In June we read the first-quarter call reports for every bank and credit union in the country and wrote that history rhymes. The second-quarter filings are now in for both sectors, and the rhyme has a subject. It is people — what they cost, how many of them there are, and how institutions are choosing to pay for them.
The median bank now spends $105,400 a year in compensation and benefits for each employee. The median credit union spends $84,900. Both figures are roughly 5% higher than a year ago and about 50% higher than a decade ago. Earnings recovered again this quarter, margins widened, and the industry kept consolidating, and we cover all of that below. But the line that moved most steadily, at institutions of every size, is the one Valhalla works on every day.
Valhalla Business Advisors brings expertise to depository institutions throughout the United States. Our team’s experience partnering with financial institutions over 20+ years enables Valhalla to differentiate for its clientele, including NCUA-insured credit unions and FDIC-insured banks, not only by bringing employee benefit and HR solutions, but also supporting their business through custom data resources.
Valhalla Business Advisors is a consultancy that focuses upon employee benefits advisory and brokerage services. Our platform includes:
- Employee Benefit Brokerage and Advisory Services
- Benchmarking Support || Including from national cross-cutting datasets and bank and credit-union specific data
- Support of strategies leveraging CUSO structures to enable credit unions to “collaborate to compete”
- Brokerage of Bank Owned Life Insurance (BOLI) and Credit Union Owned Life Insurance (CUOLI) products
We have written previously about specific work for credit unions and banks.
Models || How We Built This
Valhalla maintains its own consolidated analytics layer for depository institutions. We draw bank data directly from the FDIC’s developer API and credit-union data from the NCUA’s quarterly 5300 Call Report files, then clean, normalize, and benchmark both on a common footing. Since our first-quarter review, the credit-union history has been extended back to 2016, so this piece can show ten-year trends for both sectors side by side. That earlier post also explains why banks and credit unions differ structurally on taxes, capital, and lending; we do not repeat it here.
One note on measurement: Call-report income is filed year-to-date, so a second-quarter filing covers six months. To keep this quarter comparable with the last one, returns, efficiency, and compensation below are for the second quarter alone, annualized. Net interest margin is shown for the first half of the year in both sectors. Every figure is a median across actual filers. Bank return on assets is after tax, except at the roughly one-third of banks organized as S-corporations; credit unions pay no income tax. Counts include every institution filing a call report, a small number of which are not federally insured.
What a Person Costs
Start with the long view. In the second quarter of 2016 the median bank spent $70,000 a year per employee on salaries and benefits. Today it spends $105,400, an increase of 51%. The median credit union went from $55,300 to $84,900, an increase of 54%. The two lines have climbed almost in parallel for ten years, and the climb has steepened since 2022.
The most recent year looks the same wherever you cut it. Follow each institution from the second quarter of 2025 to the second quarter of 2026 and the median bank’s cost per employee rose 5.1%; the median credit union’s rose 4.8%. Among banks the figure runs from 4.6% at the smallest institutions to 5.3% at those between $1 billion and $10 billion. Among credit unions it runs from 4.3% to 6.3%. No size tier in either sector came in below 4%.
Paying More, Not Hiring More
At the typical institution, the rising cost per person is not a hiring story. At the median bank and the median credit union, headcount did not change at all over the past year. Total compensation expense still rose 6.5% at the median bank and 5.6% at the median credit union. Institutions are paying more for the people they already have.
Over ten years the two sectors part ways. Banks employ about 2.04 million people, almost exactly what they employed in 2016, while bank assets grew 60%. In aggregate, bank employment is flat. The typical bank that existed in both years added about 12% to its headcount while its assets grew by three-quarters. Credit unions took the other road: their full-time-equivalent workforce grew by a third, from 277,000 to 366,000, as their assets roughly doubled. That is the branch-and-member service model at work, and it is why compensation weighs more heavily on a credit union’s income statement.
The Load Institutions Carry
Measured against the balance sheet, the median bank spent 1.45% of assets on compensation and benefits in the second quarter and the median credit union spent 1.77%. Both are within a basis point of the first-quarter figures. The ratio held steady only because assets grew about as fast as pay: total bank assets rose 6% over the year and credit-union assets 5%. An institution whose balance sheet is not growing gets no such help. For it, a 5% increase in the cost of people comes straight out of earnings.
