Shinhan and SBI Want Yen Stablecoins for Korean Tourists. The Real Prize Is an $80 Billion Trade Lane

Can a stablecoin beat a credit card at a convenience store in Tokyo? Shinhan Bank and SBI Holdings think it can, at least for the nearly 9 million Koreans who fly to Japan every year.

The two firms are teaming up to test yen-denominated stablecoins for tourist spending between South Korea and Japan. On paper the scope is modest. They'll look at safeguards for travelers and study whether a tokenized yen can speed up settlement between the two countries. That second clause is the one worth reading twice.

Because tourist payments are the demo. Trade settlement is the business.

What the numbers say

Start with the traffic. Japan hosted a record 36.8 million inbound visitors in 2024. South Korea was the single largest source market, at roughly 8.8 million arrivals. Seoul to Fukuoka is a 90 minute flight. Seoul to Tokyo is under three hours. That's not really tourism. It's a commuter lane with duty-free shops bolted on.

Now the money. Korean visitors spend billions of dollars a year in Japan, and every one of those transactions passes through a card network that takes a cut, an FX spread that takes another, and a foreign transaction fee that usually runs somewhere between 1% and 3%. On a 12,000 yen dinner, nobody notices. Across a national tourism flow, it adds up to real money moving in the wrong direction.

Then there's the part nobody photographs. Bilateral trade between South Korea and Japan runs around $80 billion a year. A wire between a Busan importer and an Osaka supplier still goes through correspondent banks. It takes a day or two. It costs $25 to $50 in fees before anyone touches the FX spread, and the paperwork is worse than the pricing.

Trade finance is a $5 trillion market running on fax machines and PDF attachments. That's not a knock on the people doing the work. It's a description of the rails.

Why this corridor, and why now

Two regulatory doors opened at almost the same moment, and that's not a coincidence you can ignore.

South Korea spent nine years keeping corporations out of crypto. That ban is gone, phased out in stages, and banks are now allowed to think out loud about digital assets without their compliance teams reaching for the antacids. Japan went a different route. Its revised payment rules made stablecoins a licensed product, restricted to banks, fund transfer providers and trust companies rather than anyone with a smart contract and a dream.

So you've got a Korean bank with a distribution base and a Japanese financial group with licenses, a brokerage arm and a crypto trading venue already in house. SBI has been in digital assets longer than most of the people covering them. Shinhan has been running blockchain pilots since the days when every bank press release said blockchain and meant nothing.

Put those two together and the yen stablecoin pitch writes itself. A Korean tourist lands at Narita, loads yen into a wallet, pays a merchant, and the merchant settles without waiting on a card network. No FX spread on the consumer side. No three day float on the merchant side.

That's the pitch. Here's my problem with it.

The skeptical read

A closed-loop stablecoin between two banks is a prepaid card with extra steps. If a Korean traveler can only spend at merchants that both Shinhan and SBI have signed up, the product competes with a travel card, not with money. The container doesn't care about your consensus mechanism, and neither does a ramen shop owner who already accepts three credit cards and a transit card.

Merchant acceptance is the whole ballgame. Not the token. Not the chain. Not the consensus layer. Whoever signs up the most shops in Myeongdong and Shibuya wins, and right now neither bank has said a word about how many that's.

My second take is less popular. The banks are framing this as a consumer convenience story because consumer stories get press. But consumer payments are a low-margin, high-friction business where the incumbent has 60 years of habit on its side. The actual margin sits in the B2B leg, where a tokenized yen moving against a tokenized won can compress settlement from two days to minutes and cut the document handling that eats up back offices on both sides. The ROI isn't in the token. It's in the 40% reduction in document processing time.

If this pilot stays in the tourist lane for two years, it will quietly die. If it graduates to bills of lading and letters of credit, it becomes the most interesting thing happening in Asian trade finance.

What to watch

Three concrete things, and none of them are price charts.

First, merchant count. When the two firms publish pilot results, look for a hard number of accepting locations. Anything under a few thousand and this is a marketing exercise. Anything over 10,000 in a single metro area and the card networks should start paying attention.

Second, the settlement leg. Watch for language about invoices, purchase orders or bills of lading. If the pilot scope mentions documents rather than just payments, the banks are aiming at trade. If it only mentions wallets and point-of-sale terminals, they're aiming at tourists and hoping nobody asks about unit economics.

Third, timing on the Korean side. The corporate account rollout is moving in phases, and each phase adds a new class of institution that can hold digital assets on balance sheet. Every phase is a permission slip for a bank to move from pilot to product. Japanese regulators, meanwhile, have been predictable and slow, which is exactly what a bank wants when it's building settlement infrastructure.

Enterprise blockchain is boring. That's why it works. A pilot that quietly moves yen and won between two regulated institutions, with attestation logs and auditable provenance on every transaction, won't trend on crypto Twitter for an afternoon.

It might still be the most important thing either bank does this decade. And if it fails, it'll fail for the oldest reason in payments. Not enough merchants, not enough volume, and a consumer who'd rather just tap the card already in their wallet.