid is optically expensive on the earnings basis, albeit not so on price-to-book basis, meaning that strong synergies - either via cost savings or via aggressive (Croatian) market expansion - will be needed to justify the earnings premium. However, I would note again that clearly Addiko acquisition is being pursued for Croatian market access purposes, which is inaccessible organically. As mentioned in the main body, I put my full trust behind the management on this one.

11.6 Technology and Digitalization Risk

As a mid-sized regional bank, NLB does not have the R&D budget of a Goldman Sachs or JPMorgan - or Unicredit & Revolut - to build cutting-edge banking technology from scratch. It relies primarily on vendor solutions and partnerships. The risk is that fintech disruption accelerates faster than NLB can respond, particularly in retail banking (payments, consumer credit, investment) where digital-native challengers like Revolut are growing fast across Europe. Management’s EUR 200m+ IT investment budget through 2030 is substantial for NLB’s size but may not be sufficient if the pace of disruption accelerates.

As a defence to this, the EUR 50bn total assets target by 2030 was set by management itself, with the explicit view that this represents a floor below which a bank struggles to remain competitive given today's prevailing platform and regulatory cost structures. Scale will therefore only grow in strategic importance for banks, as the development costs of digital infrastructure and compliance need to be spread across an ever-larger customer base to remain economically viable

1. Executive Summary

Zavarovalnica Triglav is a Slovenia based insurance company, with a leading market presence in the Adria region. After navigating the “perfect storm” in 2023 - comprising the worst floods in Slovenian history and a sudden regulatory cap on health insurance - the company has emerged with record profitability, a leaner portfolio, and is currently undergoing a structural pivot from a domestic incumbent into a diversified international financial group. To this end, the management has set itself ambitious goals for 2030, and focused on international expansion, with the aim of almost doubling the pre-tax profit over the course the next 4 years.

Investment Pillars:

• Dominant market position: Triglav holds a 35% market share in Slovenia and a 21% share across the Adria region, providing a defensive, cash-generative core

• Post-crisis earnings inflection: The 2023 earnings dip to EUR 21.1m in EBT was driven by exceptional external shocks, namely the CAT (catastrophe losses). While dealing with disasters is a normal part of the insurance business, events of this scale are uncommon. With those shocks behind us, earnings are now returning to their normal, stronger levels

• Organic growth runway: Balkan markets’ insurance density (premium per capita) is roughly EUR 450 compared to EUR 2,000+ in Western Europe (e.g., Austria). Considering also the faster overall GDP growths of these markets, the company enjoys a clear runway of continued long-term organic growth, even without considering international expansion

• Capital strength & dividend visibility: The company is currently offering 4% dividend (EUR 64m paid out in 2025), with the management’s strategic plan to distribute a combined EUR 400m between 2025 and 2030 (28% of current market cap), in line with payout target of minimum 50% of net income. Combined with a strong solvency ratio of 219% and credit rating “A+,” investors can be expected to collect strong dividend yield over the next 5 years

On the back of secular tailwinds - specifically the region’s ongoing economic catch-up with Western European markets - Triglav currently trades on the lower end of my fair value estimate, hence the stock can be considered a Buy at current price levels.

2. Company Profile & Regional Dominance

Founded in 1900, Triglav has evolved from a state-supported insurer into a listed multinational present in Slovenia, Croatia, Bosnia & Herzegovina, Serbia, Montenegro, and North Macedonia.

Circa 45% of Triglav’s total revenue comes from Slovenia, with the rest of Adria countries adding only 15%. The remaining 40% constitutes of the 2 current growth areas: reinsurance and other markets (namely Italy) - which are expected to also be the biggest growth driver going forward, in turn decreasing reliance on the domestic market & region.

While being a composite insurer, non-life is currently Triglav’s by far most important segment, representing 84% of total gross written premiums (“GWP”). With the core of the company’s revenue stack being on annual-renewals, this minimizes the risk of company entering into long tailed risk policies. Please see below breakdown:

Triglav keeps a leading 35% market share position both in Slovenia, as well as the Adria region (SI, CRO, MN, SR, BiH, MCD), where is holds a 21% market share - a 9% lead to the runner up - Generali. See table below for Adria region market share breakdowns between the key players:

Furthermore, the company’s market shares in the Adria region historically have been relatively stable:

3. Financials

Over the past 10 years, the company has consistently grown its GWP by high single digits. This growth was primarily driven by the region’s fast economic development and the resulting expansion of insurance markets, as Triglav’s market share throughout the period remained stable. With the exception of 2023, Triglav also displayed consistently profitable combined ratios of 93.5% on average. The EBT over the period has also doubled, from EUR 80m annually in 2010s to EUR 160m in 2024-25. More on the 2022-2023 revenue and EBT blip in the next sections.

