TL;DR. Net interest margin (NIM) is the spread a bank earns between what it collects on its loans and securities and what it pays on its deposits and borrowings, expressed as a percentage of its average interest-earning assets. It is the single most-watched profitability metric for a traditional bank. FDIC-insured institutions reported an aggregate NIM of 3.31% in Q1 2026, eight basis points below the prior quarter, after asset yields fell faster than funding costs according to the FDIC’s Quarterly Banking Profile.
What NIM actually measures
A bank’s core business is intermediation: take in short-dated deposits, lend out longer-dated loans and bonds, and pocket the difference. NIM puts a number on that difference and scales it by the bank’s interest-earning assets, so a $10 billion regional bank and a $3.5 trillion universal bank can be compared on the same metric.
The formula is straightforward:
NIM = Net Interest Income ÷ Average Interest-Earning Assets
Where Net Interest Income (NII) = interest received on loans, securities, and cash reserves minus interest paid on deposits and borrowings. Average interest-earning assets are the period-average of everything on the asset side that throws off interest — primarily loans, investment securities, and balances held at the Federal Reserve. Non-earning assets (premises, goodwill, certain receivables) are excluded from the denominator.
Two things follow from that definition. First, NIM is not the gap between the loan rate and the deposit rate; that’s the interest spread, which ignores the bank’s funding mix. A bank funded 80% by deposits and 20% by debt has a different NIM than a deposit-only bank even at identical loan yields. Second, NIM excludes everything that isn’t interest: trading revenue, advisory fees, card interchange, and service charges all live in non-interest income. A diversified universal bank can run a thin NIM and still post a strong ROA because half its revenue arrives outside this calculation.
A worked example
Consider a stylized regional bank with $10 billion in average interest-earning assets, allocated as follows:
- $7 billion in commercial and residential loans yielding 6.50% on average
- $2.5 billion in agency mortgage-backed and Treasury securities yielding 3.80%
- $0.5 billion in reserves at the Federal Reserve earning the interest-on-reserves rate of 4.40%
Gross interest income = (7.0 × 6.50%) + (2.5 × 3.80%) + (0.5 × 4.40%) = $455M + $95M + $22M = $572 million.
Now the funding side. The bank funds the same $10 billion balance sheet with:
- $6 billion of interest-bearing deposits at an average rate of 2.40%
- $2 billion of non-interest-bearing checking deposits (no interest expense)
- $1.5 billion of FHLB advances and wholesale borrowings at 4.80%
- $0.5 billion of subordinated debt at 6.00%
Interest expense = (6.0 × 2.40%) + (1.5 × 4.80%) + (0.5 × 6.00%) = $144M + $72M + $30M = $246 million.
Net interest income = $572M − $246M = $326M. Plug that back into the formula:
NIM = $326M ÷ $10,000M = 3.26%
Almost exactly in line with the FDIC industry-wide level of 3.31% in Q1 2026.
What drives NIM up or down
NIM is the residual of two large numbers, so even small moves in either side compound into noticeable changes. Four forces explain most of the variation.
1. The level and slope of interest rates
Loans and many securities reprice as benchmark rates move. Deposits reprice too, but more slowly and less than 1-for-1. When the Federal Reserve raises policy rates, asset yields move first and funding costs follow with a lag — NIM widens. When the Fed cuts, the opposite happens, especially if deposits are already paying close to zero and have nowhere lower to go. This is precisely what happened in Q1 2026: “earning asset yields declined faster than funding costs,” per the FDIC, and industry NIM gave back eight basis points.
2. Deposit beta
The fraction of a benchmark rate move that a bank passes through to depositors is called deposit beta. A bank with a deeply loyal, fee-and-relationship-driven deposit base — community banks, branch-heavy super-regionals — has a low deposit beta and sticky funding. A bank that competes with money-market funds and high-yield online savings accounts has a high deposit beta. Deposit beta is the single biggest reason two banks with identical loan books can have very different NIMs across a rate cycle.
3. Asset mix
Loans yield more than securities, which yield more than reserves. A bank tilted toward credit cards and consumer loans (high yield, high charge-off) will report a higher NIM than a bank dominated by jumbo residential mortgages and Treasury holdings. NIM is therefore not directly comparable across banks with different mixes — a sub-3% NIM at a custody-and-trading shop like Bank of New York Mellon is structurally different from a sub-3% NIM at a generalist regional.
4. Funding mix
Non-interest-bearing deposits — your checking account — are the most valuable liability a bank owns. They cost zero, and the bank earns the full asset yield on whatever they fund. Wholesale funding (FHLB advances, brokered CDs, subordinated debt) is the opposite: it must be paid for, often at higher than retail-deposit rates. Banks that lost low-cost deposits to money-market funds during 2022–2024 saw NIM compressed not because asset yields fell, but because their funding mix shifted toward more expensive wholesale lines.
