A contrarian read on the next fintech cycle, and why the US narrative is missing the story.

Silicon Valley is not building agentic commerce for Europe. Europe is building it, quietly, on top of infrastructure the US spent the last decade trying to replicate. MiCA gives European players a regulatory perimeter US-primary stablecoin infrastructure cannot cross without local CASP authorisation. SEPA and PSD2 give European merchants payment rails Silicon Valley has spent a decade approximating. And Europe's enterprise-first B2B commerce mix is exactly the customer shape agentic payments need. The 2015–2020 fintech cycle rhymed the same way. Europe was underrated then. It won by 2020. Same story, same underrating, next cycle.

What is the current narrative getting wrong?

The default framing in 2026, repeated at every fintech conference, in every VC deck, in every Silicon Valley podcast, is that the US builds agentic commerce and Europe eventually implements it. Coinbase publishes x402. OpenAI + Stripe ship ACP. Google donates AP2 to the FIDO Alliance. Amazon Bedrock adds AgentCore Payments. Circle IPOs. And the European story reads as “MiCA slows us down; we will catch up later.”

That framing has three problems, and they compound.

First: MiCA is not a drag on European agentic payments. It is a moat around them. US-primary stablecoin infrastructure cannot serve significant European B2B volume without CASP authorisation under MiCA, meaning US players either become regulated European entities (18–24 months of legal work) or they route through European partners who already are. In either case, Europe holds the choke point.

Second: European B2B payment infrastructure was not built for consumer crypto. It was built for enterprise payouts, marketplaces, EOR contractor payments, and SaaS billing, which is exactly the buyer profile agentic B2B payments actually needs. The US is retro-fitting from consumer-crypto to B2B; Europe was already there.

Third: The tech is the easy part. The commercial part is Europe's game, and that means regulatory clarity, enterprise sales motion, and existing corporate treasury relationships.

How does MiCA actually work in Europe's favour?

MiCA, in force since December 2024, does three things that reshape the competitive landscape:

• Non-EU stablecoin caps. MiCA restricts the use of non-EU-denominated stablecoins for EU flows above certain thresholds, effectively capping them at €200M/day. For any European corporate treasury operating at scale, USDC alone is not a compliant settlement instrument. EUR-denominated MiCA-compliant alternatives (EURC via Circle Europe SAS, EURAU via AllUnity) are.
• CASP authorisation as market entry. Any entity providing stablecoin services in the EU needs a Crypto-Asset Service Provider (CASP) licence. This is a 12–24 month process with real capital requirements. It is not a paperwork exercise. It is a gate.
• EU-based settlement + reserves. MiCA requires that a significant share of the reserves backing EU-denominated stablecoins sits in EU-regulated banks. This forces settlement infrastructure into Europe, not out of it.

Together, these are Europe's regulatory moat. US-native players either become EU-regulated (structural cost) or they partner with EU-regulated players (structural dependency). In neither scenario does the US player own the flow.

Compare to the US. The federal stablecoin frameworks in progress, including the GENIUS Act, focus primarily on issuer requirements. They do not restrict European stablecoins in US flows. This asymmetry is what tilts the ground. Europe restricts US stablecoins for EU flows; the US does not restrict European ones in US flows.

What does European infrastructure already do that the US is still building?

Silicon Valley has spent the last decade trying to approximate SEPA. SEPA is what US fintech thinks of when it says “instant payments.” It is the single euro payments area, covering 36 countries with instant euro-denominated bank-to-bank transfers at near-zero cost. FedNow launched in 2023 and reaches roughly 1,000 US banks as of mid-2026. SEPA reaches every European bank and has since 2014.

PSD2 layered strong customer authentication and open-banking APIs on top of SEPA, mandating that European banks expose the payment initiation and account information APIs that US banks are still fighting Plaid over. For agentic commerce specifically, PSD2's account-holder authentication requirements combine with AP2-style agent mandate verification to give European merchants a cryptographically clean chain of authorisation that a US-only merchant cannot yet build without third-party middleware.

The Digital Markets Act, in force since 2023 and hitting Booking.com's rate parity clauses in July 2024, is the other quiet advantage. DMA-driven unbundling of European commerce marketplaces creates room for agent-payable endpoints that US antitrust has not yet approached at the same scope.

None of this is dramatic. None of it makes headlines. It is boring, foundational, cumulative. And it is why the 2020s European fintech cycle wins.

Who is actually the customer for agentic B2B payments?

Consumer crypto was the US framing for the 2020–2024 cycle. B2B commerce is the framing for 2026–2030.

Look at the customer mix: European platforms (marketplaces, EORs, SaaS billing), European corporate treasury (multinationals with cross-border payout obligations), and European agentic-commerce endpoints (hotels, publishers, tour operators moving direct against Booking.com and Amazon; see “The 20% question” for the hotel P&L version). None of these is a retail crypto buyer. All of them care about compliance, reconciliation, VAT reporting, and corporate treasury policy. All of them are already customers of the European regulated banking infrastructure.

