The moonstone Pokémon Market Index closed August 2, 2026 at 249.8, up 149.8% over the eighteen months the index spans, which begin at 100 on February 3, 2025, up 89.4% over the past year, and still climbing at +5.0% over the past 30 days. Over the same year the overall crypto market, the risk benchmark we chart beside it, fell 45.2%. That is the headline. The more useful story is underneath it: this is no longer one rally. The market has split into a closed-print market that keeps rising and a new-product market that is deflating in real time, and the gap between them is now the widest we have measured.
Eighteen months, two markets
Closed print runs are absorbing the money that new product is losing
Rank all 208 tracked sets by 30-day change and the pattern is unmistakable. Every one of the top gainers is a set whose print run closed years ago, most of them small, oddball supply: 2016 McDonald's promos (+56.8% in 30 days, +97.2% over 90), EX Trainer Kits (+29.8%), 2004's Team Magma vs Team Aqua (+21.3% on a $4,277 set value), Diamond & Pearl promos (+16.0%). Every one of the top decliners is current-era product still coming off the presses: ME04 Chaos Rising −23.0%, ME03 Perfect Order −18.5%, and SV 151, the most reprinted set of the Scarlet & Violet era, −14.3% on a $2,059 set value. This is the print-clock thesis behaving exactly as published: scarcity premiums hold only where the supply switch is off.
The four largest 30-day gainers and the three largest decliners, set values as of the July 21 reading
| Set | Set value | 30 days | 90 days |
|---|---|---|---|
| McDonald's Promos 2016 | $831 | +56.8% | +97.2% |
| EX Trainer Kit: Latias & Latios | $56 | +29.8% | +80.4% |
| Team Magma vs Team Aqua (2004) | $4,277 | +21.3% | +50.6% |
| Diamond & Pearl Promos | $3,855 | +16.0% | +35.4% |
| SV: Scarlet & Violet 151 | $2,059 | −14.3% | — |
| ME03: Perfect Order | $686 | −18.5% | — |
| ME04: Chaos Rising | $778 | −23.0% | — |
The single-card version of the same story
Liquidity is deep, and it is concentrating in modern chase art
The strongest graded demand among the cards with a deep sold tape is not vintage, it is the 2026 Victini twins, measured between our July 21 and August 3 readings. White Flare's 172/086 trades at a $1,499 PSA 10 median on 250 recorded sales and is up 12.1% on recent momentum; Black Bolt's 171/086 sits at $1,522 on 294 sales. Sample sizes that deep mean these are real two-sided markets, not thin-listing artifacts. The softness is equally specific: Tag Team alternate arts are giving back ground (Blastoise & Piplup GX −9.9% to $3,050, Arceus & Dialga & Palkia −8.7% to $4,750), and Prismatic's Umbreon ex 161/131, the modern era's flagship chase, eased 8.1% to $6,225. None of these are collapses; they are the market catching its breath where prices ran furthest ahead.
Graded movers · recency-weighted eBay sold medians
The 30th Celebration line is halving, and vintage packs don't care
The sharpest sealed moves this month are all one product line: 30th Celebration collections are down 46 to 49% in 30 days, the Ultra-Premium Collection to $755, the Sylveon ex Box to $102. Say plainly what those are, because it matters: 30th Celebration does not release until September 16, so every one of them is a preorder ask with no sold tape behind it. A preorder book repricing downward is not post-launch deflation of a product on shelves. It is a market discovering it overpaid for something that does not exist yet. At the other end, genuinely closed sealed supply keeps grinding up: a single Mysterious Treasures booster pack (2007) is $338, +38% in 30 days. Same lesson as the singles market, in starker form: launch-window sealed prices are a countdown, closed-print sealed prices are a clock that only runs one way.
All figures pulled August 3, 2026 from live moonstone market data. The moonstone Pokémon Market Index is our equal-weight index of tracked singles, each normalised to its own first reading inside a TRAILING 550-day window, which on this date begins February 3, 2025 (= 100). The base slides forward as that window does, so 249.8 is an eighteen-month reading rather than a fixed epoch and it will not reproduce from a later rebuild. The crypto benchmark is the Bitwise 10 Crypto Index, indexed to the same scale. Set values and 30/90-day changes are computed across all 208 tracked sets from daily TCGplayer market prices, which are ASKS; on this pull the provider had not updated most sets since July 21, so the set table is as of that reading rather than as of the pull date. Graded figures are recency-weighted eBay sold medians with the number of recorded sales stated in plain terms, and every graded move quoted here is measured between our July 21 and August 3 readings; per-grade medians are computed from individually verified sold listings after filtering to the exact card and printing. Sealed prices are live TCGplayer listings. The weekly top-picks book began July 30, 2026 and is too young to report returns on, it will appear in these reports once it has a measurable history. Single-figure moves on thin samples are excluded rather than smoothed. Linked product pages price live and will drift from captured figures. Nothing here is financial advice.