The US banking trade is moving from a rate story to a credit-selection story. For most of 2022 and early 2023, higher short rates created a mechanical earnings tailwind: floating-rate loans repriced quickly while legacy deposits lagged. That phase is over. The industry is now digesting the second-order effects of the Federal Reserve's 525 basis points of tightening: higher deposit betas, weaker loan demand, commercial real estate refinancing stress, and a consumer credit cycle that is no longer benign.
The important point for equity investors is that net interest margin, or NIM, is not collapsing across the system. It is normalizing from unusually favorable levels, and the dispersion between banks is widening. The FDIC reported an industry NIM of 3.17% in the first quarter of 2024, down from the prior quarter but still above the pre-pandemic average for many institutions. The market, however, is no longer willing to capitalize peak net interest income at peak multiples. Banks that can defend low-cost deposits, preserve tangible book value and absorb credit costs without cutting buybacks deserve premium valuations. Banks relying on wholesale funding or sitting on concentrated office exposure do not.
NIM Is Near the Trough, But the Path Is Uneven
The central margin question is not whether rates fall; it is whether deposit costs fall faster than asset yields. In the tightening cycle, most large banks enjoyed a positive asset-beta shock first, then a funding-cost shock later. By 2024, that lag had largely caught up. Interest-bearing deposit costs for many regional banks moved from near zero in 2021 to 3%-plus, while noninterest-bearing deposits continued to shrink as corporate treasurers and households shifted cash into money market funds yielding more than 5%.
That is why reported NII trends diverged sharply. JPMorgan Chase remained the sector's earnings anchor, with quarterly net interest income running above $23 billion in early 2024 and management still guiding to very high full-year NII relative to its pre-First Republic baseline. Wells Fargo, by contrast, saw NII decline as deposit costs and lower loan balances offset asset yields. Bank of America carried a different issue: its large fixed-rate securities book delayed asset repricing and created a visible drag, though it also gives BAC more upside as securities mature and are reinvested at higher yields.
For regionals such as PNC, Truist, U.S. Bancorp, Fifth Third and Regions Financial, the NIM discussion is more balance-sheet specific. Banks with stronger commercial operating deposits and less reliance on brokered deposits should see funding relief earlier if the Fed begins easing. Banks that paid aggressively to retain deposits in 2023 may find that deposit betas are sticky on the way down. The market is correctly assigning a higher multiple to deposit franchises that behave like core funding, not rate-sensitive wholesale liabilities in disguise.
My base case: industry NIM stabilizes before credit quality fully normalizes. That means the first improvement in bank earnings revisions may come from funding costs, but the durability of the rally will depend on charge-offs and reserve builds.
The Deposit Franchise Is Now the Main Valuation Asset
Before the 2023 regional bank shock, investors often treated deposits as a stable input. That was a mistake. The failures of Silicon Valley Bank, Signature Bank and First Republic demonstrated that deposit composition matters as much as deposit size. Uninsured deposits, concentrated industry deposits and rate-sensitive wealth deposits deserve a higher liquidity haircut than granular retail checking accounts.
The market has internalized that lesson. JPMorgan trades at a premium to tangible book not only because of scale, but because its diversified deposit base lowers funding volatility and supports structurally higher returns on tangible common equity. Wells Fargo's upside is tied to expense discipline and asset-cap relief, but its deposit franchise remains valuable. Bank of America trades with a discount partly because investors are waiting for tangible book accretion and AOCI recovery to translate into higher distributable capital.
For super-regionals, the valuation spread reflects confidence in deposits and credit, not just earnings-per-share estimates. A bank earning a 12%-14% return on tangible common equity with a stable deposit base can justify 1.3x-1.6x tangible book value in a 10%-11% cost-of-equity environment. A bank earning 7%-9% ROTCE with higher funding volatility should trade closer to 0.7x-1.0x tangible book. The math is straightforward in a residual income framework: with 3% long-term growth and a 10.5% cost of equity, a sustainable 13% ROTCE supports roughly 1.3x tangible book, while an 8% ROTCE supports only about 0.7x.
This is why reported NIM alone is not enough. A high NIM funded by expensive CDs, brokered deposits or Federal Home Loan Bank advances is lower quality than a slightly lower NIM funded by operating accounts and sticky retail balances. Equity investors should look at noninterest-bearing deposit mix, loan-to-deposit ratio, uninsured deposit concentration and accumulated other comprehensive income, not just headline spread income.
Credit Quality: Normalization Is Not the Same as Crisis
The credit outlook is deteriorating, but from unusually strong post-pandemic levels. Fiscal transfers, excess savings, tight labor markets and conservative underwriting kept losses artificially low in 2021-2022. The normalization is visible first in credit cards and commercial real estate. That does not imply a systemic banking crisis, but it does mean 2025-style earnings power must include higher through-cycle losses than the market assumed two years ago.
