On October 16, Japan’s Mitsubishi UFJ, Sumitomo Mitsui, and Mizuho banks announced the successful completion of a proof-of-concept for a joint yen-stablecoin issuance and corporate cross-border payment system. The author believes that for stablecoins to become a long-term payment tool for enterprises, the key lies in backend processes such as identity verification, reconciliation, and anomaly handling, and that Taiwan should first validate these processes through concrete corporate cross-border payment scenarios. This is a guest article by Associate Professor Jian-Ping Zhong, who also serves as the Program Director of the Master’s Program in FinTech and Information Security at National Taipei University of Technology.
(Prior context: Japan's Three Major Banks Team Up to Launch Yen-Stablecoin, Igniting New Competition in Cross-Border Payments)
(Context: The Financial Supervisory Commission has opened a pilot program for banks to tokenize deposits; only Shin Kong Bank has indicated it will first focus on its own settlement system.)
One cross-border payment shows as successfully transferred on the blockchain—does that mean the bank’s work is done? The answer lies in the identity verification, fund allocation, compliance reviews, and account reconciliation behind the transaction. As stablecoins begin to integrate into the banking system, market focus is gradually shifting from “can it transfer funds” to “can it become a payment tool businesses are willing to adopt long-term.” The application of stablecoins in supply chain finance brings together the three pillars of traditional supply chains—funds, information, and logistics—by leveraging blockchain’s real-time settlement, peer-to-peer (P2P) transfers, low costs, and 24/7 operational capabilities to address key pain points in current cross-border payments: slow speed, high costs, low transparency, and excessive intermediaries.
Japan's three major banks complete a proof of concept for a Japanese yen stablecoin.
On October 16, 2026, Japan’s MUFG Bank, Sumitomo Mitsui Banking Corporation (SMBC), and Mizuho Bank announced the successful completion of a proof-of-concept for a joint yen-stablecoin issuance and corporate cross-border payment system. The test simulated fund transfers between Mitsubishi Corporation’s domestic and overseas offices, validating the issuance, transfer, and redemption processes, as well as reviewing related compliance and operational arrangements. According to previously disclosed plans, the three banks aim to launch live transactions by the end of March 2027; however, the official launch date, transaction fees, and commercial scale remain to be confirmed.
What makes this collaboration noteworthy is that previously competing banks are now jointly building a payment network accessible to enterprises. For multinational corporations, fund transfers often span multiple banks, countries, and time zones. If digital payment tools operate only within the closed ecosystem of a single bank, their effectiveness is inherently limited. A jointly issued arrangement has the potential to integrate the customer bases of different banks, expanding the network’s reach. According to the design unveiled by Japan’s three major banks, this yen-backed stablecoin operates under a trust structure, with trust assets backed by bank deposits, and is intended for circulation among users who have completed identity verification. This design reflects banks’ efforts to extend their existing trust, customer management, and regulatory frameworks to on-chain payments.
Corporate payments require looking at the entire process, not just on-chain transfers.
From an application perspective, internal fund allocation within a group is a logical starting point. For example, Japanese companies need to transfer operating funds to overseas branches; their finance departments not only care about when funds will be credited but also need visibility into payment status, foreign exchange costs, and available funds in each location. If a shared stablecoin system can be seamlessly integrated with bank accounts and corporate financial systems, it could reduce information gaps and redundant tasks. In the future, further evaluation could extend to supplier payments and cross-border trade settlements. However, these benefits must still be validated through actual operations. A bank completing a proof of concept indicates that the technical and operational model is feasible under specific conditions, but there remains a long way to go before large-scale commercial deployment. Whether companies adopt it ultimately depends on whether the overall cost, ease of use, and risk profile are superior to existing alternatives. While on-chain transfers can be executed 24/7, this does not mean that currency exchange, fiat disbursements, or bank receipt processes can all operate around the clock.
Therefore, when evaluating stablecoin payments, it is essential to examine the entire process. Before payment, the bank must verify the customer’s identity, transaction authorization, and recipient address; arrange funds and exchange rates; and conduct necessary anti-money laundering and sanctions screening. During transaction execution, wallet authorization, key security, and network status are involved. After the transaction, confirmation of receipt, accounting, reconciliation, and exception handling are required. On-chain confirmation is a critical execution point, but it cannot be equated with the completion of all settlement operations on the bank’s side. For businesses, these seemingly tedious backend processes are precisely what determine the reliability of the payment service. If funds are delayed or discrepancies arise, who is responsible for verification? How are funds handled in the event of a failed transaction? What procedures are used to resolve inconsistencies between on-chain records and bank accounting? Without clear answers to these questions, even the fastest transfer speeds cannot support daily business operations.
Taiwan's regulatory progress: Virtual Asset Service Act and Deposit Token pilot program
Has Taiwan already started preparing? From the perspective of regulatory frameworks and banking practices, the answer is yes, but the progress and nature of various initiatives must be clearly distinguished. The Financial Supervisory Commission has announced that the Virtual Asset Services Act was passed in its third reading on June 30, 2026, incorporating regulations on stablecoin issuance licenses, reserve assets, and information disclosure, while continuing to advance subordinate authorization rules. Legislative passage, regulatory implementation, and individual business approvals remain distinct phases. Another notable development is the FSC’s opening of pilot deposit token services to banks in September 2026, requiring applicants to detail issuance and redemption processes, accounting reconciliation, and security controls. Deposit tokens remain legally classified as bank deposits and must not be conflated with stablecoins; however, both involve bridging traditional accounting with on-chain records, offering financial institutions an opportunity to build relevant technical expertise and risk management experience.
AI Agent payments require clear boundaries of authorization.
The next variable to consider is the AI agent. When AI agents enter the real world to perform tasks, the traditional financial system becomes the biggest bottleneck—AI cannot open bank accounts, lacks identification (KYC), and cannot use credit cards. Blockchain and crypto tokens, however, provide a permissionless, programmable, and real-time settlement financial infrastructure. Together, they enable AI agents to evolve from mere “text generators” into autonomous digital economic entities.
If companies later authorize AI agents to assist with procurement, initiate payments, or issue fund transfer instructions, stablecoins may become one of the available payment options. However, the higher the level of automation, the more clearly defined the authorization boundaries must be—such as how much an agent can pay, to whom payments can be made, and under what circumstances human review is required. Model decisions cannot directly replace payment authorization, and transaction records must be sufficient to trace decisions and assign responsibility.
Where can you start in Taiwan?
For Taiwan, which is deeply integrated into the global supply chain, a more practical approach is to start by addressing specific challenges in cross-border corporate payments and receipts, select clear use cases, and validate settlement times, overall costs, reconciliation efficiency, and anomaly handling capabilities before deciding which digital payment tools to adopt. Interoperability between banks, integration with overseas financial institutions, and compatibility with existing enterprise systems will all influence final adoption decisions.
The collaboration among Japan's three major banks reminds us that for stablecoins to enter mainstream finance, banks must jointly invest in building payment networks and operational systems. As AI Agent applications advance, agentic commerce is ushering in new transaction models; in the future, under compliance with regulatory standards and risk management requirements, this will further expand into agentic finance. Taiwan has already begun preparing—next comes the challenge of aligning regulations, technology, and enterprise needs to make every fund transfer more efficient, with clear accounting records and defined accountability. This is the crucial step for stablecoins to move from a financial topic to everyday financial services.
📍Related Articles📍
BlackRock partners with SBI and Startale to tokenize fund, acquire JPY stablecoin