
From Printing Money to Building Roads: The Stablecoin Wars Enter the Era of Interface Battles
Author: Fu Gui
October 6
On October 6, 2026, roughly three things happened in the stablecoin ecosystem.
That day, Arbitrum One officially launched USDG natively, joining the Global Dollar Network led by Paxos. ArbitrumDAO simultaneously put forward a governance proposal to add another 100 million ARB to the USDG incentive plan, which is currently in the voting stage. The on-chain stablecoin market cap on Arbitrum is approximately $3.78 billion. The interest generated annually by these funds sitting idle in wallets previously went entirely to Circle and Tether. The official selling point of the Global Dollar Network is straightforward—“Stop funding their profits.” This move by Arbitrum isn’t about adding support for another stablecoin; it’s an attempt to turn idle deposits of on-chain stablecoins into its own revenue source for the first time.
Same day, OKX announced the completion of a new round of strategic financing with a pre-money valuation of $25 billion. Investors in the list included Circle, QRT, Ripple, and SC Ventures under Standard Chartered Bank. OKX stated it doesn’t lack capital but wants to find long-term partners. Translating that means issuers are actively paying to buy access to channels, with three issuers joining forces to back a single exchange, afraid of their tokens getting ranked lower on its order book.
Also within this timeframe, Stripe’s head of stablecoins, Henri Stern, announced that by year-end, stablecoin-backed cards would be deployed in over a hundred countries, allowing users to spend stablecoins directly from their wallets at Visa’s 175 million merchants. The card processed $1.2 billion last month, tripling year-over-year. Stripe doesn’t issue stablecoins itself; it simply handles delivering them to places where they can actually be spent.
Chains, exchanges, and payment companies all reached out in the same direction on the same day. They were all saying one thing: the rules of the stablecoin game have changed.
The Most Profitable Issuers
Tether is the most profitable stablecoin issuer in the crypto space; this statement hardly needs qualification.
In 2025, Tether’s net profit exceeded $10 billion, with roughly 300 employees, generating approximately $33 million in profit per person—nearly 100 times that of Goldman Sachs. Its profit logic is straightforward: users mint 1 USDT with $1, and Tether uses that dollar to buy US Treasuries, earning around 5% annual interest while token holders receive nothing. The spread belongs entirely to Tether. By the end of 2025, Tether held over $122 billion in direct US Treasury positions. Including indirect exposure from repurchase agreements and money market funds, its total Treasury exposure reached approximately $141.6 billion. If treated as a country, it would rank near the 18th largest US Treasury holder globally. It also hoarded 127.5 tons of gold and 96,185 Bitcoins. Valued at year-end market prices, gold reserves amounted to about $17 billion, Bitcoin holdings to about $8.4 billion, with unrealized paper gains on Bitcoin reaching approximately $3.5 billion.
According to DeFiLlama-tracked protocol revenues, Tether’s 30-day income was approximately $510 million, ranking first; Circle ranked second with about $210 million, less than half of Tether’s; below them, Hyperliquid and Pump.fun each accounted for less than one-eighth of Tether’s.
Tether earns so much not because it prints USDT well, but because it distributes USDT far and wide. In convenience stores in Central America, rural markets in Africa, and top-up kiosks in Southeast Asia, you won’t find Coinbase or Circle’s compliance channels, but you will find people using USDT. 2026 research shows that roughly 93% of USDT on Tron is held in regular wallets, not exchanges; the Andersen Institute even estimated that about 95% exists outside centralized exchanges. Its advantage lies not in the four words “decentralization,” but in highly fragmented distribution where users, use cases, and liquidity aren’t locked behind a single super-entry point. No single channel can choke its supply, meaning no one can siphon its profits.
The Other Side of Profits
Circle also makes money, with FY2025 total revenue hitting $2.747 billion, a 64% year-over-year increase, but its margins are getting increasingly squeezed.
The squeeze isn’t due to low absolute earnings, but because an ever-smaller fraction trickles down to Circle itself. Circle’s core profitability metric is called the RLDC Margin—the percentage of total revenue left after deducting distribution costs—which stood at roughly 39% for all of 2025. That might sound decent, but it means over 60% of revenue goes to channels helping distribute its tokens. Circle’s distribution, trading, and other costs reached $1.664 billion in 2025, up ~64% year-over-year; alone, distribution costs paid to Coinbase totaled ~$1.4 billion, up ~51%. The biggest drain is Coinbase—under their agreement, Coinbase takes the lion’s share of reserve interest generated by USDC on its platform, leaving Circle with only crumbs, and channel fees escalate every year.
