Stablecoin cards spent $1.076bn in August out of a float that shrank 3.6%

A monthly spend of $1.076bn against a float of roughly $304bn is 0.35 percent ^[1][17][1], which is the number that makes the broken correlation less impressive than it reads. Work the other side of it: a 3.6 percent fall from the May peak implies a peak near $315bn and a contraction of about $11.4bn ^[2], so August's entire card volume is under a tenth of the supply that left ^[3]. Two series can diverge at those magnitudes for reasons that have nothing to do with each other, and $11bn of net redemptions is not explained by cardholders buying groceries.

The better evidence for payment utility is the ticket. RedotPay's $390.1m across more than 6.34m transactions works out near $61.50 ^[5][6][8], and if you take it out of the sector, the remaining $685.9m over 4.33m swipes averages about $158 ^[4]. So the dataset holds two businesses, one doing high-frequency small purchases and one doing something closer to occasional larger transfers, and the record was set by the first one growing.

Divide 10,671,980 transactions by 283,653 active addresses and you get 37.6 a month ^[7], more than one a day, which lands almost exactly on Kraken's reported 8.3 weekly payments per Krak Card user ^[19][8]. That is the shape of a card that pays for rides and food delivery, which is what Binance says its Brazil users do with theirs ^[20]. The same division on volume gives $3,793 per address per month ^[9], and that is either a small population of heavy spenders or an address count that is not one-to-one with people; the source does not say which.

Then the plumbing under the number. Three programmes are 55.6 percent of volume ^[13], and RedotPay's figures are self-reported by the issuer rather than observed onchain ^[14], which puts about a third of the total on the word of what Cryptopolitan calls the sector's loudest promoter. The same article gives RedotPay 54 percent of transactions in its text and 59 percent in its summary ^[6][7]; 6.34m of 10.67m is 59.4 percent ^[5], so the lower figure looks like the slip, but the reader has to do that division themselves. And the $472.8m Hong Kong claim brought by Binance affiliates ^[15] is 1.21 times the volume RedotPay processed in its best month ^[11] (volume, not revenue, so treat that as a comparison of scale rather than of solvency).

This is probably wrong, but the more interesting version of the utility thesis has nothing to do with the float: settlement got cheap enough that a $12 purchase is worth clearing onchain ^[12], Mastercard switched it on across eight chains on 3 June ^[10], Visa counts more than 160 programmes live or in development ^[11], and cheap rails buy you transaction counts long before they buy you volume. Annualised, $12.9bn is 0.06 percent of the $20tn-plus card market ^[16] and 4.2 percent of the stablecoin float ^[12], so nobody at a network has to respond to this yet. What would prove the utility read wrong: September volume slipping back under $1bn while supply recovers, or paymentscan restating RedotPay's self-reported share downward. Either would put the three-year link back where it was ^[18], and the cards back where the pool puts them.