Frontier Developments plc (FDEV.L)

Deep Dive Research Report

Listing: London Stock Exchange, AIM (ticker FDEV, ISIN GB00BBT32N39) Sector: Communication Services / Interactive Home Entertainment Fiscal year end: 31 May. Reports half-yearly. Most recent reporting period used in this report: FY26 (year ended 31 May 2026), announced 9 September 2026; FY26 Annual Report and Notice of AGM published 23 September 2026. Report date: 25 September 2026

1. What the company does

Frontier Developments builds and sells a very particular kind of video game: a management simulation in which the player is handed a blank plot of land, a set of tools, and a set of living things or machines with needs, and is then left to construct and run something. A theme park. A zoo. A dinosaur park. The genre has existed since RollerCoaster Tycoon and Theme Hospital in the 1990s. Frontier's contribution has been to industrialise it, brand it, and turn each attempt into a franchise that keeps selling for a decade.

The company calls this genre "creative management simulation", or CMS, and has spent the last three years deliberately shrinking itself back down to it. In FY26, CMS games produced 89% of group revenue, up from 77% the year before and 62% two years before that (FY26 results, 9 September 2026).

How the money actually works

The commercial mechanic is not "sell a game". It is "launch a game and then keep selling to the same people for years". A Frontier CMS title launches at a standard premium price, and then receives a long sequence of paid downloadable content packs (PDLC) sitting alongside free feature updates. The paid packs are mostly content: new animals, new rides, new scenery sets, new dinosaur species. The free updates keep the community active so there is somebody left to sell the packs to.

The economics of that are visible in the split of the Planet Zoo franchise's lifetime revenue, which management disclosed at the FY26 results as roughly 55% base game and 45% PDLC (FY26 results presentation, September 2026). Nearly half of everything a successful Frontier franchise earns arrives after the launch press cycle has ended. Chief Executive Jo Cooke put the whole strategy in one sentence on the FY26 analyst call:

"The greatest value in our business comes from creating games and franchises that players engage with for many years, not just at launch." (FY26 earnings call, 9 September 2026)

The corollary is that a back catalogue is not a residual. It is the business. In the first half of FY24, before the recovery, 72% of all revenue came from games already released in prior years (H1 FY24 results, 17 January 2024).

Founding, and the two pivots that matter

Frontier Developments was founded by David Braben, born 1964, who studied Natural Sciences with a specialisation in Electrical Science at Jesus College, Cambridge, and who co-created the 1984 space trading game Elite with Ian Bell while both were Cambridge undergraduates (Wikipedia, David Braben). The company was incorporated on 28 January 1994 and has always been headquartered in Cambridge (Wikipedia, Frontier Developments). Braben was awarded an OBE in 2014 and a BAFTA Academy Fellowship in 2015, and co-founded the Raspberry Pi Foundation in 2011.

For its first twenty years Frontier was a work-for-hire studio. It built RollerCoaster Tycoon 3 for Atari in 2004, Kinectimals and Zoo Tycoon for Microsoft, Thrillville for LucasArts, Screamride for Microsoft again. The craft was there. The economics were not: a work-for-hire studio owns nothing it makes.

Pivot one was self-publishing. Frontier listed on AIM in July 2013 and used the capital, alongside a 2012 Kickstarter, to fund Elite: Dangerous, which it released itself in December 2014. Planet Coaster followed in 2016, Jurassic World Evolution in 2018, Planet Zoo in 2019. Frontier stopped renting out its capability and started compounding it.

Pivot two was the strategic reset of 2023 and 2024, and it is the one that explains the company as it exists today. Between 2019 and 2023 Frontier had widened aggressively in two directions at once. It opened a third-party publishing label, Frontier Foundry, which released Lemnis Gate, FAR: Changing Tides, Stranded: Alien Dawn, Deliver Us Mars and The Great War: Western Front. And it pushed into genres it did not own, most conspicuously the F1 Manager series under an exclusive Formula 1 licence signed in March 2020, and Warhammer Age of Sigmar: Realms of Ruin, a real-time strategy game, in 2023.

Both directions failed. Frontier shut Frontier Foundry on 14 June 2023, stating that most of the titles released under it had failed to become profitable within a year (Wikipedia, Frontier Developments). Realms of Ruin underperformed so badly at launch that it was fully impaired, a £16.9m write-off recorded in the first half of FY24 (H1 FY24 results, 17 January 2024). F1 Manager revenue fell 34% year on year in the same half. In October 2023 the company announced an organisational review; headcount came down to about 700 and annual operating costs were cut by roughly 20% (FY24 results, 11 September 2024). The F1 licence itself was voluntarily terminated before the next game entered full development, disclosed in the FY24 accounts as commercial discussions with an IP partner resulting in the voluntary termination of a contract for a future game.

Then-CEO Jonny Watts framed the half honestly at the time:

"2023 was a challenging year for Frontier and for many companies across the games industry." (H1 FY24 results, 17 January 2024)

What emerged from the reset was a company with one genre, one engine, two studios, and a published commitment to ship one CMS game a year. That is the company being analysed here.

A concrete walk-through: what Frontier actually delivered in FY26

On 21 October 2025 Frontier released Jurassic World Evolution 3 on PC, PlayStation 5 and Xbox Series X|S. The player runs a chain of dinosaur parks across ten locations, with 99 prehistoric species that now have sex and age variants, juveniles, and genetic inheritance between generations. Jeff Goldblum reprises Dr Ian Malcolm in the campaign. Critics scored it 78 to 82 on Metacritic depending on platform, with 90% of OpenCritic reviewers recommending it (Wikipedia, Jurassic World Evolution 3).

It sold more than 500,000 base-game units inside two weeks. Then the second phase began. Frontier shipped the Wetlands Pack, the Rebirth Expansion tied to Universal's 2025 film Jurassic World Rebirth, the Crocodilia Coast Pack and the Prehistoric East Pack. The Rebirth Expansion, released in June 2026, became the best-performing piece of paid content in the franchise's history (FY26 results presentation). The Jurassic franchise, which had been 27% of revenue in FY25, became 58% of revenue in FY26.

That is the entire model in one year: a big launch, then twelve months of content sold into an installed base that has already spent forty hours building something it does not want to abandon.

2. Business segments

Statutory position first: Frontier reports as a single operating segment, the development and publishing of video games. There is no segmental profit disclosure. The company also does not disclose a geographic revenue split, on the stated grounds that the majority of revenue flows through third-party distribution platforms which hold the consumer sales data, and that building the information internally would cost more than it is worth.

That said, the business runs as four distinct franchise families with genuinely different economics, different risk profiles and different owners of the underlying intellectual property. The sub-sections below treat them as segments because that is how capital is allocated inside the company and how management talks about the roadmap.

2.1 Own-IP creative management simulation: the Planet franchises

Revenue share FY26: Planet Zoo 17%, Planet Coaster 14%. Combined 31%.

What it does. Planet Coaster (2016) is a theme-park builder. Planet Zoo (2019) is a zoo builder with a strong animal-welfare and conservation framing. Both are sold on PC and on PlayStation and Xbox consoles. Both run on a long tail of paid content packs. Planet Coaster 2 launched 6 November 2024 and Planet Zoo 2 launches 13 October 2026.

The core capability. Three things took years and are hard to copy. First, the free-form construction toolset: Frontier's piece-by-piece building system, which lets a player place individual roof tiles and fence panels and then save and share the result, is the reason its games generate enormous volumes of user-created content. Second, creature and crowd simulation: Planet Zoo animals have individual needs, welfare states and genetics, and Planet Zoo 2 adds an emotion system and dynamic behaviour driven by habitat satisfaction (Planet Zoo 2 announcement, 28 May 2026). Third, the Cobra engine, discussed in Section 3, which lets a small company ship the same simulation on PC and two console families without licensing Unreal or Unity.

Why it exists separately. These are the only major franchises Frontier owns outright. That matters commercially, not just emotionally: there is no royalty. Management guided that FY27 gross margin should recover by roughly a point precisely because Planet Zoo 2, an own-IP title, replaces the royalty-bearing Jurassic World Evolution 3 at the top of the mix (FY26 earnings call, 9 September 2026). Own IP is also permanent. A licence expires.