Rates make matters worse. Over 90% of Circle’s revenue comes from reserve interest, so every Fed rate cut directly dents its top line. Q2 2026 data best illustrates this: USDC on-chain transaction volume surged 151% YoY, yet reserve income grew by only 5%. The reason is clear: average circulating supply grew by just 25%, while reserve yields dropped by 66 basis points. The standout takeaway isn’t the 151% volume spike, but the decoupling of volume and reserve income. Circle profits from “how many dollars stay deposited in the system long-term,” not “how many times those dollars change hands annually.” Bustling on-chain turnover doesn’t equate to more idle cash earning interest on Circle’s balance sheet.
Then there’s long-term pressure from extended rate cuts. As the Fed entered its easing cycle in 2025, Circle reported adjusted EBITDA of $582 million for the year, but booked a net loss of $70.1 million due to $424 million in equity-based compensation tied to its IPO. Capital markets are equally worried about competition: on September 1, 21 of the world’s largest financial institutions, including Goldman Sachs, Citigroup, and Bank of America, announced plans to form a joint venture to issue their own stablecoins, sending Circle’s stock down 6% that day. On June 30, when Stripe, Visa, Mastercard, BlackRock, Google, and others teamed up on OpenUSD, Circle’s stock plummeted roughly 17% in a single day. On August 3, Morgan Stanley slashed its price target from $106 to $38, citing slowing USDC growth, competitive pressure from OpenUSD, and the fragility of its reserve-revenue model.
Circle’s problem isn’t poor compliance or opaque reserves; it’s excessive reliance on third parties to distribute its tokens. Once channels dictate terms, Circle has zero bargaining leverage.
The Last Mile
Saying distribution channels matter isn’t empty rhetoric. The stablecoin space uses the term “last mile” to describe how successful on-chain settlement does not equal successful commercial payment. Technically, moving funds instantly across chains from Venezuela to Vietnam is trivial; the hard part is reliably converting that on-chain USDT into Vietnamese Dong per local regulatory requirements, depositing it into the correct bank account, and reconciling the accounts.
This final mile breaks down into four steps: cross-border transfers are solved, FX conversion is largely manageable, but what truly bottlenecks operations is integrating local payment rails and compliance reconciliation. For a service provider targeting ten markets, dozens of independent banking and mobile payment integrations must be maintained. Each API could fail independently, and KYC fields plus reporting requirements vary by country. And that’s just outbound flows. Inbound flows bring their own headaches: users might pay the correct fiat but select the wrong chain, send USDT when the merchant expects USDC, or forget the memo field. Binance disclosed in 2023 that it processes roughly 4,000 monthly recovery requests for misplaced deposits, recovering a cumulative 7 million USDT; Kraken charges up to $200 per ticket to retrieve misdirected tokens.
More disheartening are the raw numbers. After cleaning publicly traceable $62 trillion in stablecoin transfers, BCG and Allium removed bot activity, internal sweeps, and speculative trading, concluding that actual real-world economic activity amounts to roughly $4.2 trillion, with only $350–550 billion identifiable as payments for goods and services—less than 2%. An ECB survey found that only 0.2% of online merchants in the Eurozone accept crypto assets, and physical retail acceptance has remained below 1% for two years. Mastercard’s Chief Product Officer bluntly stated that roughly 90% of stablecoin volume relates to crypto trading, noting stablecoins “lack a clear value proposition” in conventional consumer-to-business payments. Field visits to Yiwu are even more telling: most merchants had never heard of stablecoins, and those who tried them discovered they forfeited the 6%–13% export tax rebate available through traditional settlement. Using USDT literally cost them that revenue. On-chain systems solve “how money moves,” which does not equal “how goods get delivered.”
A cohort of companies attempting to bridge this last mile has already faltered. BitPay halted new card applications in June 2023 and hasn’t resumed; it sits at a 1.2/5 rating on Trustpilot, relies solely on email support, and reduced weekend hours to just 4. Between 2025 and 2026, Dupay shut down over “unsolvable compliance issues and capital flow restrictions,” OneKey U Cards were discontinued, Binance terminated card services in Europe and Latin America, Paris-based Kulipa collapsed overnight into insolvency affecting over 20 wallet clients including Solflare and Ready—its sponsor bank Monavate was fined €270,000 by Lithuania’s central bank, and once its license was revoked, the entire product died. Cambodia’s Huiewang Pay saw its on-chain balances drained down to just 990,000 USDT, with its license actually revoked a year prior. Polish regulators revoked Quicko’s license in January 2026, and on February 3rd, CEX.IO, Trustee Plus, and IN1—all operating different brands—simultaneously lost card functionality.
