Cardian Research · Grading · Buyer's guide · Japanese cards
Should you grade it? The 2026 maths, era by era
A PSA 10 can be worth 17 times the raw card or about three times, and PSA's cheap tiers have been shut since June. Here's the expected-value sum, which cards it works for, PSA against CGC, and when buying the slab beats grading.
The same Charizard ex from the 151 set gets a PSA 10 on 85% of the Japanese copies sent in, and on 29% of the English ones. That gap explains most of what you need to know about grading in 2026. The easier a 10 is to get, the less it's worth over the raw (ungraded) card.
On 23 August 2024 we wrote: "High-value, rare, or vintage cards are often worth grading, while lower-value, modern cards might not see a significant return on investment". A lot has changed since then: PSA raised prices in February, paused its cheap tiers in June, and its cheapest service still open now quotes a wait of 90 to 100 business days. So this piece puts numbers on the question: what a grade is worth on average, which eras and languages it works for, how PSA compares with CGC, and when you're better off buying the slab.
The maths
A grade is a bet with a fixed entry fee
Sending a card in is a bet with three outcomes: a 10, a 9 or something lower, and each has a price. The expected value is the chance of a 10 times the 10 price, plus the chance of a 9 times the 9 price, plus the rest, minus about 13% in fees for selling the slab online. From that you subtract what you put in, which is the raw card plus grading and shipping. We use S$31 a card, which sits between CGC's and PSA's bulk prices before the pause.
We ran this on 2,091 cards worth at least US$10 raw with at least 100 graded copies, grouped by era, since a 25-year-old card and this year's card are different bets. For each card the question was simple: what chance of a 10 does your copy need just to break even? Then we compared that with how often graded copies of the card actually got a 10, using PSA's population report (its public count of how many copies got each grade).
Median share of graded copies that got a PSA 10, against the chance of a 10 needed to cover S$31 in fees and 13% to resell
On vintage a 9 alone covers the cost, so the 10 you need is 0%. On modern Japanese cards the cushion is only 10 to 27 points.
Source: Cardian grading model: cards worth US$10+ raw with 100+ graded copies; odds from PSA population reports (Oct 2026), prices from sold listings; S$31 grading and shipping, 13% to resell
On vintage the answer is zero, because a PSA 9 of a WOTC-era card (the early sets made by Wizards of the Coast) sells for about 3.5 times raw, so a 9 alone covers the costs. English modern cards need a 10 between 15% and 32% of the time and get one 34% to 67% of the time. Modern Japanese cards are the close call. A Japanese Mega card needs a 10 74% of the time, graded copies get one 84%, and that leaves a cushion, or safety margin, of just 10 points.
That cushion matters, because the population report makes your odds look better than they are. People check cards carefully before sending them, and plenty are taken out of their slabs and resubmitted until they get a 10. So the copies in the report are better than the average copy in a binder. If your card is a little worse than the ones other people chose to send, a 10-point cushion disappears quickly.
Share of cards where sending a raw copy is expected to make money after fees. In brackets: the median expected return on the raw price. Black: Japanese; grey: English
A typical Japanese Mega card returns 5% on paper, before the wait. A typical English WOTC card returns about 300%.
Source: Cardian grading model: cards worth US$10+ raw with 100+ graded copies; odds from PSA population reports (Oct 2026), prices from sold listings; S$31 grading and shipping, 13% to resell
Overall, grading is expected to make money on 97% of English WOTC cards and on 89% to 96% of English modern ones. For Japanese Sword & Shield and Scarlet & Violet cards it's 71% to 72%, and for the Mega era only 57%, with a median return of 5% on the raw price before you count the wait.
Why Japanese is different
Easy 10s, cheap 9s
The reason is in the factory, not the grading room. Japanese cards come out of the pack well centred and clean far more often. Across 59 cards printed in both languages, the median Japanese copy gets a 10 85% of the time and its English version 33%.
Share of graded copies that got a PSA 10, same card in each language (Japanese printing vs its English twin)
Japanese printing and pack quality make 10s common, which is exactly why a Japanese 10 is worth less over raw.
Source: PSA population report (October 2026): 59 cards printed in both Japanese and English, worth US$100+ as a PSA 10
When most copies get a 10, a 10 isn't scarce, so it can't carry much of a premium. A modern Japanese PSA 10 sells for about 3.2 to 3.5 times raw in our data, against roughly 5 to 7 times for English modern cards and 15 to 30 times for vintage. The 9 is where Japanese grading really hurts. If 85% of copies get a 10, a 9 tells the buyer your copy had a flaw, and the market prices it at about 1.3 times raw.
