Misprint Is Building a Stock Exchange for Pokemon Cards

The Macro: Trading Cards Are a Real Asset Class Now

The global trading card market crossed $30 billion and it is still growing. Pokemon cards alone are a multi-billion dollar segment, driven by a combination of nostalgia, genuine collecting interest, and a speculative layer that looks more like financial markets than hobby shops. When a first-edition Charizard sells for six figures at auction, that gets headlines. What does not get headlines is the millions of transactions happening every week at the $5 to $500 level, where the real volume lives.

The problem at that volume level is pricing. If you want to sell a PSA 9 Umbreon Gold Star, what is it worth right now? eBay will show you completed listings, but those are noisy. TCGplayer has market prices, but they lag behind actual demand. Facebook groups and Discord servers have active buyers, but price discovery happens through DMs and negotiation. The information asymmetry is enormous, and it consistently favors the people who spend the most time tracking prices over the people who just want to buy or sell a card at a fair number.

This is the same problem that financial markets solved decades ago with order books, real-time quotes, and transparent bid/ask spreads. The question is whether that market structure can work for physical collectibles where condition grading, authentication, and shipping add friction that stocks do not have.

StockX took this approach with sneakers and proved the model works for physical goods. PWCC and Alt have applied similar logic to cards and collectibles. Whatnot built a massive business on live auction mechanics for the same market. The space is active, the money is real, and the infrastructure is still catching up to the demand.

The Micro: Goldman Sachs Meets Graded Cards

Eva Herget and Jon Jenkins founded Misprint with backgrounds that map directly onto this problem. Eva was an equity analyst at Goldman Sachs and managed to hit $500K in annual revenue selling Pokemon cards within three months. That is not a casual hobby stat. That is someone who understood market dynamics and applied them to an inefficient market. Jon is a mathematician and ML researcher who left his PhD to build pricing algorithms for the platform.

The product works like a peer-to-peer marketplace with real-time pricing transparency. Every card has visible bid/ask spreads, similar to what you would see on a stock trading platform. If you are selling a PSA 10 Moonbreon, you can see the highest current bid and the lowest current ask, and you can price your listing accordingly. That visibility changes the dynamics of every transaction because neither side is guessing.

The platform supports graded cards from PSA and CGC, sealed products, and individual card listings from sellers. There is a subscription program that offers free shipping, store credit, and access to member-only sales, with the marketing claiming $376 per year in savings for active collectors. The fees are positioned as the lowest in the category, which is a direct shot at TCGplayer and eBay’s take rates.

They came through YC’s Winter 2025 batch, and the product is live with active listings across Pokemon card categories. The community component includes Discord integration, which is where this market’s most engaged buyers already spend their time.

The ML-powered pricing is the technical moat. Jon’s background in mathematics and machine learning drives a system that processes millions of data points to provide accurate valuations across card conditions, editions, and grades. Getting that right is genuinely hard because card pricing depends on factors that are difficult to quantify: centering, surface quality, the difference between a PSA 9 and a BGS 9.5.

The Verdict

I think Misprint is solving a real problem with the right team. The trading card market has been underserved by technology that matches its actual sophistication as a market. Collectors and investors are already thinking about cards in financial terms. Giving them financial market infrastructure is a natural fit.

The competitive risk is real. TCGplayer has massive market share and recently got acquired by eBay, which gives it even more distribution. Whatnot has raised hundreds of millions and dominates live selling. PWCC handles the high-end consignment market. Misprint needs to carve out the segment that values pricing transparency and low fees over the network effects that those incumbents offer.

In 30 days, I would want to see average time from listing to sale. Speed is the proof that the bid/ask model works for physical goods. In 60 days, the question is repeat transaction rate. Do sellers come back, or do they list once and leave? In 90 days, I would want to know if the ML pricing is accurate enough that people trust it as a reference price even when they are not buying on the platform. If collectors start saying “check Misprint for the real price,” the network effects follow. The founding team has the domain knowledge and the technical skills. This comes down to execution on marketplace liquidity, which is the hardest problem in any marketplace business.