Every graded-card investor carries the same mental model: population is supply, supply is gravity, and a PSA 10 with fifty thousand siblings should be worth less tomorrow than it is today. We finally have the dataset to test that model properly. Moonstone tracks daily PSA registry snapshots for 539 cards; we sampled 115 of them, from 512-gem sleepers to the 77,674-gem Umbreon ex, and priced each one through the same verified eBay comp filter that powers our card pages. The question is simple. Do the giants underperform, and is what we see expected?
The twelve biggest PSA 10 registries we track
Ranked by PSA 10 population · prices from verified eBay sold comps, August 4, 2026
| Card | PSA 10 pop | PSA 10 price | Premium over raw | 75-day trend |
|---|---|---|---|---|
| Umbreon ex · JP Terastal Festival | 77,674 | $650 | 1.5x | +23% |
| Charizard V · SWSH Promos | 50,825 | $89 | 7.4x | +11% |
| Pikachu with Grey Felt Hat · SVP | 49,647 | $2,999 | 2.7x | −6% |
| Charizard VMAX · Shining Fates SV | 36,012 | $290 | 2.0x | +14% |
| Charizard · Celebrations Classic | 35,423 | $580 | 3.3x | +10% |
| Charizard · JP 25th Anniversary | 33,205 | $540 | 1.5x | — |
| Mewtwo VSTAR · JP VSTAR Universe | 33,159 | $260 | 1.8x | +50% |
| Eevee · SVP Black Star Promos | 30,763 | $140 | 10.1x | +36% |
| Umbreon · JP SV-P Promo | 28,635 | $260 | 1.2x | +63% |
| Charizard ex · 151 | 28,275 | $1,518 | 4.2x | +9% |
| Blastoise · Celebrations Classic | 27,625 | $135 | 10.5x | +13% |
| Mewtwo VSTAR · Crown Zenith GG | 25,975 | $550 | 1.9x | +30% |
Read the premium column before the price column. The registry king, Umbreon ex from Terastal Festival, commands just 1.5 times its raw price in a PSA 10, because 87% of submissions gem and everyone knows it. Meanwhile the SVP Eevee, sitting on a 30,763-gem registry of its own, trades at 10.1 times raw. The difference isn't demand. It's that only 38% of Eevee submissions come back as 10s, so the slab genuinely filters condition while the Umbreon slab mostly certifies the obvious.
Population is gravity, but it pulls on the premium, not the price
Across the 48 sampled cards with at least 20 verified PSA 10 comps, the correlation between registry size and the gem premium is −0.46. Sort the sample into population quintiles and the compression is unmistakable: cards with populations under 3,724 carry a median 12.7x premium over raw, the next quintile 7.4x, and everything above 13,695 flattens to roughly 2 to 3x. The grading-difficulty channel behaves exactly as theory says too: cards that gem at an 80%+ rate carry a median 2.2x premium, while sub-50% gem rates command 4.2x. So yes, this half is expected. The market prices supply and prices grading risk, with a discipline that would look familiar to any bond desk.
Median PSA 10 premium over raw, by population quintile
The giants are not falling
Here is the half that should surprise you. Comparing each card's verified PSA 10 median over the last 75 days against its prior six months, 42 of 44 cards with enough comps in both windows are up, and the median gain is 16%. The biggest-population quintile, the cards drowning in supply, still posted a median gain of 23%. The correlation between population and momentum is only −0.31, a mild drag rather than a ceiling. Part of this is the tide: our Pokemon index has been climbing all summer, and a rising market lifts high-pop slabs along with everything else. But the naive model predicts the giants should lag badly or fall outright while fresh gems mint by the hundreds every week. They aren't.
The exception proves the mechanism. Pikachu with Grey Felt Hat is the only card in the top twelve that declined, off 6% over the window at $2,999. It has the worst combination in the table: a 49,647-gem registry that is still growing fast, a 43% gem rate too high to make the slab scarce, and a price tag that priced in Van Gogh Museum nostalgia at the top. When supply pressure finally shows up in this dataset, it shows up exactly where theory says it should: the card where minting is fastest relative to the demand that remains.
The cards doing the arguing
Population figures are PSA registry snapshots taken August 3-4, 2026 across the 539 cards Moonstone tracks daily; we sampled the top 30 by PSA 10 population plus a stratified sixth of the remainder (115 cards). Prices are medians of individual eBay sold listings passed through Moonstone's verified comp filter (lot, reprint, autograph, and wrong-set listings removed; bulk-flood collapse applied). The 75-day trend compares the verified PSA 10 median of the last 75 days against the prior 165 days and is shown only where both windows hold 8+ comps; aggregate correlations use the 48 cards with 20+ verified comps overall. Premium is the PSA 10 median divided by the card's TCGplayer raw market price. One top-12 card (the JP 25th Anniversary Charizard) has too few recent comps to trend and is shown unrated.