Cantillon Deep Value · Nordics · 11 November 2026 ★ EVO ★ Track A · Valuation

Evolution is growing again. The price still assumes it will not.

The world's largest live casino supplier turns 57 of every 100 euros of sales into operating profit and trades at about 12 times that profit. Revenue rose 3.7% last quarter, the first rise on a year earlier in five quarters. The shares are up 66% from their low, and nearly half of the analysts who cover them still say sell. This issue asks why, and whether the price is low enough for what the bears are right about.

The conclusion, before anything else
Is it cheap on adjusted, normalised figures?Members
Is the balance sheet safe enough?Members
Do the accounts survive the forensic pass?Members
Is there a catalyst with a clock?Members
Is management aligned, and will value reach minorities?Members
Is it liquid enough to enter and to leave?Members
Does it clear the return hurdle for its tier?Members
Does the thesis survive the stress test?Members
Do I hold it?Members
Current actionMembers
Nine questions, one action. The answers are members-only, because the answers are the report.
The setup · Why Evolution, why now

Evolution is the first Swedish company in this series that is not an industrial or a financial, and the first whose problem is not the business but the rules around it. For two years revenue went nowhere: about EUR 520m a quarter, give or take 3%.

Nine quarters, one number: about EUR 520m
Evolution net revenues per quarter, EUR m. The gold bar is the third quarter of 2026, the first year-on-year rise in five quarters
0200400600Q3 2024: EUR 519.4m519Q32024Q4 2024: EUR 533.8m534Q42024Q1 2025: EUR 520.9m521Q12025Q2 2025: EUR 524.3m524Q22025Q3 2025: EUR 507.1m507Q32025Q4 2025: EUR 514.2m514Q42025Q1 2026: EUR 513.0m513Q12026Q2 2026: EUR 517.8m518Q22026Q3 2026: EUR 526.1m526Q32026
Source: Evolution interim report January to September 2026, key ratios by quarter (23 Oct 2026).

In the third quarter it moved. Every region grew on the quarter before for the first time since 2024, and the group grew 3.7% on a year earlier. Four weeks earlier, Brazil had banned online betting and casinos by provisional decree, and Evolution closed its studio there on 5 October. The shares rose 2.6% on the day. The banks did not change their minds.

The price is above most of the people who cover it
SEK per share. Gold line: close on 10 November 2026. Grey dots: the latest price targets we could check
500600700800900range of the five bank targets below52-week low: SEK 51552-week low 515JPMorgan: SEK 660JPMorgan 660Berenberg: SEK 700Berenberg 700Barclays: SEK 750Barclays 750Jefferies: SEK 780Jefferies 780BNP: SEK 885BNP 88552-week high: SEK 91152-week high 911SEK 855
Sources: The Fly, MT Newswires and Business Insider reports of each bank target, 25 Sep to 6 Nov 2026; price range via Bigdata.com.

That is the disagreement this issue tests. A stock that rose from SEK 515 to 855 in a year is no longer forgotten. But the banks' targets sit mostly below the price, and their reasons are not about this quarter. They are about where the revenue comes from.

1 · The thesis Cheap + Good, risk real

The market prices Evolution as a business that will not grow again. Outside Europe and Asia, it already is.

Two regions stand still. Everything else grows fast
Change in net revenues on a year earlier, third quarter 2026, by where the players are
Africa and otherEUR 27.1m, 5% of revenue
+23.7%
Latin AmericaEUR 49.0m, 9%, Brazil still open
+23.1%
North AmericaEUR 84.3m, 16%
+13.6%
AsiaEUR 192.1m, 37%
+1.6%
EuropeEUR 173.7m, 33%
-4.7%
GroupEUR 526.1m
+3.7%
Source: Evolution interim report January to September 2026, net revenues by customers' players' IP addresses.

North America, Latin America and Africa reached record revenues in the quarter. Together they are 30% of revenue and growing between 13% and 24% a year. Europe and Asia are 70% and roughly flat. If the small regions keep growing at two thirds of their current pace and the large ones hold, the group grows about 4% a year. At about 12 times operating profit, that is not in the price. Section 5 shows how little growth the price needs.