The Triglav Group maintains a strong capital position of above 200%+, staying within the management’s target range of 200-250% also during 2023 floods. The Group’s credit rating was recently upgraded by S&P Global Ratings from A to A+.

3.1 Operational resilience: the 2023 “perfect storm”

In 2023 and 2024, the Slovenian insurance market faced a “perfect storm” of historic floods and a government overhaul of the health insurance system. These external shocks caused a dramatic, though temporary, dip in metrics and unusual distortions in Triglav’s results. While impacts were partly mitigated by reinsurance, and Triglav proved its resilience by breaking even despite these pressures, the events exposed Triglav’s significant geographic concentration risk.

3.2 2023 catastrophic floods

In August 2023, Slovenia was struck with severe floods - the most catastrophic natural disaster in Slovenia’s modern history, with total national damage estimated at EUR 7bn (c 10% of GDP).

The Triglav Group, being the dominant insurer in the country, recorded EUR 212m in gross CAT claims for the year, almost 10x the average CAT claims from previous years. However, Triglav’s reinsurance protection absorbed a large share of these disaster costs, with the net negative effect on its pre-tax profit from all the 2023 CAT estimated at only about EUR 40-50m. Still, together with health insurance reform (chapter below) this resulted in the company’s EBT plummeting to EUR 21m in 2023, well below their initial plan of over EUR 100m in profits, but imporantly still remianing in the green. The combined ratio (claims + expenses vs premiums) for non-life and health rose to 101.9% in 2023, meaning the company was technically paying out more in costs than it was taking in from those premiums.

However, Triglav’s financials have fully rebounded since. Higher reinsurance recoveries and adjusted premium pricing allowed them to post a EUR 159m EBT in 2024, with 2025 target being at EUR 160m.

Thus, Investors can take some comfort in management’s proven ability to re-price policies without losing market share following disasters; however, climate change continues to increase the risk of high-frequency, high-severity CAT events characterized by “power law” distributions. Ultimately, while the business model has survived this stress test, at the moment the company remains heavily (over) exposed to the volatility of a single market.

3.3 Abolition of supplemental health insurance in Slovenia

On April 14, 2023, the Slovenian Government adopted a decree that immediately limited the maximum monthly premium for supplemental health insurance to EUR 35.7. At the same time premium prices were frozen, medical costs and claims were surging due to high inflation. The premium amount fixed by the government was insufficient to cover these actual costs. This created a situation where the claims ratio exceeded 100%. For the full year 2023, the Health segment’s earnings before tax (EBT) dropped to EUR -30m, which was only partially off-set by one-off compensation payment by the state of EUR 11m, paid out in 2024.

Nevertheless, effective January 1, 2024, Slovenia abolished the “Voluntary Supplemental Health Insurance” (VHI) system. Previously, private insurers (Triglav, Vzajemna, and Generali) sold VHI to cover co-payments in the public health system. The government replaced this with a mandatory flat-rate health contribution (EUR 35/month) managed by the state-run Health Insurance Institute (ZZZS).

Because health insurance accounted for a large portion of the market, this move caused a negative EBT contribution of EUR 30m for 2023. Following the segments’ discontinuation, this also resulted in ~90% drop in health GWP for 2024 across the private sector. As a result, Triglav’s consolidated gross written premiums fell roughly 11% in 2024 (to EUR 2.7bn), even though other segments like property and motor grew strongly, which is the primary reason for the negative growth rate in 2024. On the positive side, the elimination of supplemental health insurance also removed a source of underwriting losses (as it had become unprofitable under the price cap).

Triglav has since largely offset this loss of volume by expanding their non-life and international reinsurance segments. They are also shifting focus to “complementary” health products (e.g., specialists and fast-track services) that sit outside the public system.

3.4 2022 net income loss

The negative net income of EUR -7m for 2022 was a result of the restatement of financial data following the Group’s transition to new international accounting standards (IFRS 17 and IFRS 9, with profit recognition being delayed under IFRS 17) in 2023. While Zavarovalnica Triglav originally reported a net profit of EUR 110m for 2022 under the previous standards, the retroactive application of the new rules converted this into a technical loss. As such, this was a one-off accounting policy driven event event.

4. Management team, ownership & listing

Triglav is listed on Ljubljana stock exchange (“LJSE”), being considered a rather liquid stock for LJSE standards - with a total of EUR 80m in trading volume throughout 2024, representing 9.6% of LJSE’s total (c EUR 400k per day).