Competitive position. Within park and zoo management specifically, Frontier has no listed competitor of comparable scale. Its real rivals are SEGA's Two Point Studios in the adjacent comedic-management space, Paradox in city building, Bandai Namco's Park Beyond, and a long tail of small independents such as Texel Raptor's Parkitect and Blue Meridian's Prehistoric Kingdom. Frontier wins on production values, animal and crowd simulation fidelity and content cadence. It loses, periodically, on polish at launch: Planet Coaster 2 shipped to a Steam user rating of 57% against the original's 91%, with sustained criticism of the interface, bugs and a 30fps cap on current-generation consoles (Metacritic user reviews; GamesRadar review).

How it fits the group. This is the compounding core and the margin engine. Planet Zoo reached one million monthly active users in July 2026, seven years after launch (FY26 results presentation). Management disclosed that the Planet Zoo franchise has generated a cumulative cash profit of roughly the same order as the far older Jurassic franchise, at a materially higher return on invested development cost. A third, unannounced Planet-branded CMS title is scheduled for FY28.

2.2 Licensed-IP creative management simulation: Jurassic World Evolution, and now Disney

Revenue share FY26: 58%, up from 27% in FY25.

What it does. Jurassic World Evolution (2018), Jurassic World Evolution 2 (2021) and Jurassic World Evolution 3 (2025) are dinosaur-park management games built under licence from Universal Products and Experiences. Across those three base games the franchise has sold 13.0 million units, making it Frontier's largest by a wide margin (FY26 results presentation).

On 3 September 2026 Frontier announced an agreement with Disney to develop and publish a new CMS game using intellectual property from Disney's portfolio. The game is fully funded by Frontier. No title, property, release window or platform set has been disclosed (Frontier announcement, 3 September 2026; Stevivor).

The core capability. The thing Frontier sells to a licensor is not code. It is the ability to take an entertainment property with a mass, family-skewed audience and turn it into a construction toy that respects the property's canon. That requires the CMS toolset, but it also requires a decade of demonstrated ability to pass a major studio's brand approvals and ship on time. The Jurassic relationship has run since 2018 through three base games and dozens of content packs, and Frontier has shipped content synchronised with Universal's own film releases, most visibly the Rebirth Expansion alongside Jurassic World Rebirth.

Why it exists separately. Licensed IP buys audience reach that own IP cannot. The Jurassic franchise is the single largest thing Frontier has ever built, and it exists because a dinosaur park is both a globally recognised idea and a natural management game. The trade is margin and control. Royalties are payable to the licensor, which is one of the two reasons FY26 gross margin fell by around three percentage points against FY25 (the other being a higher share of physical disc distribution at the JWE3 launch). And control sits partly with the licensor: the May 2026 trading update disclosed that "delays beyond Frontier's control impacted the release of further content" for JWE3 (Trading update, 12 May 2026).

Competitive position. For a licensor, the set of studios that can credibly deliver a AAA-adjacent management simulation on a fixed schedule is small. Frontier's realistic peers for this kind of mandate are SEGA's Two Point Studios, Paradox, and the larger publishers' internal teams. The Disney deal is the clearest external validation the company has: Disney chose a 727-person AIM-listed studio over anyone else in the genre, and Frontier is funding it, which means Frontier keeps the upside.

How it fits the group. This is the growth bet and the scale lever. It is also, as of FY26, the single biggest concentration risk in the business, for the obvious reason that one licensed franchise produced well over half of group revenue.

2.3 Elite Dangerous: the live-service outlier

Revenue share FY26: approximately 7%.

What it does. Elite Dangerous, released December 2014, is a persistent multiplayer space simulation set in a 1:1 scale model of the Milky Way. It is not a management game. It is the direct descendant of the 1984 game its founder wrote, and Frontier acquired all rights to the Elite name in 2008 (Wikipedia, Frontier Developments).

The core capability. The procedurally generated galaxy, the flight model, and eleven years of accumulated systems. Nothing else in Frontier's portfolio looks like it, and nothing in the market looks quite like it either.

Why it exists separately. Partly history, partly economics. Elite is a live service with an engaged, long-tenured player base rather than a launch-and-content-pack product. Frontier tried to retire it from active development during the difficult years and then reversed course when the numbers improved. In FY25, Elite Dangerous revenue grew 76% (FY25 trading update, 11 June 2025). In H1 FY26 it grew more than 50% again, driven by new content and new ships (H1 FY26 results, 19 January 2026). The Trailblazers update, which let players colonise their own star systems, was the specific catalyst; SteamDB average concurrent players roughly doubled over two years, and the game sold about 300,000 copies over the last financial year (Massively Overpowered, 12 June 2026).

How it fits the group. A small, cheap, high-loyalty annuity that management has committed to keep feeding. It is not where the capital goes, and it is not where the growth is, but it costs little and it keeps producing.

2.4 Complex Games and the Warhammer 40,000 tactics line

Revenue share FY26: inside the 4% "other" line.

What it does. Complex Games, based in Winnipeg, Canada, develops turn-based tactics games under licence from Games Workshop. It made Warhammer 40,000: Chaos Gate - Daemonhunters, which Frontier published in 2022 and which performed well enough that Frontier bought the studio outright on 2 November 2022 (Wikipedia, Frontier Developments). The direct sequel, Warhammer 40,000: Chaos Gate - Deathwatch, was revealed at Warhammer Skulls in May 2026 for PC, PlayStation 5 and Xbox Series X|S, with an Xbox launch stated for winter 2026 (Xbox Wire, 22 May 2026).

Why it exists separately. It is a different genre, a different country, a different licensor, and it arrived through acquisition. Frontier kept it when it was cutting everything else, which is a statement about how the acquisition has performed relative to the Foundry label and Realms of Ruin.

How it fits the group. A small, contained second line that fills the FY27 release slate alongside Planet Zoo 2. Management guided a contribution of £5m to £10m for Deathwatch in FY27 (FY26 earnings call, 9 September 2026). It is a supporting act, not a pillar.

Segment comparison

3. Products and business detail

The Cobra engine, and why an AIM-listed company writes its own

Frontier's games run on Cobra, a proprietary cross-platform engine and toolset the company has been developing and evolving since 1988. The current generation is the fourth. Cobra presents a platform-neutral core API and resource pipeline, so code and assets authored on PC compile and run on Xbox, PlayStation and mobile targets, and it is designed around multi-processor, multi-threaded execution. The fourth generation added cloud-based analytics for data-driven live operation (Frontier technology page; MCV/Develop interview).

The strategic argument for this is specific rather than generic. A creative management simulation is an unusual technical problem: thousands of simultaneously simulated agents, a construction system that must accept arbitrary player-authored geometry, and a camera that can go from a whole-park view to a single animal's face. Commercial engines are optimised for other things. Because Frontier owns its engine team, it adds the features it needs when it needs them rather than waiting for a third-party roadmap. The cost is that Frontier carries a permanent engine-engineering overhead that a Unity or Unreal licensee does not, and the risk is that the gap between an in-house engine and the commercial engines widens on things like rendering and console certification tooling.

The full product catalogue

Live and revenue-generating

• Jurassic World Evolution 3 (21 October 2025, PC / PS5 / Xbox Series X|S). Dinosaur park management under the Universal licence. 99 species with sex and age variants, juveniles with genetic inheritance, terrain tooling, a ten-location campaign, sandbox with generated islands. Over 500,000 base-game units inside two weeks. Content released to date: Wetlands Pack, Rebirth Expansion (tied to the film Jurassic World Rebirth), Crocodilia Coast Pack, Prehistoric East Pack.
• Jurassic World Evolution 2 (2021) and Jurassic World Evolution (2018). Both still selling. The franchise sustained 96% of its prior-year revenue in FY24 and declined only 4% in FY25, a year with no new base game at all (FY25 trading update, 11 June 2025).
• Planet Zoo (2019). Zoo construction and animal welfare simulation. Reached one million monthly active users in July 2026. Roughly 45% of its lifetime revenue has come from paid content.
• Planet Coaster 2 (6 November 2024, PC / PS5 / Xbox Series X|S). Theme park builder with water-park gameplay added. Passed 400,000 base-game units within two months (H1 FY25 results, 15 January 2025). Post-launch content has included the Sorcery Pack and Toybox Pack.
• Planet Coaster (2016) and its Console Edition. Still generating meaningful revenue nine years after release; a Steam "Mega Sale" in May 2024 brought over one million new players into the game in a single promotion (FY24 results, 11 September 2024).
• Elite Dangerous (December 2014) with the Horizons and Odyssey expansions and the Trailblazers colonisation update.
• Warhammer 40,000: Chaos Gate - Daemonhunters (2022, Complex Games), plus console ports released February 2024.

