# CU Intelligence · What it means

how to read every measure: what it means, when high, when low, versus peers, our line, limits; the four credit unions you will meet; what the desk cannot tell you.

Live page: https://getbookiq.com/cu-intel/#meaning

---

HOW TO READ THIS DESK
A credit union's call report answers four questions if you ask them in order: is the franchise growing or quietly dying, is it earning enough to exist, is the auto book, which is most of the loans at small institutions, telling the truth yet, and how much of that book was written by someone else. The percentile chips place the number against credit unions of similar size in the same quarter. Red means the worst fifth of the band for that measure, green the best fifth.
IS THE FRANCHISE ALIVE · GROWTH AND SCALE
Member growth, year over year
083 vs the same quarter one year earlier
MEANS
Whether the institution is gaining or losing the people it exists to serve. For a credit union this is the franchise itself; there is no other customer.
WHEN HIGH
Growth above 3% at a small credit union usually means a field-of-membership expansion or a merger absorbed. Check which.
WHEN LOW
Shrinking 3% or more a year is the quiet killer. The credit unions that stopped filing during our window were losing members at 2% a year eight quarters out and 4.25% a year at the end. Most of them had adequate capital the whole way down. Members leaving is what the capital cannot fix.
VS PEERS
System median is roughly flat to slightly negative; more than half of all credit unions with falling membership are under $50 million.
OUR LINE
< -3% = flag
LIMITS
A single-sponsor credit union tied to one employer inherits that employer's headcount. The decline can be entirely outside management's control and still fatal.
Asset and loan growth · loans / shares
010 and 025B year over year · 025B / 018
MEANS
Balance sheet direction and how fully deposits are deployed. Loans/shares is the credit union version of the bank's loans-to-deposits.
WHEN HIGH
Loans above 95% of shares leaves little deposit headroom; funding stress arrives fast if shares run. Fast loan growth at a small credit union with flat membership is usually indirect paper bought from dealers, not members walking in.
WHEN LOW
Assets shrinking 3%+ while costs stay fixed squeezes earnings mechanically. Loans/shares under 60% means members are depositing and not borrowing; the institution is a savings account with a lobby.
VS PEERS
Credit unions under $200M run loans/shares near 61% and their median loan book shrank 1.5% over the year, with the under-$50M median down 3.2%; a small CU growing loans fast is the exception and worth understanding.
OUR LINE
assets < -3% YoY = flag · loans/shares > 95% = flag
LIMITS
Mergers absorbed distort growth. Loans/shares ignores borrowings, which small credit unions rarely have and large ones sometimes hide behind.
IS IT EARNING ENOUGH TO EXIST · EARNINGS AND CAPITAL
Net worth ratio
997 / 010 · the prompt-corrective-action measure
MEANS
Capital as a share of assets. Regulatory categories attach directly: 7% well-capitalized, 6% adequately, below that a net worth restoration plan.
WHEN HIGH
Small credit unions run 13-14% on average. High net worth with negative earnings means the cushion is being spent, not built; the runway math (net worth divided by annual loss) tells you how many years remain, and it is almost always longer than the franchise will last.
WHEN LOW
Below 9% deserves attention; below 7% is the regulatory line; the departed cohort ran roughly 0.8 points below survivors the whole way down. Direction matters more than level: falling two straight quarters is the trend that precedes the rest.
VS PEERS
A credit union under 9% while its band sits at 13% is running with a third less cushion than everyone its size.
OUR LINE
< 9 = watch · < 7 = critical, double weight · falling two straight quarters = flag
LIMITS
Net worth is a book figure. It did not catch Unilever FCU (9% reported, $8.4M insurance fund loss, fraud) or Aldersgate (10% reported one quarter before negative 166%). Capital that depends on the honesty of the filing is only as good as the filing.
Return on assets
661A annualized over 010
MEANS
Whether the operation covers its own costs. It is the difference between a credit union that builds capital and one that spends it.
WHEN HIGH
Above 0.8% is strong for a credit union. Very high ROA at a small institution usually means an unusual income item; check the year-ago quarter.
WHEN LOW
Negative ROA eats net worth every quarter it persists. Small credit unions as a group earned 0.13% in early 2026 and the $50-100M band posted negative ROA. A lender with no credit losses that still loses money has a cost problem or a truth problem; that combination, not delinquency, was the visible signal at every fraud failure in 2025.
VS PEERS
The chip strips out the rate environment by comparing against the same-size cohort in the same quarter.
OUR LINE
< 0 = flag · zero-loss unprofitability (delinquency < 0.25%, charge-offs < 0.10%, ROA < 0 two quarters running) = proposed rule R-18
LIMITS
Annualized from year-to-date; first-quarter figures are noisy. One-time items swing small institutions by a full point.
Net interest margin · operating expense / assets
(115 − 350) annualized / 010 · 671 annualized / 010
MEANS
The spread engine and the cost base. When the spread cannot cover the expense line, no amount of clean credit saves the institution; this is the structural failure mode of small credit unions.
WHEN HIGH
Operating expense above the band p90 is the cost trap: the institution is too small for its overhead. A margin above 4% at a small CU usually reflects a consumer-loan-heavy book, which carries the credit risk that pays for it.
WHEN LOW
Margin under 2% cannot carry a branch and a compliance officer. Margin minus expense (plus fee income) going negative and staying there is the exact path Unilever walked for eight quarters.
VS PEERS
Compare the structural spread (margin plus fees minus expense) within the band; it is the single most reliable separator between institutions that continue and institutions that disappear, in our data and in every published study of the question.
OUR LINE
expense > band p90 = flag · NIM down 50bp+ YoY = flag
LIMITS
Ratios of flows; a shrinking balance sheet can hold the margin while dollars fall.
IS THE AUTO BOOK TELLING THE TRUTH · ASSET QUALITY
Total and vehicle delinquency
041B / 025B · (041C1 + 041C2) / (370 + 385) · 60+ days
MEANS
Loans where the borrower has already stopped paying. This is the measure everyone watches, and it has one structural weakness: it reports a promise already broken. It confirms trouble a quarter or two after the decision that caused it.
WHEN HIGH
Above twice the band median is a problem book. Above the band p90, the book is in the worst tenth of its peers regardless of absolute level.
WHEN LOW