What the industry NIM has done lately
Aggregate FDIC data shows a steady climb through 2025 followed by a Q1 2026 give-back. The pre-pandemic five-year average (Q1 2015 through Q4 2019) was 3.25%, which the FDIC routinely uses as a reference point in its quarterly reports.
| Period | Industry NIM | QoQ change | FDIC narrative |
|---|---|---|---|
| Pre-pandemic avg (Q1'15 – Q4'19) | 3.25% | — | FDIC reference baseline |
| Q1 2025 | 3.25% | −2 bp | “Equals pre-pandemic average” |
| Q2 2025 | 3.26% | +1 bp | “Relatively flat” |
| Q3 2025 | 3.34% | +9 bp | “Above pre-pandemic average” |
| Q4 2025 | 3.39% | +5 bp | “NII rose 2.2% QoQ” |
| Q1 2026 | 3.31% | −8 bp | “Asset yields declined faster than funding costs” |
Community banks have run noticeably above this industry-wide level. The FDIC reported a Q3 2025 community-bank NIM of 3.73%, a full 39 basis points above the all-industry average in that quarter’s press release. That gap is structural: community banks lean more heavily on relationship-driven, low-cost deposits and fewer wholesale funding lines, and their loan books skew toward higher-yielding commercial real estate and small-business credit.
Why NIM matters to investors
For a traditional bank, NIM is the engine of profitability. Multiply NIM by average earning assets, and you have net interest income — typically 50% to 70% of a regional bank’s total revenue. Subtract loan-loss provisions and operating expenses, and you’re most of the way to pre-tax earnings. A 10-basis-point change in NIM at a $100 billion balance sheet is roughly $100 million of pre-tax income — enough to move a quarter’s EPS materially.
That makes NIM a primary lens for bank-stock analysis: when consensus expects a rate-cut cycle to begin, regional bank stocks usually trade off ahead of it, because investors are pricing in NIM compression. The reverse held in 2022–2023, when an aggressive hiking cycle widened spreads and lifted bank net interest income to record levels before deposit beta caught up.
Common mistakes when reading NIM
- Confusing NIM with interest spread. NIM normalizes by earning assets; spread is just rate − rate. Two banks with the same spread can report different NIMs because of funding mix.
- Comparing across business models. A trust-and-custody bank, an investment-banking-heavy universal, and a community lender all calculate NIM the same way, but the number means different things. NIM comparisons are most informative within a peer group of similar business mix.
- Ignoring the loan-loss line. A bank can boost NIM by reaching for risk — subprime auto loans, leveraged lending, lower-grade commercial real estate. That NIM is real, but the provisions taken later eat the difference. NIM net of credit costs (sometimes called “risk-adjusted NIM”) is the cleaner read.
- Treating NIM as the only revenue line. A large diversified bank can deliver double-digit ROE on a sub-2.5% NIM if fee income, trading, and wealth management carry the load. JPMorgan and Bank of New York Mellon are the clearest examples of this.
- Forgetting that NIM is backward-looking. The number reported today reflects yields and funding costs already locked in. Forward NIM depends on what is being booked at the margin: today’s new-loan rate, today’s new-CD rate, today’s swap hedges. Bank guidance usually leads the next quarter’s reported NIM by 60 to 90 days.
Related concepts to learn next
NIM is one of three pillars of bank profitability analysis. The other two are efficiency ratio (non-interest expense divided by revenue), which measures cost discipline; and provisions and net charge-offs, which measure credit quality. NIM also sits next to the capital framework — CET1, Tier 1, and Tier 2 ratios — and the liquidity framework (LCR, NSFR). Once you have all five concepts in hand, you can read any bank earnings release in about ten minutes.
Sources
- FDIC — Quarterly Banking Profile (primary source for industry NIM)
- FDIC — Q1 2026 press release (industry NIM 3.31%, −8 bp QoQ)
- FDIC — Q4 2025 press release (NIM 3.39%)
- FDIC — Q3 2025 press release (NIM 3.34%; community-bank NIM 3.73%)
- FDIC — Q2 2025 press release (NIM 3.26%)
- FDIC — Q1 2025 press release (NIM 3.25%; equals pre-pandemic average)
- Federal Reserve — H.8 Assets and Liabilities of Commercial Banks (banking-system size, weekly)
Disclosure: This article was produced with AI assistance and reviewed before publication. It is for informational purposes only and is not investment advice.