Announced and in development

• Planet Zoo 2. Launches 13 October 2026 on PC, PS5 and Xbox Series X|S, with physical discs a week later on 20 October. Standard edition £39.99 / $49.99 / €49.99; deluxe £54.99 / $64.99 / €64.99 with six extra species. New systems: aviaries and aquariums for the first time in the franchise, Wildlife Reserves for releasing animals outside the zoo boundary, and an emotion system driving dynamic animal behaviour (Planet Zoo 2 announcement, 28 May 2026).
• Warhammer 40,000: Chaos Gate - Deathwatch. Winter 2026 on Xbox Series X|S, also PS5 and PC. Up to nine playable classes, seven enemy factions.
• An unannounced own-IP Planet-branded CMS title, scheduled FY28.
• The Disney CMS title, FY29 or later.

Retired, sold or written off

• RollerCoaster Tycoon 3 (2004). Publishing rights sold back to Atari in March 2024. Frontier had sued Atari in January 2017 over unpaid royalties.
• Zoo Tycoon (2013), Screamride (2015), Kinectimals (2010-11), LostWinds (2008-09), Thrillville. Work-for-hire era.
• F1 Manager 2022 / 2023 / 2024. Licence voluntarily terminated before the FY26 instalment entered full development. The series is finished (Traxion; Gamereactor).
• Warhammer Age of Sigmar: Realms of Ruin (2023). Fully impaired in H1 FY24.
• Frontier Foundry catalogue: Lemnis Gate, FAR: Changing Tides, Stranded: Alien Dawn (publishing rights sold in FY25), Deliver Us Mars, The Great War: Western Front. Label closed 14 June 2023.

Delivery, not manufacturing

Frontier is a digital-first business. The overwhelming majority of units are sold as downloads through third-party storefronts. Physical discs exist for the console editions of the largest launches, and the step-up in physical distribution around Jurassic World Evolution 3 was explicitly cited as one of two reasons FY26 gross margin fell against FY25.

The production process is a single-studio pipeline: design, engine and tools, art, animation, audio, QA and live operations all sit in Cambridge, with the Warhammer tactics line in Winnipeg. Development costs are capitalised during production and amortised against the game's revenue after release, with a steeper profile in the first year. The capitalisation rate is a disclosed operating lever: it ran at 70-80% historically, fell to 59% of gross R&D spend in H1 FY24 when F1 Manager development was expensed as incurred, and management guided 60-65% for FY27 (H1 FY24 results; FY26 earnings call). A lower capitalisation rate means more cost hits the income statement immediately and less is deferred, which makes reported profit more conservative.

Geographies

Frontier sells globally through the same handful of storefronts and does not disclose a geographic revenue split, because the platforms hold the consumer data and management judges the cost of reconstructing it internally to be excessive. Physical distribution partners handle disc retail. Localisation is standard across the major European and Asian languages.

Two studios: Cambridge Science Park in the UK, which is the headquarters and holds the great majority of the 727 staff recorded as at March 2026 (Revelio Labs), and Complex Games in Winnipeg, Canada. A North American office in Halifax, Nova Scotia, opened in August 2012 and closed in January 2015.

Milestones that changed the business

• 1994: incorporation.
• 2004: RollerCoaster Tycoon 3 establishes Frontier in park management, but as a contractor.
• July 2013: AIM listing. The capital that made self-publishing possible.
• December 2014: Elite Dangerous, the first self-published title.
• 2016 and 2019: Planet Coaster and Planet Zoo create the own-IP CMS spine.
• 2018: Jurassic World Evolution, the first major licensed CMS.
• July 2017: Tencent takes a 9% stake for £17.7m (MCV).
• November 2022: Complex Games acquired.
• June 2023 to 2024: Foundry closed, organisational review, ~20% cost reduction, F1 licence terminated.
• July 2025 and February 2026: the first two share buyback programmes in company history.
• September 2026: Disney agreement signed; first dividend in company history declared.

4. Customers

Frontier has two layers of customer and one layer of gatekeeper, and confusing them is the fastest way to misread the business.

Layer one: the platform storefronts

Frontier's invoiced counterparties are Valve (Steam), Sony (PlayStation Store), Microsoft (Xbox store and Game Pass) and, for some titles, Epic. These are the entities that collect money from players, take a cut, and remit the balance. The industry-standard storefront cut is 30%.

The buying decision here is not really a decision. There is no procurement committee; a premium PC game that meets technical certification gets listed. What the platforms genuinely control is discovery and promotion: front-page placement, seasonal sale inclusion, bundle positioning. Frontier's May 2024 Planet Coaster "Mega Sale" on Steam brought more than a million new players into an eight-year-old game, which is a demonstration of how much economic value sits in a promotional slot the company does not own.

Subscription deals are a second, discrete platform relationship. Frontier disclosed Game Pass receipts as a component of its cash inflow in December 2023 (H1 FY24 results). These are negotiated, lump-sum arrangements, and they are lumpy.

Layer two: the players

The end customer is a person who wants to build something. Frontier's audience skews toward creative, systems-minded players, often family-adjacent given the dinosaur and zoo subject matter, and heavily PC-led with a growing console cohort.

Why they buy Frontier specifically, in order of what the evidence supports:

1. Depth of the construction toolset. Frontier's games let a player build at a granularity almost nothing else offers. For the segment of the audience that treats these games as a creative medium rather than a puzzle, this is decisive.
2. Simulation fidelity of the living things. Planet Zoo animals have welfare states, genetics and, in the sequel, emotions. This is the differentiator against a pure builder.
3. The licence. A material share of Jurassic World Evolution buyers arrive because of the film franchise, not because they were looking for a management game.
4. Content cadence. A player choosing between two builders will pick the one still receiving updates.

Switching costs

There is no contractual lock-in, no qualification testing, no regulatory approval. What exists instead is accumulated creative investment. A player who has spent 200 hours designing a zoo, bought nine content packs, and published blueprints to the Steam Workshop has built something they cannot export to a competitor. That is a real switching cost and it is why the back catalogue sustains: in FY24, with no new CMS release, Planet Coaster held 96% of its prior-year revenue.

But the switching cost protects the installed base, not the next sale. It does nothing to guarantee that the same player buys the sequel. Planet Coaster 2's mixed launch reception is the demonstration: loyalty to the franchise did not stop a large fraction of buyers from posting negative reviews.

Concentration

Concentration is severe and it is at the franchise level, not the customer level.

• One franchise, Jurassic World Evolution, was 58% of FY26 revenue.
• Three franchises were 89% of FY26 revenue.
• One title, Planet Zoo 2, is management's stated primary revenue driver for FY27, with guidance of £25m to £30m minimum.

On the platform side, concentration is also high: Steam is the dominant PC storefront and Sony and Microsoft are duopolists in console, so three commercial relationships intermediate essentially all revenue.

This is best read as an inherent property of premium single-title game publishing rather than as a company-specific failing. But it does mean that in any given year the business is a bet on one launch, and the residual risk is not diversifiable by anything management can do inside a twelve-month window.

Contract structures and revenue predictability

• Storefront sales: spot, no commitment, revenue recognised on sale. Unpredictable at the unit level, reasonably predictable in aggregate for a mature back catalogue.
• Subscription and bundle deals: negotiated lump sums, lumpy.
• Licence agreements (Universal, Games Workshop, Disney): royalty-bearing, with approval rights held by the licensor over content and schedule. The May 2026 disclosure that content delays were "beyond Frontier's control" is the clearest public evidence of how binding those approval rights can be.
• UK tax credits: the Video Games Expenditure Credit is, in practice, a large and recurring cash inflow. FY26 tax credits rose 88% year on year and management guided £8m to £9m for FY27.