2 · The business

Evolution does not take bets. It films them. It runs 27 studios with real dealers at roulette, blackjack and game-show tables, streams them to more than 800 online casino operators, and takes a share of what players lose plus fees for dedicated tables. Since buying NetEnt in 2020 it also sells slot games, now 16% of revenue and up 10% on a year earlier last quarter.

A table costs roughly the same to run whether 50 or 5,000 people watch it, so almost every extra euro of revenue drops to profit. That is why the adjusted EBITDA margin has stayed between 65% and 69% for the last nine quarters, with 17,000 full-time staff.

Where the players are is not where the customers are based
Share of net revenues, third quarter 2026, split two ways
By where the players are (IP address)Asia: 37%37%AsiaEurope: 33%33%EuropeN. America: 16%16%N. AmericaLatAm: 9%9%Africa: 5%By where the operator customers are basedEurope: 64%64%EuropeN. America: 15%15%N. AmericaLatAm: 15%15%LatAmAsia: 4%Asia 4%Africa: 2%
Source: Evolution Q3 2026 earnings presentation (23 Oct 2026). The company estimates 49% of revenue comes from locally regulated markets.

This chart is the whole debate in one picture. Evolution only sells to licensed operators. But 64% of its revenue comes from operators based in Europe, while only 33% comes from European players. The largest group of players, 37% of revenue, sits in Asia, where online casino is mostly not licensed locally. The company estimates that 49% of its revenue comes from locally regulated markets. The other 51% is legal for Evolution, and a long-term question for every regulator.

3 · Why it is mispriced, if it is

Not overlooked. Discounted for a risk that is real but has no date. Many investors will not own gambling, and many who do will not own a supplier whose largest player market is unregulated. That limits the buyers. Since the start of 2025, the business has also had a run of events that each looked like the start of a decline: stream theft in Asia, voluntary ring-fencing in Europe that, with lower channelisation, cut revenue there by 12% in 2025, a UK licence review, a lawsuit with a competitor, and now Brazil.

Profit flat, owners fewer
Three ways to see 2023 to 2026. Gold marks the latest figure

Net revenues, EUR m

2,063
2,067
2,071
20242025LTM

Earnings per share, EUR

5.01
5.94
5.24
5.39
202320242025LTM

Shares outstanding, m

214
209
199.2
179.8
202320242025Nov 26
Sources: Evolution annual report 2025 and Q3 2026 report; buyback notice 8 Nov 2026. 2023 and 2024 EPS and share counts from the Nordic Cantillon File; 2024 EPS includes EUR 151m of earn-out income. LTM = the twelve months to September 2026.

The company's answer has been to buy back its own shares. It dropped the dividend for 2025 and has spent EUR 1.32bn on buybacks since May, about 10% of the company, at an average of roughly SEK 747. The share count is down 16% since 2023. Profit has been flat, but each share owns more of it.

4 · The other side The bear case, argued properly
The steelman short. Evolution's margin is not only a scale story. More than half of revenue comes from players in countries without a local licence system, through operators licensed elsewhere. That revenue is worth less than it looks, because any year a regulator can do what Brazil just did, or what the UK did in July when it settled with Evolution for GBP 4.75m over supply to unlicensed operators. Europe, the regulated core, fell 4.7% last quarter and the CEO himself said "we should do better". Asia, the largest region, is flat and the stream theft is "not solved". Headcount is up 8% against revenue up 4%, so the margin is under quiet pressure. The buyback that lifted earnings per share ends by April 2027, and there is no dividend behind it. At 12 times profit you are paying a fair price for a business whose best years were 2021 to 2024.

Who is selling, and are they stupid? The sell-side mostly says Sell or Hold: JPMorgan Underweight at SEK 660, Barclays Sell at 750, Berenberg and Jefferies Hold at 700 and 780. Jefferies applies a 100% premium to the stock's asset beta for the Playtech lawsuit alone. Inside the company, the chairman's wife sold SEK 87m of shares at about SEK 830 in August. We looked for insider buying in 2026. Our search found none. The largest owner, Kenneth Dart's investment company, held 23.4% at the end of 2025.

