Bank of Korea paper mapped a bank-run risk in retail CBDC design

The Bank of Korea registered a research paper on February 7, 2019, finding that one form of central bank digital currency could make commercial banks more vulnerable to panics by drawing money away from their deposits. The result was conditional, not a prediction: the authors modeled an interest-bearing, account-based CBDC available directly to individuals and found that its effect changed when the central bank returned the displaced funding to commercial banks.

The paper mattered because it moved the CBDC debate beyond payment technology. It treated a digital claim on a central bank as a possible competitor to the demand deposits that help fund private credit. That placed bank balance sheets, reserve management and financial stability at the center of the design question.

A model, not an issuance decision

Bank of Korea Working Paper No. 2019-6 was written by Young Sik Kim of Seoul National University’s Department of Economics and Ohik Kwon of the Bank of Korea’s Economic Research Institute. Its publication did not announce a digital won, a pilot program or a policy decision by the central bank.

The modeled CBDC was denominated in the national currency, paid interest and represented an account-based claim on the central bank. Individuals could hold it through direct deposits at the central bank. Those assumptions are important: the paper did not analyze every possible retail or wholesale CBDC architecture, and it did not require the asset to resemble bitcoin or another privately issued cryptoasset.

The researchers constructed a monetary general-equilibrium model in which commercial banks provide liquidity, operate with fractional reserves and compete with CBDC accounts for deposits. The results therefore describe the internal behavior of that theoretical system. They are not measurements from Korean banks or observations from a functioning CBDC network.

How deposit substitution created risk

Inside the model, moving funds from commercial-bank demand deposits into CBDC accounts reduced the funding available for private credit. The smaller credit supply raised the nominal interest rate and lowered commercial banks’ reserve-to-deposit ratio. Banks consequently had less cash available relative to deposits, increasing the modeled likelihood of a panic in which withdrawal demands exhausted their reserves.

That chain is the paper’s central finding, but it was not unconditional. The authors also modeled the central bank lending all funds placed in CBDC accounts back to commercial banks. Under that arrangement, the loss of private credit was avoided. The additional funding lowered the nominal interest rate and could improve financial stability rather than weaken it.

The contrast made institutional design more important than the digital label. A retail CBDC that removed funding from banks produced a different result from one paired with a mechanism that recycled the funding into the banking system.

Why the question was consequential in 2019

The Bank of Korea paper appeared during widespread but cautious central-bank research. A Bank for International Settlements survey published on January 8, 2019, covered 63 central banks representing close to 80% of the world’s population and more than 90% of its economic output. The survey was conducted in the latter part of 2018.

Seventy percent of respondents were engaged, or expected soon to engage, in CBDC work. About half of those doing such work had advanced to experiments or proofs of concept, but only five central banks reported pilot projects. More than 85% considered issuing any form of CBDC somewhat unlikely or very unlikely within the survey’s one-to-three-year horizon.

The Korean paper supplied a reason for that caution. Issuing digital central-bank money was not merely a software deployment; under the modeled assumptions, it could rearrange deposits, lending and liquidity across the financial system.

What the record does not establish

No market-price claim is necessary to document this development, and the cited records do not demonstrate that the paper moved bitcoin, the won or bank shares on February 7, 2019. The study did not test a production ledger, measure depositor behavior or determine an optimal CBDC design. It also did not settle questions involving privacy, cybersecurity, legal authority, access limits or non-interest-bearing alternatives.

The verified development is narrower: Bank of Korea-affiliated research formally identified a bank-funding risk in one retail CBDC model and showed that central-bank lending back to commercial banks could reverse the result.