The Narrow-Banking Drift: Signal Scan

The Narrow-Banking Drift: How Stablecoins Are Quietly Reshaping Who Creates Credit

A weak signal in financial services: beneath the post-GENIUS "war for deposits", stablecoins are converting credit-creating fractional-reserve bank deposits into fully-reserved "narrow" money backed by Treasury bills, migrating credit creation out of relationship banking even at banks that capture the flows, while the GENIUS Act yield ban is already being arbitraged via exchange rewards programmes.

The consensus after the GENIUS Act is a "$6 trillion war for deposits" in which payment stablecoins drain money from community banks toward whichever institutions issue or partner with issuers. The non-obvious signal is structural: a growing share of US dollar money is being recomposed from fractional-reserve deposits, which create credit, into fully-reserved Treasury-bill claims that do not. Even banks that win the deposit race end up lending a smaller share of their assets. The 2026-2028 consequence is not which bank holds the deposit but how much aggregate bank credit the system supports, and the yield ban Congress wrote to slow the migration is already leaking through exchange rewards programmes.

Signal Identification

This is a structural shift in money composition, not a payments-rails story. Stablecoins resemble narrow banks: deposit-like liabilities matched one-for-one by short-dated, safe assets, with no maturity transformation and no lending. As the share of dollar money held in that form rises, the credit-creating capacity of the banking system contracts, regardless of where issuers park reserves. The unresolved question is whether the supporting institutions can keep extending credit at scale once a meaningful slice of money sits outside the fractional-reserve system.

Time horizon: 2-3 years (GENIUS Act yield-ban arbitrage already live in 2026; Standard Chartered projects a $2 trillion stablecoin market cap by end-2028) Plausibility band: Medium-High Geographic / Jurisdictional Scope: Primary: the United States, where the GENIUS Act, dollar stablecoins and a fractional-reserve banking system intersect. Spillover: the euro area and UK (MiCA, ECB); emerging markets via dollar-stablecoin adoption. Sectors exposed: commercial and community banks, stablecoin issuers and crypto exchanges, payments and fintech, Treasury markets, non-bank lenders and private credit, prudential and securities regulators.

What's Changing

Stablecoin aggregate market capitalisation reached $317 billion as of 6 April 2026, more than 50% growth since early 2025, though the market flattened in late 2025 and early 2026 (Federal Reserve Board, April 2026). The BIS frames the phenomenon directly: "A shift towards narrow banking through stablecoins would loosen the tight link between deposit-taking and lending that exists in a fractional reserve system," with credit supply migrating to non-bank financial institutions (Bank for International Settlements, April 2026).

The empirical core is a New York Fed staff paper using transaction-level data linking on-chain transactions to wholesale interbank payments. It documents the first evidence of liquidity-driven bank disintermediation: stablecoin partner banks face large increases in payment demand, hold substantially larger reserve balances, and "even as beneficiaries of stablecoin growth within the banking system, partner banks' loan share of assets contracts relative to peers" (Federal Reserve Bank of New York, February 2026). The winners of the deposit race lend less of what they win.

The yield ban is already being arbitraged. GENIUS prohibits issuers from offering yield but does not explicitly prohibit affiliate or third-party arrangements, and exchanges have moved to "rewards" programmes (White House Council of Economic Advisers, April 2026). Brookings reaches the same conclusion and notes banks responding with tokenised deposits (Brookings Institution, March 2026). Standard Chartered projects the stablecoin market cap to reach $2 trillion by end-2028, generating roughly $1 trillion in fresh Treasury bill demand (CoinDesk, February 2026).

Stablecoin market cap vs US bank deposits

Global stablecoin market cap (early April 2026) against roughly $8 trillion of US bank deposits (BIS; directional).

Disruption Pathway

The pathway runs in three stages. First, composition: a dollar moved from a checking deposit into a stablecoin is recomposed from a fractional-reserve liability funding loans into a fully-reserved claim backed by T-bills. Second, balance-sheet response: partner banks capturing stablecoin reserves see their loan share of assets contract, as the New York Fed shows empirically (Federal Reserve Bank of New York, February 2026). Third, credit migration: bank credit contracts on the margin and the adjustment runs through non-bank intermediaries, private credit and securitisation (Bank for International Settlements, April 2026).

Stress concentrates at three points. Community banks: the ICBA estimates community banks would lose $1.3 trillion in deposits and $850 billion in loans if stablecoins pay interest, an outcome rewards programmes inch toward (Brookings Institution, March 2026). Treasury markets: new T-bill demand could reach $2.2 trillion against $1.3 trillion of supply, a $0.9 trillion gap reshaping the marginal investor base (CoinDesk, February 2026). Regulatory arbitrage: MiCA bans interest-bearing stablecoins more broadly than GENIUS, creating a transatlantic incentive to route euro-area liquidity into US dollar stablecoins (Bruegel, May 2026).

Why This Matters

For bank boards, asset managers, Treasury debt managers and prudential regulators, the assumption that needs revising is that stablecoin growth is principally a payments and competitive-deposit story. On the available evidence, it is also a money-composition story with first-order consequences for credit creation, Treasury-market structure and the bank versus non-bank credit mix. Banks that build tokenised-deposit and stablecoin-distribution capability now will defend their position; those that wait will watch deposits leak through rewards arbitrage. Treasury debt managers face a procyclical structural buyer base; prudential regulators face a credit-supply question sitting outside the traditional bank perimeter. The next two cycles decide whether the GENIUS yield ban is enforced or quietly arbitraged into irrelevance.