While each case failed for specific reasons, they collectively expose a structural weakness: frontend brands capture user attention, but backend sponsor banks, licenses, clearing houses, and compliance nodes ultimately dictate survival. The vast majority of crypto-card projects aren’t Visa Principal Members; they rely on BIN sponsorship models to “rent” a banking membership. Sponsors handle compliance and settlement, while projects manage branding and customers. Sponsor names rarely appear in cardholder agreements, but if the sponsor fails, the card immediately dies. Wirecard’s 2020 collapse took Crypto.com and Binance’s original Visa products offline, and Metropolitan Bank’s 2023 exit from the crypto sector effectively killed BitPay.
Regarding Visa, there’s a frequent misunderstanding. Many assume card networks are merely toll-collecting intermediaries, but they’re not. What Visa actually sells is an entire road-building infrastructure suite: transaction authorization, network routing, rule frameworks, anti-fraud engines, inter-institutional clearing, chargeback dispute resolution, and global acceptance networks. For a $100 purchase, the merchant pays a merchant discount to the acquiring bank, the issuing bank takes the bulk via interchange fees, and Visa’s own network assessment is merely a rounding error—but Visa only captures that fraction because it built a trust system over fifty years enabling billions of cards and tens of millions of merchants to trade daily without hesitation.
Today, Visa is itself slamming the accelerator down this extended path. By September 2026, over 160 stablecoin-linked card programs were live on the Visa network, driving a stablecoin-clearing annualized run-rate exceeding $20 billion—up more than 15x YoY. The CFO didn’t carve this out separately during earnings calls because, relative to Visa’s yearly multi-trillion-dollar baseline, it represents a mere 0.12%. But the trajectory is unmistakable: the crypto industry hasn’t dismantled Visa; it’s simply paying tolls on a new stablecoin highway Visa is actively paving.
This last mile is fundamentally not a technical challenge, but an institutional, infrastructural, and trust issue. Blockchains solved “distance,” but “delivery” requires domestic banking relationships, FX liquidity, regulatory capability, and dispute resolution mechanisms—assets Visa and legacy banks spent fifty years cultivating, which the crypto industry still lacks cheap ways to replicate. Whoever successfully bridges this kilometer becomes the true roadbuilder. Precisely because this segment remains notoriously difficult to traverse, platforms that have already conquered their own domains now dare to demand premiums from the actual minters.
Platforms Strike Back
What truly punctured this veil was Hyperliquid.
In September 2025, Hyperliquid did something unprecedented: it publicly bid out issuance rights for its ecosystem stablecoin, USDH. Paxos, Frax, Ethena, Sky, Agora, and several others submitted bids, with a team named Native Markets ultimately winning. At the time, Hyperliquid’s monthly trading volume neared $400 billion, with $5.7 billion in stablecoins resting on-chain generating hundreds of millions in annual interest—a revenue stream previously funneled entirely to Circle and Tether. Dragonfly partner Haseeb Qureshi publicly alleged bidding irregularities, suggesting Native Markets was pre-selected. Whether rigged or not matters less than the signal Hyperliquid sent: my platform processes this volume daily; whoever mints the governing token answers to me.
USDH operated for eight months, peaking at a circulation of just $100 million—a rounding error compared to the $5+ billion in USDC circulating on Hyperliquid. Without user adoption, even flawless mechanics yield nothing. On May 14, 2026, Hyperliquid, Circle, Coinbase, and Native Markets issued simultaneous announcements halting USDH development, restoring USDC as the platform’s sole settlement asset. Native Markets sold USDH brand asset rights to Coinbase before walking away.
