The buyback is 84.6%, not 90% — and it changed recently

On Courtyard, the number on your pull is not the number in your pocket. Every card carries an FMV — a fair market value estimate — and the instant buyback pays a fraction of it. That fraction is widely described as 90%, including in an earlier version of this page. It is currently 84.6%, and the way it got there is more interesting than the number.

It is a flat rule, not a range

We index the buyback offer attached to every pull Courtyard publishes. Across 33,980 pulls with a recorded offer, the most common value is not a cluster around 90% — it is exactly 84.6% of FMV, on the large majority of all offers ever indexed. Sort those offers by card value and the median is 0.846 in every single band: under $20, $20–40, $40–80, $80–200 and $200-plus.

A number that constant across tens of thousands of observations is not a market outcome. It is a pricing rule somebody wrote down.

The buyback used to be tiered. Now it is flat.

This is the part no product page will tell you, and it only becomes visible with enough pulls to slice by card value. In July, the buyback rewarded cheap cards: a card under $25 drew 95.7% of its FMV on average, while a card over $100 drew 84.6%. By September every band had converged on the same flat rate.

Read the pairs. The top pair collapses by ten and a third percentage points. The bottom pair does not move at all. Whatever happened here was not a broad rate cut — it was the removal of a discount that used to apply only to small cards.

And that matters more than it first appears, because of where the typical pull sits. The catalogue median pull is $24.6. The median buyer's median card is squarely inside the band that lost the most. Somebody ripping for grails barely felt this change; somebody ripping regularly absorbed essentially all of it.

How fast it happened

The transition took about eight weeks, and it shows up cleanly as the share of offers landing on the flat rate.

In practical money: a $100 FMV card that cashed out near $91 in July cashes out near $85 now. Over a $1,000 run of pulls that is roughly $60 of value that quietly stopped arriving, with no change to the packs, the odds, or the prices.

We saw no announcement. This is measured from offers attached to real pulls, not read off a policy page — which is the entire reason for indexing the feed rather than trusting the marketing.

Sometimes the offer is zero

Worth knowing before you rely on the instant exit: about 3.8% of September pulls carried no offer at all — the platform declined to bid on that card rather than bidding low. That share has risen from 1.5% in July. For those cards your only exits are listing on the marketplace or paying to redeem, so the buyback is best understood as a very good default, not a guarantee.

What this does to every other number

The buyback rate is not a footnote; it is the exchange rate between the numbers on a pack page and the money in your account. Three consequences follow directly.

Break-even moves. A pack at a 1.00x EV ratio does not return what it costs — it returns 84.6% of it. Real break-even sits at 1 ÷ 0.846 = 1.18x, which is why our EV guide leads on that number rather than on 1.0.

ROI shrinks by about fifteen points. Counting FMV as cash overstates return by the size of the haircut. The corrected formula is in the ROI guide.

Redemption gets relatively more attractive on big cards. The buyback haircut scales with card value while shipping stays roughly flat at $16 in the US. On a $400 card the haircut costs about $62 against $16 of shipping; on a $30 card it costs $4.60 against the same $16. See shipping and redemption fees, which also covers the import tax that changes this arithmetic outside the US.

Your three exits, ranked

Instant-sell. Fastest, 84.6% of FMV, no shipping, no waiting. The right default for anything below roughly $60, and the exit most rips will take whether or not the buyer planned it.

List on the marketplace. Beats the instant offer by the full 15.4% haircut whenever a buyer exists, because Courtyard charges 0% marketplace fees. Costs time and needs a buyer, and is not practical for every card. Worth the effort on genuine chase cards and rarely worth it below a few hundred dollars.

Physically redeem. You get the real slab, shipped from the Delaware vault by FedEx. Worth it when the card is worth several times the landed cost — and outside the US, remember that import tax stacks on top of shipping, at 20% in the UK and 25% in Sweden with no tax-free threshold.

Check any pack's live cash-out numbers on the rankings, which already apply each pack's own measured ratio rather than the platform average.