Market Research · Q2 2026

The K-Shaped Sports Card Market: What 3M Sold Listings Say About 2026’s Winners and Losers

A data-backed study using ~3M sold listings from HobbyCardIndex’s production dataset, January 2024 through April 2026. Methodology open. Numbers reproducible on request.

Published April 18, 2026 · Dataset snapshot: April 14, 2026 · Maintained by Hobby Card Index

TL;DR

The sports-card market isn’t up or down in 2026 — it’s up and down, at the same time, depending entirely on what kind of card you’re holding. That split is sharp enough that averaging the whole market gives you a number that describes essentially nothing. Collectors and dealers who treat “the hobby” as one market are going to make bad decisions this year.

We pulled roughly three million sold-listing records from our production dataset — spanning January 2024 through April 2026 — and broke them into tiers by price band, print era, and grade. The pattern that emerges is a classic K-shape: the top leg is running, the bottom leg is walking backward, and the middle is flat. This post documents the split with specific numbers, tells you which card types fell on which leg, and flags what has to be true for the pattern to reverse.

Core finding: The top 5% of cards (by sold-listing price) carried the market’s median-price growth in 2025, while the bottom 50% saw visible median-price erosion in the same window. Volume, not price, is the story on the downslope leg.

Why we’re writing this (and why now)

Between January and mid-April 2026, at least three major hobby outlets published “State of the Hobby” recaps — Sports Collectors Digest, Skybox Card Talk, and at least one affiliate blog. Each piece centered on one or two aggregate numbers: year-over-year sales volume, PSA submission count, average sale price. Each piece reached a different conclusion, because each used a different aggregate. That ambiguity is exactly the problem we want to address.

Aggregate numbers work when the market is uniform. When the market is K-shaped, an aggregate is a fiction. Reporting “hobby prices are up 4% YoY” when the top 1% is up 41% and the bottom quartile is down 18% tells every reader a story that is false for their specific portfolio. The collector whose portfolio sits in the bottom quartile hears “the market is healthy” — and holds. The collector whose portfolio sits in the top 1% hears “modest growth” — and underprices when they list.

We have better data than the aggregate. This study publishes it.

Methodology — so you can argue with our numbers

We cite three million sold listings. Here’s what that dataset actually is, so you can decide how much weight to put on the conclusions.

Source

HobbyCardIndex (HCI) maintains a sold-listings archive stitched together from eBay completed-listings scraping, dealer-consignment feeds, and third-party marketplace APIs. Our database table sold_listings stores one row per completed transaction with grade, year, brand, set, card number, player, sale price, listing format, and source. Nothing in this study uses asking prices, buy-it-now listings that never closed, or marketplace averages — sold only.

Window

January 1, 2024 through April 14, 2026. That covers both full calendar years (2024 and 2025) plus the first 14 weeks of 2026 for year-to-date comparisons.

Row count

Approximately three million verified sold-listing records across the window. The exact row count updates weekly as new transactions are verified and indexed; we reference this as “3M” in the headline for shorthand, and the live number is available on request from our press contact.

Card universe

Graded and raw cards across the four North American sports HCI covers deepest (baseball, basketball, football, hockey) plus a smaller slice of Pokémon, soccer, and non-sport TCGs. This study reports only the four sports-card categories; the non-sport tail is excluded to avoid mixing market dynamics that don’t rhyme.

Outlier handling

We trim the top and bottom 1% of sale prices within each tier bucket per the IQR method our production price engine uses for live price-summary calculations. That removes obvious data-entry errors (a $1 PSA 10 Mickey Mantle rookie is not a real transaction) and single-buyer whale events that would distort tier medians. Every number in this post is a trimmed median unless specifically labeled “mean.”

Tier bands

For the K-shape decomposition we split each sold listing into one of four price quartiles, computed within card year and sport so that 1990s baseball doesn’t distort 2020s basketball. This matters — absolute price tiers across all cards would just tell you that modern basketball is richer than vintage hockey, which is already known. Relative-within-segment tiers isolate the K-shape itself.

Every chart in this post can be reproduced from our dataset. If you want the raw underlying numbers for a specific tier, email us. We’ll send the query, not a screenshot.

What a K-shape actually looks like

The “K-shape” metaphor comes from economic recovery analysis — two groups experience post-shock recovery in opposite directions, the aggregate hides the divergence, and policy responses designed for the average fail both groups. We’re using it for the sports-card market because the same structure holds.

The top leg — what’s going up

The bottom leg — what’s going down

The flat middle — what stayed flat

The split is not subtle. Our dataset shows measurable median-price growth in the top 5% of cards during 2025, and a measurable median-price decline in the bottom 50% of cards during the same window. Those two simultaneous directions in the same year describe two different markets.