Among the 3,417 community banks between $100 million and $10 billion that we plot, efficiency separates the earners. Banks above the median efficiency ratio earned a median 0.95% on assets; those below it earned 1.67%. Compensation is the largest single piece of that ratio, but a high compensation load is not itself the problem: the best-earning group in the sample carries above-median compensation and earns 1.79%. The 1,193 banks where a high load meets weak efficiency earn 1.00%. For them the benefit line is the first place to look.
How Institutions Are Funding It
If the cost of people rises 5% a year, the practical question is how to offset it. Call reports show what institutions have already done, because benefit pre-funding assets sit on the balance sheet.
Among banks, the tool is bank-owned life insurance (BOLI). Two-thirds of banks hold it — 67% in the second quarter, up from 61% a decade ago — with $217 billion of cash value in total. The share rose because banks without it disappeared faster; the number of holders actually fell. Adoption climbs with size: 36% of banks under $100 million, 68% between $100 million and $1 billion, and 82% between $1 billion and $10 billion. The median holder carries BOLI equal to about 14% of Tier 1 capital plus reserves, well inside the 25% interagency supervisory guideline.
The faster-moving story is on the credit-union side. Federal rules (12 CFR 701.19) allow a federal credit union to hold investments it otherwise could not, when they fund employee benefit and deferred compensation obligations. (Most states have comparable rules.) In 2016, 21% of credit unions reported such assets. Today 40% do, and the total has nearly tripled from $12.4 billion to $36.6 billion. Split-dollar life insurance grew from $2.1 billion at 523 credit unions to $11.6 billion at 1,101. Among credit unions between $1 billion and $10 billion, 91% now pre-fund benefits. Among those under $100 million, 14% do.
Read together, the two charts say that pre-funding has become standard practice at larger institutions in both sectors and remains uncommon at smaller ones.
Earnings and Margins: Banks Pull Further Ahead
The second quarter was a good one for banks. The typical bank earned a 1.27% return on assets, up from 1.19% in the first quarter and 1.13% a year ago. The first-half net interest margin reached 3.85%, 22 basis points wider than a year earlier. The median efficiency ratio improved to 61.5%, and only 4.3% of banks lost money in the quarter. Before tax, the median bank earned 1.52%.
Credit unions were level with last year; the rise from the first quarter is largely seasonal.
The typical credit union earned 0.78% on average assets, up from 0.66% in the first quarter and level with the 0.77% of a year ago. The median efficiency ratio was 77.2%, slightly worse than the 76.5% of the second quarter of 2025, and 13.2% of credit unions were unprofitable. Banks widened the earnings gap by 13 basis points over twelve months.
Inside the credit-union sector, size decided the outcome. The largest credit unions, those above $10 billion, lifted their median return from 0.64% to 0.98% over the year. Those under $100 million went the other way, from 0.79% to 0.72%, and nearly one in five of them lost money in the quarter.
Net interest margin is measured differently in the two sectors. The bank figure is the FDIC’s reported margin, net interest income as a share of average earning assets. The credit-union figure is net interest income as a share of average total assets, a larger denominator that produces a lower number. The two bars are therefore not directly comparable. Measured on total assets for both, the median bank’s first-half margin is about 3.6%, slightly below the median credit union’s 3.71%.
Fewer, Bigger: The Consolidation Speeds Up
With credit-union history now reaching back to 2016, the decade is visible in one chart. Institutions filing bank call reports have gone from 6,129 to 4,313, a decline of 30%. Credit unions have gone from 6,011 to 4,299, a decline of 28%. The two lines are nearly indistinguishable.
The past year stands out on the bank side. There are 180 fewer banks than a year ago, the largest twelve-month decline since 2020 and well above the 104 to 125 of each of the prior three years, though still below the pace of 2017 to 2020. Credit unions lost 161, in line with their recent pace. The assets have not gone anywhere: the 165 banks above $10 billion hold 87% of all bank assets, and the 479 credit unions above $1 billion hold 79% of credit-union assets.
A Regional Lens: PA, NY, NJ, MD, and CT
Across our home footprint the quarter-to-quarter movement was larger than the national figures suggest. Pennsylvania banks had the strongest quarter of the five states: the median return on assets rose from 0.87% to 1.08%, and the efficiency ratio improved more than four points to 62.2%. Connecticut improved on both sides of the ledger. New Jersey credit unions recovered from a very weak first quarter but remain the thinnest earners in the region, at a median 0.47% return, an 85.0% efficiency ratio, and a 1.82% compensation load. Maryland credit unions, at 0.51%, are close behind. In every one of the five states, credit unions carry a higher compensation load and a higher efficiency ratio than the banks they compete with.