Very low delinquency is good news only if the rest of the filing agrees. Aldersgate reported 0.00% delinquency for years while $8M of fraud ran underneath it. Zero can mean clean or can mean nobody is looking.
VS PEERS
The system sits near 0.8-1.0%; small credit unions run higher delinquency but lower charge-offs than large ones. Comparing across bands without the chip is meaningless.
OUR LINE
> 2x band median = flag · up 50%+ YoY = flag
LIMITS
Delinquency separated the credit unions that stopped filing from those that continued by only about 0.1 percentage point, in our data and in a published 13-cohort study. It points the right way and carries little distance.
Repossessed vehicle inventory / vehicle book
AS0024 / (370 + 385) · cars taken back, not yet sold
MEANS
The loss in transit. A car in the lot behind the branch will not cure; it will sell for less than was owed and the gap becomes a charge-off next quarter. This line speaks earlier than delinquency and partly independently of it.
WHEN HIGH
Above 0.40% is top-quartile territory. Across eleven consecutive annual cohorts, credit unions in the top quartile charged off at roughly twice the rate of those reporting none a year later, and the ordering held every single time. Among credit unions whose delinquency looked healthy, high repo inventory preceded 1.67 times the losses of their apparent peers.
WHEN LOW
Reporting a trivial amount is indistinguishable from reporting none; the signal begins at the second quartile. Zero can mean nothing in the lot or nothing on the line: 69% of credit unions leave it blank, and small ones with the highest delinquency report it least.
VS PEERS
Read it in the two-by-two: comfortable delinquency plus elevated inventory is the one combination the standard measure gets wrong.
OUR LINE
> 0.40% = flag
LIMITS
It does not beat delinquency or concentration as a standalone predictor; it adds to them. It is timing a loss already committed, not foreseeing an unrelated one. The value is lead time and the blind spot, not surprise.
Net charge-offs, total and vehicle
(550 − 551) annualized / 025B · (550C1 + 550C2) annualized / vehicle
MEANS
Realized losses. The rearview mirror, but the honest one: charge-offs are hard to argue with.
WHEN HIGH
Above the band p90, losses are being realized, not just brewing. Vehicle charge-offs above 1.5% at a credit union are expensive; the system runs near 1%.
WHEN LOW
Low charge-offs with rising delinquency and rising repo inventory means the losses are queued, not absent.
VS PEERS
Small credit unions post lower vehicle charge-offs than billion-dollar ones (0.48% versus 1.32%) because the large ones run the indirect books. Band context is everything.
OUR LINE
> band p90 = flag
LIMITS
First-quarter figures annualized by four are the noisiest read of the year. Direction across quarters, not a single print.
WHO WROTE THE PAPER · SOURCING AND CONCENTRATION
Vehicle share of loans
(370 + 385) / 025B
MEANS
How much of the institution is a car lender. At credit unions under $200 million, vehicles are 42% of all loans; at the smallest, more than half.
WHEN HIGH
Above 60% the credit union is a monoline: 1,309 of them exist, median $15.7 million in assets and four employees. At 80%+ delinquency runs 1.12% against a system near 0.8%. The product is the institution.
WHEN LOW
Low vehicle share at a small credit union usually means a mortgage book or a share-secured book; different risks, not fewer.
VS PEERS
A monoline with above-peer delinquency is a concentrated bet going wrong. That combination is its own flag.
OUR LINE
> 60% with vehicle delinquency at p75+ = flag
LIMITS
Concentration describes what is at stake if something goes wrong. It says little about whether something is going wrong.
Indirect share of vehicle · third-party paper / net worth
618A / vehicle · (618A + SL0036) / 997
MEANS
How much of the auto book was originated by a dealer rather than across the credit union's own desk, and how large that outside-written paper is relative to the capital behind it.
WHEN HIGH
Third-party paper above 100% of net worth is above the line NCUA capped until August 2026, when the rule was removed with delinquency at a decade high. 799 credit unions sit there. Tested fairly, concentration does predict next-year losses (it ordered correctly in all eleven cohorts), but the gradient is modest and it works about as well as the repo signal, not better.
WHEN LOW
Small credit unions are direct lenders; only 24% of their vehicle book is indirect, against 77% at billion-dollar institutions. The small end writes car loans, the large end buys them.
VS PEERS
Indirect paper carries the dealer's incentives. The credit union never met the borrower. That is the risk the line measures.
OUR LINE
> 100% of net worth = flag
LIMITS
The 5300 fields that once split indirect into point-of-sale versus outsourced-servicing arrangements are retired with zero filers. The exact category the removed cap governed is no longer measured. This line is a proxy and an upper bound.
Purchased vehicle participations
SL0036 outstanding · SL0037 purchased year-to-date
MEANS
Paper someone else wrote, sold in slices. The purest form of 'you did not underwrite this'. NCUA rule 701.22 requires documented independent due diligence and ongoing monitoring from every buyer.
WHEN HIGH
Participations above 25% of net worth is a due-diligence obligation under 701.22 that scales with the balance, not with the staff. 1,005 credit unions sit above that line, median 50 employees and $265M in assets, so this is not only a small-institution problem. System participation delinquency ran above the overall loan book in 2025.
WHEN LOW
Zero is the norm for small credit unions; most never buy.
VS PEERS
The participation market itself is invisible to its participants: no filing names a counterparty. Read the Participation Desk for the market view.
OUR LINE
no standalone flag · surfaced on the participation buyers screen
LIMITS
The filing shows the buyer's balance, not the originator, the collateral, or the servicing arrangement. Everything that matters about a participation is off the tape.
PUTTING IT TOGETHER
The composite score
one point per flag, net worth below 7% counts double · 17 flags across capital, earnings, asset quality, shrinkage, structure
MEANS
How many things are wrong at once. Credit unions rarely disappear over one number.
WHEN HIGH
Score 7 or above: 22% stopped filing within four quarters, against a 4.2% base rate, a 5.3x lift. Score 5-6: 8.8%. The lift is monotonic across every bucket.
WHEN LOW
Zero to two is ordinary life.
VS PEERS
Most component flags are band-relative, so the score compares reasonably across sizes.
OUR LINE
5-6 = elevated · 7+ = danger
LIMITS
Departure means stopped filing: mostly voluntary mergers, some liquidations, a few charter conversions that look like departures and are not. The claim is fragility, not death. Fraud failures scored 2-4 before the end; the score reads the filing, and a false filing reads as healthy.
FOUR CREDIT UNIONS YOU WILL MEET
THE AUTO MONOLINE

