Do Sealed Pokemon Boxes Go Up in Value?

Sealed Pokemon boxes are older on average than they are expensive, and the two facts are related, but not in the way the question assumes. Before any number: everything below compares different sets on a single day, and the old boxes still being quoted are the ones that were never opened, which were the minority. Nothing here follows a box through time or measures a return. With that said, across the 56 priced 36-pack English booster boxes in our sealed catalog on 2026-09-22, the rank correlation between a set's age and its box ask is 0.824. The median box ask runs from $290.44 for sets under a year old to $10,000 for sets past twenty. And the gap between the dearest and cheapest box inside a single age band widens from 2.26x in the youngest band to a peak of 31.91x at six to ten years, which means age tells you roughly where a band sits and almost nothing about where a box sits inside it.

Read down the median column and the answer to the headline question looks like an emphatic yes. Read across to the two end columns and it stops looking like an answer at all. In the six-to-ten year band the cheapest box is Darkness Ablaze at $367.47 and the dearest is a Team Up box at $11,725. Both sets are from the same corner of the hobby's history. One of them is a set people want sealed and one is not, and nothing about the calendar separates them.

The two youngest bands are tight, because a box that is still being printed is priced against a live wholesale supply and there is nowhere for it to go. The under-one-year band spans 2.26x end to end and the one-to-three band spans 1.78x. The three-to-six band spans 8.49x and the six-to-ten band spans 31.91x. That is the shape of a market where a handful of sets are collected sealed and the rest are simply old. The widening is not monotone, and we do not claim it is: the sequence across the six bands runs 2.26x, 1.78x, 8.49x, 31.91x, 14.42x, 28.00x. It peaks in the fourth band, not the last, and the last band's 28.00x is an artifact rather than a finding, because its floor is one of the two broken Base Set rows discussed below. Drop those two and the oldest band spans 1.40x, the tightest of all six.

Evolving Skies at five years old asks 1.46 times the median box in the six-to-ten year band, one band older than its own, and the reason is one card: Umbreon VMAX at a $2,283.14 ask. The box is priced against its chase card, not against its calendar, and that mechanism is measured separately in In 41 of 48 Sets, the Best Card Costs Less Than the Box.

The chase card is not the whole story, though, and the fourth row is the reason we print the full ranking rather than the cases that fit. Plasma Freeze asked $4,800 on 2026-08-24 and 2026-08-31 and then $14,999 from 2026-09-07 onward, a 3.12x step inside our own window, held flat for fifteen straight readings since. No card in that set explains it. A single relisting at a new number is indistinguishable in our data from a revaluation, and one such row is enough to put a box fourth on this table.

There is a second reason to be careful with the top of the age table, and it is measurable from our own daily sealed tape. Take the 56 boxes of this study and ask, for each age band, how many of them are quoted often enough to enter a movement panel at all. The answer is the survivorship warning restated as a number.

Read the last row. Not one of the five boxes in the twenty-and-over band is quoted often enough to enter the panel, and three of the five were last read before our window opened, on 2026-08-19, 2026-08-21 and 2026-08-24. The $10,000 median that answers the headline question is computed from five asks, not one of which our own tape can show changing. In the ten-to-twenty band only three of nine boxes clear the panel rule, two have no history row at all and three more fell off the tape before the window. The age curve is real in the sense that the numbers are what the catalog says, and it is least verifiable exactly where it is most dramatic.

The same pattern shows up across the whole sealed catalog rather than just these boxes, which is the wider version of the finding. Over the 23 days from 2026-08-30 to 2026-09-21, sealed products asking $2,000 or more repriced on a median 4.5% of day-to-day steps and 28.6% of them never changed price once. Sealed products under $25 repriced on 19.0% of steps and only 6.3% never moved. That table is about price, not age, and the two are different questions; we report it because price and age are correlated here at 0.824, and because it is computed on a much larger population than 56 boxes. Note that it is a population of sealed products of every format: 9 of the study's 18 boxes asking $2,000 or more are absent from it, and five of the fourteen products in its top band are not boxes at all.

Three named checks make the point concrete. The Neo Genesis box has six readings inside the window and all six are $14,000, and those six run 2026-09-14 to 2026-09-20 rather than spanning the window. The Gym Heroes box has nineteen readings and all nineteen are $10,000. Evolving Skies, by contrast, has twenty-two readings carrying ten distinct prices. The age curve is what the catalog says, but the far end of it is made of the least quoted asks in the catalog.

One pair is worth resolving rather than leaving open, because our own tape resolves it. The Base Set Revised Unlimited box asks $499.99 while a single loose pack from that same edition asks $888.91, and those two cannot both be live. The pack carries 29 readings from 2026-08-23 to 2026-09-20 at eight distinct prices, rising steadily from $823.35 to $888.91. The box carries exactly one reading ever, $499.99 on 2026-08-21, and has not been read since, two days before the catalog-wide tape begins. The live figure is the pack and the stale figure is the box. The Base Set (Shadowless) pair is the same shape and starker: that box has two readings, both $499.99 on 2026-08-18 and 2026-08-19 and nothing since, against a 1st Edition pack quoted at $15,000 on all 23 days of the window. Both boxes carry the identical figure to the cent, which is itself the tell. We leave them in the age table because removing rows for looking wrong is how a cross-section becomes a curated one, but no reader should treat either as a market price.