The net picture: roughly half of Frontier's revenue in a steady year comes from a back catalogue that behaves like an annuity with a slow decay, and the other half comes from one launch whose outcome is genuinely uncertain until it happens.

5. Competitive landscape

Where Frontier actually sits

Frontier does not compete with the games industry. It competes for a specific slice of PC and console spend from players who want to build and manage. That slice is small relative to shooters and sports, and it has an unusual structure: the biggest publishers largely ignore it, and the specialists who serve it are mostly either private, sub-scale, or owned by a Japanese conglomerate.

The consequence is that Frontier is one of the largest pure-play companies in its own niche, while being a rounding error in the industry as a whole. That is the asymmetry to hold onto.

The named competitors

Two Point Studios, owned by SEGA. The closest analogue in spirit. Two Point Hospital (2018), Two Point Campus (2022) and Two Point Museum (March 2025) are comedic management sims descended directly from Bullfrog's Theme Hospital. SEGA acquired the studio in May 2019 (Wikipedia, Two Point Studios). Two Point Museum reviewed well and reached Nintendo Switch 2 in October 2025. Where Frontier wins: simulation depth, construction freedom and production scale. Where it loses: Two Point ships on Nintendo hardware, which Frontier's CMS titles do not, and it has SEGA's marketing behind it.

Paradox Interactive. The dominant listed name in strategy and city building, with Cities: Skylines, Cities: Skylines II, Prison Architect, Crusader Kings and Stellaris. Paradox is genuinely adjacent rather than overlapping: city building draws from the same player, but Paradox's core competence is deep systemic strategy, not free-form construction. Paradox has had its own difficulties, notably with Cities: Skylines II, which was handed from Colossal Order to Iceflake Studios at the start of 2026 (Wikipedia, Cities: Skylines II).

Microsoft / Xbox Game Studios. Owns the historical Zoo Tycoon brand (Frontier built the 2013 version for them) and runs Age of Empires and Microsoft Flight Simulator. A latent rather than active competitor in park and zoo management, but it could enter at will and it controls one of the three storefronts Frontier sells through.

Bandai Namco. Published Park Beyond, developed by Limbic Entertainment, a direct theme-park competitor that did not establish itself against Planet Coaster.

Kalypso Media (private, Germany). The Tropico series. Overlaps on management and city building at a lower production tier.

everplay group (formerly Team17). UK-listed indie publisher, renamed in February 2025. Not a CMS specialist, but the nearest listed UK comparator for a games business of similar shape and the obvious reference point for how the UK market prices this kind of company.

tinyBuild and Devolver Digital. The other two AIM games listings. Both are cautionary tales rather than competitors: tinyBuild's market capitalisation has fallen roughly 92% from its 2021 IPO level, and Devolver Digital cancelled its AIM listing on 16 September 2026 after a 96% share price decline (TechTimes).

Independent specialists. Texel Raptor (Parkitect), Blue Meridian (Prehistoric Kingdom). Small, well-regarded, no capacity to fund a Disney-scale licence.

Barriers to entry

The honest answer is that they are moderate, not high, and they are not where an outsider would guess.

What is not a barrier: capital. A competent twenty-person team can ship a park builder on Steam, and several have.

What is a genuine barrier:

1. Licensor trust. Universal has renewed with Frontier across three base games and eight years. Disney chose Frontier in 2026. This is the single hardest thing on the list to replicate, because it is earned only by shipping on schedule through a major studio's approval process repeatedly.
2. Simulation and tooling depth accumulated over a decade. Planet Zoo's animal systems and Frontier's piece-by-piece construction toolset are not weekend projects. They sit on top of an engine started in 1988.
3. Installed base and Workshop content. Seven years of community-created content is an asset that accrues to the incumbent and compounds.
4. Content cadence economics. Sustaining a paid-content pipeline for years after launch requires a team and a cost base that a one-game studio cannot carry.

What is not protected at all: the genre concept itself, and the next player's first purchase. Frontier's moat protects its existing franchises. It does not automatically protect the next launch, which is exactly what Planet Coaster 2's launch reception showed.

Structural shifts

Three worth naming. First, consolidation at the top of the industry has continued, with EA taken private in August 2026 by a consortium led by the Saudi Public Investment Fund at around $55bn enterprise value (Variety). This removes a listed comparator and concentrates ownership of large IP. Second, the small-cap listed games sector in London has largely broken, with Devolver delisting and tinyBuild down roughly 92% from IPO, which makes Frontier's position as a profitable, cash-generative, dividend-paying AIM games company genuinely unusual. Third, entertainment licensors are increasingly willing to hand properties to genre specialists rather than internal studios, which is the trend the Disney deal sits on.

Where Frontier is strong, and where it is exposed

Strong: genre leadership in park and zoo simulation, a licensor relationship record that competitors cannot buy, an owned engine tuned to the specific technical problem, a back catalogue with demonstrated multi-year decay rates, and a balance sheet with no debt.

Exposed: a single franchise at well over half of revenue, dependence on licensors for both margin and schedule, a one-release-per-year cadence that leaves no slack if a title slips, and a demonstrated ability to disappoint its own community on a sequel launch.

6. Industry

What drives demand

Demand for Frontier's products is driven by four things, in rough order of importance.

Installed hardware base and platform health. PC is the company's primary platform, and PC gaming has been the fastest-growing platform segment. PC grew 12.0% in 2025 to $43.6bn, Newzoo's strongest recorded growth rate for the platform, and is forecast at $45.9bn in 2026, up 5.3% (PC Gamer, citing Newzoo; Gaming and Media).

Licensed-property cultural cycles. Jurassic World Evolution 3 launched in October 2025, three months after Universal released Jurassic World Rebirth in July 2025, and Frontier then shipped a Rebirth Expansion tied to the film. Film release schedules are a genuine demand driver for the licensed part of the portfolio and are outside Frontier's control.

The creative-builder audience. A structurally growing cohort, adjacent to the Minecraft and Roblox generation, that treats games as a medium for making things and shares the results on social video. This is the audience that makes a Frontier game's marketing partly free.

Promotional and discount cycles. Deep seasonal discounting on Steam and console stores is how back catalogues recruit new cohorts. It also compresses realised price per unit.

Size and growth

The global games market is forecast at roughly $205bn in software revenue for 2026, up about 4.6% from $188.8bn in 2025 on Newzoo's methodology, with Midia Research putting the figure nearer $236.9bn on a hardware-inclusive basis (Udonis industry report; Gaming and Media). Mobile is the largest platform at roughly $107bn, console around $48bn, PC around $46bn.

Frontier addresses PC and console only, so its addressable market is the roughly $94bn PC-plus-console pool, and within that a management and simulation slice that is a single-digit percentage share. Newzoo's PC genre data puts shooters at $7.4bn (17% of PC segment revenue) and adventure at $5.8bn (13%); simulation and management sit meaningfully below both. Frontier is not playing in a large pond. It is playing in a small, defensible one.

Position in the value chain

Frontier occupies the developer-publisher position: it funds development, owns or licenses the IP, and takes publishing risk, but does not own distribution. Upstream of it sit the IP licensors (Universal, Games Workshop, Disney) who take a royalty. Downstream sit the storefronts (Valve, Sony, Microsoft) who take a cut of the gross. Frontier keeps the middle. That is a structurally better position than pure work-for-hire, which is what it used to be, and a structurally worse position than owning a storefront, which it will never do.

Regulation and government policy

The material regulatory item for Frontier is UK creative-sector tax relief, and it is currently in transition.

The Video Games Expenditure Credit (VGEC) replaces the older Video Games Tax Relief (VGTR). VGEC carries a headline rate of 34%, taxable at the main corporation tax rate, giving an effective net benefit of about 25.5% of qualifying expenditure, slightly above VGTR's effective 25%. VGEC became mandatory for all new games from 1 April 2025; games already in development before that date may continue to claim VGTR until April 2027. The critical change is scope: VGTR allowed claims on European Economic Area expenditure, while VGEC rewards UK expenditure only. Subcontractor spend caps have been removed (BDO; Alliotts).