Under $50 million, four employees, more than 60% of loans in vehicles, direct lending to members it knows. Its risk is total concentration in one product and one local economy, and it has no analytics capacity at all. Delinquency runs above the system; charge-offs run below it. It needs a read it cannot produce.

THE INDIRECT BUYER IN THE MIDDLE

$200 million to $1 billion, 60 to 120 employees, half or more of the vehicle book bought from dealers, third-party paper above net worth. Big-institution exposure with small-institution staffing: enough people to run an indirect program, not enough to audit one. The fifteen auto-focused credit unions between $500M and $1B carry the worst delinquency and charge-offs of any cohort we measured.

THE QUIET FADE

Fortress capital, delinquency below the band, members leaving 3% a year, assets shrinking. Nothing on the credit side will ever flag it. It merges into a larger institution from a position of strength, if it moves early, or from weakness if it waits. Member growth is the only line that saw it coming.

THE BOOK THAT LOOKS TOO CLEAN

Zero delinquency, zero charge-offs, negative earnings, a stable balance sheet, and a filing that looks a lot like last year's. That was Unilever FCU and Aldersgate in 2025, and it cost the insurance fund $16 million. A lender with no credit losses that cannot make money is either mispriced, overstaffed, or misreported. All three are board emergencies, and the read does not need to know which.

