ROGUEQUANT_

$ cd ..

2026.07.17 // DIGITAL ASSETS // 4 MIN

Stablecoins Ate the Payment Rails While You Were Watching AI Stocks

A $322B market cap, on-chain volumes rivaling Visa, and tokenized Treasuries at their 10th consecutive monthly record. The most underpriced infrastructure story of 2026—in three charts.

ALESSIO ROCCHI ·

Every financial headline in July 2026 is about AI: chip demand, hyperscaler capex, the SOX at its most overbought level in three years. Fine. But while everyone stares at the AI trade, a quieter infrastructure story crossed three thresholds that would have been front-page news in any other year.

Stablecoins stopped being a crypto story. They became a payments story—and increasingly, a market-structure story.

The Base Layer: $322 Billion and Two Issuers

Start with the stock, not the flow. Total stablecoin market cap crossed $322 billion in June 2026—an all-time high, up from roughly $150 billion in January 2024. More than doubling in 30 months, through a rate cycle, an election cycle, and two crypto drawdowns.

FIG. 01 // MARKET CAP

Stablecoin market cap, June 2026 all-time high

$322B

+115% vs Jan 2024 (~$150B) · +5% vs Dec 2025 ($308B)

58%
23%
19%
USDT $187BUSDC $75BOther $60B
Two issuers still control ~81% of the market: Tether (USDT) at ~$187B and Circle (USDC) at ~$75B.

DATA: COINDESK RESEARCH · TRANSAK · JUN 2026

The concentration is the part that should bother you. Tether alone is ~58% of the market. If you believe stablecoins are becoming systemically relevant payment infrastructure—and the volume data below suggests they are—then a single offshore issuer holding $187B of that infrastructure is a concentration of risk that no payments regulator would ever have designed on purpose.

The Flow: Rivaling Visa, With an Asterisk

Here's the number that gets quoted everywhere: total on-chain stablecoin transaction volume in 2025 exceeded $27 trillion, more than Visa and Mastercard's combined annual volume.

Now here's the honest version of that number.

FIG. 02 // SETTLEMENT VOLUME

2025 annual volume: stablecoins vs the largest card network

StablecoinsCard network
$0T$10T$20T$30TStablecoins — raw on-chain: $27T+STABLECOINS — RAW ON-CHAIN$27T+Visa — payments volume: $14.2TVISA — PAYMENTS VOLUME$14.2TStablecoins — adjusted*: $10.9TSTABLECOINS — ADJUSTED*$10.9T
*Adjusted volume filters bots, MEV and self-transfers (Visa-comparable methodology). Adjusted stablecoin volume grew 91% in 2025.

DATA: BESSEMER VENTURE PARTNERS · VISA FY REPORTS · 2025

Raw on-chain volume counts bots, MEV, exchange shuffling and wash activity. Bessemer's adjusted figure—filtered to be Visa-comparable—is $10.9 trillion for 2025. That's still below Visa's $14.2 trillion.

So no, stablecoins have not "flipped" Visa. What matters is the derivative: adjusted volume grew 91% in one year. Visa's grows single digits. You don't need a spreadsheet to see where those two lines cross if the growth holds—and "if the growth holds" is doing real work in that sentence.

The New Collateral: Tokenized Treasuries

The third chart is the one I'd argue is most relevant for anyone building trading systems.

FIG. 03 // TOKENIZED RWA

Tokenized real-world assets, May 2026 record: $28.9B

10TH CONSECUTIVE MONTHLY RECORD

$0B$6B$12B$18BUS Treasuries: $16.1BUS TREASURIES$16.1BStocks: $2.4BSTOCKS$2.4BOther RWA: $10.4BOTHER RWA$10.4B
Tokenized Treasuries dominate at $16.1B. Tokenized stocks are the fastest-growing slice: +20.4% month-over-month.

DATA: COINDESK RESEARCH · MAY 2026

Tokenized real-world assets hit $28.9 billion in May 2026—the tenth consecutive monthly record. More than half of that is tokenized US Treasuries: yield-bearing, on-chain, T+0 collateral. Tokenized stocks are small ($2.4B) but growing 20% per month.

CoinDesk's analysts project tokenized assets could reach $400 billion in 2026. The sell-side goes further on stablecoins themselves: Citi's base case is $1.9 trillion by 2030, Standard Chartered sees $2 trillion by end-2028, BNY projects $1.5 trillion by 2030. Apply the usual discount to sell-side crystal balls—but even the bear case implies a multiple of today's market.

Why a Quant Should Care

This is not (only) a payments story. Three things change for market participants:

  1. Settlement becomes a parameter, not a constraint. T+0, 24/7 settlement in tokenized collateral collapses the funding and margining assumptions baked into most execution and treasury systems.

  2. On-chain flow is public alt-data. Every stablecoin mint, burn and transfer is observable in real time. Aggregate stablecoin flows into exchanges have been a serviceable risk-appetite signal for years; the dataset gets richer as volumes migrate on-chain.

  3. The arbitrage surface is expanding. Tokenized Treasuries vs. their off-chain twins, tokenized stocks vs. the primary listing, stablecoin FX pairs trading around the peg—every new tokenized instrument ships with a basis to trade.

The Failure Modes, As Usual

The honest assessment requires the other side of the ledger:

  • Concentration risk: 58% of the base layer is one issuer whose reserves are attested, not audited, and whose regulatory domicile is a moving target.
  • Depeg history: USDC traded at $0.87 in March 2023 over a bank run at a reserve bank. The mechanism that broke then has not fundamentally changed.
  • Regulatory bifurcation: the US GENIUS Act and EU's MiCA created workable frameworks, but they also created jurisdictional arbitrage—and the next crisis will find the seam.
  • Metric games: the industry loves quoting raw volume. Any analysis that doesn't lead with adjusted figures is marketing.

The bottom line: the AI trade is crowded and priced for perfection. The stablecoin infrastructure build-out is real, measurable on-chain, growing 91% a year on honest numbers—and almost nobody at a traditional trading desk is positioned for what it does to market structure. That asymmetry is worth more of your attention than another Nvidia earnings preview.


Building on-chain data into a trading pipeline, or fighting with tokenized collateral? I'd genuinely like to hear what's working—the research reports tell one story, the trenches tell another.