This study answers the sealed half only. The singles half, whether a set's cards gain value with age, is a separate measurement on a separate population, and the two halves are worth reading together because they need not agree. Sealed supply is destroyed by opening and singles supply is created by the same act, so a set can have an expensive box and a cheap checklist at the same time. Vivid Voltage and Evolving Skies are nine months apart in age and 7.69x apart in box price.

Survivorship is the central limit and it cannot be corrected from our data. The boxes we price are the boxes still sealed and still listed. Every box opened between release and today has left the sample, and opening rates are highest on the sets whose contents were most wanted, which is the same population that would show the largest gains if it had survived. The cross-section therefore measures the price of surviving supply, not the return on a purchase.

These are asks from one marketplace on one day, and on the oldest boxes they are asks that have not moved in weeks or are not on the movement tape at all. The dearest box in the study, Legendary Treasures at $18,750, has no sealed history row on file of any kind, so we cannot say when its ask last changed or whether the listing is still live. Three others have a row but no reading inside our 23-day window: the Base Set Revised Unlimited box last read on 2026-08-21, the Base Set (Shadowless) box on 2026-08-19 and the Gym Challenge box on 2026-08-24. They did not fail to be recorded; they fell off the tape. The sample is 56 SKUs across 54 distinct sets, because two sets carry two qualifying box SKUs each, Journey Together and Mega Evolution, each of which sells a standard and an Enhanced box.

The 20-and-over band holds five boxes and two of them are the stale Base Set pair at $499.99. Removing them raises the band's median from $10,000 to $13,999.99 and collapses its range from 28.00x to 1.40x, so the headline understates the level and badly overstates the dispersion in that one band. We report the band's range alongside its median for that reason, and readers should treat the oldest band as five asks rather than as a measurement.

Medians take the upper-middle value on even samples throughout, and in one place that convention moves a headline. The under-one-year band holds six boxes, so its published median of $290.44 is the upper of the middle pair; the lower-middle value is $241.16. The 34.4x multiple would be 41.5x on the other convention. We state the convention rather than switch it, but the reference denominator of this study is the one figure it materially affects.

We have no MSRP series, so the under-one-year median of $290.44 is an observed current-print ask, not the price a box sold for at launch. Every multiple in this study is computed against that observed reference and is a statement about the cross-section, never about a return. On windows: our sealed tape holds scattered readings from 2026-08-01, reaches catalog-wide coverage on 2026-08-23, and the longest per-product series in it is 29 readings. The 23-day window used for every movement figure here, 2026-08-30 to 2026-09-21, is the window the singles tape also covers, chosen so both studies in this wave measure the same days. It is a choice, not a limit imposed by the sealed tape.

A recorded series of sealed box prices running years rather than weeks, showing that individual boxes did not in fact appreciate as they aged even though older boxes ask more today, would overturn the reading of the level curve while leaving the cross-section itself intact. That is the single most valuable observation anyone could add here and we will have it ourselves in a few years. Separately, if the within-band dispersion narrowed toward the 2.26x of the youngest band as more boxes entered the six-to-ten band, the claim that age sets a level and not a price would fail. And a sold-price panel on vintage boxes landing far below the stale asks would change the far end of the age curve materially.

Universe: every row of sealed_products with price_usd > 0 and packs = 36 that classifyProduct() calls a Booster Box and does not flag as a lot. That is 57 SKUs on 2026-09-22. Release dates come from the api_cache row report:spread:setsmeta, which holds 220 TCGplayer set names of which 201 carry a release date; the join is exact on sealed_products.set_name against that name, with no normalising, and 56 of 57 SKUs matched. The one exclusion is the Delta Reign Booster Box, whose set releases 2026-11-06 and would carry a negative age. Age is (2026-09-22 minus release date) divided by 365.25. Bands are half-open in years, and no box in the sample lands on a band boundary. Nothing in the pipeline enforces a language filter; "English" describes what all 56 matched rows turned out to be, verified by inspection, rather than a predicate in the query. The correlation is Spearman's rho on the 56 matched rows. Medians take the upper-middle value on even samples, which matters once: the under-one-year band's median is $290.44 on that convention and $241.16 on the lower-middle. The staleness tables are computed from a different pipeline: the per-product daily sealed histories in api_cache under sealedhist:<tcgPlayerId>, restricted to NON-LOT products with a reading on at least 20 of the 23 days from 2026-08-30 to 2026-09-21, counting a step as moved when consecutive readings differ at all. That restriction takes the 1,800-product panel to the 1,428 non-lot products the price-band table counts. The by-age-band table applies the same reading rule to this study's own 56 boxes instead, which is why its rows are small. Both pipelines carry ASKS. Scripts: apps/web/scripts/sealed-c2-outofprint.mts and sealed-d4-final.mts. All figures and queries are in apps/web/.scratch-sealed/figures.json.