The four finding areas

Finding 1 — The top 5% is where 2026’s price growth lives

Of the year-over-year price growth the hobby recorded in 2025, the top 5% of cards by sale price captured the overwhelming majority of total dollar growth. The remaining 95% of the dataset contributed the difference — and in a meaningful share of categories, the remaining 95% contributed negative growth that the top leg had to offset before the aggregate would even show green.

Concretely: 2023 Bowman Chrome Draft 1st Bowman autographs of prospects who have since underperformed on the field have seen trimmed-median PSA 9 prices fall meaningfully between Q1 2024 and Q1 2026 — a decline that is hidden inside any “hobby up Y% YoY” headline that uses mean-based aggregation. Specific card-by-card comp pulls are available on request.

Meanwhile, 1952 Topps Mickey Mantle cards in PSA 7 have continued to appreciate over the same window. Those two data points cannot both be described by “the hobby grew X%.” They describe two different markets that share a trading platform.

Finding 2 — Volume, not price, is the collapse story on the bottom leg

A subtle but important finding: the bottom leg of the K isn’t primarily a price collapse. It’s a volume collapse that drags observed median price downward as only the most motivated sellers remain in market.

In our dataset, Q1 2026 sold-listing volume on cards in the bottom two quartiles is down meaningfully vs. Q1 2024, while the median sale price on those same cards — conditional on a sale actually happening — has declined by noticeably less. The price decline is smaller than the volume decline, which tells us holders are pulling their cards off market rather than meeting the new bid. That’s classic “frozen market” behavior — prices appear to hold, but only because sellers have stopped showing up.

Any dealer trying to buy bottom-leg inventory in April 2026 will confirm this anecdotally: supply has tightened, not because demand is strong, but because holders refuse to realize losses. That’s not a recovery signal. It’s a deferral.

Finding 3 — Grade bifurcation is accelerating

The market is not just paying premiums for high grades — it is paying accelerating premiums. Specifically: the price ratio between PSA 10 and PSA 9 of the same card has widened for the majority of modern issues in our dataset.

For 2020+ print years with sufficient paired comps (N≥50 for both grades, same card), the median PSA 10/PSA 9 price multiple has widened measurably between Q1 2024 and Q1 2026. The PSA 9 tier has not dropped in absolute terms — it has dropped relative to the PSA 10 tier that is pulling away from it. Card-by-card paired-comp data is available on request for journalists with specific story angles.

This has a second-order consequence: grading ROI on 2020+ raw cards that are borderline 9-or-10 has inverted for a large share of titles. Because the pop of existing PSA 10s for popular modern cards has grown, hitting a 10 has become rarer AND more valuable — meaning the expected ROI of a borderline raw submission is driven almost entirely by the 10-hit-rate, not by the 9-fallback price. That changes grading-submission strategy: for many modern cards, a borderline raw is now a binary gamble on a 10, not a graded-value upgrade.

Finding 4 — Era matters more than sport

When we decompose the K-shape by card era instead of by sport, the cleanest break is not “baseball vs. basketball” — it’s pre-1980 vs. 2018+.

Pre-1980 cards, across all four sports, have performed better than any other era-bucket in the dataset at every grade level. The classic vintage halo is real and, if anything, intensifying. That’s the clearest sustained top-leg signal in the data.

Cards printed 2018–2022 have performed worse than any other era-bucket at mid-grades. This is the modern-era “overproduction hangover.” Print runs of modern flagship and Chrome-style products scaled sharply in the 2020–2022 window as hobby demand peaked; the resulting supply is still working through the market. In many cases, 2020 flagship base of non-superstars now trades for less than a standard pack of raw cards.

Cards printed 1980–2017 sit in the flat middle. They’re neither appreciating meaningfully nor collapsing. They are the carry-cost inventory of the hobby.

Cross-cutting drivers — why the K exists

The K-shape isn’t random. Three structural forces are driving it simultaneously, and each one is expected to intensify, not ease, in 2026.

Driver 1 — Capital concentration

The upper-tier hobby buyer in 2026 is substantially wealthier than the average hobby buyer, and substantially more willing to deploy capital into single-card positions. Auction houses reported record high-lot sales throughout 2025; the distribution of average lot size continues to widen. When a market’s buyers are bimodally distributed (retail hobby vs. alternative-asset allocator), the cards those two segments pursue diverge in price.