This matters to Frontier more than to most. FY26 tax credits rose 88% year on year, benefiting from the VGEC transition, and management guided £8m to £9m of credits for FY27 (FY26 results presentation). That is a large number relative to the company's operating profit. It is also policy, and policy can change.

Beyond tax, the relevant regulatory surface is age rating (PEGI, ESRB), consumer protection rules around in-game purchases, and data protection. None of these is currently a binding constraint on Frontier's model, which sells premium games and content packs rather than loot boxes or gacha mechanics.

Cyclicality

Premium PC and console games are mildly counter-cyclical at the point of purchase, in the sense that a £40 game that delivers hundreds of hours is a cheap entertainment substitute in a downturn. The bigger cycle for Frontier is not economic but internal: the release cycle. Revenue in any year is dominated by whether a major title shipped and how it landed. FY24 had no significant CMS launch and the year was poor. FY26 had Jurassic World Evolution 3 and the year was a record. This is the dominant rhythm of the business and it is not a macro phenomenon.

A secondary cycle is the console generation. Frontier's CMS titles now ship on PS5 and Xbox Series X|S only, having previously supported the prior generation, which lifts what is technically possible but shrinks the addressable console base until hardware penetration builds.

Tailwinds and headwinds

Tailwinds: PC is the fastest-growing platform; creative-builder content performs disproportionately well on video platforms, reducing effective user acquisition cost; UK tax relief on games is structurally supportive; licensors are increasingly outsourcing to genre specialists; and consolidation has removed several mid-tier competitors.

Headwinds: development costs per title continue to rise industry-wide; discovery on Steam is harder every year as release volume grows; the VGEC change narrows qualifying expenditure to the UK only; and the console install base for current-generation-only titles is still building.

7. Growth triggers

Drawn from the six most recent reporting periods and their accompanying analyst presentations and calls.

• One new creative management simulation game per year, on average, as a standing cadence commitment. First set out as a three-game roadmap for FY25, FY26 and FY27 at the H1 FY24 results (17 January 2024), and restated as a permanent operating cadence at FY26 (9 September 2026). Repeated across all six periods.

  "The strongest and most stable pipeline in Frontier's history." (FY26 earnings call, 9 September 2026)

• Planet Zoo 2 launches 13 October 2026 on PC, PlayStation 5 and Xbox Series X|S, with physical discs on 20 October, introducing aviaries, aquariums, Wildlife Reserves and an animal emotion system. Management described pre-orders as "very, very positive" and guided it as the primary FY27 revenue driver (FY26 results, 9 September 2026). First flagged as a Planet Zoo sequel at H1 FY26 (19 January 2026), dated at the FY26 trading update.

  "Very, very positive." (Management on Planet Zoo 2 pre-orders, FY26 earnings call, 9 September 2026)

• Warhammer 40,000: Chaos Gate - Deathwatch ships in FY27, developed by Complex Games, launching winter 2026 on Xbox Series X|S with PS5 and PC versions. Revealed at Warhammer Skulls in May 2026 (FY26 results, 9 September 2026). First referenced as an unannounced Complex Games title for FY27 at H1 FY26 (19 January 2026).

• A new own-IP Planet-branded creative management simulation is scheduled for FY28. (H1 FY26, 19 January 2026, as an unannounced FY28 CMS title; named as a new Planet franchise entry at FY26, 9 September 2026.) Repeated.

• The Disney agreement signed 3 September 2026, under which Frontier will develop and publish a new CMS title using Disney intellectual property, fully funded by Frontier, targeted for FY29 or later (FY26 results, 9 September 2026). Management noted Disney offers shared marketing opportunities across its assets and platforms, though marketing support is not contractual.

  "A great program for shared marketing opportunities." (Jo Cooke on the Disney relationship, FY26 earnings call, 9 September 2026)

• Continued content support for Jurassic World Evolution 3, with the content roadmap resuming after third-party delays. Following the Rebirth Expansion becoming the best-performing paid content in franchise history in June 2026, management committed to further releases (FY26 results, 9 September 2026; delay disclosed at the trading update of 12 May 2026).

• Further new content for Elite Dangerous, following revenue growth of over 50% in H1 FY26 driven by new ships and the colonisation systems (H1 FY26, 19 January 2026; reaffirmed at FY26, 9 September 2026). Repeated.

• Gross margin recovery in FY27 as the mix shifts back to own IP. Management guided that margin should recover as Planet Zoo 2 replaces the royalty-bearing Jurassic World Evolution 3 at the top of the revenue mix (FY26 earnings call, 9 September 2026).

• Continued share buybacks, with scope flagged beyond the two programmes completed in FY26, subject to a stated 10% share count limit per period (FY26 results, 9 September 2026).

  "We really do believe we are significantly undervalued and it makes sense therefore to use that surplus cash to do those buybacks." (CFO Alex Bevis, FY26 earnings call, 9 September 2026)

Trigger summary

8. Key risks

1. Single-franchise concentration at an extreme level

Mechanism. In FY26, the Jurassic World Evolution franchise produced 58% of group revenue, up from 27% the prior year. Three franchises produced 89%. In FY27, management has guided that a single title, Planet Zoo 2, is the principal revenue driver. If Planet Zoo 2 lands badly, there is nothing else in the FY27 slate large enough to absorb the gap: Warhammer 40,000: Chaos Gate - Deathwatch was guided at a contribution roughly one-fifth the size.

Calibration. High probability of some variance around the outcome, moderate-to-severe consequence when it goes wrong. This is not a tail risk; it is the shape of the business. The mitigating factor is that a mature back catalogue now provides a substantial base, and the FY24 experience shows what a bad year looks like: revenue fell but the company survived and cut costs.

2. Licensor dependency on both margin and schedule

Mechanism. Two of Frontier's three largest future opportunities sit on someone else's IP. A licensed title pays royalties, which is one of the two identified causes of FY26's gross margin decline. More importantly, the licensor holds approval rights over content and timing. Frontier disclosed at its May 2026 trading update that "delays beyond Frontier's control impacted the release of further content" for Jurassic World Evolution 3 and that it was working to return to its planned content roadmap (Trading update, 12 May 2026).

That is the mechanism in its mildest form: a delayed content pack. The severe form is non-renewal. Licences expire. Frontier has not disclosed the term of the Universal agreement, and it lost the Formula 1 licence entirely in 2024, albeit by mutual agreement after poor performance.

Calibration. Moderate probability of recurring schedule friction, low-to-moderate probability of a licence loss in any given period, high consequence if the Jurassic licence were lost given its 58% revenue share.

3. Sequel execution risk on own IP

Mechanism. Frontier's model requires the sequel to inherit the original's audience. Planet Coaster 2 tested that and the result was uncomfortable: a Steam user rating of 57% against the original's 91%, with criticism of the interface, bugs and a 30fps cap on current-generation consoles (Metacritic). The franchise still grew revenue sharply in FY25 because the launch itself sold units. But a sequel that alienates its community shortens the paid-content tail, which is where nearly half the lifetime revenue sits.

Planet Zoo 2 is the immediate live test of whether Frontier learned from this. Planet Zoo is a more beloved product than Planet Coaster was, which raises both the opportunity and the downside.

Calibration. Moderate probability, moderate consequence. It is not company-threatening, but it directly attacks the compounding mechanism that makes the business model work.

4. A one-release-per-year cadence with no slack

Mechanism. Management has committed to an average of one CMS game per year, and the pipeline is fully specified: Planet Zoo 2 in FY27, a new Planet title in FY28, the Disney game in FY29 or later. There is no bench. A slip of one quarter moves a launch across a fiscal year boundary and creates a year with no major release, which is precisely what FY24 looked like. The company's own history contains a fully impaired title and a terminated licence, so slippage and cancellation are not hypothetical.

Calibration. Moderate probability over a multi-year horizon, high consequence in the affected year.

5. Reliance on UK tax credits for a large share of reported profit

Mechanism. FY26 tax credits rose 88% to a level that is very large relative to adjusted operating profit, and management guided £8m to £9m for FY27. The VGEC regime that generates them changed on 1 April 2025, narrowing qualifying expenditure from EEA-wide to UK-only and running transitional VGTR claims out to April 2027. Any further tightening of the regime, or a shift of development work outside the UK, would reduce this inflow directly.