Raw price above which a PSA 9 still covers S$31 in fees and 13% to resell, by era (median PSA 9 ÷ raw in each row)
On vintage a 9 is a good result. On modern Japanese cards it's a small loss on almost anything you'd grade.
Source: Cardian grade ladder: median PSA 9 price ÷ raw price, cards worth US$10+ raw, 13 Oct 2026; S$31 grading and shipping, 13% to resell
After 13% in selling fees and S$31 for grading, that 9 only covers its costs on a Japanese Scarlet & Violet card worth more than about S$237 raw. Most of the Japanese cards people grade are worth far less, so for them a 9 is a small loss and the 10 is the only result that pays. On a WOTC card the same sum works from about S$15 raw.
What's changed in 2026
The cheap tiers are shut, and the wait is a cost
On 29 November 2025 we wrote about a bulk order sent on PSA's cheapest option: "in a major deviation from the norm, our grades were returned in a mere 30 days". Since then PSA raised Value Bulk from US$21.99 to US$24.99 on 10 February, and by May the wait for it was 140 to 160 days. PSA Japan paused most of its Value tiers for online orders on 15 May, and on 2 June PSA paused all four Value tiers with a queue of about 10 million cards.
The cheapest open service is now Standard, launched on 14 September at US$59.99 with an estimated 90 to 100 business days, which is about S$80 a card with shipping. The queue was 8.2 million cards on 6 October. What all this did to prices is covered in our piece on the sell-off.
The wait is a cost in its own right. Since 2024, if you pick any day and look 140 days ahead, our Japanese PSA 10 index was lower a quarter of the time, and more than 20% lower one time in ten. Cards sent in the week PSA Japan paused its cheap tiers would have come back 140 days later to a market about 34% lower.
A higher fee also changes the sum for the other option, which is buying the finished slab. A Japanese PSA 10 currently costs a median 2.92 times raw. Grading your own copy is cheaper only if the fees, spread over the copies that come back as 10s, add up to less than that.
Japanese Scarlet & Violet: what one PSA 10 costs you through grading, as a multiple of the raw price, against buying the slab
Assumes your copy gets a 10 as often as the market's (86%) and the misses sell as 9s. At 60%, the lines cross at S$26 and S$66.
Source: Cardian PSA 10 premium, Japanese Pokémon (13 Oct 2026); Cardian grading model; PSA price list (Oct 2026), US$59.99 at S$1.2787 per US$ plus S$3 shipping
At a S$31 bulk rate, grading beats buying the slab on anything above about S$19 raw. At S$80 a card the line moves to about S$48, and if your copy gets a 10 only 60% of the time it moves to S$66. Below that line, with no five-month wait and no price risk, buying the PSA 10 is simply the cheaper way to own the card. English cards work out much the same, at about S$43.
It's the same logic as opening packs. When the result is a lottery, the finished card is often the cheaper way to get the one you want.
PSA vs CGC
Only a Pristine 10 keeps up with PSA
Which company graded the card matters almost as much as the grade. Using each card's own recent sales, we compared other graders' 10s with the PSA 10 of the same card. We used English cards only, since that's where enough copies sell under both labels.
English Pokémon: median price of each label's 10 ÷ the PSA 10 of the same card, from each card's own recent sales
A Pristine 10 outsold the PSA 10 on 43% of cards; a plain CGC 10 on 6%. BGS and TAG rows are only 14 cards each.
Source: Cardian grader comparison: cards with 5+ sales of both labels since Oct 2024, median of each label's latest ten sales, 13 Oct 2026
A plain CGC 10 sells for a median 0.23 times a PSA 10 and beat it on only 6% of cards. A CGC Pristine 10, CGC's stricter top grade, sells for 0.92 times and beat the PSA 10 on 43% of cards. So not all 10s are equal, but a Pristine beating PSA is closer to a coin flip than a sure thing.
CGC is cheaper to use, at US$17 a card in bulk with about 120 days' turnaround in May, and it gives 10s more often on English Scarlet & Violet cards (56% of graded copies against PSA's 41%). The catch is that you're betting on a Pristine, because a plain CGC 10 usually sells for a fraction of the PSA price.
In that November 2025 post we described PSA Offer as "An immediate sales channel where third-party partners (including entities like crypto funds) make offers on your cards directly through the PSA app". PSA now lists those offers for US-based collectors, paid into banks in the US, Japan, Canada or the UK, so from Singapore the real choices are getting the slabs back or consigning them, meaning someone sells them for you.
If it were my card, I'd grade English vintage and modern cards without much hesitation. I'd grade a modern Japanese card only above about S$50 raw after checking the centring, and below that I'd buy the PSA 10 and let someone else sit through the wait. That's my opinion from the numbers above, and I'll redo the sums when PSA reopens its cheap tiers.