Deposit growth at stablecoin-exposed regional banks slowed only mildly since the GENIUS Act — and the signal is not statistically distinguishable from noise

Summary

Weak evidence of a slowdown, not a clear one. Using FDIC call-report deposits for five regional banks Standard Chartered's January 2026 research named as most exposed to stablecoin-driven deposit outflows (M&T, Truist, Citizens, Regions, KeyBank), year-over-year domestic deposit growth ticked down slightly, from +2.87% in the year before the GENIUS Act was signed (July 18, 2025) to +2.08% in the year after (dollar-weighted, June 2024-June 2026). Over the same window, four diversified/money-center banks (JPMorgan, Bank of America, Wells Fargo, Citibank) accelerated, from +3.94% to +5.89%. That gap is directionally consistent with the "exposed banks feel it more" thesis, but at the individual-bank level the pattern is inconsistent (3 of 5 exposed banks slowed, 2 accelerated) and a two-sample t-test on the nine banks' growth deltas is not statistically significant (p=0.48). Two of the most commonly cited "exposed" banks, Huntington and Fifth Third, had to be excluded because 2026 mergers (not stablecoins) roughly doubled their reported deposits. Overall: there is a real, small deceleration among exposed regionals and a real, larger acceleration among diversified banks in the same window, but with only 14 months of post-Act data and n=9 banks, this cannot be called a confirmed stablecoin effect — it is consistent with a mild slowdown, but just as consistent with ordinary bank-to-bank noise.

What "clearer regulatory shape" means here

The event used as the before/after pivot is the GENIUS Act (Guiding and Establishing National Innovation for U.S. Stablecoins Act), signed into law July 18, 2025 — the first comprehensive federal framework for payment stablecoins, requiring 1:1 reserves in cash/Treasuries and banning stablecoin issuers from paying yield directly to holders. That yield ban is exactly why banks lobbied for the bill: without it, stablecoins would compete even more directly for interest-sensitive deposits. A separate market-structure bill governing stablecoin custody/exchange rules was still stalled in the Senate as of late January 2026 per Standard Chartered, so "clearer shape" as of today means GENIUS Act implementation, not full regulatory closure.

Which banks are "most exposed" and why

Standard Chartered's Geoff Kendrick (Jan. 27, 2026 research note, via CoinDesk) screened 19 U.S. banks and securities firms by reliance on net interest margin (NIM) — the metric most directly threatened when sticky retail/transactional deposits migrate to stablecoins. The banks flagged with the largest deposit-outflow risk: Huntington Bancshares, M&T Bank, Truist Financial, and Citizens Financial Group. Standard Chartered's broader conclusion: "regional U.S. banks are more exposed on this measure than diversified banks and investment banks, which are least exposed." The firm projects roughly $500B could exit developed-market bank deposits into stablecoins by 2028, versus Citigroup's wider $182B-$908B range through 2030. I added Regions and KeyBank (comparable super-regional, NIM-heavy franchises) to the exposed group to raise the sample past n=4, and used JPMorgan, Bank of America, Wells Fargo and Citibank — the classic diversified/money-center comparison group — as the control.

Method and the M&A confound

For each bank I pulled quarterly domestic deposits (FDIC call-report financials, banking.financials) at three anchor dates: 2024-06-30, 2025-06-30 (approximately when GENIUS was signed), and 2026-06-30 — giving two equal-length, one-year windows straddling the signing, per the standard before/after comparison discipline (matched window lengths, not an arbitrarily long baseline). Two commonly-cited "exposed" banks had to be dropped from the analysis: Huntington National Bank's reported deposits jumped from $180.6B (Dec. 2025) to $227.5B (Mar. 2026) — a completed bank acquisition, not deposit inflow — and Fifth Third Bank's deposits jumped 43% in the same window for the same reason (its pending Comerica acquisition). Including either would misattribute M&A-driven deposit jumps to stablecoin competition running the wrong direction. The remaining diversified-bank comparison also carries one known confound: Wells Fargo's post-Act acceleration (-0.74% to +11.62%) reflects the Federal Reserve lifting its 2018 asset-growth cap in June 2025, unrelated to stablecoins, and by itself explains most of the diversified group's apparent acceleration.

