The Spread puts a whole corner of financial services on one chart and asks whether the thing everyone says is true. No. 1: credit unions.
Every credit union merger comes with a reason, and the NCUA writes it down. Of the 762 mergers it approved between the first quarter of 2021 and the first quarter of 2026, 572 (75%) gave the same one: to offer members expanded services. In plain English: we need to be bigger.
So we checked whether bigger pays. We put every federally insured credit union on one chart, all 4,214 of them, from Navy Federal's $204.4 billion down to the 124 that hold less than $1 million, and measured each one's return on assets for the first half of 2026.
Source: NCUA, Call Report data, June 2026. Data as of 2026-06-30. Net income for January to June 2026, doubled, over assets on June 30, 2026: a comparison, not the NCUA's own ratio. Returns beyond −2% or +3%, and assets under $1 million, are drawn at the edge.
Show the data
| Return on assets | Count | |
|---|---|---|
| Under $100M (median) | 0.65% | 2,438 |
| $100M to $500M (median) | 0.74% | 1,028 |
| $500M to $1B (median) | 0.75% | 277 |
| $1B to $10B (median) | 0.80% | 446 |
| $10B and over (median) | 0.87% | 25 |
| Navy Federal | 1.38% | – |
| First Technology | 0.28% | – |
| ESL | 2.26% | – |
The typical return barely moves
Read the orange lines from left to right. The median credit union under $100 million earned 0.65%. The median one over $10 billion earned 0.87%. Climbing from the smallest band to the largest buys 0.22 percentage points at the median.
If scale is a strategy for earning more, it is a patient one.
What scale does buy: a much smaller chance of a losing year
Now look below the zero line. That is where the size bands really differ.
469 of the 2,438 credit unions under $100 million, 19%, lost money in the first half of 2026. Not one of the 25 over $10 billion did. On these numbers, scale works less like a growth plan and more like a seatbelt. It does not make the car faster. It changes how the bad days end.
The two that did what the deck says
Two of the largest credit unions in the country took the advice at full strength last year. Digital ($12.7 billion) and First Technology ($17.1 billion) merged into First Technology. Ent ($10.3 billion) and Wings Financial ($9.5 billion) merged into Wings. The NCUA approved both in the third quarter of 2025, and both gave the same reason: expanded services.
In the first half of 2026 they posted the two lowest returns of the 25 credit unions over $10 billion.
Before anyone drafts the obituary: on our reading this is what the first year of a merger of equals tends to look like, with two of everything (core systems, branches, brands) still being paid for while they become one. One half-year is not a verdict on either deal. It is, though, the opposite of what "we need scale" is usually taken to promise, and a board should know which year of a merger it is looking at.
Size is not destiny
The spread inside each band is wider than the gap between them.
Source: NCUA, Call Report data, June 2026. Data as of 2026-06-30.
Show the data
| Figure | What it counts |
|---|---|
| 933 | of the 2,438 credit unions under $100 million out-earned the median one over $10 billion |
| 2.26% | ESL, $10.8 billion in assets: the highest return of any credit union over $10 billion |
| 0.28% | First Technology, $27.9 billion: the lowest of the same 25 |
ESL and First Technology sit in the same size class, and one earned about 8 times what the other did. Size predicts whether a credit union loses money. It says very little about how well one does.
The part the merger deck skips
Bigger credit unions hold less capital for every dollar of assets.
None of that is a warning sign on its own. A large, diversified balance sheet can run on a thinner cushion. But a merger that is sold on safety should be asked what happens to the cushion.
If you sit on a board
Three questions to take into the next strategy meeting:
- Which problem are we solving? A thin return and the risk of a losing year are different problems, and on this chart scale mostly fixes the second.
- Where do we sit in our own band? The free credit union peer benchmark places any credit union against those nearest it in size, on return, net worth and delinquency.
- What happens to our capital? Ask for the combined net worth ratio before the vote, not after.
Our take
"Scale" does two jobs in merger conversations, and they get blurred. It is a good answer to "how do we survive a bad year?" It is a weak answer to "how do we earn more for members?", because at the median it barely does. If someone presents a merger as a path to better returns, ask them to show where on this chart the evidence sits.
How we did this, and what it leaves out
- Return on assets is not a credit union's score. A credit union is owned by its members. A low return can be a choice: better rates, lower fees.
- One half-year, measured against period-end assets. A credit union that merged during the half carries the larger balance sheet against only part of a year's combined income, which lowers the ratio. This is a comparison, not the NCUA's own figure. The NCUA approved both near-equal mergers in the third quarter of 2025; the approval date is not the date the two balance sheets were combined, and we have not assumed one.
- Very small credit unions swing on small sums. Some are community development lenders with grant income. We name only credit unions over $10 billion.
- Merger reasons are approvals as filed, not completed mergers.
The full list of credit unions by assets is in largest credit unions, and every merger approval is in credit union mergers. Both tables are free to download. To see one credit union's return, net worth and delinquency against its peers, use the peer benchmark.