How the EverySlab Pokémon Card Index works

What it measures

The index follows the typical price change of English Pokémon cards from one month to the next and chains those changes together, starting at 100 in January 2021. There are two versions: graded (PSA, BGS, CGC and others at grade 9 or higher) and raw (ungraded cards in near-mint condition).

Where the data comes from

Monthly prices come from PriceCharting's price history for each card and grade, converted to euros at the month-end exchange rate. How often each card sells comes from the completed sales we record on eBay, Fanatics Collect and Goldin. A month's value is published on the first day of the next month, once the month is complete.

How it is calculated

Each month we compare every card's price with its price the month before, then take a weighted average of those changes.

• A card stays in the index for up to 12 months after its last real price update. When it is updated again, the whole change since the last update counts.
• A card counts in a month when the average of its two prices is at least €30. A fixed cut-off on either end would drop cards just as they fall below it, or add them just as they rise above it.
• Each card is weighted by how many recorded sales it had in the 12 months before. Cards nobody trades do not move the index; when too few cards have sales, all cards count equally.
• Changes are averaged on a logarithmic scale and the extreme 1% at each end is capped, so a single bad price cannot move the index.
• Each month's average change is applied to the previous level. The index is recalculated from 2021 every month.

The sales index A second index built only from actual eBay sales of graded cards, comparing each card with its own earlier sales. It checks the EverySlab Index against real transactions, but it is noisier and past months can change as new sales arrive.

Next to the EverySlab Index we publish a second index built only from completed eBay.com sales of English graded cards at grade 9 or higher. Each sale is compared with other sales of the same card and grade, so a month with many cheap cards does not look like a fall. Single months are noisy, so each point is the average of three months. The model is refitted every month, so past values can change slightly as more sales arrive. It covers graded cards only: there are too few raw sales before 2024.

Drawdown and risk, in plain words

Drawdown

How far the index is below its highest point so far. At −25%, prices would have to rise by a third to get back to the peak.

Max drawdown

The deepest fall from any peak to a later low. It is the worst loss someone who bought at the top would have had to sit through.

Months under water

How many months the index has been below its last peak. It resets to zero at a new high.

Single-card stress

The same measures for each card on its own, using at least two years of real price updates. We show the typical card, the worst 10%, and how many cards lost more than half.

Limits

PriceCharting prices are estimates from US sales, converted to euros, so currency moves are part of the index. Sales weights only reach back as far as the sales we have recorded, which are thinner before 2024. Japanese cards are not included yet.

Questions