The variant perception
The market believes the unregulated half of revenue is a slow leak that ends growth for good. I believe it is a risk that arrives market by market, as Brazil did, and that the regions growing 13 to 24% a year now outweigh a flat Europe and Asia. The gap closes if the year-end report on 4 February shows group revenue still growing on a year earlier, with Brazil out of the numbers. If it does not, the market was right.
5 · The stress test Assume it already went wrong

The thesis holds, but it rests on a part of the business no one can see clearly. Section 4 argues against Evolution. This section argues against us: if a long bought at SEK 855 is down 20% a year from now, at about SEK 684, what is the most likely way that happened? Two banks already sit near that level, JPMorgan at 660 and Berenberg at 700.

Three ways a long loses 20% within a year
The premortem: assume the trade already failed, then find the most likely reasons
STORY 1
Asia slips back
Probability: Medium

Asian players are 37% of revenue and have grown 1.6% in a year. The stream theft that cut Asian revenue in 2025 is, in the CEO's words, not solved.

Earliest sign: Asia revenue below EUR 185m in the Q4 report on 4 February.

STORY 2
The rules close in on the other 51%
Probability: Medium

Only 49% of revenue comes from locally regulated markets. Brazil shut overnight. The UK fined Evolution in July for supplying unlicensed operators.

Earliest sign: A second large market ring-fences, or a regulator acts on the unregulated share.

STORY 3
The biggest buyer leaves
Probability: Low to medium

The company has bought about 80,000 shares a day. EUR 675m is left, due by the April 2027 AGM. After that, no dividend and no buyer of last resort.

Earliest sign: No new buyback or dividend plan with the Q4 report.

Probability bars: one = low, two = medium, three = high; one and a half = low to medium. Sources listed at the end of the issue.

Ruled out: a balance sheet story. Evolution has no bank debt, a EUR 300m credit line and generates more than EUR 1bn of cash a year. If Evolution loses 20%, it will not be because of its debt.

The load-bearing assumption

Of everything the thesis assumes, one breaks it on its own: the revenue from unregulated markets shrinks slowly, as regulation arrives, not suddenly by enforcement. A slow shift towards regulated markets is something Evolution has lived with for a decade. A sudden closure of Asian or Latin American revenue is not in any of our worlds above the bear case.

Regulated share is rising slowly. Asia is flat
Per quarter, Q3 2025 to Q3 2026, by where the players are. Gold marks Q3 2026

Regulated share, %

46
47
48
49
49
Q3Q4Q1Q2Q3

Asia revenue, EUR m

189
194
198
190
192
Q3Q4Q1Q2Q3

Europe revenue, EUR m

182
178
167
173
174
Q3Q4Q1Q2Q3
Source: Evolution interim report January to September 2026, net revenues by customers' players' IP addresses.

What the price assumes. Our price assumption tool says SEK 855 needs owner earnings to grow about 0.7% a year for ten years, at a 9% required return. That is close to no growth at all. The stock is priced for the flat years to continue, not for decline.

What the data says. The slow path is visible: the regulated share rose from 46% to 49% in a year, and North America and Latin America, mostly regulated, are where the growth is. The sudden path is visible too: Brazil went from a regulated market to a banned one in a day, and the UK settlement shows the regulated core is watching the rest. Today the evidence leans slow, with one sudden event already in the quarter.

Second-order effects · Who else the buyback and the bans touch
WhoWhat changes, and how it comes back to Evolution
Kenneth DartHis 47.9m shares were 23.4% at the end of 2025. The buyback has lifted that to 26.6% and will take it to about 28% when the programme ends. Sweden's mandatory bid rule sits at 30% of the votes. That caps how long buybacks can run at this pace unless he sells, and it makes him the holder whose next move matters most.
Operators licensed elsewhereThey carry most of the unregulated revenue. Every new local licence regime asks them to choose: pay local tax or leave. When they leave, Evolution loses the revenue until local operators replace it.
Pragmatic Play and PlaytechTwo of the live casino challengers. A Brazil ban hits all suppliers equally. Ring-fencing and slow licensing in new markets favour suppliers that already hold many licences, as Evolution does.
Brazilian playersWell below half of Latin American revenue, by the company's estimate. Brazil is not ring-fenced, so some play may move to operators licensed elsewhere. That would turn regulated revenue into unregulated revenue, the wrong way.
If you only remember one fact. 37% of Evolution's revenue comes from Asian players, but only 4% from operators based in Asia. The size of that gap is the size of the risk, and the reason the stock trades at 12 times profit.
The Cantillon File

