Ryan Cohen just bid $56 billion for eBay. With $9B in cash and a meme premium.

Cohen is trying to build Berkshire in real time. The market hasn't decided if he can.

GameStop in May 2026 is two companies stapled together. One is a shrinking video-game retail chain doing $3.63B in annual revenue, down 5% YoY, that just printed $418M of net income mostly from interest on its cash hoard. The other is a holding company with $9.4B in liquid assets, $368M in bitcoin, a CEO whose 171.5M stock options vest only between $20B–$100B market cap, and an active non-binding offer to acquire eBay for $56B. The retail business is the past. The holdco is the bet. On May 4, 2026, Cohen made the bet visible — submitting a $125-per-share offer for eBay that values the e-commerce platform 4x larger than GameStop itself. The market is now pricing two opposite outcomes simultaneously: a Berkshire-style compounding machine that 5x's from here, or a retail meme with $9B that Cohen is about to spend on overpriced acquisitions.

$56 billion. Half cash, half stock. And a $20B debt letter from TD Bank.

On May 4, 2026, GameStop submitted a non-binding proposal to acquire 100% of eBay at $125 per share — a 20% premium to Friday's close, a 46% premium to February 4 (when GME began accumulating eBay shares). Cohen disclosed a roughly 5% stake in eBay already, a $20B "highly confident letter" from TD Bank, and a pledge to find $2B in annual cost cuts within 12 months of closing. The deal would target eBay's $2.4B sales and marketing budget, which Cohen argued is producing flat net active buyer growth.

The Structure

The Math Problem

GameStop's market cap is $11.1B. eBay's is $46.2B. The target is roughly 4x larger than the acquirer. Even with $9B cash and a $20B TD letter, the cash half of the deal is ~$28B — leaving GME ~$1B short on cash before the stock issuance. The stock half (~$28B) at $24.75 means issuing roughly 1.13 billion new GME shares on top of the current ~448M float — a ~2.5x dilution if the deal closes at the offered price.

That dilution math is why GME stock fell ~7% Monday despite the headline. The market is pricing the optionality of Cohen pulling off a transformational deal AND the near-certainty that completing this specific deal at this specific price would massively dilute existing holders.

$9.4B in cash. 4,710 bitcoin. No debt to speak of.

If you do not understand GME's balance sheet, you cannot value the stock at $24.75 — because the operating retail business does not justify it on its own. Here's what's actually there as of January 31, 2026:

The Bitcoin Pivot — and Pivot-Back

In March 2025, the GameStop board approved bitcoin as a treasury reserve asset. In May 2025, the company purchased 4,710 BTC for ~$428M, financed via convertible notes. That made GME the 14th-largest corporate bitcoin holder. Then in January 2026, blockchain trackers caught GameStop transferring the entire BTC position to Coinbase Prime. Cohen did not confirm a sale but pivoted publicly — calling the new acquisition strategy "way more compelling than bitcoin."

The treasury currently sits at $368M, ~$60M underwater. If GameStop sells, that's $368M added to the deal-funding war chest at a realized loss — material at the trade level but not at the deal level (the eBay cash component is $28B, not $400M). The bitcoin holding's biggest impact at this point is narrative: Cohen flagged it as expendable, signaling that "treasury optionality" gets reallocated to whatever produces the highest IRR.

The Cohen Compensation Plan

In early 2026, the GME board approved an "all-at-risk" compensation plan for Cohen. No salary. No cash bonus. His entire pay package is 171.5 million stock options that vest in tiers as GameStop's market cap crosses $20B, $40B, $60B, $80B, and $100B. At the current ~$11.1B market cap, none of those tiers have triggered. Cohen literally does not get paid until the stock roughly doubles, and his maximum payout requires a ~9x from here.

That structure is rare and informative. It tells you Cohen is not a steward — he is a swing-for-the-fences operator with a single binary outcome. It also tells you why the eBay bid is the size it is: $56B is the kind of deal that would put GME's market cap into the first vesting tier almost overnight if it works.

This isn't a retail acquisition. It's a vertical lock on the trading card secondary market.

The eBay bid looks crazy if you read it as "shrinking video-game retailer buys legacy e-commerce platform." It looks very different if you read it as what Cohen has actually been building for twelve months: a complete, end-to-end pipeline for trading card origination, grading, authentication, vault custody, and secondary-market resale. That pipeline already exists. Three of its four nodes are GameStop assets. The fourth is eBay. Buying eBay closes the loop.

The Pokémon & TCG Market is in a Mania Phase

The numbers everyone is quoting on social media are not quite the numbers in the analyst reports — but the analyst reports are themselves staggering. The global trading card games market hit roughly $13B in 2025 and is projected at $15B in 2026, with consensus CAGRs in the 7–10% range through 2031. Pokémon alone holds ~22% global TCG share. The broader trading card market (TCGs + sports cards + collectibles) is closer to $21B in 2024 with $58B projections by 2034 at 13% CAGR. Average Pokémon card prices rose +46% YoY into January 2026. Specific chase cards moved 200–500%.

