CBDC Architecture Choice: the 4 Major Tradeoffs

Wiki route

This entry sits under fintech index. Read it with Japan Financial Regulation — Legal Framework for Tokens, Crypto Assets, and Payments for adjacent context and Three-Layer Structure of Japan's Stablecoin Regulatory Regime (JPYC, USDC, Project Pax) for the broader system boundary.

[!info] TL;DR When a central bank chooses a CBDC architecture, it faces 4 core tradeoffs: survival of financial intermediation vs central-bank direct (decides the fate of commercial banks), monetary sovereignty vs cross-border interoperability (decides the relationship with wholesale-CBDC corridors), privacy vs anti-money-laundering (decides the central bank’s visibility), and room for innovation vs system stability (decides programmability). Every live CBDC design is a different combination of these 4 tradeoffs.

Key facts

• Direct CBDC marginalizes commercial banks → politically unacceptable → adopted by none of the G10 central banks ^•
• The two-tier architecture is strong domestically, weak cross-border → requires wholesale-CBDC corridors (mBridge / Agorá) ^•
• token-based multi-tier is naturally cross-border-friendly, but threatens the sovereign-currency hierarchy ^•
• Direct CBDC gives the central bank the strongest AML capability, but zero privacy ^•
• In the two-tier model commercial banks handle KYC, and the central bank does not see users directly ^•
• token-based multi-tier can tune privacy at the cryptographic layer, but compliance cost is high ^•
• The programmability of smart contracts is the main room for innovation in token-based multi-tier ^•

Mechanism / How it works

1. Survival of financial intermediation vs central-bank direct: Direct CBDC marginalizes commercial banks → collapse of financial intermediation → the central bank is forced to take on credit allocation (lending directly to companies) → politically unacceptable. All major central banks protect the bank-deposit base (two-tier or multi-tier). This is why none of the G10 central banks has adopted direct CBDC.

2. Monetary sovereignty vs cross-border interoperability: The two-tier architecture is strong domestically (a complete central bank → commercial bank → user control chain) but weak cross-border (it must borrow wholesale-CBDC corridors such as mBridge). token-based multi-tier is naturally cross-border-friendly (interoperability of token standards) but threatens the sovereign-currency hierarchy (a country’s tokenized deposits may circulate overseas). This is the logic behind e-CNY choosing two-tier + mBridge cross-border supplementation.

3. Privacy vs anti-money-laundering: Direct CBDC = the central bank is fully visible, AML is strongest but privacy is zero (greatest opposition in Europe). Two-tier = commercial banks handle KYC, the central bank does not see users directly, consistent with current financial-privacy norms. token-based multi-tier = privacy can be tuned at the cryptographic layer (zero-knowledge proofs / threshold signatures), but compliance cost is high and the regulator’s learning curve is steep.

4. Room for innovation vs system stability: token-based multi-tier (smart contracts) has the greatest room for innovation, but requires next-generation infrastructure (EVM / token standards / chain-native development stack). Two-tier is the most stable but has a low innovation ceiling (an account system is inherently not programmable). e-CNY chose two-tier prioritizing stability; DREX chose token-based prioritizing innovation.

Origin & evolution

2017-2020 The early CBDC frameworks of the BIS / IMF discussed almost only the “direct vs two-tier” dichotomy, and the innovation dimension had not yet surfaced. 2020-2022 Enterprise DLT such as Hyperledger / Corda verified programmability at the PoC stage → “token-based multi-tier” surfaced as a 3 option. 2022-2024 Each central bank weighed the tradeoffs based on its own political / regulatory preferences: China prioritized stability (e-CNY two-tier), Brazil prioritized innovation (DREX token-based), the eurozone compromised between intermediation protection and programmability (the digital euro’s hybrid + holding caps). 2026+ the three-way MRA (EU MiCA + US GENIUS + HK HKMA) + wholesale-CBDC corridors (mBridge + Agorá) further constrain the architecture-choice space.

Related

• Wiki Index
• CBDC Multi-Tier Architecture Overview
• 3 Major Active CBDC Paradigms
• Five-layer unbundling of central-banking functions
• GENIUS Act Section 18: Comparison of Foreign Regulatory Regimes

Sources