Card Grading Explained

On the night of February 21, 2026, a T206 Honus Wagner in a PSA holder marked “1” sold for $5,124,000 at Goldin. The grade, the lowest on the scale, described a card with rounded corners, creasing and a century of handling. The price described something else: a documented chain of custody back to a 1909 cigarette pack, a label that said the card was real, and a plastic case that let a buyer in another city trust both without touching it (Sports Collectors Daily; MLB.com, February 2026). The number on the label was almost irrelevant. The label itself was worth millions.

Two months earlier, in December 2025, PSA had graded a single card 58,000 times in one month: a Japanese McDonald’s promotional Pikachu given away with children’s meals, submitted at $16–25 a card through PSA’s Value Bulk tier and GameStop drop-offs by people who had worked out that a PSA 10 of it would sell for more than the fee (GemRate via SI Collectibles, January 2026). Roughly two in five of everything PSA touched in 2025 came back a 10, and about half of its Pokémon. Each of those 10s was, in the language of the market, “gem mint,” the same word applied to the Jordan rookie that gems about one time in a hundred.

Those two labels are the whole of this guide. One is a certificate of authenticity on an object whose supply was fixed in 1909; the other is a manufactured unit of a product whose supply is fixed only by the grader’s throughput and the submitter’s willingness to pay. Both say PSA, and both trade as if the plastic and the digit were the asset, because in a real sense they are: the grade is what buyers price, the population report is what defines scarcity, and the holder is what a stranger will pay for sight unseen. The hub’s flagship guide, Investing in Sports Cards, covers the market from the T206 Wagner to the $1.25 eBay median; this one goes deeper on the machine that assigns the digit.

Third-party grading did not describe an existing scarcity, it created one, and that mechanism is what the rest of this guide depends on: what you buy in a slab is a bundle of four services whose value is set by the market’s trust in the company that sold them.

A trading card is printed cardboard whose condition varies continuously: a shade of centering, a soft corner, a print line under the gloss. Before 1991 that variation was priced by negotiation between two people who could both see the card. Grading replaced the negotiation with a digit. A company examines the card, assigns a number from 1 to 10, seals it in a tamper-evident holder with a label carrying the number and a serial, and adds the result to a public census, the population report. From then on the card trades as “a PSA 8,” not as “a 1957 Topps Mays with slightly soft corners,” and the buyer never opens the case.

You are paying for four things at once. Authentication: a competent stranger has said the card is genuine and unaltered. Condition: the same stranger has placed it on a scale every other participant uses. Custody: the holder makes the card harder to damage and its condition harder to dispute. And liquidity: because the first three are standardised, the card can be sold to a buyer who has not seen it, at a price discoverable from other sales of the same card at the same grade. The fourth service turned cards into an asset class, because it is what allows price databases, indexes and funds to exist. None of them can price soft corners. All of them can price a PSA 8.

The consequence is the guide’s central claim: grading manufactures the scarce thing. The raw supply of a 1986 Fleer Jordan was fixed in 1986. The supply of the asset the market actually prices, a Jordan certified Gem Mint 10 by PSA, is decided one card at a time by a company in California, published on a website, and has grown every year since 1991. For a modern card the point is sharper: the raw print run is undisclosed and effectively unlimited, so the gem population is limited only by how many people pay the fee and how strict the grader is that day.

The digit is not a measurement of the card so much as a unit of supply, and the company assigning it is a central bank for a currency whose issuance it need not disclose. That is why the market’s trust in the grader is the entire asset: a PSA label is worth more than anyone else’s not because PSA is more accurate but because PSA holders are what the most buyers bid on, and liquidity begets liquidity.

The marketing treats “a PSA 10” as one kind of asset when it is at least three, so what a slab has done for its owner over a full cycle has to be read in parts, separating the premium that survives from the premium that evaporates.

Start with what has held. On cards whose raw supply was fixed decades ago and whose gem population is measured in hundreds, the grade premium has been the most durable source of return in the market. The 1952 Topps Mantle in PSA 9 went from $282,588 at Memory Lane in December 2006 to $5.2 million in a January 2021 private sale, about 23% a year for fourteen years; the SGC 9.5 copy made $12.6 million at Heritage on August 28, 2022. The same card in PSA 8, population 35, traded between $1.02 million and $1.83 million in the year to June 2026 and reached $3,721,000 at Heritage on August 21, 2026 for a copy with an added quality designation (Heritage; Goldin; Sports Collectors Daily).

The 1986 Fleer Jordan PSA 10 sold between $260,000 and $317,000 through 2025 (Athlon), while Card Ladder’s sales record for the PSA 9 runs from $29,100 to $51,730, an average of about $36,800 and a last sale of $43,900 on June 2, 2026; the $8,000–12,000 sometimes quoted for a Jordan PSA 9 is the 1986 Fleer sticker, not the #57 rookie. On these cards the digit is the asset, and it has compounded.

Now what has not. A 2019 Prizm Silver Zion Williamson PSA 10 sold for more than $6,000 in late 2020, by a breaker’s account; by late 2025 the same card in the same holder was clearing between $185 and $400 on eBay (PSA Auction Prices Realized; Ultimate Sports Breaks, 2025). A 2018 Prizm Silver Luka Dončić PSA 10 crossed $2,600 at the 2021 peak and last sold for $1,532 in May 2026 (SlabStox; Card Ladder).

