Crypto is the first asset banks trim when they pull back risk

Breaking down that 26% changes the picture. Inside the largest tier, digital assets are the third cut rather than the first: 33% of banks above $100bn are pulling back from private credit, 33% from commercial real estate, and 31% from digital assets ^[3]. Crypto tops the aggregate because it is the category small institutions single out, 26% at community banks and 29% at credit unions, more than any other option put to them, while core consumer lending stays put across the sample at 11% to 20% ^[10][11]. Which follows, if you think about it from the other side of the balance sheet: a community bank with no private credit book can only express caution where it has something to exit.

Set the 26% against the 60% cutting somewhere and roughly 43% of the derisking cohort names digital assets ^[12]. Set that 60% against the third holding exposure steady across every category ^[15] and about seven points go unexplained ^[13], which is probably rounding, though the release does not say.

The preference number deserves a closer look. Of the 80% agreeing that a stable, predictable regime is best for the long-term health of US banking, 56 points are strong agreement, so 70% of the agreement is emphatic rather than polite ^[5][14]. American Banker reads the caution as a demand for fixed guardrails rather than leniency, sharpest in digital assets, where shifting rules compound the market's own volatility ^[18]. That maps onto where the statutes actually sit: GENIUS has passed with implementation details still open, attention has moved to the more contentious CLARITY Act, and banks and fintechs both object to how that bill handles yield and deposits, with more implementation friction expected even if it is enacted ^[7][8].

There is reason to read this as a softer story than it appears. The survey does not report baseline exposure by tier or define what counts as derisking ^[20], so 26% might be most of the institutions that ever had a crypto relationship, or it might be the tidying of correspondent lines nobody was using; if it's the first, that's worse for crypto distribution than the aggregate implies, and if it's the second, it's just noise. And the respondents are risk and compliance professionals ^[1], the most conservative function in any building, though also the function that signs the memo.

This is probably wrong, but my read is that the binding constraint on crypto's banking access over the next year is a risk committee's tolerance for rules changing mid-underwriting rather than the count of bills passed, and American Banker's own prescription points the same way: fortify internal risk management and stress-test volatile portfolios, because external clarity may take time ^[6], and broad deregulation reduces neither compliance cost nor operational complexity ^[9]. Budget spent stress-testing a book you already own is budget not spent onboarding an asset class you do not. What would prove me wrong is a later wave in which digital-asset pullback drops beneath the 11% to 20% band that consumer lending occupies ^[11] while the biggest banks are still cutting private credit at 33% ^[3].