As is typical in most large caps on LJSE, Republic of Slovenia maintains a significant stake of 63% through two sovereign funds (ZPIZ and SDH). The remaining 37.4% free float is held by a mix of international institutional investors and local retail shareholders. While providing stability, the Slovenian government’s high state’s means that the management occasionally has to balance “social responsibility” with “profit maximization.” This can result in slower decision-making and suboptimal capital allocation.

The average tenure of the management team is over 12 years, which is exceptionally high for a listed financial institution. CEO Andrej Slapar, who has led the company since 2013, has been the primary architect of the Group’s modern strategy, steering the organization through the Eurozone crisis, the pandemic, and the 2023 floods. This continuity has fostered a highly predictable corporate culture focused on technical discipline. Management maintains a strict target to keep the Combined Ratio below 95% and has consistently delivered results between 91% and 94%. This track record suggests a sophisticated ability to price risk and a refusal to chase market share through unprofitable policies, a discipline that was particularly evident in the Group’s rapid recovery following the extreme catastrophe events of 2023 - and which will be vital as Triglav pushes into new & foreign markets.

5. International expansion & management plans

As of 2026, the Triglav is in the early stages of its most ambitious strategic period to date. With the launch of its 2025-2030 Strategy, the company has shifted its vision from being the “leading regional player” to becoming a truly “international insurance and financial group.” This shift is underpinned by a goal to double its earnings before tax to between EUR 250m and 300m by 2030.

Triglav aims to increase its total business volume to over EUR 2.4bn in 2026, with a long-term goal of reaching EUR 3.0bn by 2030.

To this end, Triglav’s growth focus has moved West and North, namely via reinsurance and non-life growth. During 9M 2025, Triglav Re (reinsurance) saw a 50% YoY increase in premium volume (reaching EUR 279m), while the growth in direct international non-life insurance was even higher at 388% (reaching EUR 412m). This surge was primarily fueled by the Group’s entry into the Italian motor insurance market through its partnership with the insurtech Prima and Ageas Re, which alone generated EUR 308m in new premiums. Together, these two shifted the Group’s revenue structure so significantly that Slovenia’s share of total business volume dropped from 57% to 43% (67% in 2021) in just a single year - a “fall” caused by the rapid expansion of the international denominator rather than a loss of domestic competitive standing. Despite this aggresive expansion, the combined ratio for 9M 2025 remained low at 91.3%, indicating a healthy market entry strategy - however with this too-short of a time period to draw and confident conclusions.

Furhtermore, while the combined ratio of Western European reinsurance business can be expected between 89-94%, which is below Triglav’s 93.5% historical trend, the opposite is true for the Italian Auto insurance market, which can be expected to deliver 97-101% combined ratios. Thus, Triglav will need to carefully manage its further expansion and profitability of the Italian market.

5.1 Organic growth run-way

Besides the international expansion, the key pillar of Triglav’s investment thesis is the significant structural growth opportunity in the Balkans. Currently, insurance penetration - premiums as a percentage of GDP - ranges between 1% and 5% across the Adriatic region, far trailing the 6% to 11% typical of Western European markets. This disparity is even more pronounced in insurance density; while annual premiums per capita in Austria exceed EUR 2,000, the average across Southeastern Europe remains approximately EUR 450 (being as low as EUR 130-160 in Macedonia and Bosnia). This underdeveloped landscape reflects lower historical insurance awareness and lower disposable incomes, but it fundamentally provides a long-term runway for expansion as these markets mature and converge with broader European standards. Similarly, as these countries develop and become richer, Triglav stands to benefit also via asset management arm.

With its established presence and market share, Triglav is uniquely positioned to capitalize on this secular trend as regional economies develop and consumer uptake increases. Beyond mere penetration gains, Triglav’s target markets are projected to see weighted average GDP growth of approximately 2.7%, significantly outstripping the 1-2% expected in Western Europe. Adding to this an insurance convergence estimate - assuming Adria regions insurance penetration converges with Western Europe over 30 years (for the next 5 years) - this adds an incremental 3% p.a. to the top line. Together, this brings a “floor” organic revenue growth of 6% for the Adria region. Assuming the international & reinsurance, following explosive growth in 2025, flatlines but still remains at somewhat elevated 3% p.a., the combined topline growth would stand at 5% p.a.

6. Valuation

Considering aforementioned 5% annual growth, the insurance premiums collected and GWP by 2030 would stand at EUR 2.0bn and EUR 3.0bn, respectively. The historical average 93.5% combined ratio is assumed. While in the first 9M of 2025, no deterioration of ratios could be noticed, hence no adjustment is made to the historical combined ratios, as noted previously previously the impact of Italian expansion shall be closely monitored going forward. Nevertheless, this is partially considered by our relatively low further growth assumption in the market of 3% p.a.

