Zombie Corp · Investigation #2

Gaming hit an all-time high.
So why are the companies dying?

The industry just posted its biggest year ever — and the layoffs, studio closures, and billion-dollar write-downs won't stop. The analyst Matthew Ball framed the paradox. This is one forensic answer the macro data can't give: a balance-sheet answer. We check the leaves, not just the canopy.

Investigation #2 Embracer · Unity · Take-Two Live · June 2026
Investigation #2 · Verdict Three companies — three verdicts

Embracer caught early. Unity flagged — a mask without the debt trap. Take-Two cleared — scary optics, sound core.

What these verdicts do not say
  • They do not say the industry is failing. 2025 was the biggest year the games industry has ever had, and that record is genuine. The paradox is the point.
  • They do not say Take-Two is a buy or Unity is a sell. Take-Two is the control case: it impaired its goodwill, it generates cash, it has a dated catalyst. That is what a clean read looks like, not a recommendation.
  • Embracer is not a live call. It is the retrospective test — what the scorecard would have flagged before the collapse. The instrument is being scored here as much as the companies are.

A record year. And a body count.

2025 was the best year the games industry has ever had. Global content sales reached a new all-time high — and yet new-company funding fell off a cliff, layoffs ran for a third straight year, and several of the biggest names took write-downs measured in billions. Two facts that should not sit next to each other, sitting next to each other.

2025 content sales
$195.6B
All-time high (Ball / industry data)
Year-over-year
+5%
~$10B above the 2021 peak
Private games funding
−55%
Year-over-year
Q4 2025 funding deals
~40
Down from ~200 in Q4 2021
The question

Industry analyst Matthew Ball put the paradox plainly in his annual State of Video Gaming: if gaming is at an all-time high, why are so many game companies still struggling? His report answers it from the top down — geography, platforms, attention, funding. This investigation answers a narrower slice of it from the bottom up, on the one surface where weakness has nowhere to hide: the balance sheet.

You can build growth. Or you can buy it.

Between 2020 and 2022, a lot of the industry chose to buy it. Studios and publishers went on an acquisition spree, funded either with debt or with their own inflated stock. Every one of those deals created goodwill — the accounting word for the premium you pay above the hard value of what you bought. Goodwill is a promise: that the thing is worth what you paid for it.

When the promise breaks, the rules say you must write the goodwill down. That is exactly what is happening across gaming right now — Embracer took a SEK 6.0 billion goodwill hit last quarter, Take-Two wrote off $3.5 billion, even Sega wrote off roughly $200 million on Rovio. The Goodwill Mask Scorecard exists to ask one question with evidence, not opinion: is a company hiding genuine weakness behind acquisition goodwill it has never honestly written down?

The three layers it checks.  01 — The zombie core: can't cover its interest from earnings, and isn't young enough for that to be a growth-phase excuse.  02 — The mask: goodwill exceeds tangible equity, so the balance sheet only looks positive because of intangibles.  03 — Stale recognition: the acquisitions were debt-funded and the goodwill has never been honestly written down. Scored on the Goodwill Mask Scorecard v1.2 — the same framework calibrated on Roper (cleared), Wayfair (screened out), and B&G Foods (17/20 — Acute) before any live case. Full method →

Three roll-ups. Three very different answers.

Take three of the most acquisitive names in gaming and run them through the same three layers. They do not come out the same — and that is the point. A scorecard that flags everything is just a mood. A useful one tells the genuinely masked from the merely scary.

The test Embracer · peak '21–'23 Unity · now Take-Two · now
01 — Zombie core
Can't cover interest from earnings
● PresentDebt-funded roll-up; leverage covenant capped at 2.5× EBITDA, breaching its own targets. ◐ PartialYears of GAAP losses — but the debt is low-coupon convertibles, so interest coverage isn't the trap. ○ AbsentEBITDA $760M, up 282%; recurrent spending is 78% of bookings.
02 — The mask
Goodwill > tangible equity
● PresentGoodwill far exceeded tangible equity at the peak; the balance sheet stood on intangibles. ● PresentGoodwill of $3.17B ≈ its entire equity; strip intangibles and tangible book is negative. ○ AbsentCarried goodwill, but already impaired it and backs it with real, cash-generative IP.
03 — Stale recognition
Goodwill never honestly written down
● Present → now clearingStale for years; now writing it down — SEK 6.0B in goodwill last quarter alone. ● Present$3.17B goodwill unchanged 2024→2025 — even as the ironSource ad network that made it is shut down. ○ ClearedTook the medicine: a $3.5B goodwill write-down already booked in FY2025.
Verdict Caught early
Would have flagged it before the collapse
Flagged
A mask — without the debt trap
Cleared
Scary optics, sound core
Scroll the table to see all columns →
Caught early

