Goodwill and Impairment — What Sony Paid For at Bungie

On 21 May 2026, Bungie published a blog post announcing the end of Destiny 2's live-service development. Monument of Triumph, the final content update, would arrive on 9 June. After that: no more expansions, no more seasons, no Destiny 3. Twelve years of one of gaming's most sustained franchise relationships, concluded under the heading "Every End is a New Beginning."

Bungie has used that phrase before. When they left Microsoft in 2007. When they parted from Activision in 2019. When Sony acquired them for $3.6 billion in January 2022. At every transition, the language has been identical — forward-looking, structurally optimistic, pointing somewhere ahead. The new thing this time is Marathon. The irony is sharper than the branding implies: Marathon is not a new thing at all. It was Bungie's original franchise, the Mac FPS trilogy from 1994 that preceded Halo by seven years. The studio's new beginning is its oldest game. A company that has spent twenty-five years becoming something different has navigated its way back to the start.

Marathon, the 2026 extraction shooter, peaked at approximately 88,000 concurrent Steam players at launch in March and was sustaining 11,000 to 16,000 by mid-May. Development cost: over $250 million.

In the same fiscal year, Sony's Game and Network Services segment posted record operating income: ¥463.3 billion, up 12% year-on-year. On the same balance sheet, ¥120.1 billion in Bungie-related impairment charges. These two facts belong together. One shows the resilience of Sony's gaming business. The other shows the cost of an assumption about live service as an acquirable capability — an assumption still embedded in enough balance sheets across the sector that examining how it failed here is worth more than the accounting line it produced.

A Studio That Kept Becoming Something New

Understanding what went wrong at Bungie requires going further back than the operational failures of 2023 and 2024. The conditions were set earlier, through a corporate history that kept redefining what the studio was supposed to be optimising for.

Bungie has spent twenty-five years serving four distinct corporate agendas. Under Microsoft, they were the architects of Xbox's cultural identity. Halo was the credential that made a challenger console credible, the franchise Microsoft kept when Bungie negotiated their independence in 2007. What Bungie walked away with was creative autonomy and no IP to show for it. The transition established a pattern: each ownership change would leave something behind.

The Activision years (2010 to 2019) produced Destiny and its sequel, but within a publishing structure that had already begun to pull against the game's design ambitions. The initial agreement between Bungie and Activision committed the publisher to $500 million in funding across a slate of four Destiny titles over a decade. That roadmap dissolved early. Activision wanted faster sequels; Bungie had designed a live-service model that required sustained investment in ongoing content rather than packaged releases. When the ten-year deal ended in 2019, framed as an amicable creative separation, Bungie emerged self-publishing for the first time, with Destiny 2 as their sole commercial asset and a franchise whose direction had been contested from its earliest planning stages.

Independence arrived alongside genuine commercial momentum. Destiny 2 was generating meaningful revenue, and Sony arrived in January 2022 with a $3.6 billion acquisition offer and a new set of priorities: live-service expertise was what PlayStation needed, and Bungie would be its vehicle. The combination of creative freedom, financial resource, and an acquiring parent with strategic ambitions of its own produced a period of rapid internal expansion. Senior talent was seeded into multiple incubation projects: Project Payback, a third-person action-adventure spin-off; an internal project referenced in coverage as "Gummy Bear"; a growing portfolio of concepts, each staffed by experienced developers who were therefore no longer staffing Destiny 2.

The contraction that followed was costly. Technical debt accumulated as development resources dispersed. The content cadence that live-service audiences require slowed and grew erratic. Monetisation choices that extracted from rather than served the player base eroded trust accumulated over years. Content vaulting — removing paid expansions from the active game — was the most visible and contested of these decisions. The taxonomy of monetisation quality that separates durable from fragile recurring revenue is examined in the companion Price of Play analysis; the short version is that Bungie had moved from building goodwill to spending it, and the account was not bottomless.

The layoffs came in rounds. Approximately 100 staff in October 2023. A further 220 in July 2024, with significant transfers to Sony Interactive Entertainment. A third wave was announced on 21 May 2026, simultaneous with the Monument of Triumph announcement — the roles being cut were those that had existed to produce Destiny content: sandbox design, expansion narrative, the long-cycle content factory that generated quarterly seasons. No approved successor project exists for those teams. Destiny 3 has not been greenlit.

