For more than five months, as the U.S. Congress has debated the Digital Asset Market Clarity Act (CLARITY Act), banks and crypto firms have clashed over a seemingly simple question: should stablecoins be allowed to pay interest?

Wall Street argues that interest-bearing stablecoins could trigger deposit outflows from banks, while Silicon Valley contends that prohibiting interest payments would stifle innovation. The White House has attempted to mediate between the two camps, while President Donald Trump has publicly urged the banking industry to loosen its stance.

Yet the debate between deposit tokens and stablecoins is fundamentally misguided. Advocates of deposit tokens portray stablecoins as unregulated shadow banks, while the stablecoin supporters dismiss traditional banks as dinosaurs with little relevance in a blockchain-based world. Both sides overlook the critical reality that banks create monetary credit, while stablecoins move money efficiently. The financial system needs both.

The real question is not who wins, but how three forms of digital money should divide responsibilities in an era when financial transactions increasingly operate at machine speed. Should stablecoins, tokenized deposits, and central bank digital currencies (CBDCs) compete, or should they complement one another?

On April 7, 2026, the International Monetary Fund (IMF) provided perhaps the clearest answer to date in its Tokenized Finance report. The report outlined a future digital monetary system built upon a three-layer architecture.

According to the IMF, tokenized finance is giving rise to three distinct forms of digital money. Tokenized Deposits are digital representations of traditional bank deposits that preserve the existing two-tier monetary system while adding programmability. Regulated Stablecoins are privately issued digital tokens backed by high-quality reserve assets and designed for global circulation; and wholesale CBDCs (wCBDCs) are digital claims on central banks used for settlement among financial institutions and market infrastructures.

Each serves a different purpose and allocates risk differently. Tokenized deposits carry bank credit risk beyond insured limits while still relying on existing prudential supervision and deposit insurance frameworks. Stablecoins offer global accessibility and technological innovation, yet their stability depends on both reserve quality and the issuer’s ability to honor redemptions. Wholesale CBDCs eliminate settlement risk, but require central banks to build entirely new digital infrastructures.

The three layers can be understood as follows.

Layer One: CBDC as the Sovereign Settlement Anchor

CBDCs should serve as the ultimate settlement asset for systemically important tokenized transactions. Their role is to preserve the singleness of money: the principle of equal value across institutions and platforms. As the lowest-risk settlement asset available, CBDCs represent the final line of sovereign monetary authority, and cannot be delegated to private entities.

Layer Two: Tokenized Deposits as the Engine of Credit Creation

Tokenized deposits preserve the core logic of fractional-reserve banking. Deposits remain within the banking system, allowing banks to extend loans and support economic growth. Tokenized deposits enable payment, settlement, and liquidity management to occur within a unified framework, facilitating atomic settlement and reducing reconciliation costs by bringing deposits onto the blockchain. Most importantly, meanwhile, tokenized deposits retain the banking system’s unique ability to create credit – a function stablecoins cannot replicate.

Layer Three: Stablecoins as Borderless Distribution Networks

Stablecoins such as USDC and USDT are backed by fully reserved assets and offer something different: unrestricted, global money movement outside traditional banking networks. Their advantage lies in bypassing correspondent banking chains, SWIFT delays, and high transaction fees. A business can pay a supplier in Argentina at 11 p.m. on a Saturday and have funds delivered within seconds. Each layer performs a distinct function, forming a complete monetary architecture.

Consider a practical example. A Fortune 500 company holds tokenized deposits at JPMorgan through JPMD. In return, it receives access to low-cost credit facilities because its deposits remain within the banking system. When it needs to pay a supplier in Argentina over a U.S. holiday weekend (when traditional SWIFT transfers are unavailable) the supplier prefers to receive USDC. The solution is straightforward: the company converts JPMD into USDC through an atomic transaction on-chain with no settlement risk.

The significance of this example lies not in the technology itself, but in the architecture behind it. Tokenized deposits create low-cost credit, functioning like the heart and arteries of the financial system. Stablecoins distribute liquidity to the economic periphery, much like capillaries reaching the furthest reaches of the body. Neither replaces the other; they are complementary components of a single ecosystem.

The emergence of the AI agent economy has made this three-layer architecture increasingly critical. Goldman Sachs has already embedded Anthropic engineers within its technology teams to develop autonomous agents capable of handling transaction reconciliation, compliance reviews, and customer authorization – processes long considered difficult to automate due to their reliance on document-intensive judgment.

The banking industry is moving from AI-assisted operations toward AI-authorized transactions. As agents begin independently executing financial activities, the requirements for monetary infrastructure change dramatically. Humans may be able to tolerate T+1 settlement cycles and banking hours, but AI cannot.

The IMF report highlights how tokenized systems transform settlement from discrete events into continuous, real-time processes, eliminating the temporal buffers that regulators traditionally relied upon to intervene before settlement became final. Liquidity stresses that once unfolded over days could emerge within minutes—or even seconds.

The three-layer monetary model aligns naturally with the needs of autonomous AI agents. Stablecoins provide always-on, global, permissionless payment rails that support machine-to-machine micropayments. Tokenized deposits provide credit and liquidity backing for larger business transactions executed by AI systems; and CBDCs provide trusted final settlement assets and preserve monetary stability.

Under this framework, innovative private institutions continue to serve customers, while central banks maintain ultimate responsibility for settlement and monetary unity. Agents operate at the front end, while the three-layer monetary architecture supports them at the back end. Remove any layer, and the system becomes less resilient.

The rise of tokenized finance and AI-driven economies presents Taiwan with three strategic pathways. First, the Central Bank should advance the design of a wholesale CBDC framework as the foundational settlement layer upon which tokenized deposits and stablecoins can operate. Recent statements by Governor Yang Chin-long have reinforced the position that CBDCs should serve as the digital form of final settlement money, providing an anchor for public trust in Taiwan’s emerging tokenized economy.

Second, regulators should consider allowing domestic banks to issue tokenized deposits on open ledger infrastructure. Global institutions such as HSBC, Citi, and JPMorgan have already developed tokenized deposit systems within private networks. Taiwan’s banks possess strong deposit bases and customer trust, positioning them to become significant issuers of tokenized deposits in the Asia-Pacific region. Opening regulated pathways to permissioned public blockchains could connect Taiwanese businesses to international tokenized liquidity networks.

Third, Taiwan must define the strategic role of a New Taiwan Dollar stablecoin. If an NTD stablecoin serves only as a domestic payment tool, it will struggle to compete internationally against dollar- and euro-based stablecoins. If it becomes integrated into Taiwan’s globally significant semiconductor and AI supply chains, however – potentially serving as a settlement currency for tokenized capital markets – it could gain a distinctive competitive advantage.

The IMF concludes its report with a simple warning: “The window for shaping the architecture of the tokenized financial system is open, but it will not remain so indefinitely.”

That observation applies equally to Taiwan. The three-layer architecture of CBDCs, tokenized deposits, and stablecoins is not an optional experiment, but is rapidly becoming the foundational infrastructure of the agentic economy. Those who actively design this architecture will become the architects of the future monetary system, while those who wait will merely adapt to frameworks designed by others.

Taiwan’s opportunity lies not in waiting for the United States to resolve its own debates, but in proactively developing a clear framework that defines how CBDCs, tokenized deposits, and stablecoins can coexist within Taiwan’s financial ecosystem to support the next generation of AI-driven economic activity.

The author is a co-founder of TokenBacon Taiwan Co., Ltd.