Neither player got worse. The PSA 10 population of each grew every week people submitted, and a card that gems a third of the time from an unlimited raw supply has an unlimited PSA 10 supply. The label did not lose its meaning; it was applied to more objects. The hub’s guide to Investing in Modern Basketball and Football Cards walks through five of these round trips; the mechanism is this guide’s subject.

The third kind of PSA 10 was never scarce even on the day it was graded: the McDonald’s Pikachu, the 2023 151 Charizard ex (about 96,000 graded, roughly 27,400 tens, per a 2026 community population snapshot), any card whose submissions are counted in tens of thousands. For these the slab is a $25–80 retail product, and the “PSA 10 premium” is close to the fee plus a margin for the wait.

Against the index and against stocks

There is no long-run index of grade premiums as such; Card Ladder’s indexes, which track specific cards at specific grades, are the closest thing. Its CL50 fell 23% in 2022 and 9% in 2023, then rose about 28% in 2025, while the S&P 500 returned about +53% on a total-return basis over 2022–2025 (Card Ladder via SI Collectibles and Yahoo Sports; S&P Dow Jones Indices). A blue-chip card index roughly matched cash over four years in which stocks returned half again what you put in, and it tracks “the PSA 10” of cards whose 10-population grew throughout, so it measures a larger asset class at the end of the window than at the start. That is the baseline against which every grade premium should be judged.

A slab is a claim on the reputation of one company, and the five reputations are not equal, so it matters who assigns the digits, where they came from, who owns them and how the volume divides between them.

PSA, Professional Sports Authenticator, was launched in July 1991 by David Hall as a card-grading arm of his coin grader, PCGS, and by the accounts of people who worked there nearly closed in its first years for lack of submitters; Joe Orlando, who had his own first card graded around 1992–93, became president in 2002 (Sports Collectors Digest; Wax Pack Gods). Its parent, Collectors Universe, was listed until February 8, 2021, when a group led by the collector and entrepreneur Nat Turner, with D1 Capital Partners and Cohen Private Ventures, Steve Cohen’s family office, completed a tender at $92 a share, about $853 million, after raising an initial $75.25 offer; co-investors included the Chernin Group and athletes including Kevin Durant (SEC filings; Front Office Sports; Sports Collectors Daily, February 2021). The company, renamed Collectors, is private and publishes no financials.

SGC, Sportscard Guaranty, was founded in 1998 in Florida and built its name on vintage: the $12.6 million Mantle and the three most expensive Wagners on the public record sit in its black holders. On February 29, 2024, Collectors announced it was buying SGC for undisclosed terms, with SGC keeping its Boca Raton office, management and brand; at the time GemRate had PSA at 78% of the market, CGC at 10%, SGC at 7% and Beckett at 4% (Sports Collectors Digest; SI Collectibles; ICv2, February 2024). Since then the two most trusted names on vintage have had one owner.

BGS, Beckett Grading Services, opened in 1999 as an extension of the price-guide publisher founded in 1984, and for a decade its four sub-grades and Black Label were the premium name in modern cards; it graded 824,000 cards in 2025, 3% of the market.

CGC Cards is the trading-card arm of the Certified Collectibles Group, the comic and coin grader majority-owned since 2021 by funds managed by Blackstone’s Tactical Opportunities business, in a deal that valued the group above $500 million (Blackstone); it took the Pokémon overflow when PSA’s queues lengthened and graded 4.92 million cards in 2025, up 121%, for 18%. TAG Grading, the newest entrant, grades with computer vision plus human verification and reports a 1,000-point score beneath its grade; it graded 440,000 cards in 2025, up 83%, for 2% (GemRate, January 2026; Athlon, 2026).

Three things about that chart matter more than the ranking. The concentration: PSA had 72% of all grading in 2025, 76% of sports and 69% of games, and with SGC’s 13% of sports under the same owner, Collectors assigned roughly nine of every ten sports-card grades (GemRate shares, January 2026). The mix: 16.8 million of the 26.8 million cards were games and non-sport, up 95%, while sports fell 12%; by volume the grading industry is a Pokémon industry that also grades sports cards, and its fees and queues are set by Pokémon demand.

And the direction of the second tier: SGC fell 24% in 2025 and in August 2026 graded about 39,000 sports cards, 66% fewer than a year earlier (GemRate via Bleacher Nation, September 2026), while CGC more than doubled. The market is choosing a grader for liquidity, and liquidity is choosing PSA. The two companies that assign nine in ten sports grades have one owner, a private group whose backers include a hedge-fund founder’s family office, and the third-largest grader belongs to a private-equity firm. None is listed, none publishes revenue or its gem-rate policy, and each earns its fee whatever grade it assigns.

The price cliffs described later sit on definitions most buyers have never read: what the digit measures, how the companies’ scales differ, and what the letters after a grade mean.

Every grader scores four attributes: centering (the ratio of the borders, front and back), corners, edges, and surface (creases, print lines, scratches in the gloss, stains, dents). The grade is a judgment, by a person, on the combination, and the companies publish standards rather than formulas. PSA’s published standard for a Gem Mint 10 is centering within roughly 55/45 on the front and 75/25 on the back with corners, edges and surface essentially perfect under magnification; a Mint 9 allows about 60/40 front and 90/10 back with one minor flaw (PSA grading standards, which add that graders may allow a small variance for eye appeal). Below 9 the standards loosen a step at a time down to 1, Poor, the grade on the $5.1 million Wagner; PSA also awards half-points from 1.5 to 8.5.