Note also that the company changed its headline adjusted measure from Adjusted EBITDA to Adjusted Operating Profit from FY26, explicitly so that the new measure includes video game tax credits (FY25 trading update, 11 June 2025). A reader comparing FY26's record adjusted profit to pre-FY26 figures should understand that the definition moved, and that it moved in a direction that flatters the comparison.

Calibration. Low probability of near-term policy reversal, moderate-to-high consequence if it happened.

6. Key-person transition, concentrated in a twelve-month window

Mechanism. Frontier has changed its chief executive, its chief marketing officer and the status of its founder inside one year. Jonny Watts stepped down as CEO on 1 January 2026 after 27 years with the company and left the board on 31 May 2026. Jo Cooke, who had rejoined as Chief Marketing Officer in August 2025, became CEO. Dan Lazarides became CMO on 1 August 2026. And David Braben, the founder, president and largest shareholder, moves from Executive Director to Non-Executive Director on 1 October 2026 while retaining the President and Founder titles (GamesBeat; H1 FY26 results, 19 January 2026).

The transition has been handled with unusual care: Watts stayed five months past handover to support it. But the creative direction of a studio is not a fungible asset, and the founder is stepping back from executive involvement at the exact moment the company is taking on its largest-ever licensed commitment.

Calibration. Low-to-moderate probability of disruption, moderate consequence and slow to become visible.

7. The Rule 9 concert-party constraint on capital returns

Mechanism. David Braben controls 32.6% of voting rights including shares held by his spouse. Because buying back shares mechanically increases a continuing holder's percentage, further buybacks would ordinarily trigger a mandatory offer obligation for the "David Braben Concert Party" under Rule 9 of the City Code on Takeovers and Mergers. The Takeover Panel granted a waiver, approved by shareholders at a general meeting on 27 June 2025, permitting the concert party's holding to rise but not above 36.31% (Share buyback waiver circular).

Frontier's practical solution has been for Braben to give the executing broker an irrevocable instruction to sell into each buyback so that his percentage stays flat. That works, but it means every buyback is accompanied by a stream of founder-sale announcements, and it means the capacity for buybacks is bounded by a shareholder-approved headroom that must be periodically renewed.

Calibration. High probability of continued optical noise from founder sales, low probability of an actual constraint binding in the near term.

8. Platform and discounting dependence

Mechanism. Essentially all revenue flows through three storefronts that control discovery. Frontier's back catalogue is heavily dependent on promotional events it does not control: the May 2024 Planet Coaster Mega Sale brought more than a million new players into an eight-year-old game. Deep discounting also compresses realised price. If Steam changes its promotional architecture, or if a platform deprioritises the genre, the back-catalogue annuity decays faster.

Calibration. Low probability of a discrete shock, high probability of slow structural pressure.

9. Small-cap AIM listing dynamics

Mechanism. Frontier is an AIM-listed company with a founder holding roughly a third of the votes, Tencent historically holding around 9%, and institutional ownership well below half. The free float is thin. That amplifies price moves in both directions and makes the shares vulnerable to forced selling. It is also the context in which the CFO stated on the record that management believes the shares are "significantly undervalued", which is itself a statement that the market is not clearing efficiently. The broader AIM games cohort has deteriorated badly, with Devolver delisting in September 2026 and tinyBuild down roughly 92% from IPO, which affects sentiment toward the whole group regardless of Frontier's own results.

Calibration. Certain, ongoing, moderate.

9. Walk the talk

The six reporting periods used

Frontier reports half-yearly. The six most recent reporting periods, and the dates of their results announcements and accompanying analyst presentations, are:

1. H1 FY24 (six months to 30 November 2023) - announced 17 January 2024
2. FY24 (year to 31 May 2024) - announced 11 September 2024
3. H1 FY25 (six months to 30 November 2024) - announced 15 January 2025
4. FY25 (year to 31 May 2025) - announced 10 September 2025
5. H1 FY26 (six months to 30 November 2025) - announced 19 January 2026
6. FY26 (year to 31 May 2026) - announced 9 September 2026

A disclosure on source quality: full verbatim transcript-level material was obtainable for the FY26 call. For the other five periods this analysis works from the results announcements themselves, the accompanying results presentations, the interim trading updates that fall between them, and contemporaneous coverage. The management statements quoted below are drawn from those primary announcements.

The narrative

Start at the bottom, January 2024. The H1 FY24 results were the worst set of numbers in the company's listed history. Revenue was down 17%, there was a full impairment of Warhammer Age of Sigmar: Realms of Ruin, a restructuring provision, and a loss per share. Management's response was three specific commitments: a 20% reduction in annual operating costs with full effect from 1 June 2024; a refocus on creative management simulation; and three named CMS games for FY25, FY26 and FY27. They also held the FY24 revenue guidance range they had set in November.

"Refocussed on CMS games" delivering "more predictable and recurring returns." (H1 FY24 results, 17 January 2024)

That is a large number of falsifiable promises to make in a single bad announcement. Track each one.

September 2024, FY24 results. Revenue came in at the middle of the guided range, so the guidance held. The cost reduction was delivered: headcount came down to about 700 and operating costs were cut by roughly 20% as promised. The strategic reset was real, not rhetorical: Frontier sold the RollerCoaster Tycoon 3 publishing rights back to Atari, and it terminated the F1 contract before the next game entered full development rather than sinking more capital into a line it had already decided against. The forward commitment was modest and specific.

"Confident of delivering profit in FY25 as the next step to improved financial performance." (FY24 results, 11 September 2024)

CEO Jonny Watts's framing at the same results was notably unadorned: "This year has been a challenging period for Frontier, so I'd like to thank our incredible team for rising to the challenges following our strategic reset."

January 2025, H1 FY25. Delivered. Planet Coaster 2 shipped on 6 November 2024, the first of the three promised CMS titles, and it shipped in the stated window. Operating costs were down 25% year on year, ahead of the 20% commitment. The company swung from a substantial loss to a profit. Guidance was reaffirmed rather than raised.

"The Board remains confident of delivering FY25 revenue and profitability in line with expectations following the strong performance achieved in the first seven months." (H1 FY25 results, 15 January 2025)

June and September 2025, FY25. Delivered again. FY25 revenue grew, CMS revenue grew 25% year on year, and the CMS share of revenue rose from 62% to 77%. Adjusted EBITDA came in an order of magnitude above the prior year. Cash grew substantially. The second of the three promised CMS titles, Jurassic World Evolution 3, was announced on 6 June 2025 with a firm date of 21 October 2025. And management started returning capital for the first time in the company's history, launching a £10m buyback on 8 July 2025 after obtaining the Rule 9 waiver at a general meeting on 27 June 2025.

Two things at this point deserve scrutiny rather than applause. First, part of the FY25 improvement came from a £3.5m gain on selling the Stranded: Alien Dawn publishing rights, which is a one-off disposal, not operations. Management disclosed it clearly. Second, from FY26 the company changed its headline adjusted measure from Adjusted EBITDA to Adjusted Operating Profit, explicitly to include video game tax credits, on the stated grounds that this gives a "more accurate representation of the business' performance". That is a defensible argument and it was flagged a full year in advance. It is also a change that makes the following year's headline number look better than it would have under the old definition, and any comparison across the boundary needs to account for it.

The forward commitment was again deliberately vague: "further annual growth in FY26".

January 2026, H1 FY26. This is the period where management's conservatism becomes visible. Jurassic World Evolution 3 had launched on 21 October 2025 and sold over 500,000 units in two weeks. H1 revenue was up 26% and adjusted operating profit up 76%. CMS revenue was 90% of the total. December trading had been strong enough to set a company Christmas Day sales record for a single game. And with all of that in hand, management raised FY26 guidance to approximately £100m of revenue and approximately £11m of adjusted operating profit.

