When I set up Propex's HK Ltd back in 2023, the banking question was simple: try for a real Hong Kong bank, fall back to an EMI like Airwallex or Statrys, call it a day.
Three years later that mental model is dead.
In the last twelve months, three things happened that most founders are still sleeping on:
The US passed the GENIUS Act in July 2025, giving stablecoin issuers a real federal framework.
The OCC granted national trust bank charters to Circle (USDC), Paxos, and Ripple — three more in early 2026.
On April 20, 2026 — exactly a week before this hits your inbox — the Bank for International Settlements published a warning that USDT and USDC expansion threatens Asian banking stability.
Translation: stablecoins are no longer a "crypto thing." In the US, they're bank-grade. In Asia, they're a regulatory pressure point. And if your company structure quietly assumes the old world — one EMI, one currency, one happy compliance officer — you're setting yourself up to be repriced or de-risked at exactly the wrong moment.
I'm not a lawyer. But I run three companies across HK, the BVI, and Indonesia, and I've watched my own banking stack get rewired twice in eighteen months. Here's what I'm doing about it — and what I'd tell any founder spinning up an entity in the next 90 days.
The new question isn't "where do I bank" — it's "what rails do I run on"
For most of the 2010s, structuring a company meant: pick a jurisdiction, get a real bank account, stress about wire fees, move on.
Today the question splits into two parallel rails.
Rail 1 — Fiat. A real bank or EMI in your operating jurisdiction. Mercury for US LLCs. Airwallex or Statrys for HK Ltd. Aspire or a clean local bank for Singapore Pte Ltd. Wise as the connector across everything else. This rail handles invoices, payroll, government filings, and the suppliers who still only speak SWIFT.
Rail 2 — Stablecoins. A treasury account that holds USDC or USDT, settles in minutes, and lets you transact globally without banking-hour holds. Until 2025, this rail was duct-taped together with personal exchange accounts and a prayer. After the OCC charters, you can now hold stablecoins inside a regulated US trust bank with the same legal status as a deposit. That changes the conversation completely.
You need both. Not one.
What this means for jurisdiction choice in 2026
The jurisdictions I'd evaluate right now look different than they did even six months ago. A few moved up the list, a few moved down.
Up:
US LLC (Wyoming, Delaware, New Mexico). Now with a clear federal stablecoin framework and OCC-chartered trust banks issuing USDC. If your customers are global and pay you in stables, this is the cleanest legal home for the dollar leg of your treasury.
Hong Kong Ltd. The territorial tax regime is unchanged. HK's own stablecoin licensing regime under the HKMA is now operational. Real banks here are still hard to land as a foreigner, but EMIs like Statrys remain the workhorse and HKD/USD/CNY conversion is essentially frictionless.
UAE Free Zone. Aggressive on Web3 licensing, predictable cost base, and Emirates NBD will actually onboard a foreign founder if your activity makes sense on paper.
Sideways:
Singapore Pte Ltd. Still excellent, but onboarding slowed in 2025 after MAS tightened beneficial-owner KYC. Plan for 6–10 weeks, not 2.
Down:
Anything where your stablecoin activity puts you in a gray zone with the local central bank. The BIS warning will be cited by Asian regulators for the next 12 months as cover to slow-walk approvals. If you bank in a country whose central bank is jumpy about USDT outflows, expect surprise account reviews.
The single biggest mistake I see founders make right now is picking a jurisdiction based on tax rate alone. In 2026, the binding constraint isn't tax — it's whether your fiat rail and your stablecoin rail can co-exist without one of them getting frozen.
The setup I'd run if I were starting today
Two-entity, two-rail. Same logic I designed for Companiz clients, refined after watching what actually survives contact with reality.
A holding entity in a jurisdiction with low friction on stablecoin treasury. US LLC or HK Ltd are the two I keep landing on.
An operating entity wherever you actually need to invoice, hire, or hold a local license. PT PMA in Indonesia if you're building locally. UAE Free Zone if you want a residence permit attached. Singapore Pte Ltd if you sell into Southeast Asia and want regional credibility.
The banking stack:
A real account at a stablecoin-friendly EMI — Mercury for the US LLC, Airwallex or Statrys for HK.
A treasury wallet (multisig, in the operating company's name) holding USDC for working capital. Convert to fiat only when a SWIFT invoice forces your hand.
A corporate spend card that runs on stablecoins for daily ops. Elephants is what I'm using right now for HK and SG flow — useful when you want to spend USDC without round-tripping through a bank wire.
This setup costs more to maintain than the old "one EMI account" model. Maybe $200–400/month (MercuryPro $299) in extra subscriptions and accounting overhead. But it gives you something the old model never did: optionality.
If your EMI freezes you — and EMIs freeze accounts more than people admit — your treasury keeps moving.
If a regulator slows your fiat onramp, your stable coin rail still settles.
If you want to pay a contractor in Buenos Aires at midnight, you don't wait for Monday morning in London.
What I'd actually do this week
If you're forming a new entity:
Don't pick a jurisdiction based on tax alone. Pick based on banking + stablecoin compatibility first. Tax is the third question, not the first.
Open both accounts on day one — fiat and stablecoin. Don't wait until you "need" the second one. The day you need it, you can't open it.
If you already have an entity:
Audit whether your current EMI lets you receive stablecoins directly. Most still don't. A growing number now do.
Stress-test the setup: if your fiat account froze tomorrow morning, could you keep paying people for 30 days without changing anything else?
The founders who get squeezed in 2026 are the ones who treated banking as a checkbox during incorporation. The ones who built two rails — fiat and stablecoin, in jurisdictions chosen for that compatibility — will move faster than everyone else for the next three years.
I learned this the hard way watching one of my EMIs hold a routine wire for two weeks because a compliance officer in a timezone I'd never met wanted "more context." If I'd had the second rail working then, I would have routed around it in an afternoon.
Build the second rail.
Need help figuring out which jurisdiction + banking combo fits your situation? Talk to our AI advisor at agent.companiz.xyz — it walks you through structure, banking, and tax in about ten minutes, no sales call.
The Nifty Founder is a bi-weekly newsletter about building, structuring, and owning in the new economy. Subscribe free at niftyfounder.com.