Beckett reports the four attributes separately as sub-grades and stratifies its top: a BGS 9.5 is Gem Mint, a BGS 10 is Pristine, and a Black Label 10 scored 10 on all four, rare enough to command a large premium over a plain 10 (Beckett; CardGrade.io, 2026). Every price database translates a BGS 9.5 as a PSA 10; the translation is imperfect and the two do not trade at parity. SGC uses PSA’s scale and half-points; CGC grades on a 10-point scale with a Pristine 10 above Gem Mint; TAG adds a 100-to-1,000 score on its higher tiers to separate a low 10 from a high one (TAG; PreGradeCards, 2026).

Qualifiers, Authentic and the autograph loophole

When one attribute is far below the others, PSA can assign the grade the rest of the card earned and append a qualifier: OC (off-centre), MC (miscut), MK (marks), ST (stained), PD (print defect). The market treats a qualified grade as two or three points lower; secondary guides put it at 30–50% of the clean price, and on trophies the discount is larger. The $2.1 million Wagner sold at Goldin in 2013 was a PSA 5 (MC).

Two labels carry no number: “Authentic” means genuine but ungraded, usually because the card was trimmed or recoloured; the $5,856,000 and $3,599,000 Wagners sold at Heritage in 2026 were SGC Authentic. And for years PSA’s autograph service would encapsulate a signed card, authenticate the signature and label the card simply “Trading Card” without authenticating the card; counterfeit 1990 Topps Frank Thomas “No Name On Front” error cards with genuine signatures sold in PSA holders for more than $1,000 each before PSA said it would close the loophole (SI Collectibles, 2024). A holder tells you only what the label says it examined.

The fee schedule and the queue are the supply constraint on every gem grade in the market, so here is the cost of getting a card into a holder as of September 2026, then the two episodes, 2021 and 2026, in which the largest grader could not process what it was sent.

Grading is priced by tier, and the tier is set by the declared value of the card, which caps the grader’s liability, and by the turnaround you pay for. PSA raised its five entry tiers by $5 on February 10, 2026, to $24.99 Value Bulk, $32.99 Value, $49.99 Value Plus, $64.99 Value Max and $79.99 Regular, which covers a declared value to $1,500 (DraftKings Network; SI Collectibles). Above that, 2026 fee summaries put Express at $149, Super Express around $299–349 and Walk-Through at $599, with Premium tiers into four figures, plus about $10 handling per order and insured return shipping from about $15 (CardGrading.app; The Card Shop Finder; the summaries disagree on where each declared-value band ends, so read PSA’s own page before submitting).

On June 2, 2026 PSA paused all four Value tiers, and Regular became the entry price for a single card. On September 14, 2026 PSA renames Regular as Priority, still $79.99 with an estimated 70–80 business days and a $1,500 insured maximum, and opens a Standard tier at $59.99 with an estimated 90–100 business days and a $1,000 maximum; the Value tiers stay paused, and a reholder costs $12.99 (PSA, September 2026; SI Collectibles).

The others are cheaper on paper. CGC’s tiers are $17 Bulk (25-card minimum) and $20 Economy from March 24, 2026, and $55 Standard, $100 Express and $300 WalkThrough from January 6, 2026, with turnaround targets of 40, 20, 10, 5 and 2 business days (CGC Cards; CardGrade.io). Beckett lists Base at $14.95, or $17.95 with sub-grades, at 75-plus business days, Standard $34.95, Express $79.95 and Priority $124.95, but paused new online Base and Standard submissions on August 5, 2026, citing submissions up 102% in a year, with reopening expected from September 15 (Beckett; Sports Collectors Daily; CardGrade.io).

SGC’s base tier was $15 until September 2, 2026, when it raised it to $50 to cut turnaround, calling the structure temporary and saying further changes should be expected between late 2026 and early 2027; it reported 2,000 new submitters since July 1, and Bleacher Nation added that submissions were up 24% (Sports Collectors Daily; SI Collectibles; Bleacher Nation). TAG’s economy tiers were frozen at capacity in mid-2026; Priority was $149 at 15–30 business days (TAG; Athlon).

Cheaper is not the same as liquid: a modern card in a CGC or SGC holder sells into a smaller buyer pool than the same card in a PSA holder, and the price gap is usually larger than the fee saving, which is why the flow keeps going to the most expensive grader. And the fee is only part of the ticket; the rest is the time value of a card that spends 40 to 100 business days in a queue while its price moves.

The 2021 and 2026 backlogs

In March 2021, at the top of the pandemic boom, PSA suspended its Value, Regular and Express services, about 98% of its submissions, because it had received more cards than it could unbox; its backlog had grown from about 3 million cards to roughly 12 million, and the company said it would be disingenuous to keep accepting cards it could not process (PSA customer update, April 2021; Sports Collectors Daily; SI Collectibles). The suspension lasted about three months. Cards submitted at the peak came back into a market that had already turned, so a whole cohort of modern PSA 10s arrived as sellers into the 2022 bust, and the overflow went to SGC, CGC and Beckett, which is when CGC’s Pokémon business began.

