September 2026 Card Prices: Old Games Rose, Promos Fell
Every month we freeze a snapshot of how the trading card market moved, built from our own nightly price record. The September 2026 report covers the 30 days from 23 August to 22 September across 21,513 cards in 15 games. The headline number was dull. What sat underneath it wasn't.
A flat market that wasn't flat
Our composite index fell 0.8% over the window. Taken alone, that reads as "nothing happened". It hides a market that was moving in both directions at once: 7,664 cards fell by more than 2% and 6,292 rose by more than 2%, while 7,557 were broadly unchanged. One card in ten fell more than 16%, and one in ten rose more than 14%.
That spread is the useful lesson. An index tells you what a basket did. It says very little about what your card did, which is why it's always worth checking a specific card rather than reading the market headline.
The longest-running games rose
All three of the longest-running games in our record went up: Pokémon (+2.5%), Yu-Gi-Oh! (+2.2%) and Magic (+2.0%). Among the twelve newer games, nine fell. One Piece fell furthest at −4.9%, followed by Gundam (−4.4%), Grand Archive (−3.8%) and Digimon (−3.1%).
Among cards worth at least $20, five of the ten biggest risers were Pokémon cards from sets released between 2007 and 2011 — a δ (delta species) Pikachu from POP Series 5, a Call of Legends Snorlax, a Majestic Dawn Mewtwo, a Platinum Shaymin LV.X and a Pokémon Rumble Zapdos — each up more than 50% over the 30 days.
Promos gave back their premiums
The clearest pattern was on the way down. Nine of the ten biggest fallers were promotional or alternate-art printings: tournament "Round 1" promos of Sanji, Roronoa Zoro and Monkey D. Luffy from One Piece, alternate-art cards from Dragon Ball and Digimon, and a Digimon regionals champion promo. Each lost more than half its price over the window.
What we can and can't tell you
Those are the numbers. Explaining them is a different job, and we want to be clear about which part is which.
Our interpretation, not a finding: recent promo and alternate-art printings tend to launch at a premium, when supply is thin and attention is high, and to drift down as more copies reach the market. The pattern above fits that, and it's the same "spike versus drift" behaviour we describe in when to sell a trading card. But our data records prices, not causes. We can't see how many copies were distributed, and a single month is a short window. If you hold cards like these, treat this as a reason to check them, not a prediction about where they go next.
How the numbers are built
For each game, the index is the equal-weighted mean of each card's price relative to the start of the window, trimmed by 5% at each tail so a handful of wild swings can't carry a whole game. A card counts only if it has at least 15 clean daily readings, a game is reported only if at least 150 cards qualify, and moves beyond ±100% are excluded as data errors. Movers are limited to cards worth at least $20. The full report has the per-game table, the complete risers and fallers lists, a CSV download and the method in detail. The numbers on it are frozen and won't change, so it's safe to cite.