With the above assumptions, the EBT from non-life insurance would stand at EUR 128m. Assuming life grows at the same 5% p.a., if shall contribute EUR 32m in EBT by 2030, up from EUR 25m currently. In line with the premium growth, the investment result is assumed to also increase from the current run-rate of EUR 45m p.a. to EUR 57m p.a., while the non-insurance result increases from the current run-rate of EUR 10m to EUR 16m, in line with the management’s forecast of AUM growth. Together, this gives us a combined total 2030 run-rate EBT of EUR 232m, and a net income of EUR 181m.

My estimates align with the top end of management’s plan for total business volume at EUR 3bn. However, they fall short of management’s pre-tax profit (EBT) guidance of EUR 250-300m. This suggests management is significantly more optimistic about the profitability of other segments, such as life insurance, asset management, and their own investment results.

As Triglav’s life insurance portfolio matures and expands, there is indeed room for it to contribute more to profits. Conversely, I have excluded any potential turnaround or contribution from the health segment, as the company is currently restructuring it. Finally, as awareness of the importance of personal savings grows across Europe, the asset management arm could see accelerated growth.

However, these optimistic projections must be weighed against the risks of rapid international expansion. This pace of growth could force Triglav into aggressive pricing to gain market share, which would hurt its profit margins (the combined ratio). Furthermore, the frequency and severity of catastrophe (CAT) events are expected to increase due to climate change. Despite that those can technically be fully reflected by re-pricing the policies, and the company has demonstrated its strong risk management during the 2023 floods, the threat to profits - and the risk of a single “knock-out” event - remains.

Below my NPV fair value valuation matrix separately for market cap & share price, assuming 10% COE and various 2030 EBT and PE ratios (dividends between 2026-2030 assumed as per my model for all scenarios). Note that the current market cap of EUR 1.46bn (EUR 63.6/share) is assumed as an entry point:

All-in all, in my view, Triglav’s current implied fair value stands between EUR 1.4-2.0bn or EUR 62-87/share. This is assuming a 10-15x PE and my 5% base case topline growth scenario. With the current market cap of EUR 1.45bn, the stock is currently trading within that range, hence can be considered as a Buy.

In addition, were one to believe the management’s guidance of EUR 250-300m EBT by 2030, Triglav remains a compelling buy, with an implied fair value of EUR 1.5-2.5bn or EUR 66-109/share at a PE of 10-15x, respectively.

On the flip side, were one to assume that the company’s growth and profitability fully stalls due to margin compression, and thus EBT would remain at the current EUR 150m run-rate, the implied valuation would be EUR 1.0bn or EUR 45/share. I would note that this was also my purchase price for the stock, however I still see it as a buy at the current levels - but with the downside risk now less covered.

1.1 Executive Summary

Cetis d.d., headquartered in Celje, Slovenia, is a European small-cap industrial company in security printing and identity management business. The company operates in a high-barrier niche defined by deep government relationships and technologically complex production. My investment thesis is built on the following pillars:

• Contracted existing revenue base: A core Slovenian business with a ~EUR 20m run-rate (virtual monopoly) supplemented by ~EUR 40m in contracted international revenues

• Foreign expansion track record: since pivoting toward international markets in 2021, foreign revenues have grown 2.5x, proving Cetis can compete and win tenders globally at scale

• The “New Factory” catalyst: existing facilities are currently running at full capacity, creating an optical stagnation in recent revenue. The completion of a new EUR 40m automated facility this year serves as a primary catalyst, designed to double production capacity while drastically improving efficiency

• High-margin digital pivot: the industry shift from physical documents to integrated phy-gital identity solutions (biometric databases, eGovernment platforms) expands high-margin recurring service revenues and increases customer stickiness

• Hidden subsidiary value: the wider Group generates ~EUR 120m in revenue. With the parent trading at ~13x P/E based only on the core security business, the adjacent packaging, commercial printing, and digital subsidiaries are essentially “free” additions to the current valuation

The stagnant or slightly declining revenues over the past three years are, in my estimation, entirely capacity-driven. The ongoing EUR 40m capital commitment to the “New Factory” - an automated manufacturing and office space - signals strong confidence from the private owners (MSIN) and the management, who also have stake in the company, in long-term demand. Beyond doubling capacity, this facility will lower marginal costs through automation. This deep-dive considers the mother entity only; the subsidiaries should be viewed as a margin of safety or a “bonus” to the investment. Thus, unless specified otherwise, all numbers will refer to the mother company’s EUR 60m run-rate security printing business only, not the EUR 120m Group’s run rate business.