The second episode was larger and better documented, because this time PSA published the number. On May 14, 2026 Collectors announced a $200 million investment in capacity; submitters responded with a 20% spike that added about 1.6 million cards to a backlog already approaching 10 million, and on June 2 PSA paused the Value tiers, extended Collectors Club memberships and launched a public tracker updated every two weeks (PSA service-level update, May 2026).

The tracker’s first full count, on June 9, put the backlog at about 14 million cards rather than 10 million; it read 12 million on June 30, rose to 12.4 million on July 28 with convention-season submissions, then fell to 11.85 million on August 11 after a record grading month, 10.9 million on August 25 and 9.9 million in early September, against a target of 5 million that PSA said would take five to six months, with one paused tier possibly reopening before the end of September (PSA Backlog Tracker via Athlon, Yahoo Sports, SI Collectibles, Heavy, PreGradeCards and Sports Collectors Daily). Behind it: PSA graded 2.17 million cards in March 2026, up 47% on a year earlier, and 2.065 million in May, two-thirds of them games (GemRate via SI Collectibles).

The lesson of both backlogs is the same. The gem supply of every modern card is rationed by one company’s throughput; when that throughput fails, the cheapest tiers close first, the entry fee triples for the marginal card, and the cards already in the queue come out months later into whatever market exists then. A grader’s queue is a supply pipeline with a lag, and the lag is longest exactly when demand is highest.

Every grader sells a set of secondary services, each legitimate, each a way to run the same card through the grader more than once, and together they are why a population report over-counts the low grades of every card that has ever been worth the fee.

A crossover is a card in one company’s holder submitted to another, almost always PSA, with a minimum grade stated. PSA grades the card in its existing holder; if the result meets the minimum it cracks the case and re-encapsulates; if not, the card returns untouched, and the full fee is charged either way (PSA crossover terms, 2026). Because the market pays more for a PSA label than for a BGS 9.5 or SGC 10 of the same card, high-grade cards cross over steadily, and a card that fails to cross is one PSA has said it would grade lower than its label claims.

A reholder replaces an older PSA case with the current one without regrading, for $12.99 a card at PSA’s 2026 price; buyers pay a little more for the newest holder, partly because a few generations of slab are rumoured, never confirmed, to have been graded more loosely. A review asks PSA to look again at a card in its own holder for an upgrade; the fee is charged whether or not the grade moves.

The fourth service is not sold by anyone: the submitter cracks the case, discards the label and sends the card in raw. Nothing prevents it. The card gets a fresh look from a different grader on a different day, and because grading is a human judgment on a continuum, a card at the top of a 9 has a real chance of returning a 10. On a card where that step is worth tens of thousands of dollars it is rational to try several times at $80–150 a try.

The consequence is in the census. The population report does not decrement when a holder is cracked, so every 9 that became a 10 on its second trip is still counted as a 9. The reported population of 9s of any expensive card is an over-count of unknown size, the reported population of 10s is accurate, and the ratio between them, which is what the market uses to price the cliff, is tilted toward making the 10 look rarer than it is.

PSA’s technical answer is fingerprinting. In April 2021 it acquired Genamint, whose software photographs and measures every submission and matches it against prior scans, so a card that was a 9 on one trip and arrives raw a year later is recognised, along with a trimmed edge or a recoloured corner (PSA; PR Newswire, April 2021). How PSA uses the recognition, whether a resubmission is graded independently, held to its prior grade or flagged, is not published, and the fingerprint exists only for cards photographed since the system went in, so the census of every card graded before 2021 carries the over-count permanently.

The population report is the only prospectus this market issues, and reading it takes three numbers, a decade of one card’s census, and the figure that decides whether a gem grade is scarce at all, the gem rate.

A population report is the grader’s public count of every card it has graded, by set, card and grade; PSA’s is free and searchable. Three readings matter. The total is how many copies have ever been worth the fee to someone, which for a modern card is a floor on the print run and for a vintage card is most of the surviving supply. The count at your grade is the size of the asset class you are joining. The count at the grade above caps your price, because every buyer of your PSA 8 is a potential buyer of a 9, and the 9’s price is the ceiling on what an 8 can be worth.

The census only grows. In 2013 a collector posted PSA’s count for the 1986 Fleer Jordan: 13,302 graded, 2,096 of them PSA 9 (Collectors Universe forums, 2013). By 2025 the total was above 25,000, the PSA 9 count about 2,700 and the PSA 10 count about 330 (PSA population report via Athlon and collector forums, 2025); by September 2026 Card Ladder’s reading of the report was above 28,000 graded with 326 tens. The total nearly doubled in twelve years on a card printed in 1986; the 9s grew about 30%; the 10s stayed a rounding error.

That is what fixed supply looks like in a census: raw copies come out of closets at a rate set by the price, the low and middle grades absorb them, and the gem count barely moves because the gem rate on a forty-year-old card is about 1%.

The gem rate is the print run

The most useful number in grading is the share of a card’s submissions that come back a 10, the gem rate; the graders do not publish it per card, and GemRate, an independent tracker, compiles it from the population reports. Across everything PSA graded in the first half of 2025 the gem rate was 43%, down from 45% in 2024; games ran near 50% and sports 34%, with about 77% of game-card submissions printed in the 2020s (GemRate via cllct, 2025). By April 2026 a compilation of GemRate’s data put games at 60.8% and sports at 41.2% (CardGrade.io).

