Last week two university credit unions announced they would combine into a $13.5 billion institution with more than 850,000 members, pending member and regulatory approval. One of them is in Texas. The morning after news like that, somebody in every executive meeting asks the same question. Did anyone see this coming?
Until recently the honest answer was no. The numbers that could have answered it were public the whole time, sitting in quarterly filings nobody had the tools to read at scale. That has changed, and it changes what a marketing or growth leader can bring to the table.
What we built
Every credit union files a detailed report with NCUA four times a year. Those filings go back decades, and NCUA also publishes a list of every merger it approves. That gave us something most predictions never get: a way to test against what already happened.
This summer we built a merger propensity model from those filings. It scores every credit union in the country from 0 to 100, where a higher score means the institution looks more like the ones that later merged out. This piece is not about the inputs. It is about whether the model works, and what you can do with it.
The test
We scored every credit union on its June 2024 filing, before any of the mergers in question were announced. Then we checked NCUA's merger lists to see which ones were gone by the end of December 2025. That is 4,591 credit unions and an 18-month window.
Credit unions scoring 70 or higher: 395 of them, and 57 merged out. That is 14.4 percent, or roughly one in seven.
Credit unions scoring under 40: 1,141 of them, and 23 merged out. That is 2.0 percent, or about one in fifty.
The national rate across all 4,591 was 5.4 percent.
A high score meant a credit union was about seven times as likely to be gone within 18 months as a low score. We ran the same test one year earlier, on June 2023 filings, and got 19.2 percent at the top against 4.1 percent at the bottom. Different year, same pattern.
Where it is strongest
We also asked a harder question. Does the model beat the simplest rule every executive already knows, which is that small credit unions merge and large ones do not?
Under $500 million in assets, the answer is yes, in every size group. Above $500 million, size alone does better. Of the 717 credit unions at $500 million or more in June 2024, only 12 merged out, and the model had flagged two of them.
That matches last week's news. Two large, healthy institutions with a shared mission chose each other. Filed numbers measure condition. They do not measure strategy, board chemistry or a CEO who is ready to retire. Above $500 million, the press release is the data.
Below $500 million is where the model does its work, and that is where the volume is. In our test, 236 of the 248 credit unions that merged out were under $500 million.
Texas right now
On June 2026 filings, 380 Texas credit unions get a score. Twenty-four score 70 or higher, and 15 of those are under $500 million. For reference, NCUA's merger lists show 33 Texas credit unions merged out across 2023, 2024 and 2025, and the national count of federally insured credit unions fell by 156 in the year ending June 30, 2026, from 4,370 to 4,214.
We are not naming the 24. A score is a reason to look closer. It does not know what a board knows.
What a growth leader does with it
Three uses, in the order we would take them.
Your own score. Whether or not a merger is anywhere in your plans, your filed numbers are telling the market something about you. Walking into the next leadership meeting knowing what they say, before anyone outside the building does, is worth an hour.
The institutions near your branches. When a high scorer sits a few miles away, a conversation now, long before anything formal, is worth more than a call after the announcement. When a low scorer near you starts closing branches, that is a growth opportunity with a date on it, and the marketing plan should know about it first.
Your partner short list. If growth by combination is on the table, the model narrows a list of hundreds to a handful worth a real look, and the public record tells you which of them are likely to be receptive.
The model runs inside FI Snapshot, the platform we built to read any bank or credit union from its filings. If you want your own score and the watch list for the institutions around your branches, book a 30-minute working session and bring your branch ZIP codes. You keep what we find.
One more thing
We could have led with the seven-to-one ratio and stopped. The size split is the finding we would want to know before trusting anyone's merger model, including ours.