The transition from manufacturing physical passports to establishing overall digital identity infrastructures presents a massive tailwind. This move into software-type solutions, which compliment the physical documents, provides pathway to even better margins, lower long-term Capex, and significantly higher switching costs compared to simple printing contracts. By expanding capacity during this critical industry shift, Cetis is well-positioned to capture market share in emerging economies as they also start digitalising their bureaucracies.

However, investors must account for the inherent risks of the Ljubljana Stock Exchange (LJSE), primarily extreme illiquidity, which is exemplified in Cetis. With a highly concentrated ownership structure and a free-float of only ~1.6%, the stock should be treated as a de-facto private equity investment. It is suited only for a disciplined buy-and-hold approach, as divesting during a downturn may be impossible; position sizing must reflect this.

Furthermore, disclosure regarding specific contracts, future strategy and the New Factory is very limited, requiring a degree of informed assumption. This deep dive presents my own interpretation of company’s direction, and should in now way be relied on for you own investment purposes.

With a market cap of EUR 140m, Cetis trades at current run-rate PE ratio of approximately 12.5x for the mother entity alone - before accounting for the New Factory’s impact or the value of the broader Group / Subsidiaries. Considering the value of Subsidiaries, the current PE stands at 9.5x, and the company can be expected to continue providing c 5% dividend yield in the following years. However, assuming the increased production following the opening and full ramp-up of the New Factory - assumed by 2029 - Cetis currently trades at forward 2029 PE of only 4.4x on a stand-alone basis, and PE of 3.3x considering also the value of Subsidiaries.

As such, the stock price is poised for re-rating once the New Factory’s increased business volume starts ramping up, and Cetis should be on path to becoming EUR 350+m market cap (2.5x from today’s valuation) company, all the while an increase in dividend payouts can be expected as the production ramps up. Thus, the stock continues to offer a compelling entry point for long-term investors, willing to trade liquidity for deep value and growing yield.

2. Company History and Business Model

2.1 Company history

Following Slovenia’s independence, Cetis was incorporated as a joint-stock company in 1990 and listed on the Ljubljana Stock Exchange in 1997. This era marked the transition to market capitalism, requiring the company to compete for contracts rather than relying solely on state allocation. The establishment of Cetis-ZG in Croatia in 1991, shortly after Croatian independence, demonstrated early foresight of foreign expansion. MSIN d.o.o., a Slovenian investment and consultancy firm, acquired the dominant stake in Cetis through a series of transactions, primarily culminating in a major takeover in early 2018. Following the ownership consolidation by the MSIN Group, the largest graphic-packaging conglomerate in Slovenia was built. This era is characterized by increased capex investments, in turn modernising the company and allowing its expansion into foreign markets, which culminated in 2022, resulting in de-facto doubling of run-rate revenues from ~EUR 30m to ~EUR 60m. A series of add-on acquisitions also provided for vertical integration (e.g., Zarja Elektronika, Sotech), transforming Cetis from a standalone printer into a diversified industrial holding providing security and printing solutions, as well as positioning the company for the upcoming industry’s transition towards digitalised identity solutions.

2.2 Product mix

The security identity division is the company’s “moat” and the mother company’s main business. Its a high barrier to entry business due to it’s trust, certification, and technical complexities. To this end, Cetis produces biometric passports, identity cards, visas, voting ballots, driving licenses, vehicle registration cards, and tachograph cards (used to monitor truck drivers’ hours).

In line with industy’s trends, the company is also moving up the value chain from a “printer” to “system integrator”, providing full turnkey service of identification solutions, including:

• Public Key Infrastructure (PKI): this is the “brain” of digital identity. It manages digital certificates and encryption keys & allows a citizen to digitally sign a document or log into a government portal securely

• AFIS (Automated Fingerprint Identification System): this software ensures “one person, one identity.” It checks biometric data (fingerprints, facial features) against a central database to prevent people from registering under multiple names

A printing contract is a one-off event. A system integration contract involves software maintenance, data management, and long-term service agreements, creating recurring revenue streams that are stickier and higher margin. To this end, one of Cetis’ primary focus areas is an establishment of eGovernment platform, which serves as a central point for all public administration services.

The shift towards services can already be seen also in Cetis’ revenue structure, with services having increased from c 5% of total revenues to c 20% today.

