From Information to Liquidity: How Stablecoins Reshape Bank Intermediation

From bank accounts to stablecoins: information, credit allocation, and aggregate risk

Christine A. Parlour, Uday Rajan, and Haoxiang Zhu, in this paper, develop a stylized theoretical model comparing two financial systems:

1. An account-based bank system, in which banks observe firms’ payment-account histories and use them to assess the probability of low cash-flow outcomes.
2. A token-based stablecoin system, in which this relationship information is unavailable, so non-bank financial intermediaries instead focus on firms’ probability of very high cash flows.

Each firm has a project requiring an initial investment and generating one of three outcomes: high, medium or low cash flow. The authors calculate investment and aggregate welfare under four information regimes: no information, full information, information about the low state, and information about the high state. They then compare the set of projects financed, aggregate investment, expected value and the distribution of realized risks.

Findings

• The authors show that, in a frictionless environment, it should not matter whether payments are made through bank deposits or stablecoins: financial intermediation can be reorganized without changing aggregate investment. The key difference arises because bank accounts generate valuable information about customers’ cash flows.
• In the model, bank lending based on information about low cash-flow risk tends to screen out projects likely to fail, although it can also reject some positive-NPV projects. Stablecoin-based finance, by contrast, shifts lending toward non-bank financial intermediaries (NBFI) that identify projects with strong upside potential.
• The token-based system produces more extreme outcomes: it finances relatively more projects with high upside, but also more projects with a high probability of very low cash flows. In the numerical example, bank financing generates higher aggregate value than NBFI financing, while NBFI financing produces a higher-risk portfolio with greater exposure to both successes and failures.

Parlour et al. conclude that stablecoin adoption may require regulation beyond the stablecoin issuer itself, including oversight of non-bank lending intermediaries, mechanisms for producing reliable transaction histories in a token-based system. The paper explicitly does not model maturity or liquidity transformation.

Stablecoin Risk
Authors: Christine A. Parlour, Uday Rajan, Haoxiang Zhu
From: University of California at Berkeley, University of Michigan, MIT

Stablecoin payments and bank liquidity

Michael Junho Lee, Donny Tou develop a theoretical model of banks that manage reserves and lend while servicing stablecoin issuers. They then link on-chain stablecoin issuance and redemption transactions to wholesale interbank payments through the Fedwire Funds Service. Empirically, they use a difference-in-differences design around new stablecoin-bank partnerships formed after the 2023 banking crisis.

Findings

• The authors identify a liquidity channel of bank disintermediation in addition to the traditional deposit-substitution channel. Banks servicing stablecoin issuers experience a substantial increase in payment activity and intraday reserve volatility: in the post-partnership period, daily interbank payment value rises by approximately 67% relative to pre-partnership levels.
• Partner banks respond by holding significantly more reserves while reducing the share of assets allocated to loans by approximately 14 percentage points. The implication is that even banks benefiting from stablecoin-related deposit inflows may operate a relatively narrow, liquidity-oriented business model rather than converting those deposits into additional lending.
• The authors conclude that stablecoins can affect banks not only by replacing deposits, but also by transmitting payment and redemption-related liquidity shocks to partner banks. At larger scale, this could concentrate reserves in a small number of institutions, increase aggregate demand for reserves and complicate monetary-policy implementation.

Stablecoin Disintermediation
Authors: Michael Junho Lee, Donny Tou
From: Federal Reserve Bank of New York