Silicon Valley is trying to build the customer relationship from a consumer-crypto starting point. Europe is starting from the enterprise B2B relationship, where the actual money is, and adding the agentic layer on top.

What did we learn from the last European fintech cycle?

The 2015–2020 fintech cycle rhymed the same way. Silicon Valley was building consumer neobanks (Chime, Robinhood). Europe was building enterprise payment infrastructure (Adyen), consumer lending at scale (Klarna), regulated banking-as-a-service (Solaris), international cross-border money (Wise, Revolut), and payment orchestration (Payrails, Mollie).

By 2020, Adyen was the payment platform Uber, eBay, and Netflix ran on globally. Klarna was the buy-now-pay-later infrastructure of European retail. Revolut had more customers than any US neobank. Wise had reshaped international remittances. Stripe's Ireland base gave it the European corridor it needed. Every one of those companies was European. Every one built on European regulatory + infrastructure advantage.

Silicon Valley narrative in 2015 dismissed all of them. “Europe is over-regulated. Europe is slow. Europe will not produce a Stripe.”

Europe produced its own generation of category-defining fintech, and by 2020 those companies were setting the terms US fintech has been chasing ever since.

The same underrating is happening now. The 2026–2030 cycle follows the same pattern, and the same class of European players, with the same regulatory + infrastructure + enterprise-buyer advantage, is positioned to win the agentic-payments layer.

The clearest current signal: AllUnity's EURAU, Germany's first fully reserved MiCAR-compliant Euro stablecoin, issued by the BaFin-licensed joint venture of DWS, Flow Traders, and Galaxy. That is a European-issued, European-regulated, European-settled instrument that Silicon Valley cannot build without going through EU regulatory doors it has not yet opened.

What does this mean for European operators building now?

Three things:

• Build on the moat, not around it. MiCA-compliance is not a burden. It is a differentiator that gets sharper every quarter. Design for CASP compliance from day one; do not architect around it hoping for future exemptions.
• Own the corridor, not just the technology. European enterprise buyers care about SEPA-native fiat, EUR-denominated stablecoin settlement, and PSD2-compliant authentication. The technology stack (UCP, ACP, AP2, x402) matters less than the commercial + regulatory layer beneath it.
• The go-to-market advantage is enterprise, not consumer. The buyers are corporate treasurers, platform CTOs, EOR compliance leaders. Meet them where they already procure, through regulated banking channels and existing corporate finance networks, not in developer forums.

Bluerails is built for exactly this thesis. European-native, to-be MiCA-compliant, AllUnity-partnered, and positioned at the orchestration layer where Europe's regulatory + infrastructure moat compounds.

Frequently asked questions

Does this mean US agentic-payments companies lose? Not exactly. It means they lose the European volume. The US has its own consumer and retail market and will build stack for that. The European market is Europe's to own, covering corporate B2B, enterprise cross-border, and regulated commerce.

How does this compare to what happened with GDPR? GDPR was a compliance moat that forced US SaaS to build EU-specific infrastructure or lose access. MiCA is doing the same for stablecoin-native flows. The difference is that MiCA-compliant infrastructure is significantly harder to bolt on than GDPR compliance, which favours incumbents with regulatory patience.

What about the UK post-Brexit? The UK's FCA framework is friendlier to stablecoin infrastructure than MiCA in some respects, but it lacks EU-single-market scale. UK players either operate on the fringe of MiCA (adapter model) or build EU-regulated subsidiaries. Either way, Europe holds the reference standard.

When does this thesis play out? 2026 is proof-of-concept. 2027 is commercial validation; first significant European agentic-payments corridors moving meaningful volume. 2028–2029 is category consolidation. 2030 is the “obvious” era, when the narrative catches up to the infrastructure that has been quietly building underneath.

What should a European CTO do right now? Assume MiCA-compliance is the default requirement, not the special case. Pick European partners for the regulated legs of the stack (settlement, KYA, corridor routing). Instrument for agent discovery and agent-payable checkout now; the demand curve is 12–18 months from material, and the infrastructure work is 60–90 days.

What's next

The five articles that build on this thesis:

• Marketplaces are dead, long live the marketplace: the macro thesis
• How AI agents actually find and trust your site: the discovery layer
• Who owns the checkout?: the merchant control layer
• x402 in production: the settlement leg
• The 20% question: the hotel P&L case

If you are building or buying agentic-payments infrastructure for European operations, run your site through Bluerails' Agent Score scanner: URL in, score out, plus the UCP, ACP, and x402 endpoint code with AllUnity settlement plumbed underneath.

Europe wins from here. The question is which European player runs the layer.