2.3 Foreign expansion

While Cetis is the incumbent provider for Slovenia, the domestic market is saturated - with Cetis controlling it in its entirety. The growth story is thus fully based on continued international expansion. The company has already demonstrated ability to be competitive in the international market, with the share of revenues form foreign markets increasing from 30-50% between 2014-2021 period, to c 70% today. To this end, Cetis currently serves over 30 countries across Europe, Africa, Latin America, and the Middle East, and has ongoing contracts in Honduras, Nigeria, Mexico, Ecuador, Equatorial Guinea, Burundi, Niger, Guinea-Bissau, the Maldives, Liberia, Gambia and Kosovo. The company’s core focus is on longer term contracts - with a usual duration of c 5 years - thus, a renewal risk exists & the company provides no detailed insight into contracts’ tenors and terms.

They target “mid-sized” nations where global giants like Thales or Idemia might be too expensive or inflexible. By offering modular solutions - where a country can buy just the personalization system or just the passports - Cetis competes effectively against the “all-or-nothing” bundles of larger competitors. Nevertheless, with the construction of New Factory, Cetis will have ability to take-on also larger countries / contracts, which it was unable to do so with their current, manual & labour intensive production facilities. It will also significantly expand it’s digital capabilities and reduce unit production costs via automation, allowing it be more competitive.

2.3 Industry landscape

The global market is consolidating and evolving into a “security technology” market, with the personal ID market forecasted to grow at a CAGR of 8% through 2030 (Grand View Research). Key growth drivers include increased international travel, digitalization (e.g., mobile driver’s licenses), biometrics, and rising anti-counterfeiting requirements.

While Cetis competes in a crowded global landscape, being based in a neutral European country should present a tailwind. Especially considering the USA’s accelerated alienation of emerging markets, European players should be well-positioned to capitalize on gaining share in those regions and be perceived as stable partners. In this sense, key competitors will be large, established European companies such as Thales (France), Veridos (Germany), and IDEMIA (France). While Cetis cannot compete with these firms on brand perception, scale or R&D spend, it can compete on agility and - with relatively lean operations, private & efficient ownership, (soon) state-of-the-art facilities, and production based in a still relatively lower-priced country - especially price. Furthermore, Cetis, being a smaller player, can provide customized solutions for smaller client nations that global giants might ignore or force into standard templates. Due to its small size vs the established competitors, Cetis would rarely bid alone on massive national contracts; instead, they act as the technology partner for local consortia - by Cetis providing the technology so the local partner can keep the price lower than the giants. They are the efficient choice for governments looking for a solid product without the giant corporate overhead.

It shall be noted that, while I believe a gradual shift to a “no-passport” future should benefit Cetis (by acquiring sticky e-identity contracts in emerging markets over the next few years), a sudden or “winner-takes-all” shift presents a major risk to the investment case. If biometric facial recognition at borders - using just a camera and a backend database - replaces physical booklets entirely, Cetis’s manufacturing volume will collapse. Similarly, if all worldwide governments’ security solutions will shift to a single winning provider (such as historically was the case with many other digital products), there is also a risk that Cetis will be locked-out of digital solutions. Nevertheless, given the national security implications, I see such risk as minimal.

2.4 Financial snapshot

As mentioned before, the company can be split into two eras: before 2021, the company was operating at run-rate revenue rate of EUR 30m. 2022 was the inflection point, marked as a mass documents expiry & replacement year in Slovenia, as well as by successful expansion in foreign markets, together de-facto doubling the run-rate revenues to EUR 60m. With the company currently operating at full capacity, slight reduction in revenues, as well as decreased profitability can be seen in 2024 - a trend that also continued into H1’25. With continued increase of overhead costs, this has resulted in compression of EBITDA, and the increase in depreciation is further suppressing the net income, which has declined from a peak of EUR 19m in 2022 to EUR 14m by 2024 - including the distributions from Subsidiaries, as is shown in below table (without considering Financial income / distributions from Subsidiaries, net income has decreased from EUR 16m in 2022 to EUR 11m in 2024). Given the natural cyclicality of orders under existing contracts, and the expiration of some agreements, I believe the current revenue trends are not alarming. When viewed alongside the fact that the company is operating at peak capacity, these minor fluctuations do not signal a structural decline, but rather a temporary ceiling on output.

Following the heavy capex cycle for the New Factory, the net debt has increased from a negative EUR -4m in 2022 to EUR 1m in 2024. It shall also be noted that the company has paid out EUR 11m in dividends in 2024 (when it paid out dividends for both 2023 and 2024 – as due to received Covid subsidies it was unable to pay a dividend in 2023), without which the net debt position would have actually decreased. Similarly, while the Covid raw materials price spikes and input cost pressures have since subsided, the company still notes that long delivery lead times persist, resulting in an elevated inventory and NWC position.