What the data actually shows

A Welch's t-test on the nine per-bank growth deltas (exposed mean change +0.20pp vs. diversified mean change +3.41pp) returns t=-0.77, p=0.48 — not significant at any conventional threshold, with n far too small to separate a real effect from bank-specific idiosyncrasy (asset caps lifted, mergers, one-off wholesale deposit swings).

Bottom line and caveats

• There IS a small, real gap: exposed regionals' dollar-weighted deposit growth eased about 0.8 percentage points while diversified banks' growth rose about 2 points, in the 14 months after GENIUS Act signing versus the 12 months before.
• That gap is NOT statistically distinguishable from noise given only 9 banks and 14 months of post-Act data — several individual exposed banks actually accelerated, and the diversified group's acceleration is mostly explained by one bank-specific event (Wells Fargo's asset-cap removal), not stablecoin-driven inflows to money-center banks.
• Two of the four banks most frequently named as "most exposed" (Huntington, and by extension Fifth Third among super-regionals) had their deposit series so distorted by 2025-2026 M&A that they could not be used at all — a real limitation on how cleanly "exposure" can be tested with call-report totals alone.
• Stablecoin market cap itself is still a small fraction of the ~$18-19T U.S. bank deposit base (roughly $300-500B as of early 2026 per various estimates), so any effect through mid-2026 would necessarily be small relative to normal deposit volatility — consistent with what the data shows.
• This is a descriptive comparison of two small bank groups over one post-period, not a causal estimate; it does not control for loan demand, rate-cutting cycles, or bank-specific balance-sheet strategy, any of which move deposits as much as or more than stablecoin competition would at this stage.

Every query behind this report

7 warehouse calls ran in this session, in order. Each is reproducible against the same snapshot.

SELECT cert, name, state_abbr, total_assets_thousands, total_deposits_thousands FROM banking.institutions WHERE name ILIKE '%Huntington%' OR name ILIKE '%M&T Bank%' OR name ILIKE '%Truist%' OR name ILIKE '%Citizens Bank%' ORDER BY total_assets_thousands DESC NULLS LAST LIMIT 20

SELECT cert, name, state_abbr, total_assets_thousands FROM banking.institutions WHERE name ILIKE '%M&T Bank%' OR name ILIKE '%Manufacturers and Traders%' OR name ILIKE '%JPMorgan Chase Bank%' OR name ILIKE '%Bank of America, National%' OR name ILIKE '%Wells Fargo Bank, National%' OR name ILIKE '%Citibank, National%' ORDER BY total_assets_thousands DESC NULLS LAST FETCH FIRST 20 ROWS ONLY

SELECT DISTINCT repdte FROM banking.financials WHERE cert = 9846 AND repdte >= DATE '2023-01-01' ORDER BY repdte

SELECT cert, repdte, domestic_deposits_thousands FROM banking.financials WHERE cert IN (9846,6560,57957,588,628,3510,7213,3511) AND repdte >= DATE '2023-01-01' ORDER BY cert, repdte

SELECT cert, name, state_abbr, total_assets_thousands FROM banking.institutions WHERE (name ILIKE '%Fifth Third%' OR name ILIKE '%Regions Bank%' OR name ILIKE '%KeyBank%' OR name ILIKE '%Comerica Bank%') AND active = true ORDER BY total_assets_thousands DESC NULLS LAST FETCH FIRST 10 ROWS ONLY

SELECT cert, repdte, domestic_deposits_thousands FROM banking.financials WHERE cert IN (6672,17534,12368) AND repdte IN (DATE '2024-06-30', DATE '2025-06-30', DATE '2026-06-30') ORDER BY cert, repdte

SELECT * FROM (VALUES
('exposed', -0.54), ('exposed', -4.67), ('exposed', 6.32), ('exposed', -2.54), ('exposed', 2.44),
('diversified', -1.79), ('diversified', -3.43), ('diversified', 12.35), ('diversified', 6.52)
) AS t(group_name, delta_pp)