Trading In Your Old Game Paid for the Next One and That Loop Is Ending

The loop of trading in a finished game to buy the next one is broken

A game on a disc was priced twice. Once at purchase, once at trade-in. Michael Pachter of Wedbush Securities puts it plainly: historically at least a third of games were sold used, and the copies people traded in became the currency that paid for the next one (TheGamer). The sticker price always had a recoverable portion baked into it. Resale was the discount.

Go digital and that portion is gone. Pay $70, put 40 hours into finishing it, and what stays in the account is an entitlement flag. The price on the store page has not moved, but what a player actually spends has. Nothing in the market currently closes that gap.

The disappearing side is not small. Dataintelo valued the second-hand platform market, counting pre-owned games, consoles, and peripherals, at $7.2 billion in 2025 with a path to $13.8 billion by 2034. That forecast assumed physical media keeps shipping. Sony ends disc production for new PlayStation titles in January 2028, which breaks the assumption. Morningstar’s Kazunori Ito expects brick-and-mortar stores to keep shrinking and eventually disappear without new PlayStation titles in circulation, and Pachter compressed it to four words: brick and mortar game retail is doomed (IGN, CNBC).

One fact hides underneath all of it. Publishers never saw a cent of that $7.2 billion. Used sales settled between GameStop and the customer, which is why the industry treated resale as an enemy for two decades. If the whole market vanishes, what publishers gain is not a stack of extra full-price sales. Residual value going to zero raises the real cost of every purchase, and that shows up in how often people buy.

The legal route just closed

The legal route just closed. The CJEU’s 2012 UsedSoft ruling allowed resale of downloaded software copies, and in September 2019 a Paris first-instance court applied that exhaustion doctrine to downloaded games, holding that Valve could not stop users from reselling Steam purchases. The Paris Court of Appeal reversed it in 2023. Games bundle a program with graphics and sound into a single protected work, so they fall under the general copyright directive rather than the software directive, and that directive does not exhaust rights in a dematerialized copy (Osborne Clarke).

Forcing resale through rights is finished. The only remaining path is a publisher opening one voluntarily, and the reason to refuse just evaporated. On physical resale a publisher’s share of the second sale was exactly zero. A digital buyback and relist happens inside the publisher’s own storefront, which means a cut of every transaction plus a record of who resold what and when. The thing the industry spent twenty years blocking becomes a revenue line.

The timing helps too. January 2028 is already on the calendar, so there is room to build, and until then discs and downloads sell side by side. That overlap is the last window in which real trade-in prices and resale rates can be observed and used to seed a digital residual model.

flowchart LR
  A[Player returns license] --> B[Residual pricing engine<br/>age, discount history]
  B --> C[Store credit issued]
  A --> D[Relisted copy<br/>sold at a discount]
  D --> E[Settlement split<br/>publisher, store, seller]
  C --> F[Credit spent on a new release]

Open narrowly: one publisher, one store

Open it narrowly: one publisher, one storefront. The product has three parts.

The first is the settlement rail. A player hands back a license, store credit lands immediately, and the license goes back on sale as a discounted relisted copy. Proceeds split between publisher, storefront, and the original buyer on a ratio agreed up front. Only one piece of this is technically hard. Revoking the entitlement from the first account and granting it to the second must complete inside a single transaction, because any intermediate state is a fraud path waiting to be found.

Pricing comes next. Used cars have residual tables and games have nothing. A model that takes time since release, completion rate, historical discount depth, and concurrent player trend and returns a buyback price is the actual moat here. Start with a plain depreciation curve, then correct it as buyback and resale data accumulates. Whoever holds that data becomes the reference price.

The last piece sits on top of the other two. Once residual value is a number, products stand on it: buy a new release cheaper on the condition that it comes back, or take the projected buyback value off at checkout instead of after the fact. Auto leasing already runs on residuals, and the structure transfers directly.

The MVP is small. One publisher and a handful of indie titles, run end to end from buyback to relist. Two numbers matter. What share of returned licenses actually resell, and what share of issued credit gets spent on a new release. The second one tells you whether the Pachter loop still turns in digital.

The risk is cannibalization. If relisted copies start eating full-price sales, the publisher walks. That means caps on relisted volume, a minimum window after launch, and restrictions attached to relisted copies, and every one of those dials belongs to the publisher from day one. That is what the contract is really about.

A publisher’s signature will arrive well before a court’s. The stretch between now and January 2028 is the window to go get one.