The grading-services side of the market is the cleaner fundamental signal. Over 20 million cards were professionally graded in 2024 (+16% YoY), with PSA processing 15.34 million submissions. A PSA-10 grade routinely produces a 3–10x value uplift over the same card raw or PSA-9. PSA-10 rookie cards delivered an 18.3% one-year return — beating major equity benchmarks. The 30th anniversary of Pokémon falls in 2026 and is already pulling forward demand on Generations and 25th Anniversary product lines.

One honest caveat: a chunk of this growth is speculative. NPR ran a "greater fool theory" piece in April 2026. Modern sealed product is showing parabolic price action that historically precedes corrections. But the secondary-market infrastructure layer — grading, authentication, vault custody, and resale — captures fees on volume regardless of whether prices are going up or coming down. If the bubble pops, dealers panic-sell into the same eBay/PSA pipeline. The infrastructure wins on velocity in either direction.

The Vertical GameStop Has Already Quietly Built

This is the part that very few people writing about the eBay bid have actually walked through. Over the past 18 months Cohen has been building each piece of a complete card-collecting stack. As of April 2026, four of the five pieces are operational and integrated:

Power Packs already routes the user's choice through this exact funnel. When you open a digital pack on powerpacks.com, you get a PSA-graded physical card that lands in your PSA Vault account. Your three exit options are: (a) instant GameStop buyback at ~90% of estimated value, (b) ship the physical card to your home, or (c) list it through PSA's eBay store. Path (a) gives GameStop margin on the buyback spread. Path (b) gives GameStop ~6% transaction fee. Path (c) gives eBay a marketplace fee — currently captured by an outside company.

The Strategic Math

Today, every Power Packs user who chooses path (c) is generating revenue for someone else. The PSA Vault → eBay store integration already exists — that's a real partnership PSA spent years building, not a Cohen invention. Acquiring eBay would convert that fee leakage into recurring infrastructure revenue inside the GameStop holdco, and at meaningful scale: eBay's collectibles category alone does multiple billions in annual GMV, with TCG sub-categories growing fastest. Pokémon was the hottest mover on eBay in all of 2025 according to eBay's own year-end disclosures.

This is also the strategic logic behind eBay's existing PSA partnership — they've been building this infrastructure for the same reason. They saw the same demographic shift (children's hobby → adult investment asset class), the same grading volume growth, the same PSA-10 premium. The difference: eBay has the marketplace but no origination point and no retail footprint. GameStop has the origination, the retail footprint (~1,600 stores for in-person authentication intake and live commerce), and now the digital reveal layer. The two companies are mirror-image halves of the same vertical.

Cohen's offer letter explicitly cited this: "GameStop's roughly 1,600 locations in the US would give eBay a network for authentication, intake, fulfillment, and live commerce." That's not retail-synergy boilerplate. That's specifically describing a card-collecting pipeline.

Why This Doesn't Automatically Work

Antitrust is the obvious risk. A combined GME+eBay would control a meaningful share of the secondary collectibles market in the US. The EU and US regulators are already paying attention to platform consolidation. eBay's auction format and existing seller network would face concentration scrutiny. If the deal goes through with conditions — say, mandated open access for competing grading services or marketplaces — the monopoly thesis weakens significantly.

The bigger structural risk: the Pokémon market is overheated. Modern sealed product appreciation rates are unsustainable. PSA itself faces ongoing trust questions after the late-2025 buyback scandal. If the bubble cracks before GameStop closes the deal, the entire premise — that secondary card market volume keeps compounding at 13% CAGR — gets revisited at the same time GME is taking on $20B+ in debt. That's a sequencing risk that should not be ignored.

The retail business is shrinking. The cash hoard is doing the heavy lifting.

Where the Earnings Actually Come From

The $418M FY26 net income looks like a turnaround. Look closer: the operating retail business is declining double-digits in its two largest categories (hardware -12%, software -27%). Operating income is up because GameStop has aggressively closed unprofitable stores — the footprint is now ~1,600 US locations, down sharply from peak. The bigger driver: interest income on the $9B cash pile. At ~5% yields on Treasury bills, $9B generates ~$450M annually in interest income — almost the entire reported net income.

That's not a damning observation, but it is a defining one. GME at the operating level is barely break-even retail. GME at the consolidated level is profitable because of its balance sheet. The eBay bid is partly an attempt to convert that idle cash into an actual operating asset before someone (analysts, activists, or the next CEO) demands the company simply return the cash to shareholders.

Two analysts cover the stock. The consensus is 50% below the current price.