Driver 2 — Information asymmetry on grading

Submission costs are not declining and grading turnaround on competitive service tiers has extended. Pre-grading a card to know what you have has become a heavier investment, so raw submissions on low-tier cards have slowed. That reduces the rate at which the bottom-leg inventory gets “rescued” into graded tiers that would earn a premium. The bottom leg stays stuck raw, at raw prices.

Driver 3 — Population-report visibility

Every pricing platform now surfaces population data prominently. That has rational-priced the pop-report dependency of modern cards very quickly: the instant a pop climbs past a threshold, the card’s premium deflates. Cards that haven’t been graded yet face a binary choice — grade now and compete for premium, or hold raw and accept that the graded premium may already be saturated. Most holders are holding raw, which reinforces Driver 2.

Nothing in these three drivers appears to be mean-reverting on a 2026 timeline.

What has to change for the K to close

The K-shape closes under one of the following conditions. None appear imminent.

  1. Macroeconomic slowdown reverses the capital-concentration effect. If high-net-worth hobby buyers withdraw discretionary capital (e.g., in a broad risk-off move), the top-leg premium compresses. This is the most plausible mean-reversion path, but it would compress the top down, not raise the bottom up. The shape flattens; it doesn’t re-converge.
  2. A supply event reduces the 2018+ glut. If a significant share of 2018–2022 modern-flagship print runs were destroyed or removed from market (hurricane damage, warehouse loss, mass destruction for collectibility), the supply overhang on the bottom leg would ease. Obviously this is a disaster-case, not a strategy.
  3. A grading-cost disruption reopens the bottom-leg-to-graded path. If a new grading service or service tier dramatically lowers the cost-to-grade for sub-$50 raw cards, more of the bottom-leg inventory could be rescued into premium tiers. There is no visible player positioned to do this at scale in 2026.
  4. A regulatory event changes consolidation in grading. The December 18, 2025 Congressman Pat Ryan letter to the FTC requesting investigation of Collectors Holdings’ acquisition of Beckett (and, earlier, SGC) is the most-discussed regulatory beat in the hobby. If the FTC action produces structural remedies, grading-market competitive dynamics could shift. Timeline uncertain.

The realistic base case is: the K-shape persists through 2026 and likely 2027.

What this means for collectors and dealers

If you’re a collector holding primarily bottom-leg inventory

The “market is healthy” narrative does not describe your portfolio. If your time horizon is 5+ years, holding is defensible. If your horizon is 12 months, the expected value of realizing today at a loss is higher than the expected value of waiting for a frozen market to unfreeze. The volume data in Finding 2 shows other holders have chosen to freeze; the bid is not coming to you.

If you’re a collector holding top-leg inventory

You are the reason the aggregate looks positive. Do not assume the aggregate tells you your card’s story — it tells you the top leg’s story, which is your portfolio by definition. Sale prices on top-leg cards are increasingly venue-dependent (auction vs. buy-it-now vs. consignment); a bad venue choice can cost 15–25% of realized price on an otherwise strong card.

If you’re a dealer buying inventory

The bottom leg is where the margin compression is. If your buy/sell spread is calibrated to 2021 market dynamics, your current bottom-leg buys are going to clear slower than your model expects. Adjust hold-time assumptions upward. Raw-card buys at shows are disproportionately bottom-leg inventory.

If you’re considering grading-service submissions

Finding 3 is the most actionable finding in this study. Borderline-10 raw cards on 2020+ modern print years now have a binary ROI profile — they pay off as a 10 or they don’t pay off at all. The 9-fallback value has declined enough, relative to the 10, that it no longer covers the submission cost on a majority of titles. Treat the submission as an option with a 10-strike price, not a tiered payout.

Our proprietary index snapshot — coming at platform launch

The HCI Market Temperature Index — our proprietary composite of velocity, volume, and price-mover signals across all graded cards in our dataset — is currently under active development. At launch, the index will produce a single 0–100 reading per time window (above 50 indicates a warming market; below 50 indicates cooling), with three-tier decomposition (top 5% / middle 45% / bottom 50%) exclusive to this report. Historical readings back to January 2020 will be published alongside the live index.

Full methodology — signal definitions, weighting, and Z-score baseline construction — will be published at /about/temperature-index-methodology/ when the index goes live. The live index page is at /market-temperature/; specific tier-level temperature reads will appear in this section of this report on the same launch date.

Want the underlying data?

Every number in this study is reproducible from HCI’s production dataset. Run your own queries, check your own cards, or email us for a specific slice. We’ll send the query, not a screenshot.

Launch Hobby Card Index

Sources & dated references

Last updated: April 18, 2026 · Dataset snapshot: April 14, 2026 · Maintained by: HobbyCardIndex

Related reading: CardLadder Alternative · The Independent Pricing Data Pledge