Nevertheless, over the past 3 years, the Company has been delivering a very high ROE and ROCE of 27% and 25%, respectively.

The operational cash-flow has averaged c EUR 16m over the last 3 year period. As the current capex cycle for the construction of New Factory concludes, the c EUR 11m of annual capex spending of the past 3 years is expected to subside to c EUR 7m p.a. into 2029, as per Company’s own estimations. Net of capex (and before net financing flows and distributions from Subsidiaries), the company thus histrorically generated c EUR 5m of cash p.a., which was de-facto fully paid out in a form of dividends. Going forward, this run-rate net-of-capex cash flow shall increase to c EUR 9m, in line with reduced annual capex needs - before accounting for any revenue growth. It shall also be noted that, on top of the above, the company had also received c EUR 2m p.a. of annual dividends from Subsidiaries for the past 3 years, which alone more than off-sets the annual financial interest costs of c EUR 600k (2025F).

Below the P&L, cash waterfall, BS and key financial indicatior. Please note that below P&L, under line item “Financial Income” includes distributions from Subsidiaries, hence it differs & is higher than purely standalone net income shown in the table in the next section.

2.4 New Factory

Factory construction progress - June 2025 (top) and December 2025 (bottom)

In 2024, Cetis commenced construction of a new 10k m2 manufacturing-office building. The ground and 1st level are to be dedicated to manufacturing, while the 2nd / top floor will be dedicated to office spaces.

The total estimated capex for the New Factory is EUR 40m. In late 2024, Cetis was also awarded a government subsidy of EUR 7.8m (payable thought the period of construction), hence the net cost to the company will be EUR 32m. Construction, while still in progress, is expected to be finalised still within 2026.

While Cetis provides very limited information with regards to the New Factory, it’s impact can be implied from the environmental impact report. It reveals that the existing grid connection capacity of the company will be expanded from the current 2 MW to 3.5 MW. Compared to the existing labour-intensive production process, requiring 47 people, the new modernised work line will require only 28 workers (in 3 shifts). This will enable the company to eliminate its current scalability bottleneck (manual labour heavy process), enabling the company to secure large-scale international contracts that were previously beyond its operational reach. The automated production process will thus bring 40% lower incremental headcount requirement - however, for the purpose of incremental analysis, we also assume the overhead expenses will increase by 20% to support the increased business volume. Further to the labour savings, the new manufacturing process will also result in 63% and 56% savings in per-piece-produced energy and materials spent, as per the environmental report.

Given the inherent incentive of the environmental report contracted by the company to exaggerate the savings, for the purposes of this analysis, the above numbers will be upward adjusted by 25%. Furthermore, while headcount will be lower, salary for the new 28 employees are assumed to be 25% higher vs existing ones, as more high skilled labour will be required for the new automated process. Overall, below the adjusted key net savings / assumptions:

• Net labour savings: 25% per unit produced ; 20% increase in overall overheads on top

• Net material-per-piece savings: 45%

  • Since Cetis doesn’t breakdown electricity costs within COGS, and they should be negligible anyway (c EUR 400k p.a.), I ignore those savings and apply only material savings coefficient to reported COGS

• New Factory’s volumes are assumed to be sold at 25% lower unit price vs old factory’s to enable to secure sufficient new business

• New Factory will operate at 81% utilisation rate - in line with Slovenian average

• Services are expected to remain at the same 2022-24 % share of revenue going forward - a conservative assumption given the historical trend and industry’s & company’s transition into turnkey identity solutions

To this end, existing facility is operating at near full capacity and is producing EUR 49m (2024A) of goods per year. With the new incremental power capacity of 1.5MW (75% increase from existing level), combined with a net energy-per-piece savings of 54%, we can derive that the new factory will be able to add incremental peak output equivalent to 162% of current output. Considering aforementioned 25% per-unit sale price haircut and 81% New Factory utilisation, the implied net incremental revenue of goods would be 99% or EUR 52m. Below the implied Post-New Factory run-rate P&L, considering also the opex savings:

*) Financing income is excluded, as it mainly relates dividend distributions from Subsidiaries

**) Incremental financing costs assume that an additional EUR 10m of debt will need to be drawn in H2’25 / H1’26 for capex purposes - since, considering a large dividend payout (EUR 10m) this year, the operational cash-flow may not be sufficient for the final phase of capex needs.

As per above, once the New Factory is fully ramped up, which would likely take up to 3 years, Cetis will transform from a EUR 60m revenue and EUR 11m net income run-rate business, into a EUR 120m revenue and EUR 32m net income run-rate business.