Evolution, by the numbers

The numbers behind the business, from its own annual reports. The full File updates with every report.

Figures to31 December 2025
  1. Where the money comes fromEvolution had revenue of EUR 2.1bn in 2025. That is 9% more a year since 2023.
  2. Where the money goesOf every 100 in sales, 59 became operating profit in 2025. Over the period it has ranged from 59 to 64.
  3. Who gets itEarnings per share went from EUR 5.01 in 2023 to EUR 5.24 in 2025. Dividends took 50% of owner earnings over the last 2 years.
  4. What protects itReturn on capital employed was 30% after tax in 2025. Net cash of EUR 727m.
  5. What the price assumesToday's price needs owner earnings to grow 1% a year for ten years. Revenue has grown 9% a year.
Every figure from the company's own filings.Read the full Evolution File →
For members

The score, the numbers, the timing and the size

Members unlock the full Evolution issue: the nine answers and the action, the Cantillon Value Score pillar by pillar, the verification log against the Q3 filing, the cash bridge that turns net cash into net debt, the three worlds with targets and expected IRR, the price at which it clears the hurdle, the catalyst calendar, all twelve flags, the IVT read on the weekly and daily with the line in the sand and the flip level, the stress-test exit rules, the checkpoints and the tear sheet.

Or take the free read first.

Every Friday the complete daily brief ships open, with all levels and every terminal capture.

Includes the IVT Regime Playbook (PDF), the nine regime states and how to trade each.
What happens next

Publishing an idea starts the coverage, it does not end it. Evolution goes on the Deep Value watchlist. The Brazil vote in Congress gets a short note when it lands, because it is the first test of the load-bearing assumption. The year-end report on 4 February 2027 gets an earnings note within 72 hours and a full re-score, and the premortem is re-run against it. Each checkpoint gets an update on its date whether or not anything happened, and a timing alert goes out within 24 hours if the chart gate flips. The Cantillon File chapter above is rebuilt from the 2026 annual report when it is published.

What the words mean
Live casino real dealers at real tables, streamed to players online. RNG random number generator games, mainly slots. Operator the online casino that holds the player relationship and buys games from Evolution. Player IP the internet address a player connects from, used to estimate where revenue comes from. Regulated market a country with its own licence system for online casino. Ring-fencing blocking players from a market unless the operator holds that market's licence. Earn-out part of an acquisition price paid only if targets are met; when targets are missed, the liability is released as income. Premortem assume the investment already lost money, then work out the most likely way it happened, before deciding anything. Load-bearing assumption the one assumption whose failure breaks the thesis on its own. Adjusted net debt borrowings and leases, less free cash, plus anything else already spoken for. EV (enterprise value) market value plus adjusted net debt, what you pay for the whole business. EV/EBIT enterprise value divided by normalised operating profit. Owner earnings operating cash flow less investment and lease payments. ROIC return on invested capital. Expected IRR the probability-weighted yearly return across the three worlds. Timing gate whether the chart says buy now, start partial, or wait.
Cantillon Deep Value is research and education, not investment advice. Nothing here is a recommendation to buy or sell any security. Thesis-critical figures are checked against the company's filings before publication and the verification log is in section 6; figures that could not be checked are labelled as estimates. Disclosure: the author holds no position in Evolution at publication. No trading takes place in a covered name within three trading days either side of publication. No compensation is received from any covered company. Evolution reports in euros and its shares trade in Swedish kronor. Gambling is regulated or prohibited in many countries; this issue analyses a listed supplier and does not promote gambling. Do your own diligence.