Each round of cuts was accompanied by public acknowledgement from Pete Parsons that the studio had spread itself too thin. What went unaddressed in those statements was a governance question with its own timeline. Between September 2022 and June 2024, an auction account registered as "bngpparsons" documented $2,414,550 in classic car purchases, including a 1971 Porsche 911S acquired for $201,000 in November 2023 — the month after the first layoff round. Parsons did not comment publicly. The reaction from former employees was extensive and sharply negative. The spending was his to make; the timing made visible a question about whether the studio's incentive structure was aligned with its obligations to the people it was simultaneously letting go.

By August 2025, Parsons announced his retirement after twenty-three years, framing it as "passing the torch." The departure aligned with the vesting of equity from the 2022 Sony acquisition. The role he vacated became Studio Head, held by Justin Truman. Bungie no longer has a chief executive. That structural change is not incidental: it reflects the full integration into PlayStation Studios that the original deal had nominally deferred, and the end of the independent subsidiary arrangement that had been a condition of the acquisition.

The Final Shape and the Problem of a Farewell Tour

In June 2024, Destiny 2 released The Final Shape, the culminating expansion of its ten-year narrative arc. Peak concurrent players on Steam reached 314,634, within 1% of the all-time record set at Lightfall's launch sixteen months earlier. Former players returned. Community sentiment, briefly, recovered.

Within three months, concurrent players had fallen to 101,450. Within nine months, to 60,035. The expansion performed well. It could not sustain the audience it temporarily recovered.

Trust, once drawn down far enough, does not replenish through a single release. The architecture of accumulated grievances — the vaulted content, the monetisation friction, the seasons that arrived under-resourced — had made even a genuinely good product insufficient to reverse the direction. The peak lasted one quarter. By April 2026, Destiny 2's monthly peak on Steam had fallen to 12,948: a 96% decline from the June 2024 high.

Then the finale recast the numbers. Monument of Triumph, the final content update, launched on 9 June 2026 and drew a peak of 167,867 concurrent players — the game's largest Steam audience since The Final Shape itself. The spike does not complicate the diagnosis; it completes it. An audience that size was still reachable. What brought it back was not a season or a roadmap but an ending.

The Acquisition That Failed to Acquire

When Sony recognised ¥31.5 billion against Bungie intangible assets in Q2 of its fiscal year ending March 2026, CFO Lin Tao confirmed what engagement data had been signalling for quarters: Destiny 2's sales and user numbers had not reached expectations. When ¥88.6 billion followed in Q4, alongside ¥18.3 billion in development cost corrections, the picture was complete. Total Bungie-related impairment for the fiscal year: ¥120.1 billion, approximately $765 million in dollar terms, against an acquisition price of $3.6 billion.

What the impairment reveals about the acquisition is less in the headline figure and more in what the consideration was structured to purchase. Of the $3.6 billion, approximately $1.5 billion was upfront cash. Another $612 million was deferred. Some $304 million was contingent on employee retention vesting. And $1.2 billion in employee incentives, structured outside the formal purchase price, was designed explicitly to retain the people whose accumulated knowledge constituted the asset being purchased. Between $1.5 and $1.6 billion of the headline consideration was not paying for a business. It was paying for its workforce not to leave.

The retention premium was the hedge against the asset walking out the door. When the workforce it was designed to retain was subsequently halved through layoffs and reassignments, the hedge expired without delivering its underlying. The intangibles now written down were precisely what that premium was built to preserve.

The acquisition context is examined in detail in the gaming M&A analysis published earlier this year, which showed how the deal, announced thirteen days after Microsoft's $68.7 billion Activision announcement, broke the disciplined sequencing that had made Sony's earlier studio purchases — Insomniac, Housemarque, Bluepoint — so capital-efficient. Bungie had no prior exclusive PlayStation relationship, no platform-specific commercial proof of concept. Sony paid the premium and assumed the capability would transfer. The impairment charges now establish, in accounting terms, that it did not.

The G&NS segment's underlying profitability, driven by high-margin network services and digital content, strengthened considerably in the year that absorbed the Bungie charges. FY2026 guidance of ¥600 billion in operating income implies an approximately 13.6% margin on lower hardware revenue. Sony's gaming business is resilient. The Bungie asset does not define it. But the segment result makes the Bungie result more legible rather than less: the underlying operation was generating enough to absorb a nine-figure write-off and still post its best-ever year. The cost of the error has been borne by a business that can afford it — which is analytically distinct from arguing that the cost was acceptable.

The broader pattern at Sony is harder to absorb. In May 2022, Jim Ryan committed PlayStation to twelve live-service titles by fiscal year 2026, with live service accounting for 60% of the development budget. At least eight of the twelve were cancelled before release, among them The Last of Us Online (a Naughty Dog multiplayer project that absorbed years of development before cancellation), a live-service God of War, and the Bend Studio initiative. Of the titles that reached market, Helldivers 2 was the only one that performed with commercial conviction.