Embracer Groupthe pattern, on a delay

Embracer is the clearest goodwill-mask story gaming has produced — and the clearest proof of what the scorecard is for. At its peak it was a debt-funded roll-up of dozens of studios, valued near $11.6 billion. Every signal the scorecard reads — debt-funded acquisitions, goodwill towering over tangible equity, leverage straining its covenants — was lit up years before the market repriced it. The trigger came in May 2023, when a $2 billion investment collapsed and the stock fell 40% in a day.

What makes Embracer instructive today is that the mask is now coming off in public: a SEK 7.2 billion impairment last quarter (SEK 6.0B of it goodwill), a full break-up into separate listed companies, and a swing all the way to a net cash position. The lesson isn't "Embracer is a zombie." It's that the forensic signals were readable long before the $11.6B-to-$1.7B fall. That lead time is the whole product.

−85%
Market cap vs. 2021 peak
SEK 6.0B
Goodwill impaired last quarter
SEK 3.8B
Net cash today (de-levered)
Flagged

Unitya mask without the debt trap

Unity took a different road to the same place. It bought ironSource in 2022 for $4.4 billion in an all-stock deal — so instead of loading up on debt, it diluted its own shareholders. That choice is why it isn't a classic interest-coverage zombie: its debt is low-coupon convertible notes (~$2.24B against ~$2.06B of cash, maturities pushed out), so coverage isn't the problem.

But the mask is real, and it's live. Goodwill of $3.17 billion sits against its entire stockholders' equity — strip out goodwill and other acquisition intangibles and tangible book equity is negative. The sharpest flag: that $3.17B of goodwill is unchanged from 2024 to 2025 — even though the ironSource Ad Network that generated much of it was shut down in April 2026. The asset is gone; the goodwill stays on the books. That's exactly the question the scorecard is built to ask.

$3.17B
Goodwill, unchanged YoY
Negative
Tangible book equity
$4.4B
ironSource — all stock, no debt
Cleared

Take-Two Interactivescary optics, sound core

On the surface, Take-Two looks like the worst of the three: a $4.48 billion net loss in FY2025, driven by a $3.55 billion non-cash goodwill write-down tied to its ~$12.7B Zynga acquisition. A lazy read flags it instantly. The scorecard clears it — and the difference is the discipline that makes the tool worth anything.

Take-Two already took its goodwill medicine honestly, rather than hiding it. Underneath the GAAP noise, the business throws off real cash: EBITDA of $760 million, up 282%, with recurrent consumer spending at 78% of bookings. And it has the single biggest catalyst in entertainment dated and confirmed — Grand Theft Auto VI on November 19, 2026 — alongside its first positive net-income guidance in years. A company that impaired its goodwill and generates cash with a real catalyst is not masking weakness. It's the control case.

+282%
EBITDA growth, FY2026
78%
Recurrent spend share
Nov 19 '26
GTA VI — dated catalyst

The high is real. So is the rot.

Here is the balance-sheet answer to the paradox. The all-time high is genuine — players are spending more than ever. But a meaningful slice of the apparent value created in the 2020–2022 boom was never operating strength. It was goodwill: growth that was bought, debt-funded or stock-funded, then carried on the books at a price the underlying business couldn't justify. Now it's being written down across the sector at once.

The companies struggling are not a random sample. They are disproportionately the ones that bought growth with goodwill instead of building it — and the scorecard separates them from the ones that merely look frightening on a GAAP screen. Embracer would have been readable years early. Unity carries a live mask without the debt trap. Take-Two looks worse than all of them and is the soundest of the three. Same record-breaking industry. Three different fates, decided upstream — in how each one chose to grow.

The takeaway

A record top line can sit on top of a decade of overpaid acquisitions. The headline number tells you the industry is healthy. The goodwill line tells you which companies actually are.

Holding a company that grew by acquisition? Let's read its goodwill line together.
Request a Conversation
Investigation #2 · Ipalibo Da-Wariboko · June 20, 2026
Scored on the Goodwill Mask Scorecard v1.2. Figures from company filings and reports: Embracer Q4 FY2025/26; Unity FY2025 Form 10-K and Q1 2026; Take-Two FY2025–FY2026. Industry figures per Matthew Ball, State of Video Gaming in 2026, and Circana. Not investment advice.