Verda's stablecoin census counts 16 wholesale specialists among 494 Latin American firms

Sixteen out of 494 is 3.2% of the map ^[14], or about one company in 31 ^[15]. The report, built on Verda's Stablescape database and published in early October 2026 ^[7], puts the system's weak point in that slice. "Fragility in the system is concentrated in its thinnest layer," it says ^[17].

The figure counts companies. Chu said Stablescape does not track transaction volumes or provide market share figures ^[9]. The 16 also cover three functions, wholesale liquidity, corporate treasury and credit ^[1], so the firms doing large-scale currency conversion are some subset of that number.

The way firms are classified moves the true count in both directions. Exchanges and payment companies filed under other categories also supply liquidity ^[5]. If they hold the risk on their own books, the back end is wider than 16. Chu said Verda believes some of them ultimately depend on the same underlying desks ^[5], and to the extent that holds, the number of independent risk-holders is smaller than the headline suggests.

Where a failure would land is clearer. "The problem would be at the exits. Spreads would widen, cash-outs to local bank accounts would slow or pause, and funds in transit with the failed desk could be stuck," Chu said ^[4]. A token crosses a blockchain in seconds, but turning it into pesos or reais in a local account still runs through the desks the report counts as scarce ^[8].

The strongest case against the fragility reading comes from Chu himself. "Mature FX markets also have far fewer dealers than customer-facing firms. What matters is redundancy and capital," he said ^[10]. His standard is specific: "Each major currency should have several independent, well-capitalized desks with separate banking relationships, and each wallet should be able to route between multiple players" ^[11]. Testing it means counting desks per currency and banks per desk.

The evidence fits more than one outcome. If most conversion volume clears through a handful of the 16 and they share banks, the report's warning holds as written. Should the exchanges classified elsewhere carry their own currency risk, the region has more redundancy than the headcount shows. And where each major currency already has several well-capitalized desks, Chu's dealer comparison applies, and a 3.2% share of dealers ^[14] says little about risk on its own.

I'd expect the first outcome to be closest, at least in direction. The pass-through Chu described ties many of the 478 firms outside the back end ^[16] to the same few desks, so the dependence reaches well past the 16 the report counted. Per-desk volume data showing several independent, separately banked desks in each major currency would prove that wrong.