The same announcement carried the CEO transition. Jonny Watts's parting statement was:

"I'm incredibly proud of what Frontier has achieved in the first half of FY26 and in recent years. The success of Jurassic World Evolution 3 and the strength of our CMS portfolio are a testament to the talent and dedication of our team. I am delighted that we have successfully restored Frontier to a position of consistent, sustainable growth and profitability." (H1 FY26 results, 19 January 2026)

May, June and September 2026, FY26. Guidance was then raised twice more. On 12 May 2026 the board upgraded FY26 revenue expectations to approximately £103m. In June the trading update indicated adjusted operating profit of approximately £19.0m. The audited FY26 result, announced 9 September 2026, was revenue of £104.8m, up 16%, and adjusted operating profit of £21.4m, up 62%, with statutory operating profit up 97% to £25.0m.

Set that against the January guide of approximately £11m of adjusted operating profit. The company delivered roughly double what it had guided nine months earlier, with about two thirds of the financial year already complete at the time of guiding. Part of the gap is a genuine upside surprise in tax credits from the VGEC transition, which management flagged. Part of it is that Jurassic World Evolution 3's content tail outperformed. But a company that guides £11m and delivers £21.4m is not being precise; it is being conservative, and deliberately so.

The same results carried the first dividend in company history, a £5.0m special dividend, and confirmed £15.5m of buybacks executed during FY26 that reduced total voting rights by about 10%.

The promises that were kept

The promises that were missed or quietly dropped

Three, and they all predate or sit alongside the reset.

The F1 Manager series. Frontier signed an exclusive multi-game Formula 1 management licence in March 2020 and shipped three annual instalments. The series was then abandoned. The FY24 accounts disclosed it in language that does not use the word "cancelled": "commercial discussions with an IP partner resulted in the voluntary termination of a contract for a future game before full development started." That is an accurate description and also a euphemistic one. The honest read is that a strategic bet made in 2020 was written off in 2024.

Frontier Foundry. Launched in 2019 and rebranded in 2020 as a third-party publishing arm, closed on 14 June 2023 with the company stating most titles under it had failed to become profitable within a year. Six games, one label, four years, no lasting value beyond the Stranded: Alien Dawn rights that were later sold.

Planet Coaster 2's reception. Nobody promised a Steam score. But the strategic logic of the reset was that CMS sequels produce "more predictable and recurring returns", and Planet Coaster 2 landed at a 57% Steam user rating against the original's 91%. Management has not, in its results announcements, engaged with that gap directly. The financial outcome was fine. The community outcome was not, and the model depends on the community.

Assessment

This is management that does what it says, and that has become steadily more conservative in what it says.

The evidence for the first half is strong and it spans a full turnaround. In January 2024 this company made four specific, checkable commitments from a position of real distress: cut costs 20%, refocus on one genre, ship three named games across three years, and return to profit. Every one of those was delivered, on or ahead of schedule, and the three-game roadmap has held for nearly three years including through a change of chief executive. Companies that say what they will do while losing money usually do not deliver all of it.

The evidence for the second half is the FY26 guidance sequence. Guiding to approximately £11m of adjusted operating profit in January and delivering £21.4m in September is not a near miss in the flattering direction; it is a guide the company had good reason to know was low. Some of that was tax-credit upside outside their control. Some of it was a deliberate policy of guiding to a number that cannot embarrass them. Investors should calibrate accordingly: a Frontier guide is a floor, not an estimate, and the next guide for Planet Zoo 2 of "£25m to £30m minimum" should be read in that light, with management's own qualifier that there is "potential for higher performance".

Two qualifications keep this from being an unreserved endorsement. First, the change of headline adjusted measure from Adjusted EBITDA to Adjusted Operating Profit, made specifically to include tax credits, was announced a year in advance and is defensible, but it does flatter the FY26 headline against prior years. Second, the historical record contains two strategic bets, F1 and Foundry, that were abandoned entirely, and that history is the reason to hold the Disney deal to the same standard as everything else rather than treating it as free option value.

10. Shareholder friendliness index

Dividends. Frontier paid no dividend in FY24 and no dividend in FY25. In FY26 the board declared the first dividend in the company's history: a special dividend of 14.1p per share, totalling £5.0m, with an ex-dividend date of 17 September 2026, a record date of 18 September 2026 and payment on 9 October 2026 (FY26 results, 9 September 2026). It is explicitly a special rather than the initiation of a progressive ordinary dividend, and it follows a year in which cash rose to £44.0m at the 31 May 2026 year end and £51.4m by 31 August 2026 against a working-capital buffer requirement the CFO put at about £20m. The trend is therefore nil, nil, then a one-off distribution funded from genuine surplus rather than from borrowing, which is the right way round.

Buybacks and dilution. Two programmes ran inside the last three years, both requiring a Rule 9 waiver approved by shareholders on 27 June 2025 because repurchases would otherwise raise the David Braben concert party above the mandatory-offer threshold. Programme one was authorised at up to £10m, launched 8 July 2025 and completed by October 2025. Programme two was announced on 26 February 2026 at up to £8m and up to 1,429,327 shares, and it completed having repurchased 1.4m shares at an average of 372.51p for about £5.3m (Investing.com; LSE / Alliance News). Across FY26 the two programmes together purchased 3,947,854 shares for £15.5m, plus a further £2.0m of shares bought into the Employee Benefit Trust, a total of £17.5m of share purchases during the year. The effect was a reduction in total voting rights of approximately 10%. Nothing older than these two programmes was found: prior to July 2025, no buyback programme has been identified in Frontier's annual reports, its exchange announcements or financial news coverage, and the company had never paid a dividend, so the three-year record is nil-nil-then-both in the most recent year. Dilution is being actively suppressed rather than merely offset: the CFO's September 2026 option exercise was satisfied by transferring existing shares out of the EBT specifically so that no new shares were issued.

Verdict: Returns Capital. In a single year Frontier retired about a tenth of its share count and paid its first-ever dividend, returning over £25m in total while carrying no debt, and it did so with management stating on the record that it considered the shares undervalued.

11. Insider activities

Sources. The MoatMap cross-market disclosure database (data current 25 September 2026) is used as the spine for the last twelve months. Because the UK venue is open rather than bot-walled, that spine has been cross-checked against primary RNS "Director/PDMR Shareholding" announcements published through the London Stock Exchange, and material transactions absent from the MoatMap block have been added below with their RNS citations. Readers should note that the MoatMap block captured only the three CFO transactions and did not capture the founder's sales or the non-executive director's purchase, both of which are documented in RNS filings and are included here.

Recent transactions, most recent first

Routine sub-threshold housekeeping has been excluded. Braben also made smaller buyback-linked disposals shortly before the twelve-month window opened, including 907 shares on 23 July 2025 and 7,329 shares on 9 September 2025; the pattern is continuous from the launch of the first buyback programme on 8 July 2025.

Buys, and how to read them

There is exactly one genuine open-market purchase in the window, and it is small but not meaningless.

David Walsh, Non-Executive Director, bought 5,500 shares across three tranches on 12 and 13 November 2025, at an average of 540.55p, for approximately £29,730. Two features make this more interesting than the size suggests. First, the resulting holding disclosed in the announcement was exactly 5,500 shares, which means this was the establishment of a position from zero, not a top-up. A non-executive director choosing to become a shareholder for the first time is a different signal from one adding to an existing stake. Second, the timing: the purchase came three weeks after Jurassic World Evolution 3 launched and about two months before the January 2026 guidance upgrade, so it was made with visibility on the launch's early trajectory that the market did not fully have.

Against that, it is a single buyer, not a cluster, and the absolute sum is modest relative to a non-executive director's annual fee. No other director bought in the open market in the twelve-month window. There is no cluster-buying signal here, and no purchase by the chief executive, the chief financial officer or the chairman.

Sells, and why they happened

The volume of selling in the table is almost entirely explained by a single structural mechanism, and misreading it would be the biggest analytical error available in this section.

David Braben's sales are mechanical, not discretionary. Braben controls 32.6% of Frontier's voting rights including shares held by his spouse. Under Rule 9 of the City Code on Takeovers and Mergers, a company buying back its own shares mechanically increases the percentage held by continuing shareholders, which would have obliged the "David Braben Concert Party" to make a mandatory offer for the whole company. The Takeover Panel granted a waiver, approved by shareholders at a general meeting on 27 June 2025, capping the concert party's holding at 36.31% (buyback waiver circular, June 2025).