Shuhei Yoshida, the former Sony executive who oversaw the first-party studio strategy preceding Jim Ryan's tenure, was removed from running Sony's internal studios after resisting the live-service strategy shift. He later stated publicly that he would have "tried to resist" the push, and that the sector's competitive conditions made success unlikely: "You can't plan a success in this industry." His prescription was a return to Sony's demonstrated strengths: premium single-player titles with deep creative investment, rather than continued attempts to compete in a live-service market whose winners had established durable audience relationships years before Sony assembled its portfolio. He was not listened to. He was moved aside. The impairment charges are the delayed cost of that decision.

Concord makes the pattern most acute. Firewalk Studios' hero shooter launched in August 2024 at a $40 price point, peaked at 697 concurrent players on Steam in its opening weekend, and was pulled from sale on 6 September — less than two weeks post-launch, with full refunds issued and the studio subsequently closed. Eight years of development, at a cost estimated at $200 million at minimum. The title launched in the same year as Marvel Rivals, a free-to-play hero shooter; competing at premium pricing against zero-cost alternatives in the same genre defined its trajectory before a player logged in.

Marathon now carries the weight of what remains. The extraction shooter peaked at approximately 88,000 concurrent Steam players at launch, declined by 71% within a month, and was sustaining 11,000 to 16,000 by mid-May 2026. The extraction genre has a smaller addressable market than a mainstream looter-shooter — the performance is not catastrophic in isolation. What matters is the hypothesis it was meant to validate: that Bungie's resources were being prudently redirected from a mature live-service title to a capable successor. Marathon's first months do not establish it.

Sony's FY2026 guidance of ¥600 billion in G&NS operating income assumes the absence of further Bungie charges. That is the correct base case. What the guidance does not contain is any live-service contribution from Bungie of scale. The ¥600 billion is Sony's gaming business without the thesis it paid $3.6 billion to build.

The Unearnable Asset

Sony's failures sit within an industry pattern that should prevent any analyst from treating them as company-specific.

Five years of live-service failures, from Anthem to Destiny 2, are too varied and widespread to read as isolated mistakes. The pattern is structural. Anthem, BioWare's shared-world shooter, launched in 2019 with production quality that could not compensate for content depth insufficient to sustain a live-service commitment. Support was curtailed; an overhaul was abandoned. Marvel's Avengers invested heavily in licensed IP that guaranteed an opening audience but no foundation for ongoing engagement — the live-service element felt grafted onto a single-player framework, and player numbers reflected it. Babylon's Fall, the Square Enix and PlatinumGames collaboration of 2022, reached sub-1,000 concurrent players within weeks of launch and shut down in under a year. Suicide Squad: Kill the Justice League, released in 2024, contributed to a $200 million reported loss at Warner Bros. Games and downstream studio closures including the dissolution of Monolith Productions.

The pattern across these failures is consistent enough to generalise. Each was built on the assumption, sometimes explicit in management commentary and sometimes embedded in the business model, that sufficient investment in production quality, licensed IP, or brand recognition could substitute for the fundamental requirements of a live-service franchise. Those requirements are relational. Players choose to return every week, over years. That choice is renewed or withdrawn based on whether the studio is making decisions that feel like they serve the audience or extract from it.

The Bungie Family Tree: A Smaller World Than It Appears

Concord's failure connects to Bungie's history through a specific human thread. Harold Ryan, who served as Bungie's president during Destiny 2's troubled early development before departing in the mid-2010s, went on to found ProbablyMonsters in 2016 — a studio incubator designed around a different model for triple-A development. One of its incubated studios was Firewalk, founded by Ryan Ellis and other Destiny franchise veterans. ProbablyMonsters sold Firewalk to Sony in 2023. Ellis directed Concord.

When Concord launched and failed, the studio was closed. Institutional knowledge from Bungie's Destiny era — the same leadership vintage, the same live-service heritage — had been applied to a different game, at a different studio, acquired by the same platform holder. The result was not different.

This is worth more than a small-world observation. If the people who built one of gaming's most successful live-service franchises could not replicate the conditions for success when they tried, that is a data point about whether live-service expertise is transferable at all, even when the people themselves move.

