Price data and methodology

Price series and update dates

Trend, Average and Low are separate fields supplied by the price feed. Singles may also offer 7D MA, 30D MA and foil fields. These fields are displayed as supplied; the website does not calculate the marketplace's Trend or Average formula from individual trades.

Catalog prices and the analytical metrics below use the standard Cardmarket Trend field as their reference price. Changing a product chart to Average, Low or Foil changes that chart's price and return; it does not change the reference series used for market rankings.

Updates are scheduled daily after the source import completes. Source delays and processing can leave an older observation visible; the displayed date is the best guide to freshness, rather than a promised publication time.

Charts and current metrics normally share the latest completed analytics date for which the source guide is marked fresh. A stale guide or newer prices still awaiting analytics can cause an older date to remain visible. Use the observation or metrics date to judge freshness: refreshing a page does not create a new market observation. Missing values remain unavailable rather than becoming zero prices.

Chart returns and comparisons

The product chart uses positive, finite observations from the selected series. Where several observations share a UTC date, it uses the latest valid observation for that day. A chart return is (ending price / starting price − 1) × 100. The 1D chart uses the last two available daily observations; a gap can span more than one calendar day.

The primary product defines the comparison window. Each row states its price date and starting date. Products with later starting coverage have a shorter measured return. A missing observation on the selected date is shown as unavailable. Hovering or selecting a chart range changes the table context; the headline and range marker show the latest observation in the chosen period.

Chart windows of 1W, 1M, 3M, 6M and 1Y request 7, 30, 90, 180 and 365 calendar days, respectively, back from the latest available product observation, capped to the usable metrics date. Missing dates are not added to this history.

Ranking returns use the standard Trend reference price and a calendar-based starting point. For 1D, 7D and 30D, the starting price is the latest available observation on or before one, seven or thirty days earlier. The search goes back at most 400 days from the ending date. A missing or zero starting price gives no percentage return. Absolute change is ending price minus the same starting price.

When observations are missing, the measured interval can be longer than the label. Ranking returns can therefore differ from a chart return over its available points. Neither calculation includes selling fees, shipping, taxes or your personal purchase cost.

Large moves and small starting prices

Ranking flags prompt further inspection and do not remove or correct source values. Large move marks a reported change of at least 100% in magnitude. Small base marks a change of at least 50% in magnitude when the current reference price and reported return imply a starting price below €0.10.

The implied starting price is current price / (1 + fractional return). It is an estimate, not a dated source observation. These flags cannot establish a bad data point, a stale-to-fresh transition, a completed sale or statistical confidence. Inspect the product history before interpreting an extreme percentage as market activity.

Market breadth and average returns

Gainers share is gainers / (gainers + losers) × 100, rounded to a whole percentage, for the selected market universe. Unchanged items are excluded. When no gainers or losers are available there is no measured share. It is a count of price directions, not a survey of investor sentiment.

Average daily return is the arithmetic mean of available 1D returns on the displayed metrics date in the selected universe. Each product with a non-missing return has equal weight; zero returns are included and missing returns are excluded. Average price is the arithmetic mean of available reference prices. Neither average is weighted by sales volume, market value or your holdings.

Singles are catalog products classified as single cards. Sealed includes products with a recorded type other than single; All includes the full catalog. Different universes and missing prices change the population being averaged. Extreme returns from tiny starting prices can strongly affect the average, so it is not an investable index or the return on a typical portfolio.

Liquidity: frequency of reference-price changes

The liquidity score is a price-update proxy. For a window of N calendar days ending on the metrics date, count consecutive available reference-price observations whose prices differ, then divide by N − 1. The score runs from 0 to 1. The standard liquidity score uses 30 days, with 7-day and 90-day versions also available.

For example, 20 changed observation pairs in a 30-day window give 20 / 29, or about 0.69. A score requires at least two observations within the window and a recorded price date reaching back to at least the window's start. It does not require an observation on every date. Pairs can span gaps, and missing dates reduce the possible score because the denominator stays fixed.

Activity change is the 7-day score minus the 30-day score. A positive value means reference prices have changed more frequently recently relative to that longer baseline. This measures neither verified transaction volume nor a probability or speed of selling. Flat or repeated source prices can lower the score even when trading occurs.

Volatility, momentum and drawdown

30D volatility

The sample standard deviation of available 1D percentage returns during the last 30 calendar days, including the metrics date. At least five returns are required. It is displayed as a percentage and is not annualized. Missing returns are excluded; a return spanning a data gap still counts as one observation.

90D momentum

The percentage change from the latest available reference price on or before 90 days earlier: (ending price / starting price − 1) × 100. The starting price must be positive and no more than 400 days before the ending date. Missing coverage can make this interval longer than 90 days.

90D maximum drawdown

The largest observed percentage fall from an earlier running peak within the trailing 90 calendar days, including the metrics date. At each observation, divide its price by the highest price up to and including that observation in the window and subtract one; take the lowest result, displayed as a percentage. At least two reference-price observations and a positive running peak are required, so a recently listed product can have a value before 90 days of history exist.

These metrics describe recorded reference prices. They do not predict future prices, and gaps or source anomalies can affect them. No statistical confidence score is attached.

Market activity rankings

The default market activity ranking combines each product's price percentile with its 30D liquidity percentile within the same day and universe. Singles and All use 40% price and 60% liquidity; Sealed uses 20% price and 80% liquidity. Products need both a reference price and a 30D liquidity score to enter this ranking.

Percentiles use ascending rank: (rank − 1) / (number of products − 1), with ties sharing their first rank. A one-product group receives percentile 1. The weighted score is ranked highest first, with product ID breaking ties. This is a relative catalog score; higher price contributes to it, and it is not a count of purchases or views.