This is the single most striking valuation gap in any name we've covered. After fifteen years of broken thesis cycles, after the meme run, after the bitcoin pivot, after the Berkshire framing, the institutional research community has largely walked away. Only two sell-side analysts maintain active coverage of GME. Their consensus 12-month target is $13.50 — roughly 45% below the current $24.75 share price.

Why is the gap so wide? Because the two analysts left covering GME are valuing the operating retail business — which is dying — and not the optionality on the $9.4B cash hoard or whatever Cohen does with it. Their model is straightforward and probably correct on the retail piece. It just isn't the right model for what GameStop has become.

That said, the wide gap cuts both ways. If Cohen's eBay deal blows up or any acquisition fails to materialize, the $13.50 target is what the market would re-anchor to. The stock isn't priced like a $13.50 retail stock. It's priced like a holdco with optionality. Take away the optionality and the floor moves a long way down.

GameStop is no longer a retailer. It's a holding company in formation.

Comparing GameStop to other video-game retailers (which are mostly extinct) misses the trade. The right comp set is now meme-era cash hoards, holding companies, and crypto-treasury equities — all the structures the market has used to capitalize idle cash on a balance sheet. Here's how GME stacks up against its real peer group post-pivot:

The Honest Berkshire Comparison

Cohen explicitly invoked Berkshire in the eBay bid announcement: "Similar to Berkshire Hathaway, except what Berkshire did in decades we're attempting to do in a much shorter time." The comparison is aspirational and strained. Berkshire built its cash machine on insurance float — a structurally cheap and recurring source of capital. GameStop's $9B came from ATM equity raises during the meme run — a one-time event that diluted holders by ~30%. Berkshire compounds capital. GameStop has accumulated capital. Whether Cohen can convert accumulation into compounding is the entire trade.

The closer comp is probably Cohen's own Chewy. He built it from scratch into a $13B e-commerce business in roughly a decade. That's a real operator track record — not a marketing line. Whatever you think of the eBay price, the operator is not unproven. But Chewy was a greenfield build. GameStop is a turnaround inside a shrinking business with a CEO trying to leapfrog into M&A. The skill set required is genuinely different.

Real cash. Real intent. Real execution risk.

The range of outcomes is unusually wide. Position accordingly.

Honest scenario map. With a $9B cash floor and $20B–$100B Cohen vesting tiers as the upside ceiling, GME has a wider distribution of possible 12-month outcomes than almost any name we cover.

The asymmetry to flag: the bear case is bounded by the cash floor (~$9.4B / 591M shares = ~$15.90 in pure liquid assets per share). GME mathematically should not trade below ~$15 unless management actively destroys the cash hoard. The bull case is bounded only by Cohen's execution quality and the available pool of acquirable consumer companies. That's an unusual distribution — left tail truncated by balance sheet, right tail open-ended by operator skill.

This isn't a stock. It's a wager on Ryan Cohen.

GameStop in May 2026 is not investable as a retail business. The retail business is shrinking double-digits in its two biggest categories, has no growth narrative, and exists primarily to generate the modest operating cash flow that funds the rest of the company. If you're underwriting GME at $24.75, you're underwriting a $9.4B cash hoard, a CEO with $20B–$100B options vesting tiers, and the first concrete swing — the eBay bid — that came across the wire on May 4.

The defensible read: Cohen has demonstrated operator skill at Chewy. The all-options comp plan is the most aggressive alignment structure on a public US balance sheet. The $9.4B war chest is real and creates a hard floor under the stock. Burry — the only person who has ever been short, then long, then right on GME twice — bought back in at $25 in April. The eBay bid is structurally aggressive but it is not a trial balloon. And the strategic logic is genuinely sharper than the headline price suggests: Power Packs has already quietly built four of the five layers of a complete trading-card pipeline (origination, digital reveal, PSA grading, PSA Vault custody). eBay is the missing fifth layer — the secondary market. Buying it converts fee leakage into infrastructure revenue inside the holdco. Cohen isn't trying to compete with Amazon. He's trying to own every dollar the PSA-graded card secondary market generates.

The honest read: the eBay deal at the offered price is mathematically punishing for current GME holders. ~2.4x dilution means existing shareholders own ~30% of the post-close company. If eBay's board negotiates the price up, that ratio gets worse. If eBay rejects outright, GME is back to "what does Cohen do with $9B" with one failed swing on the record. The analyst community has effectively given up on the name. Two analysts. $13.50 consensus. Net institutional selling.

This is the kind of trade where position sizing dominates conviction. The bear case ($15) is ~39% downside. The bull case ($45) is ~82% upside. The stretched case ($70+) requires Cohen to actually become Cohen-Buffett. Reasonable people can disagree on probabilities. The one thing they shouldn't disagree on is that GME is no longer a fundamentals trade.