Decision-action posture for this signal: Prepare. The structural mechanism is documented and the yield-ban arbitrage is live, but the magnitude of credit-supply contraction is contested. Banks, Treasury debt managers and prudential regulators should build tokenised-deposit, T-bill-demand and non-bank-credit monitoring capability now; community banks with concentrated deposit franchises are closer to Decide.

Counter-Argument

The strongest objection is that the credit-contraction effect is small and largely offset by equilibrium adjustments. The White House CEA models the GENIUS yield prohibition and finds that at baseline calibration, eliminating stablecoin yield increases bank lending by only $2.1 billion (0.02%) with a net welfare cost of $800 million; even worst-case stacking produces only $531 billion in additional lending (4.4%) and at most $129 billion (6.7%) at community banks (White House Council of Economic Advisers, April 2026). Bruegel argues the fear is overblown because stablecoin issuance does not change total banking-system deposits and central banks can recycle reserves to credit-creating institutions, with euro-area bank excess liquidity around EUR 2.3 trillion at end-Q1 2026 (Bruegel, May 2026). Brookings notes banks can securitise loans and other intermediaries can extend credit (Brookings Institution, March 2026).

Yet the counter-case rests on equilibrium adjustments that are themselves the signal. If bank credit contracts and non-bank credit expands, the prudential perimeter narrows and credit creation runs through channels with less capital, less liquidity regulation and less crisis-era backstop. "Small" aggregate effect is consistent with large compositional change, and the New York Fed's bank-level evidence already shows the compositional shift inside the banking system.

Implications

Taken together, the sources point to a durable recomposition of dollar money and the credit channels it supports, not a passing competitive skirmish. The inflection window is 2026-2028, defined by whether stablecoin growth resumes its 2025 pace and whether the yield-ban arbitrage is closed, tolerated or formalised. Winners build tokenised-deposit and non-bank-credit capability while the bank perimeter shifts; losers defend a deposit franchise against a competitor also recomposing the asset side. The contest is shifting from who holds the deposit to who creates the credit.

Early Indicators to Monitor

Stablecoin aggregate market cap resumes growth above the early-2026 plateau, with monthly net issuance turning positive.
A US bank regulator (OCC, FDIC, Federal Reserve) issues guidance treating affiliate stablecoin rewards as prohibited yield, or explicitly tolerates them.
Tokenised-deposit launches by major US banks scale to material balances and capture stablecoin-style payment use cases.
Treasury Quarterly Refunding Announcements explicitly cite stablecoin reserve demand as a factor in bill-issuance plans.
A New York Fed or BIS follow-up paper extends the loan-share-contraction finding to a wider sample of partner banks.

Disconfirming Signals

Stablecoin market cap stalls or contracts through 2026-2027, leaving the share of dollar money outside fractional-reserve banks negligible.
Empirical follow-ups fail to replicate the partner-bank loan-share contraction in a broader sample.
Non-bank credit channels (private credit, securitisation) fail to expand to offset any bank-credit contraction, prompting a policy reversal.
Regulators close the affiliate-rewards loophole credibly, and stablecoin growth slows as a result.
Central banks recycle reserves to credit-creating institutions in a way that visibly neutralises the compositional effect, as Bruegel suggests is possible.

Strategic Questions

Should banks build stablecoin-distribution and tokenised-deposit capability now, or wait for regulatory clarity on the yield-ban perimeter?
At what point do Treasury debt managers treat stablecoin reserve demand as a structural buyer base, with implications for bill versus coupon issuance mix?
Do prudential regulators extend their perimeter to cover non-bank credit channels that absorb the compositional shift, or accept a smaller bank perimeter as the new equilibrium?

Keywords

Stablecoins; GENIUS Act; narrow banking; fractional-reserve banking; tokenised deposits; bank disintermediation; Treasury bills; MiCA; non-bank financial intermediation; credit creation; payment rails; financial stability

Bibliography

Source tiers: Tier 1, governments, regulators and intergovernmental bodies. Tier 2, think-tanks, academic institutes, major consultancies and quality data providers. Tier 3, quality journalism and specialist trade press. Tier 4, vendor, company and practitioner sources, used only as directional corroboration.

Tier 1 Stablecoins in 2025: Developments and Financial Stability Implications. Federal Reserve Board (08/04/2026).
Tier 1 Effects of Stablecoin Yield Prohibition on Bank Lending. White House Council of Economic Advisers (08/04/2026).
Tier 1 Stablecoins: framing the debate (speech by Pablo Hernandez de Cos, BIS General Manager). Bank for International Settlements (20/04/2026).
Tier 1 Stablecoin Disintermediation (Staff Reports, no. 1185). Federal Reserve Bank of New York (February 2026).
Tier 2 Next steps for GENIUS payment stablecoins. Brookings Institution (03/03/2026).
Tier 2 A new strategy to contain stablecoin risks in the European Union. Bruegel (20/05/2026).
Tier 3 U.S. Treasury may boost T-Bill issuance as stablecoins eye $2 trillion market cap: StanChart. CoinDesk (23/02/2026).

Analyst inferences and editorial framing

Claim-fidelity self-disclosure. The narrow-banking framing and the NBFI credit-migration conclusion are verbatim BIS positions (20/04/2026). The loan-share-contraction finding for partner banks is a direct quote from the New York Fed staff report (February 2026). Market-size anchors and Standard Chartered projections are faithful summaries of the Fed Board, BIS and CoinDesk sources. CEA and ICBA numbers are verbatim from the CEA and Brookings sources. The "already being arbitraged" framing of GENIUS affiliate rewards is analyst synthesis of the CEA and Brookings sources, both of which treat the loophole as likely to be circumvented. The "$6 trillion war for deposits" lede is analyst characterisation of the post-GENIUS public narrative, not lifted from any single cited source.