Q2 2026 CAMELS Data: The Top Has Never Looked Better, and the Bottom Is Getting Heavier

NCUA’s second-quarter 2026 CAMELS counts and June financial data are on NCUA’s website. There hasn’t been an NCUSIF (National Credit Union Share Insurance Fund) briefing yet, but the numbers are public, and I walked through them on the latest episode of With Flying Colors.

An industry pulling apart at both ends

Code 1s are 21.3% of credit unions, the highest share going back to 2016 in the data I track. Code 3s are 14.5%, the lowest in ten years. Code 4s are up seven to 100, while the industry keeps shrinking. So those 3s are either moving up to 2s and 1s or down to 4s.

Eighty-three percent of credit unions are now Code 1 or 2, up from 82.7%. Combined 4s and 5s went from 2.4% to 2.5%. The top of the distribution has never looked better. The bottom is showing stress we haven’t seen since 2017. A stronger average hides a thickening tail.

Why the middle is emptying

NCUA doesn’t have the resources, and it’s getting a little friendlier with coding. It’s also being nicer on the better credit unions and harsher on the bad ones. That may be an offshoot of the new marching orders: deal with the real problems, and don’t do warning-shot Code 3s or CYA Code 3s. Those aren’t in writing anywhere. They do happen.

All of this predates the FFIEC (Federal Financial Institutions Examination Council) proposal, which will make CAMELS ratings even simpler.

The bottom of the distribution

Code 4s were as high as 138 in the third quarter of 2024, trended down to 93 last quarter, and came back up to 100. There are six Code 5s, one over $1 billion. The count was as low as two in 2024, peaked at 10 last quarter, and is back to six. Typically, when it goes down, it’s because they were liquidated or merged, not that they got better.

Seven credit unions have negative net worth, up from three at the end of 2025 and zero at the end of 2022. Seven’s a big number. That’s bad if you’re one of them. It’s good if you’re looking for a merger partner. Get on NCUA’s merger registry if you aren’t.

Capital decides how much say you keep

Thirty credit unions are below 6% net worth. Capital is king, and it can keep NCUA at bay. Between 6 and 7% you’re writing NCUA every quarter for earnings retention. They get really ticked off if you don’t file. Between 4 and 6% you’re on a net worth restoration plan, which can be like running the gauntlet. Below 2%, NCUA has a set number of days to conserve. It’s the speed of decline that matters.

What the financials show

Margins are at ten-year highs: net interest margin of 3.49% and loan yields of 6.1%. ROA is 91 basis points, up from 63 at the end of 2024. Provision expense is 55 basis points, still above the pre-pandemic range of 0.40 to 0.48.

Delinquency has plateaued, at a higher level: 96 basis points. Participation loan delinquency is 1.11%, the highest in the data set, on $88 billion in balances. Buyers don’t always have the originator’s underwriting insight, so losses can surprise when they hit.

Certificates are 29% of shares, up from 18.4% in 2016. The core is more rate-sensitive than it was five years ago, and earnings are at risk if rates move sharply. Share growth of 5.2% is running ahead of loan growth of 4.9%. I’m not sure that continues. Probably not, would be my guess.

Code 4s and 5s are 4.5% of credit unions under $50 million and zero over $10 billion. The smaller you are, the fewer resources you have to deal with your issues.

Where the pendulum goes

Quite frankly, I don’t think they can get much better. Near term, with CAMELS being redefined and a softer touch, it’s quite possible the trend gets a little better over the next year or two. Further out, Inspector General reviews come in and NCUA has to say how it will deal with them, and two years down the road there will be a new president, one way or the other. The pendulum probably swings.

What to take to the board

Benchmark against your peers, not the aggregate. If your delinquency is twice the peer, examiners aren’t supposed to refer to peer, but they will ask what you’re doing to control it.

The problem pipeline is the fullest it’s been in years. Bad if you’re in it, good if you’re shopping for a merger partner.

Participations have been an exam focus and will continue to be.

And expect asset-liability management (ALM) conversations tied to your deposit mix. Most have been more reasonable lately. A couple of recent ones with overly aggressive specialists shocked me. Prepare for the conversation without assuming which version of it you’ll get.

If you have questions about your exam, or had one that didn’t go the way you expected, reach me on LinkedIn or at info@marktreichel.com.