This article was first published in Central Banking on January 26, 2026.
AMRO’s CEO Yasuto Watanabe argues that sustained policy attention is needed to ensure the foundation of trust remains resilient amid financial transformation
The rapid expansion of artificial intelligence (AI), digital payments, and tokenized finance is often framed as a story of efficiency: faster settlement, lower transaction costs, and broader financial access. Across Asia, these gains are already visible. But beneath the surface, a more fundamental shift is under way. Financial innovation is not eliminating the need for trust. It is reorganizing trust in ways that prove most consequential during periods of stress.
This matters because trust is the hidden infrastructure of financial stability. When confidence holds, transactions appear seamless. When it breaks, even the most technologically advanced systems revert to familiar dynamics: runs and fire sales, and the need for policy intervention. Technology may change the speed and scale of these processes, but it does not repeal them.
Digital payments have become core financial infrastructure across the ASEAN+3 region. Notably, digital wallets have seen rapid adoption in ASEAN, with hundreds of millions of active users by early 2025, embedding them firmly in everyday economic activity. In China, digital wallets account for more than 80 percent of e-commerce transactions, underscoring their central role in online commerce. US dollar–denominated stablecoins are increasingly being used for settlement and treasury management, particularly through financial hubs such as Singapore.
The benefits of digital wallets and stablecoins are clear: faster transactions, greater convenience, and lower barriers to participation. Yet from a financial stability perspective, the questions are familiar. What assets back these instruments? How liquid are the reserves? And how resilient are they when many users seek redemption at the same time?
Stablecoins, despite their technological wrapper, remain vulnerable to classic run dynamics. Their credibility ultimately rests on the soundness of issuers, the quality and liquidity of reserve assets, and the assumption that one unit can always be redeemed at par. When confidence falters, reserves must be liquidated rapidly, which can result in a fire sale of reserve assets such as government bond or commercial paper markets, potentially transmitting stress to the wider financial system.
The price dislocations observed in some stablecoins in 2022, when assumed one-to-one parity briefly failed, demonstrated how quickly confidence can evaporate. More recent episodes of market volatility have reinforced the same lesson: improved design reduces risk, but it does not eliminate it. Technology may change the transmission speed, but trust, once questioned, can erode quickly even when the underlying code is sound.
These dynamics intersect with monetary policy in increasingly important ways. Historically, US easing cycles have been followed by large and volatile capital flows to emerging ASEAN+3 markets, often amounting to hundreds of billions of dollars within a year. The growing role of token-based, non-bank dollar liquidity adds a new layer to this transmission mechanism.
Unlike traditional banking channels, tokenized instruments operate with lower frictions, near-continuous settlement and, increasingly, automated treasury and portfolio management. The result is not necessarily larger capital flows, but faster ones. Adjustment timelines compress, and market sentiment can be translated into cross-border movements with little balance-sheet absorption by regulated intermediaries.
AI further amplifies this effect. In practice, AI-driven liquidity and portfolio management tools can lead multiple institutions to adjust positions simultaneously, intensifying market movements during periods of stress. In ASEAN+3, where capital flows react quickly to US monetary policy shifts, such automation may compress adjustment timelines and magnify swings in both local and dollar liquidity. The risk is not isolated model error, but collective behavior that reduces diversity of judgment precisely when it is most needed, reinforcing the need for regulatory oversight and careful design.
As trust migrates from institutions to data and algorithms, the way it is engineered becomes a policy concern. Traditional financial systems rely on centralized data collection and management to establish credibility and compliance. While effective, this model carries systemic vulnerabilities: limited data portability, single points of failure, large-scale data breaches with wide spillovers, and excessive dependence on a small number of data custodians.
Experiences with national digital identity systems—such as in India and Singapore—illustrate both the gains in inclusion and efficiency, and the policy challenges associated with data concentration. These are not abstract concerns. Data failures can quickly become systemic failures when they sit at the core of financial infrastructure.
Privacy-preserving technologies such as zero-knowledge proofs and decentralized identity frameworks point to a different approach. They allow users to prove eligibility or compliance without revealing underlying information, reducing data concentration risks while preserving regulatory objectives such as Know Your Customer (KYC) and anti-money laundering (AML). Regulators in Japan and Europe have signaled cautious openness to such tools, emphasizing technology neutrality and user protection over rapid adoption.
Prediction markets offer another example of how trust and innovation interact. Properly designed, market prices can aggregate dispersed information and sometimes move ahead of official indicators. Poorly designed, they amplify noise. For policymakers, their value lies not in replacing judgment, but in complementing existing analytical tools, particularly as early-warning signals rather than forecasting engines.
The central question, then, is not whether innovation should proceed. Suppressing new technologies is neither feasible nor desirable. At the same time, their integration into the financial system requires careful attention to how risks evolve under stress. Without appropriate safeguards, new instruments and systems may introduce vulnerabilities that only become visible when market conditions deteriorate.
In this sense, trust functions as a form of public infrastructure. As finance becomes faster, more automated and more data-driven, the design and governance of trust increasingly matter for macro-financial stability. Ensuring that this architecture remains resilient require sustained policy attention alongside the pursuit of efficiency, inclusion and growth.