To keep his percentage flat rather than drifting toward that cap, Braben gave the executing broker an irrevocable instruction to sell shares into the buyback pro rata. Frontier disclosed this explicitly when launching the February 2026 programme: Braben "will sell into the buyback program to maintain his ownership percentage unchanged" (Investing.com, 26 Feb 2026). Every Braben sale in the table above is dated inside an active buyback programme window, and each RNS states the transaction was executed in line with the programme. His economic exposure to Frontier is unchanged by these sales. Read as a conviction signal, they are noise.

Alexander Bevis's September 2026 transaction is a cashless option exercise. On 10 September 2026, the day after the FY26 results, the CFO exercised 50,000 options under the Company Share Option Plan at a 250p strike and sold the same 50,000 shares on the same day at 515.51p, realising a gross gain of roughly £132,755 before tax. This is the standard mechanism by which an executive monetises a vested option without funding the strike price out of pocket, and the company satisfied the exercise from the Employee Benefit Trust so that no new shares were issued. It is compensation crystallisation, not a view on value. It is worth noting that it happened the day after a results announcement that sent the shares up sharply, which is the normal open window for such transactions, and it is worth noting that the CFO chose to retain none of the resulting shares. The June 2026 transaction of 5,084 shares at 354.0p is also option-related per the RNS headline of 24 June 2026; MoatMap classifies it as "Other", meaning neither a clean open-market buy nor sell, and no reason beyond the exercise itself is disclosed.

No other reason for selling has been disclosed in any of the filings reviewed, and none needs to be inferred: the buyback mechanism and the option plan account for every disposal in the window.

Net assessment

Frontier's insiders are, on the raw arithmetic, heavy net sellers: hundreds of thousands of shares disposed of against a single purchase of 5,500. On the economics, that arithmetic is misleading to the point of being useless. The founder's disposals exist solely to hold his percentage constant while the company retires stock, an arrangement disclosed in advance, approved by shareholders, and structurally required by the Takeover Code. Strip those out and the CFO's same-day option exercises out, and the window contains one transaction that reflects a decision about value: a non-executive director establishing a position for the first time with his own money, shortly after the company's biggest launch in years.

The activity is concentrated in two people, and neither is expressing a view. What is arguably more informative than any individual trade is the company's own behaviour: Frontier spent £15.5m buying its own shares during FY26 and the chief financial officer stated on the record at the FY26 results that "we really do believe we are significantly undervalued and it makes sense therefore to use that surplus cash to do those buybacks." Corporate buying at that scale, from a board that had never bought back a share before, is a firmer signal than the personal dealings.

Plain read: neutral, with one mildly positive data point. There is no bullish cluster, but there is also nothing here that should be read as insiders heading for the exit. The one thing that would change this assessment materially is an open-market purchase by the new chief executive, who has been in post since January 2026 and has not yet bought shares in the market.

12. Scenarios

Bull case

Planet Zoo 2 lands on 13 October 2026 the way the original did in 2019, not the way Planet Coaster 2 did in 2024. The aviaries and aquariums give returning players a reason to rebuild everything from scratch, the emotion system makes the animals feel alive in a way that generates the kind of short-form video that sells management games for free, and the launch review scores sit where the first game's did rather than where the Planet Coaster sequel's did. The pre-orders management described as "very, very positive" convert into a base that the content pipeline can feed for five years, and the paid packs land at roughly the same share of lifetime revenue the original achieved.

That outcome does something structural, not just cyclical. Frontier ends FY27 with two own-IP franchises both operating at scale and both in their sequel generation, alongside a Jurassic franchise whose third instalment has already proved it can out-earn its predecessor. The revenue mix shifts back toward own IP, which lifts gross margin, and the one-CMS-per-year cadence starts to look less like a promise and more like a factory. The new Planet-branded title arrives in FY28 into a market that now expects a Frontier release every autumn.

Meanwhile the Disney relationship turns out to be the beginning of something rather than a one-off project. Disney's IP portfolio contains several properties that map naturally onto construction and management, and a successful first title creates the option for a second. If Frontier can demonstrate that it can take any major entertainment property and turn it into a durable management franchise, then the company is no longer a studio with three good games. It is a platform that licensors come to, and the constraint on growth becomes development capacity rather than demand. In that world the cadence goes from one game a year to one game a year per line, and Cambridge stops being the only studio that matters.

Capital allocation compounds the effect. The company is already retiring roughly a tenth of its shares annually while paying a dividend and carrying no debt, and the CFO has said the surplus above a roughly £20m working-capital buffer is available for returns. Continued buybacks at a share count that is shrinking against a profit base that is growing does the arithmetic that management is explicitly betting on.

Base case

Planet Zoo 2 does what management has guided: it delivers the £25m to £30m minimum that has been signposted, with the usual launch friction and the usual six months of patches, and it establishes a content pipeline that runs through FY28. Warhammer 40,000: Chaos Gate - Deathwatch delivers its smaller contribution from the Winnipeg studio. The Jurassic back catalogue decays at the slow rate it has historically decayed at, holding most of its prior-year revenue in a year with no new base game, and Elite Dangerous continues its quiet recovery on the back of new ships and content. Gross margin recovers by about the point management indicated as own IP returns to the top of the mix.

FY28 brings the new Planet title and FY29 or later brings the Disney game. The cadence holds but does not accelerate, because Frontier has 727 people and each CMS title absorbs a large fraction of the studio. Revenue grows, but not smoothly: it steps up in launch years and consolidates in the years between, and a reader looking at any two-year window sees either a boom or a plateau depending on where the releases fell.

Capital returns continue at roughly the current shape, buybacks constrained by the Rule 9 headroom and topped up by occasional specials rather than becoming a progressive dividend. Management continues to guide conservatively and beat, which produces a pattern of upgrades that becomes expected and therefore stops moving the shares as much.

Nothing breaks. The company remains what it currently is: the largest specialist in a small, defensible genre, profitable, debt-free, structurally dependent on one launch a year, and carrying an acknowledged concentration in a licensed franchise it does not own.

Bear case

Planet Zoo 2 repeats Planet Coaster 2. It sells its launch units, because the franchise name does that on its own, and then the Steam reviews settle in the mid-50s on complaints about performance, interface and missing features that the original had. The community that made Planet Zoo reach a million monthly active users seven years after launch does not migrate, and the paid-content tail, which is where nearly half the lifetime value sits, is shorter and thinner than the model assumes. The financial damage does not show up in FY27, when the launch units still book. It shows up in FY28 and FY29, when the content packs underperform and the sustain rate on the franchise breaks its historical pattern.

Simultaneously the Jurassic concentration bites from the other direction. The franchise was 58% of FY26 revenue, and Jurassic World Evolution 3 is now in the back half of its content cycle. Universal's approval process, which already caused disclosed content delays in the spring of 2026, causes more of them. Or the licence comes up for renewal on terms that are worse, because Universal has watched Frontier build 58% of its business on a property Universal owns and prices accordingly. Frontier has no leverage in that negotiation and no substitute of comparable scale until the Disney title arrives, which is FY29 at the earliest.

The FY28 own-IP Planet title slips a quarter, as game releases do, and lands in FY29 instead. That creates a year with no major launch, which is exactly the FY24 shape: revenue falls, the cost base does not, and the company is back to cutting. The tax credit cushion, which has been carrying a large share of reported profit, narrows as VGEC's UK-only expenditure rule bites and the VGTR transitional claims expire in April 2027.

Around all of this the leadership transition is still settling. A chief executive nine months into the job, a chief marketing officer two months into the job, and a founder who moved to non-executive on 1 October 2026 are together managing the largest licensed commitment in the company's history. The last two times Frontier took on a large strategic bet outside its core, F1 Manager and Frontier Foundry, both were abandoned, and the Disney game is a genre fit but a new relationship, a new approval process and a new audience.

None of this is fatal. Frontier has no debt, a substantial cash balance and a back catalogue that would keep paying through a bad year. The bear case is not bankruptcy. It is that the company spends another two or three years proving it is a studio that makes good games in a small genre rather than a franchise machine, that the buyback headroom gets used up defending a share price rather than compounding value, and that the reset of 2024 turns out to have fixed the cost base without fixing the concentration.