The counterexample is Helldivers 2. Arrowhead Game Studios' cooperative shooter launched in February 2024, sold over 20 million units, reached 458,709 peak concurrent players on Steam in its launch week, and sustained meaningful engagement through 2025 and into 2026 — expanding to Xbox in August 2025, registering a 182,000 peak concurrent in September of that year. Sony's fastest-selling first-party title on record. And notably: developed by a small studio with a focused design intent and a clear understanding of its audience, not by a team carrying the accumulated institutional weight of four corporate owners' worth of competing priorities.

The gap between Helldivers 2 and the table above has nothing to do with budget or IP recognition. Arrowhead was a niche studio; Firewalk spent years and at minimum $200 million producing a game that attracted 697 players at launch. Arrowhead treated the player relationship as the product. Transparent communication, rapid response to community feedback, monetisation at the value-accretive end of the spectrum. The game succeeded because the studio maintained the agreement with its audience that a live-service franchise requires — and maintained it consistently, through every content decision, rather than spending it down in service of other objectives.

The live-service model has durable winners. Fortnite, GTA Online, League of Legends, Final Fantasy XIV's remarkable recovery after a near-catastrophic initial launch. Each built their audience the same way: a product with a defined identity, fair and transparent monetisation, and decisions over time that demonstrated the studio was in dialogue with its audience rather than extracting from it. None of them acquired that audience through an M&A transaction. They earned it.

What to Monitor: Sony's Live Service Transition

For investors following the G&NS segment, the Bungie outcome closes a chapter without resolving what replaces it.

Marathon's quarterly trajectory — Peak concurrent and monthly active users through the first two-to-three season cycles. Live-service titles that cannot stabilise their base in early seasons rarely recover. The extraction genre has a smaller addressable market; what matters is whether Bungie captures a loyal core or sustains only launch transience.

Goodwill impairment testing — Sony's auditors did not impair the goodwill component of the Bungie acquisition in FY2025. That judgment rests on Marathon's remaining potential. If Marathon's engagement data continues on its current trajectory into FY2026, the triggering condition for a goodwill charge becomes easier to reach.

Management commentary on Bungie — After two impairment disclosures totalling ¥120.1 billion, how prominently Bungie and Marathon feature in Sony's forward guidance is itself informative. Silence about a $3.6 billion acquisition is a disclosure choice.

The Helldivers 2 template — Arrowhead's success was developer-driven, with Sony as publisher. Whether Sony's forward live-service bets follow the Arrowhead model — small team, focused design, fair monetisation, community dialogue — versus the Firewalk model — large budget, long development, premium pricing into a free-to-play market — is a clearer signal than any management commentary.

Goodwill, Unimpaired

There is a small irony in Sony's accounting treatment of the Bungie acquisition. Of the ¥120.1 billion recognised in impairment charges for the fiscal year ending March 2026, none was taken against goodwill. The intangible assets — the IP, the operational value, the expertise — have been written down. The goodwill, in accounting terms, remains on the balance sheet.

This is not unusual. Goodwill impairment requires a specific triggering assessment: the carrying value of the reporting unit must exceed its estimated fair value. Sony's auditors have not reached that conclusion, and the structure of the charges suggests management would contest it. The goodwill will be tested again in future periods.

On the other ledger — the one that measures player trust, audience relationship, and community health — the goodwill was impaired years before any accountant examined it. It was impaired when paid expansions were vaulted. It was impaired when live-service seasons arrived under-resourced and over-monetised. It was impaired when the studio's communications became the language of corporate inevitability rather than the language of a developer talking to its players. By the time The Final Shape arrived to demonstrate what Bungie was still capable of producing, the account had been drawn down too far for a single good product to replenish it. The peak lasted one quarter. The decline resumed.

Sony's FY2026 guidance is for ¥600 billion in G&NS operating income, substantially recovered from the reported FY2025 result precisely because it no longer contains the Bungie drag. The segment's resilience is not in question. What is open is the question the acquisition was meant to answer: what does Sony's live-service capability look like without Bungie as its foundation? The twelve-title commitment has produced one sustained success, eight cancellations, and a last-chance-saloon extraction shooter that is currently performing below the top hundred titles on the Xbox platform.

The franchise Bungie spent twelve years building was not simply a product. It was a relationship — and relationships are not on the asset schedule. They appear in neither the intangibles nor the goodwill line. They do not transfer in acquisitions. They are not retained through vesting schedules. They are earned, and re-earned, through every content decision, every monetisation call, every moment of communication between a studio and the audience that chose to keep showing up.

Live service is not an asset. You cannot acquire it, impair it, or carry it at value on a balance sheet. Every entry in the graveyard has the same root cause.