The critical assumption with regards to above is if the company will indeed be able to find sufficient business to fill the additional volumes. Given the company’s historical track record, and the large insider ownership, I find comfort in this assumption. The overall scale of the company, even after ramp-up, still remains small, meaning a single large scale order could fill in most of the capacity (e.g. revenues with Slovenia only, a country with a population of 2 million, stand at EUR 20m, one-sixth of total ramped-up revenue - albeit having to be noted Cetis has multiple contracts with Slovenia, for various secure identity solutions) . A potential pricing impact of trying to fill the excess capacity is also reflected in the assumed 25% unit-price discount. Furthermore, I believe the assumption of % services revenues remaining at the current 18% share of overall revenues is a conservative assumption given the historical trend and company’s strategy to further expand this area. Assuming service revenues would double, such an increase of EUR 22m would mostly flow straight straight to the bottom line.

2.5 The Group / Subsidiaries

Beyond the security core at the mother company, Cetis also acts as a holding company for a portfolio of commercial printing and security firms, which have not been considered in the above evaluation (“Subsidiaries”). While majority of the c EUR 60m of additional revenues from these Subsidiaries comes from lower-margin commercial printing businesses (EGP, Cetis-Flex and Cetis Graf), the 2 latest additions / acquisitions - Zarja Elektronika and Sotech – exhibit a strong revenue growth and high profitability. For the purposes of this SOTP analysis, a simple 10x PE valuation was assigned to all major subsidiaries, representing an implied 10% cost of equity requirement - despite this arguably being a conservative assumption for especially the rapidly growing Cetis Flex, Zarja Elektronika and Sotech.

With combined earnings of EUR 4m as of 2024, considering 10x PE and adjusting for Cetis’ ownership share, these Subsidiaries combined are worth EUR 33m - representing a 24% of the current market capitalisation.

3. Governance, ownership & dividend policy

Ownership is split over 200,000 common shares. Cetis has a very concentrated ownership structure, which, while frequent in the CEE, would be atypical of US or UK markets. Nevertheless, it maintains a good balance between “rational” and strategic private long-term capital. MSIN is in the driver’s seat with aggressive capital allocation (such as the EUR 40m New Factory investment), while the Slovenian government, as the second-largest shareholder, provides a critical “governance check” that benefits the interests and security of minority investors - as well as provides further comfort on the stability of domestic business. Finally, outsized management ownership ensures a strong alignment of interests between the leadership and the owners. The executive managers, in place since 2015-2017, have a proven track record of transforming and scaling the company from a net income of roughly EUR 1-2m at the start of their tenure to a c EUR 15m net income run-rate business today.

Both the main shareholder, MSIN, and the government have a clear thirst for distributions. Consequently, following the explosion in profitability, dividend distributions - which were averaging EUR 300k p.a. between 2017-21 period - have also exploded. Dividends were increased to EUR 5.9m in 2022, and EUR 10.5m was paid out in 2024 (partially in a form of back-dividend for 2023, due to non-payment in 2023 due to covid restrictions). A further EUR 9.6m was paid in 2025, despite the company being in the middle of a heavy capex cycle. In turn, this provides comfort regarding the illiquidity of the stock; the company shall continue to distribute any excess cash via dividends, as per the preference of both existing owners. At the current share price, this entails a 5% dividend yield (last 4y avg.). Thus, in a worst-case scenario, even if a shareholder is unable to sell their position, they will still collect dividends in the process.

4. Valuation

EUR 140m market cap is considered for January 2026, with EUR 6.5m flat annual dividend payment (last 4y avg.) until 2029. Assuming the stock re-rates to 10x on operational business, with the net income reaching EUR 32m, the implied IRR stands at 26%. Zero value continues to be assigned to any Subsidiary.

Considering both the net income and PE remain at the current EUR 15m - which includes net income of Subsidiaries - and 10x (i.e. New Factory completely fails and brings no additional business), a 6% return would be reached.

5. Conclusion

For a long-term value investor, Cetis presents a compelling ‘heads I win, tails I don’t lose much’ setup. At a suppressed 12.5x PE, you are buying into a dominant, moated business with a protected downside; even if the new factory generates zero incremental business, the core remains solid and the heavy capex is already behind us. Meanwhile, you’re getting a 5% yield - well above current bond rates - to ‘wait’ for the growth story to unfold. If the new factory and orders ramp up as expected, I see Cetis as a EUR 350-500m company in four years. It is a rare combination of market dominance and a (almost) free call option on future growth, making it a strong ‘Hold’ for existing owners and a ‘Buy’ for those who can tolerate the liquidity risk.