Credit card charge-offs have moved materially higher from trough levels as lower-income borrowers face rent, insurance and interest expense pressure. Issuers such as Capital One, Discover and Synchrony are more directly exposed, but large diversified banks also show consumer loss normalization. The key variable is unemployment. If jobless claims remain contained and wage growth stays positive in real terms, card losses can rise without becoming destabilizing. If unemployment moves decisively above 4.5%, reserve builds would become a more meaningful headwind for bank EPS.
Commercial real estate is the more idiosyncratic and politically visible risk. Office properties remain the pressure point because values have reset lower, leasing demand is structurally impaired by hybrid work, and maturities must be refinanced at coupons that are often 300-500 basis points above the original debt cost. The FDIC and bank disclosures have shown rising noncurrent rates in non-owner-occupied CRE, particularly office. Importantly, the exposure is not evenly distributed. Money-center banks have diversified revenue and larger capital markets franchises, while some community and regional banks have CRE loans that exceed 250%-300% of risk-based capital.
The base case is a grind, not a cliff. Banks will extend some performing loans, force equity injections where sponsors have capacity, and recognize losses over several quarters. The risk case is a local refinancing spiral: weaker property cash flows lower appraisals, lower appraisals reduce advance rates, and borrowers with 2025-2026 maturities hand back keys. Investors should separate banks with detailed office disclosure, conservative loan-to-value marks and strong reserves from those using broad CRE categories that obscure office concentration.
Capital, AOCI and the Hidden Duration Trade
One reason bank stocks are more rate-sensitive than usual is accumulated other comprehensive income. The rapid rise in yields left the industry with hundreds of billions of dollars of unrealized losses on available-for-sale and held-to-maturity securities. The FDIC reported unrealized losses on securities above $500 billion in early 2024, still a meaningful drag on tangible common equity even after some improvement from the 2023 peak.
Rate cuts would help tangible book value by reducing unrealized losses, but the benefit is not uniform. Banks with large fixed-rate securities portfolios, such as Bank of America and some regionals, have more visible AOCI recovery potential. Banks with shorter-duration books get less mark-to-market upside but also carried less balance-sheet volatility. From an equity perspective, duration is a double-edged sword: it can lift tangible book in a bull steepener, but it also limits reinvestment flexibility and can keep asset yields below market for longer.
Capital rules add another layer. The Basel III endgame proposal, even if softened, keeps pressure on large banks to manage risk-weighted assets, securitization exposures and trading books. That favors institutions with fee income, payments scale and capital-light businesses. For regionals, the issue is less Basel complexity and more confidence: can they grow loans, absorb CRE losses and still return capital? Buyback capacity will become a key signal. A bank trading below tangible book should repurchase shares if credit visibility is acceptable; if management refuses, investors should ask what the balance sheet is telling them.
Stock Selection: Barbell Quality and Selective Recovery
The bank equity setup argues for a barbell rather than a blanket sector overweight. On one side, high-quality franchises such as JPMorgan and select custodial or fee-heavy banks deserve premium multiples because they combine deposit scale, diversified revenue and strong capital. On the other side, select regionals trading below tangible book can rerate if NIM troughs and credit losses remain manageable. The avoid list is equally clear: banks with high CRE concentration, weak deposit granularity, low reserve coverage and limited capital return flexibility.
In practical terms, investors should track four metrics during earnings season. First, sequential NII guidance: stabilization matters more than the absolute year-over-year decline. Second, deposit beta on the way down: a bank that cuts deposit costs quickly after Fed easing will show real franchise value. Third, criticized and nonperforming CRE migration: office losses must be quantified, not explained away. Fourth, tangible book value per share: in a sector where long-term value creation is book-value compounding, EPS beats funded by reserve releases or one-time items deserve a lower multiple.
- Most attractive profile: low-cost deposits, diversified loans, ROTCE above 12%, clear buyback capacity and office CRE below peer averages.
- Most vulnerable profile: high loan-to-deposit ratio, heavy brokered funding, CRE above 300% of capital and weak tangible book growth.
- Best macro backdrop: gradual Fed cuts, stable employment, a modestly steeper yield curve and no sharp decline in commercial property liquidity.
Conclusion: Margins Can Heal Before Credit Does
The next phase for US banking stocks will not be driven by a simple call on the Fed. Lower rates can relieve funding costs and repair AOCI, but they also signal slower nominal growth and can pressure floating-rate loan yields. The real alpha will come from identifying banks where deposit costs fall, credit losses remain contained and capital returns resume before consensus estimates fully adjust.
My sector view is constructive but selective. Net interest margins are closer to a trough than a peak, particularly for banks with strong core deposits. Credit quality, however, has not fully cleared the adjustment to higher refinancing costs and weaker commercial real estate values. Investors should pay up for franchises that can compound tangible book through the cycle and be disciplined with regionals until office exposure, reserve adequacy and deposit behavior are proven in the numbers. In bank investing, the cheapest price-to-book ratio is often a warning; the best opportunity is the institution that can defend book value when the credit cycle turns.