On the surface, it looks like Circle won, but scrutinizing the terms reveals the actual victor. The new arrangement, termed AQAv2, hinges on three clauses: First, roughly 90% of cost-adjusted USDC reserve yield flows to the Hyperliquid ecosystem, surpassing the 50% native markets initially committed to under the USDH model. Second, both Circle and Coinbase stake 500,000 HYPE tokens; if returns fall short of covering costs, staked tokens may be slashed. Third, Coinbase manages treasury deployment while Circle handles tech integration and cross-chain execution—their labor combined leaves the lion’s share of revenue for the platform. Accruals began August 26, with the inaugural $14.58 million payment arriving October 3. The crux of this narrative isn’t that Hyperliquid failed self-issuance—it strategically manufactured competitive tension, forcing Circle and Coinbase to commit tangible HYPE collateral and surrender ninety percent of profits merely to retain access to this distribution funnel. Hyperliquid may not have won the war of “self-issuing stablecoins,” but it absolutely secured victory in “everyone mints, but everyone pays rent to the platform.”
Polymarket pursued a completely different tactic: instead of negotiating profit splits, it internalized the entire settlement layer. In April 2026, Polymarket upgraded, replacing its platform collateral from USDC.e with its proprietary pUSD. Backed 1:1 by USDC and enforced via smart contracts, pUSD superficially resembles a wrapped token, but structurally shifts power dynamics. Users can deposit virtually any asset—ETH, DAI, WBTC on Ethereum; USDT, WETH on Polygon; SOL, USDe on Solana; even Bitcoin and Tron-native USDT—totaling over twenty supported tokens. The system auto-bridges and swaps everything into pUSD for internal settlement. While users continue seeing dollar-denominated balances, Polymarket seizes control of the on-platform settlement layer while underlying reserves remain anchored to USDC. Polymarket then monetizes these trapped reserves, generating easily over $50 million annually in interest income. Rather than displacing issuers, it demotes issuers into base-layer assets while transforming its own contextual settlement layer into the actual value-capture layer.
Exchange and prediction market operators do this, but even meme-token launchpads are adopting the playbook. Pump.fun’s treasury has already accumulated nearly $2 billion. Co-founder Noah Tweedale publicly stated in September 2026 that the team is deliberating on launching a proprietary stablecoin, bluntly noting decentralization matters less than capturing terminal-user touchpoints. Pump.fun initially migrated trading pairs from SOL to USDC to resolve revenue settlement friction, with the next step being self-issuance to permanently lock reserve yield within the platform ecosystem. On BNB Chain, Four.Meme skipped waiting for self-issuance, directly designating UXUY’s UUSD as its official launchpad asset, hardcoding UUSD as the default quoting pair into the creation workflow. Meanwhile, Solana-based LetsBonk partnered with Trump-family-affiliated World Liberty Financial (WLFI). WLFI deployed genuine capital incentives, rewarding developers who paired newly launched tokens with USD1. On January 3, 2026, LetsBonk hit 8,800 daily token creations, catapulting its market share from 3% to 30%. Even meme-launchpads recognize a simple truth: whoever embeds stablecoins natively into their own context controls the money tap.
Three Generations of Evolution
Looking backward, the stablecoin category has quietly cycled through three distinct generations.
Generation one, labeled Issuer-owned, concentrated ownership with publishers. Tether and Circle typify this era, where issuers simultaneously controlled branding, minting rights, reserve yields, and channel relationships. Distributors merely served as logistical couriers collecting modest handling fees. During this phase, issuance functioned as a licensed privilege; securing compliance status meant passively harvesting Treasury yields. Tether saturated emerging markets leveraging first-mover advantages and Tron’s negligible transfer fees, while Circle captured US institutional capital through Coinbase partnerships and regulatory pedigree. Both carved out exclusive fiefdoms, leaving distribution partners without bargaining power.
Generation two operates under the framework of Issuer-as-Infrastructure, wherein publishers devolved into pure infrastructure providers. Paxos epitomizes this shift: formerly manufacturing BUSD for Binance and PYUSD for PayPal, it now produces USDG for the Global Dollar Network, functioning essentially as a stablecoin contract manufacturer. Paxos explicitly bifurcated its portfolio into White Label Stablecoins acting purely as OEM suppliers, alongside GDN-integrated USDG distributions voluntarily sharing reserve yields with network participants. Publishers retained only regulatory licenses, custody solutions, minting mechanics, compliance frameworks, and API endpoints; branding belonged elsewhere, user acquisition occurred downstream, and by the USDG iteration, even dominant yield-allocation authority surrendered upstream. Issuers regressed from decision-makers into outsourced compliance contractors, executing whatever directives channels dictated. Paxos survives not despite this reality, but because it recognized a hard market truth: regulatory badges depreciate rapidly while user-acquisition gateways appreciate exponentially.