Set those beside the fixed-supply cards: the Jordan gems at about 1.3%, the 1999 1st Edition Base Set Charizard at about 2.3%, roughly 124 tens among 5,325 graded (Heritage, December 2025; Pikawiz, 2026), and the 2023 151 Charizard ex, the modern reprint of the same character, at about 28%.

The gem rate is the true print run of the asset you are buying. A card with a 40% gem rate and an unlimited raw population has an unlimited PSA 10 population, delayed only by the fee and the queue; a card with a 2% gem rate and a raw population fixed in 1999 will still have a few hundred tens in 2036. Two cards can carry the same holder and the same words, Gem Mint, and be opposite kinds of asset. Before you pay a grade premium, look up the PSA 10 population today and its growth over the last twelve months: a 10-population growing at 50% a year is inventory being revealed; one growing at 2% is an asset.

The population is the one input a buyer can check before bidding, and almost none do, so here is what a population of ten, of tens, of hundreds and of tens of thousands at a grade means for the price of that grade, using cards with public sales at each scale.

At a population in single digits, the grade is priced as art. The 1955 Topps Roberto Clemente has been submitted to PSA more than 7,400 times; eleven copies are PSA 9 and one is a 10. A PSA 9 sold privately through Heritage for $1.15 million in August 2025, the fourth seven-figure sale of the grade since 2022, while a PSA 7 last sold for about $20,665 (Sports Collectors Daily; Intelligent Collector; Sports Card Investor). Two points, a population of eleven against one in the hundreds, fifty-five times the price.

At a population in the tens, the grade trades a few times a year like a thin stock: the 1952 Topps Mantle in PSA 8, population 35, printed $1,380,000 (Heritage, August 2024), $1,020,000 (Fanatics Collect, 2025), $1.83 million (Goldin, June 2026) and $3,721,000 for a designated copy (Heritage, August 21, 2026), a range of nearly four to one on the same digit.

At a population in the hundreds the grade has a market: the Jordan PSA 10, about 330 of them, sold between $260,000 and $317,000 through 2025 in enough transactions to draw a line. At a population in the thousands it has a ticker, and the ticker mostly falls while the population grows: the Zion Prizm Silver PSA 10 went from more than $6,000 to under $400 as the fee was paid on copy after copy. At tens of thousands the grade is a retail product: PSA graded the McDonald’s Pikachu 58,000 times in December 2025, ten times any other card that month, and by SI Collectibles’ arithmetic about 6 million Pokémon PSA 10s were added to the market in 2025 (GemRate via SI Collectibles, January 2026).

The rule that follows is the most useful sentence in this guide. Price a graded card off two populations: the count at its grade, which is the supply you compete with when you sell, and the count at the grade above, which is the ceiling. A PSA 8 with 35 in the population and nine above it is scarce; a PSA 8 of a 1960s common with 800 in the population and 200 nines above it is not, whatever the listing says; a PSA 10 with 27,000 in the population is a $200 card whatever the character on it is worth in 1999 form. Then check the direction: a population that grew 50% in a year says the raw supply is still coming out and today’s price is a bid into a pipeline.

One step on the scale, from 9 to 10, carries most of the value in graded cards, and measuring it across vintage sports, vintage Pokémon and modern shows why the shape is the same everywhere even though the multiple is not.

Nothing about a graded card is less intuitive to a reader arriving from equities than the way price moves across the scale: not linearly with the digit, but in a cliff at the top. Below 8 the steps are modest and continuous; the 1952 Topps Mantle is $37,800 as a PSA 1 (January 31, 2026), $85,000 as a 1.5 (July 19, 2026) and around $134,000 as a 5 on PreGradeCards’ 2026 record, with sales of the grade from $108,000 (Fanatics Collect) to $198,000 (Sports Card Investor) (Sports Collectors Daily). From 8 to 9 the step becomes a multiple, and from 9 to 10 it becomes a different asset.

On vintage sports the Jordan is the reference: PSA 9 between about $29,000 and $52,000 on Card Ladder’s sales record (last $43,900 on June 2, 2026), PSA 10 between $260,000 and $317,000 through 2025, roughly seven to eight times for one point on a population of about 2,700 nines and 330 tens. On vintage Pokémon the 1st Edition Base Set Charizard traded at $30,000–50,000 in PSA 9 and $360,000–450,000 in PSA 10 across 2026 guides, a multiple near ten, with a record $954,800 for a PSA 10 at Goldin in February 2026 on a population of about 124 tens (Cards N Packs; Pokémon Pricing; Goldin; Heritage).

On modern sports the multiple compresses as the tens accumulate. Our September 2026 basket priced the PSA 10 median at 12.4 times the PSA 9, a wider step than the Jordan’s only because the medians mix cards with very different populations; on any heavily graded rookie the 10 population grows with every submission, and the step narrows with it.

Three reasons, and they compound. The population: for any card old enough to have been handled, the 10 is a different scarcity class from the 9, often by a factor of ten. The buyer: the marginal bidder on a 10 is a registry collector building a ranked set on PSA’s public Set Registry, or a fund assembling “the best,” and that buyer is price-insensitive in a way the buyer of a 9 is not. The option: a 9 carries the chance of becoming a 10 on resubmission, and the 10 is the ticket that has already paid.

When the 10 population is fixed all three hold and the multiple is stable for years; when it grows faster than the buyer pool, the first reason fails, the third inverts, and the multiple compresses. The cliff is also where the money is lost: every dollar of the multiple is a dollar a grader’s judgment can take away on a crossover, a review or a resubmission, and a dollar a forger can manufacture by changing one character on a label.