Deeper Dive: Pennsylvania, $500 Million to $1.5 Billion
State medians blend very different institutions. A custom peer cut is more useful, so we again isolated Pennsylvania institutions between $500 million and $1.5 billion in assets: 34 banks holding $29.2 billion and 14 credit unions holding $13.9 billion. In this band the credit unions closed much of the gap in one quarter. Their median return rose from 0.73% to 0.96% and their efficiency ratio improved from 77.1% to 72.2%, while the banks moved from 1.07% to 1.10%.
The scatter below places all 48 on the same two measures as the national quadrant. Seventeen of them — nine banks and eight credit unions — sit above the band median on both compensation load and efficiency ratio. We have named a handful of the strongest performers. PS Bank, Central Penn Bank & Trust, and American Bank pair lean compensation with efficiency ratios near 50% or better. Freedom Credit Union runs a 60% efficiency ratio, the best of the band’s credit unions. The Dime Bank is the instructive case: it pays more than the band median for its people and still earns a 1.77% return on a 54% efficiency ratio. Paying well is not the problem. Paying well without the earnings to support it is.
The Valhalla Advantage
Reading the data is the easy part. Acting on it is where we come in. When the cost of people rises 5% a year, the question for a board is practical: how do you keep and reward the people who drive the franchise without giving up the earnings the margin recovery just delivered? Start where the cost actually lives — employee health benefits. Medical and pharmacy spending is likely the fastest-rising and least transparent component of the compensation-and-benefits line at most depository institutions, and it is also the most addressable: through self-funded plan design, stop-loss structuring, and genuine pharmacy-benefit-manager (PBM) transparency, we help banks and credit unions take costs out of the health plan without reducing the coverage employees actually receive.
From there the conversation turns to funding and retention. For banks, that often runs through Bank-Owned Life Insurance, which two-thirds of the industry already uses. For credit unions, the analogous tools are 457(b) and 457(f) nonqualified deferred compensation, collateral-assignment split-dollar arrangements, and employee-benefit pre-funding permitted under 12 CFR 701.19 — the same tools the data shows spreading quickly among larger credit unions. In either case we can put a dollar figure on the opportunity before anyone commits to a meeting: what a smarter health-benefits strategy could recover, what closing the gap to the peer median is worth per year, and where an institution’s pre-funding stands against institutions its own size.
That is the difference between a broker who sells products and an advisor who reads the balance sheet. We operate with no contingent or supplemental commissions. The data tells us where to look; the independent relationship determines what we do once we get there.
If you run a bank or credit union — or advise one — and you would like to see exactly where your institution sits against its true peers on any of the dimensions above, reach out. We are happy to prepare a custom, institution-specific benchmark on request.
SAMPLE SCOPE OF WORK
APPENDIX
Methodology & Sources
Methodology & sources. All figures are medians computed from FDIC (banks) and NCUA (credit unions) call report data as of June 30, 2026, unless described as totals. Return on assets, the efficiency ratio, and compensation ratios are for the second quarter alone, annualized: for banks, the FDIC’s quarterly ROA and efficiency ratio and quarterly salaries and employee benefits; for credit unions, second-quarter amounts derived by subtracting first-quarter from year-to-date filings, with net income taken as the call report’s reported bottom line (account 661A), inclusive of CECL credit-loss expense, over average assets. Net interest margin is year-to-date for the first half of 2026 in both sectors. Compensation per employee divides annualized quarterly salaries and benefits by full-time-equivalent employees. One-year growth figures follow the same institution from Q2 2025 to Q2 2026 and take the median across institutions. Bank-owned life insurance is the cash surrender value reported on Schedule RC-F, compared with Tier 1 capital plus the allowance for credit losses; the 25% figure is an interagency supervisory guideline, not a regulatory limit. Credit-union benefit pre-funding is the total reported under 12 CFR 701.19(c) (account 789G), of which split-dollar life insurance is accounts 789E and 789E1. Year-over-year comparisons are against Q2 2025 and ten-year comparisons against Q2 2016. Asset tiers are harmonized across both sectors as <$100M, $100M–$1B, $1B–$10B, and ≥$10B. First-quarter figures may differ by a basis point or two from those published in June because of subsequent filing amendments. Institutions named in the Pennsylvania spotlight are identified from public call report data.