Generation three manifests as Platform-owned Settlement, where ecosystems seize complete control over settlement layers. This dynamic currently unfolds across multiple verticals. OpenUSD onboarded 140+ institutional partners, establishing shared underlying ledgers and pooled reserves. Visa, Stripe, and Shopify each govern distinct contextual use-cases while distributing revenue proportionally to traffic volume, even allocating equity stakes to partners catalyzing network expansion. PayPal adopted an even more aggressive posture via PYUSDx, enabling unrestricted enterprise self-branding of stablecoins without touching custodial reserves or navigating compliance pathways. PayPal and MoonPay transformed issuance into a plug-and-play configurable module activated merely through API ingestion. Within this architecture, platforms dominate user gateways, quote assets, contextual clearing, trade routing, and yield-sharing matrices, reducing native stablecoin issuance itself into increasingly commoditized base infrastructure.
Yet generational progression transcends simplistic narratives declaring “issuers lost, platforms won.” Issuers aggressively counter-attack by vertically integrating along upstream/downstream value chains. Circle autonomously advances its Arc blockchain initiative attempting sovereign infrastructure reclamation; Stripe simultaneously engineers Bridge stablecoin conversions, acquires Privy for wallet aggregation, deploys stablecoin debit cards, and expansively penetrates both issuance and foundational stacks; Visa architected dedicated Stablecoin Platforms integrating native settlements into established global networks; Tether aggressively expands into merchant payments, private L1 deployments, integrated wallets, and brick-and-mortar physical infrastructure. Every participant constructs parallel transportation arteries precisely terrified of collapsing back into mere minting assembly-line contractors.
Ultimately, three-generation evolution boils down to one principle: barriers to minting currency continuously deflate while thresholds for constructing payment infrastructure relentlessly escalate.
The Power of Interfaces
Gazing historically at combined USDT and USDC market-cap dominance yielding upward of eighty percent now carries zero analytical utility.
Issuance authority depreciates while distribution rights appreciate. Cross-chain bridges and intelligent routing render inter-stablecoin swaps virtually zero-cost. Whether users transact via USDT, USDC, or pUSD grows increasingly irrelevant. What genuinely matters involves securing immediate checkout consideration and fusing custom settlement layers deeply into commerce, remittance, border-crossings, and bank-clearing workflows. Tether’s defensive moat derives not from Treasury holdings but entrenched penetration across millions of wallets plus physical retail nodes throughout emerging economies; its distribution network spreads organically yet fiercely uncontainable by any singular upstream chokepoint marking its fundamental divergence from Circle. Circle aggressively purchases strategic OKX equity, surrenders ninety percent profit-share metrics on Hyperliquid, self-funds Arc chain construction—all acknowledging that pristine compliance documentation and fully backed reserves prove worthless absent direct terminal-user delivery pipelines.
Future landscapes reject singular super-stablecoins achieving universal monopoly dominance. Instead, autonomous contextual settlement tokens proliferate across infinite application boundaries. Exchanges deploy exchange-native currencies, prediction markets operate isolated betting frames, meme launchpads circulate community-specific mediums, commercial banks maintain restricted ledger formats, card networks govern proprietary credit architectures. Underlying layers potentially share consolidated reserve pools and standardized compliance registries, though visible operational channels diverge completely while revenue allocation follows proportional traffic attribution. Early-stage VC firm a16z articulated precisely that stablecoin breakthrough opportunities bypass disrupting Visa’s existing 150-million merchant footprint entirely; rather they address previously invisible merchant cohorts lacking physical storefronts possessing zero legal entity formation histories or established credit scoring profiles yet capable accepting digital transactions seamlessly via simple wallet signatures. Machine-to-machine micro-transactions between autonomous agents execute even more brutally efficiently through x402 protocols embedding native crypto-checkout commands directly inside standard HTTP packets rendering traditional processing surcharges functionally disposable taxes optimizable away algorithmically.
These represent tomorrow’s infrastructure.
This conflict transitions decisively away from “who owns dollars” toward “who controls entry-points feeding fiat-value into actual economic activity.” Currency manufacturers increasingly resemble base-tier raw material producers while gateway controllers command decisive geopolitical leverage. Entities successfully delivering tokenized units directly into consumer pockets bridging functional last-mile logistics universally dictate future regulatory frameworks. Legacy market-share percentages constitute yesterday’s historical record; interconnected interface architectures define tomorrow’s operational paradigm.