One set of numbers in this guide can be dated to the day: what Invest Alternative’s own collection engine recorded in the first week of September 2026, presented as ours and with the method stated.

We track a basket of key sports cards through PriceCharting and record the median price at each grade across the basket. On September 8, 2026 the PSA 10 median was $5,849, the PSA 9 median $473 and the ungraded median $295, a 12.4-times step from 9 to 10 and a 1.6-times step from raw to 9. Between September 1 and September 8 the PSA 10 median rose 1.0%, the PSA 9 median 4.5% and the raw median fell 3.6%. These are medians across different cards at each grade, not one card priced three ways, so the 12.4 is a property of our basket, not a market constant; the shape is what matters.

Our Pokémon basket, five pinned cards priced the same way, shows the same shape at a larger scale: a PSA 10 median of $5,257 against an ungraded median of $90 on September 8, about 58 times.

On the sales side, the TheCardAPI feed we ingest logged 84,928 eBay sports-card sales on September 7, 2026 at a median price of $1.25 and a day’s value of about $2.62 million. Four of the five largest sports-card sales we recorded from September 2 to 4 carried a grade in the listing: a 2020 Donruss Optic Downtown Tom Brady PSA 10 at $31,950, a 2009 Topps Chrome Stephen Curry /999 BGS 9.5 at $27,200, a 1996 Topps Chrome Kobe Bryant PSA 10 at $21,100 and a 1951 Bowman Mantle PSA 1 at $18,600; the fifth, a 1952 Topps Mantle at $18,769.69, was listed without a grade in its title.

The five largest Pokémon sales, $12,200 to $43,000, were all PSA 10s. Above five figures on eBay, the raw card barely exists. Our provisional IA Composite stood at 100.271 on September 8, 2026, up 5.74% over thirty days and 0.29% over a year; it is a work in progress and we quote it as ours.

Once the value moved from the card to the label, the label became the thing worth faking, and the fraud that follows the grade premium has evolved in three stages, each with named cases and dates, before what the graders have done and what you must do yourself.

The oldest fraud is the alteration: trimming edges to sharpen corners and centre borders, recolouring a corner, rebuilding a surface. Bill Mastro, the dominant auctioneer of his era, admitted in his October 2013 guilty plea that he had cut the sides of the most famous Wagner in the world, the PSA 8 Wayne Gretzky bought in 1991, with a paper-slicing machine before it was ever graded, and served 20 months (U.S. Attorney, Northern District of Illinois, 2015).

In spring 2019 members of the Blowout Forums documented, card by card, that dozens of high-grade PSA and Beckett slabs sold through PWCC had earlier sold raw or in lower grades and been trimmed or recoloured in between; the FBI subpoenaed PWCC that July (InsideHook; Sports Collectors Daily, 2019). The grader is the only line of defence, and it has been beaten, repeatedly, on six-figure cards, by a blade and patience.

The second fraud is the forged label, the direct product of the cliff. Between 2022 and May 2024, according to the U.S. Attorney for the Southern District of New York, Anthony Curcio and Iosif Bondarchuk cleaned the grade and serial off genuine PSA labels, printed new 9s and 10s and new serials, and sold the relabelled slabs, holding raw or lower-grade cards, through marketplaces, shops, auctions and card shows, using a front man to face the buyers and attempting more than $2 million of fraud.

Bondarchuk pleaded guilty; a Manhattan jury convicted Curcio of wire fraud and conspiracy on January 28, 2026 after a two-and-a-half-week trial before Judge Ronnie Abrams; sentencing was reported for May 2026 by the Justice Department and for October 2026 by KIRO 7, each count carrying up to 20 years, and no sentence had been reported when this guide was checked on September 9, 2026 (Department of Justice, January 28, 2026; KIRO 7). The scheme required no skill with cardboard. It required a printer, because the market was paying for the digit.

The third fraud is the slab itself. Community documentation in 2026 describes genuine PSA cases opened and resealed around worse cards, an acetone method that strips a real label’s ink while preserving its ultraviolet features, and counterfeit holders good enough to pass at a table; one collector documented a $28,000 loss on fake slabs at a single show. We have not verified the methods and cite them as reported (PreGradeCards; Card Source Canada, 2026).

PSA’s own count is in its 2025 Fraud Report, a 32-page document: submissions with a projected market value above $200 million intercepted as counterfeit or altered in 2025, a figure its leadership called conservative and said would exceed $1 billion at PSA 10 values; counterfeit submissions up 45.3% year on year, Pokémon up 125% and sports up 5.1%; altered genuine cards up 108.7%, altered Pokémon up more than 400%; and 112 of 181 submitted 1952 Topps Mantles fake (PSA 2025 Fraud Report via Sports Collectors Daily, SI Collectibles and ICv2). Read it as a company describing its own vigilance; it counts only what reached PSA.

What the graders have done, and what you must do

PSA bought Genamint in April 2021 to fingerprint every card it photographs and flag roughly seventy classes of anomaly, redesigned its holder, moved to a label whose ink fails if tampered with, runs a public certification lookup that returns the grade and, for recent years, a photograph of the card, and said in 2024 it would end the autograph practice that let counterfeits ride genuine signatures. Beckett and CGC run their own lookups. The graders publish these defences; they do not publish detection rates, and the fraud report’s headline counts what was caught.

The defence that works is procedural. Look up every certification number on the grader’s own site before paying and compare the card in hand or in the listing with the image there, every card, every time. Weigh a slab against a known-genuine one of the same generation. Above the eBay Authenticity Guarantee threshold, $200 from July 2026, buy through the guarantee, which routes the card through a PSA-operated authenticator. Above $5,000, buy from a house that will take it back. And on a vintage card whose high grade first appears in the population report after 2019 with no prior sales record in any holder, assume the 2019 pattern until proven otherwise. The graders’ fraud reporting is an annual report and a lookup page; the reporting that protects you is your own.

The grade premium is quoted gross, and everything you pay to reach, hold and realise it comes out of the same multiple, so the friction, the carry and the tax on a graded card add up here before the arithmetic in the next section.

Round-trip friction is the largest line. Selling on eBay costs 13.25% of the sale to $7,500 and 2.35% above, plus $0.30–0.40 an order; the buyer pays sales tax on top. At the houses the buyer’s premium is 22% with a $29 minimum at Heritage and 22% with a $19 minimum at Goldin under 2026 terms, and 20% at Fanatics Collect, which charges the seller nothing on Premier auctions; a seller’s commission of 0–10% is negotiated on top at most houses (eBay fee schedule; Heritage and Goldin 2026 terms; Fanatics Collect). Grading is a cost of entry when you buy raw and of exit when you cross over or reholder: $79.99 at PSA’s open entry tier in early September 2026 and $59.99 at the Standard tier from September 14, plus about $10 handling and $15-plus insured return shipping per order.

Carry is small but real. A slab needs 40–50% relative humidity, no sunlight and a stable temperature; the holder protects against none of those. Specialist insurance ran about $0.55–0.70 per $100 of value a year in 2026 quotes, falling with scale, and a homeowner’s policy caps collectibles at a few thousand dollars (Collectibles Insurance Services rates via Sports Collectors Digest). The eBay Vault, operated by PSA, charges 0.37% a year on cards above $50 with a 1% withdrawal fee capped at $50; Fanatics Collect stores platform purchases free.

Tax is the line most buyers discover at the end. A card is a collectible under Internal Revenue Code §1(h)(4), so the maximum federal rate on a gain held more than a year is 28%, against 20% for stocks, with the 3.8% net investment income tax under §1411 above the income thresholds and state tax after that; a gain on a card held a year or less is ordinary income.

Grading fees on a card held for investment are added to basis; keep the invoice. The wash-sale rule in §1091 covers stock and securities and does not reach cards. Form 1099-K reporting reverted for tax year 2025 onward under the One Big Beautiful Bill Act of July 4, 2025 to more than $20,000 in gross payments and more than 200 transactions, both conditions (IRS Form 1099-K FAQs, October 2025); the tax is owed whether or not a form arrives. The dealer, investor and hobbyist regimes are set out in the flagship guide.

The grading advertisements never do these sums: when it pays to grade a cheap modern card, when a vintage card must be graded whether it pays or not, and what one graded card returns after every fee, carry cost and tax over a five-year hold.

The $50 card

Take a current-year base rookie bought raw for $50, whose PSA 10 sells for $180 and PSA 9 for $45, and assume it gems at the April 2026 sports-card rate of about 40%. Selling raw on eBay nets $50 less 13.25% and $0.40, about $43. Grading at PSA Regular costs $79.99 plus a share of handling and return shipping, call it $88 a card in a ten-card order. The expected sale is 0.4 × $180 + 0.6 × $45 = $99, about $86 net of eBay’s fee. You spent $88 to turn a $43 outcome into an $86 one: a loss of about $45 against selling raw.

The general rule falls out of the same sums. With an $88 ticket, grading a $50 card beats selling it raw only when the PSA 10 comp is above about $310 at a 40% gem rate, $575 at 20%, and $220 at the 60% rate of modern Pokémon. At the paused Value Bulk tier, $24.99 plus handling, the same card needed a PSA 10 comp of only about $143. The June 2026 pause changed no card’s gem rate; it roughly doubled the price a modern PSA 10 must reach before grading it makes sense.

The $5,000 card, bought, held five years, sold

A vintage card bought raw for $5,000 is a different problem, because above a few hundred dollars the raw market barely exists: buyers assume a raw card at that price has been rejected by a grader or trimmed, and the houses will not catalogue it. Grading is not optional. A $5,000 declared value sat above PSA’s Regular limit of $1,500 in 2026 and in a tier the fee summaries put at about $300–350 (they disagree on where the Express band ends; we take the higher tier), plus insured shipping both ways, roughly $410 all-in, or 8% of the card.

The outcomes are asymmetric: a PSA 7 worth $5,000 means you paid $410 to make the card saleable; a PSA 6 worth perhaps $2,800 loses $2,600 on the digit; a PSA 8 worth $12,000 makes $7,000 on one grader’s judgment. That variance is why the dealer trade in vintage is pre-grading, and why an outsider should buy vintage already in a holder.

Assume it returns a PSA 7 the market values at $6,000 in the holder, a 20% lift for liquidity that is an assumption, not a measurement. Hold five years, insured at $0.60 per $100, about $36 a year, $180 in all. Assume 7% a year of appreciation, an assumption rather than a measurement, below Card Ladder’s vintage index in 2025 (+14%) and above its flat years: $8,415 at sale.

Sell on eBay: 13.25% and $0.40 comes to $1,115, netting $7,300. Your basis is $5,410, the purchase plus the grading fee. The gain is $1,890; at 28% the federal tax is $529, leaving $6,771, or $6,591 after the insurance. Against $5,410 out, that is $1,181, 21.8% over five years, about 4.0% a year, on a card that rose 68% on paper from raw price to sale price. With no liquidity lift the same sums give $305, about 1.1% a year; with a PSA 6 the trade loses money before tax; through a house at a 22% buyer’s premium and a 5% seller’s commission the eBay case’s $7,300 becomes about $6,550. The digit did all the work; the fees and the rate took most of the reward.

For a reader who has decided to put cards into holders, or to buy cards already in them, the sequence below is in the order that avoids the expensive mistakes.

1. Decide which side of the fee you are on. If you are buying for investment, buy graded, PSA or SGC for vintage and PSA for anything after 1980, and let the seller carry the grading variance. Submit only cards you already own raw, or where you can pre-grade.
2. Pre-grade before you pay: centering against the 55/45 and 60/40 standards with a ruler or a phone app, corners and edges under a loupe, surface under raking light. If the card is not a clear 9 candidate, do not pay $60–80 to be told so.
3. Run the arithmetic above for each card: gem rate from GemRate, PSA 10 and PSA 9 comps from PSA’s Auction Prices Realized or Card Ladder, the fee of the tier that is actually open, not the one that was open last year.
4. Pick the grader for the buyer, not the fee: PSA for liquidity; SGC where a vintage buyer prefers the black holder; CGC for Pokémon where the buyer pool accepts it; Beckett where sub-grades or a Black Label attempt add value. A cheaper holder that sells for 15% less has saved you nothing.
5. Submit correctly: a Collectors Club membership or an authorised dealer for PSA (GameStop and many card shops act as drop-off points, on their own price lists and queues), the tier matching the declared value, cards in penny sleeves and semi-rigid holders, insured shipping, and a photograph of every card before it leaves.
6. Watch the queue, not the calendar. PSA’s tracker is published every two weeks; the estimated turnaround on the day you submit is the earliest you will see the cards, and a price can move 30% in a 90-business-day wait.
7. When cards return, verify each cert number and image on the grader’s site, record cert numbers and purchase prices in an inventory, and insure at graded value.
8. Sell where the holder is worth most: eBay under the Authenticity Guarantee to about $5,000, a house’s regular sale above that, and never a private buyer you cannot verify for a card whose whole value is a label.

The case for owning cards at all is in Investing in Sports Cards, the flagship, and in the hub’s other guides: Investing in Vintage Baseball Cards, Investing in Modern Basketball and Football Cards, Investing in Pokémon and TCG, Investing in Magic: The Gathering and Investing in Sealed Product; this one ends with what would change our reading of the grading machine itself.

Seven readings would change the view in this guide; each is given with its level as of September 2026 and the threshold that matters.

• The PSA backlog tracker. 9.9 million in early September 2026, down from 12.4 million on July 28 and about 14 million in the tracker’s June count, against a target of 5 million; PSA said one paused tier could be considered for reopening before the end of September, and Standard opens at $59.99 on September 14. Below 5 million the Value tiers reopen, the entry fee falls back toward $25 and the break-even PSA 10 price roughly halves, bullish for submissions and bearish for every modern 10 already in a holder. Back above the June reading of 14 million, the 2021 pattern is repeating.
• PSA’s gem rate. 43% in the first half of 2025, 45% in 2024; sports 34–41%, games 50–61%. A fall of five points over a full year means the grader tightened, making existing tens scarcer; a rise above 50% overall means the label is being applied to more objects than the bid can absorb.
• Sports grading volume. 10.0 million cards in 2025, down 12%. A second year of decline confirms the sports census is saturating; a rebound above 11 million means closets are still emptying.
• The 10-count on the flagship modern rookies. Cooper Flagg’s 2025-26 Topps Chrome #251 and the Wembanyama base rookies are the test cases. A PSA 10 population that doubles in the twelve months to September 2027 says the 2021 mechanism is intact.
• Population growth on the vintage trophies. The Jordan’s 330 PSA 10s and the Mantle’s 35 PSA 8s. Ten new tens on the Jordan in a year, or five new 8s on the Mantle, would mean resubmission or a loosened standard is reaching the fixed-supply tier, and the multiples here would need re-basing.
• The fraud report. More than $200 million intercepted in 2025, counterfeits up 45.3%, 112 of 181 Mantles fake. A fake-Mantle ratio above three in four in the 2026 report, or a first federal case built on resealed genuine holders rather than relabelled ones, would mean both the raw and the second-hand slab markets had become unusable without a fresh review.
• Grader consolidation and pricing. SGC’s volume, down 24% in 2025 and 66% year on year in August 2026, and its September 2, 2026 service changes. If SGC’s brand is folded into PSA’s, or CGC’s economy tier moves above $30, the last cheap route into a holder closes and the entry fee becomes a permanent floor under every modern card’s PSA 10 price.
• Our tape. A PSA 10 to PSA 9 basket ratio of 12.4 on September 8, 2026. Below 8 says the tens are being revealed faster than the bid; above 16 says the fixed-supply cards are pulling away from the rest.