WHAT THIS DESK CANNOT TELL YOU
Which dealers wrote the indirect paper, or how each is performing. Whether a participation's originator is any good. Whether a blank repossession field means an empty lot or an empty line. Anything about a filing that is false. Anything after the last cycle, which trails the FPR site by about a quarter. The numbers are the institution's own claims about itself, resolved and compared; treat every flag as the question to ask next, never as the answer.
WHERE THE MONEY SITS · LOAN MIX AND FUNDING
The tape read's Loan mix and Funding groups answer two questions the four core questions skip: what kind of lender this is, and who is lending it the money. Neither group carries a flag on its own; they tell you which of the other numbers to weigh.
Loan mix concentration · real estate vs vehicle
real estate (703 + 386) / 025B · first-lien share · credit card 396 / 025B · unsecured 397 / 025B
MEANS
What the loan book is made of. A small credit union is usually one of two things: a car lender or a mortgage lender, and the two carry opposite risks. Vehicle paper is short, loses money fast and visibly, and reprices with the book. Real estate is long, rarely defaults, and reprices almost never.
WHEN HIGH
Real estate above half of loans at a small credit union is a duration bet, not a credit bet. A 4% first mortgage written in 2021 still sits on the book while shares now cost 3-4%; the margin squeeze shows up in NIM, not in delinquency, and it lasts as long as the mortgages do. Credit cards or unsecured signature loans above 15-20% of the book is the opposite exposure: high yield, high charge-off, and the first thing a stretched member stops paying.
WHEN LOW
Real estate under 10% means a consumer book: vehicles, cards, signature loans. Faster losses, faster repricing, and every asset-quality line on this desk becomes the read that matters.
VS PEERS
Under $50 million the book is vehicles first and unsecured second; real estate share climbs with size and passes half of loans somewhere above $500 million. A mortgage-heavy $30 million credit union is unusual and usually a housing-cooperative or church-sponsored charter.
OUR LINE
no standalone flag · context only
LIMITS
The 5300 gives balances, not rates or maturities, so the duration exposure is inferred from the mix, not measured. Home equity lines are in the real estate total but behave like consumer credit.
Commercial lending · % of loans and % of net worth
commercial loans (400 series) / 025B · commercial / 997
MEANS
Member business lending: loans to businesses, farms, and investors, the one category most credit unions did not exist to make. Against net worth it is the exposure the statute cares about; against loans it is how much of the book it is.
WHEN HIGH
Commercial above 100% of net worth means a single bad year in business lending can consume the capital. The old statutory member business lending cap sat at 1.75 times net worth (roughly 12.25% of assets); most small credit unions are far below it, and one approaching it has usually hired a lender from a community bank and is growing a book the board does not fully understand. Business loans are lumpy: one $2 million credit at a $40 million institution is a concentration on its own.
WHEN LOW
Zero is the norm under $100 million. It is not a weakness, only an absence.
VS PEERS
Commercial share of loans runs near zero at the small end and 8-12% above $1 billion. A small credit union with a meaningful commercial book is the exception and belongs in its own peer set, not its band.
OUR LINE
no standalone flag · context only · above 100% of net worth is surfaced on the tape read
LIMITS
The filing does not separate owner-occupied from investor real estate, or show borrower concentration. Participations in commercial loans bought from other credit unions land here too, with all the 701.22 caveats.
Core shares vs certificates · money market
(regular shares 657 + share drafts 902) / 018 · certificates 908 / 018 · money market 911 / 018
MEANS
Who is funding the institution and why they stay. Regular shares and draft accounts are relationship money: the member banks here. Certificates are rented money: the member chose the rate. Money market shares sit between, priced like a certificate and callable like a draft account.
WHEN HIGH
Certificates above 40% of shares means the credit union is renting its deposits and will re-rent them at whatever the market says next quarter; the funding cost follows rates up with a lag of months, not years, and the members holding them leave for 25 basis points. Money market above a quarter of shares is hot money in the same sense with no maturity to slow it down. Both compress the margin exactly when a long loan book cannot reprice.
WHEN LOW
Core shares above 70% of the base is sticky, cheap funding, the structural advantage a credit union has over every other lender. It is also the profile that goes with an older membership and a shrinking one; check member growth before calling it a strength.
VS PEERS
The system moved sharply toward certificates through 2023-2025 as rates rose; a credit union whose certificate share doubled in two years bought its liquidity. Compare the mix against the band and against its own history.
OUR LINE
no standalone flag · context only · certificates above 40% of shares is surfaced on the tape read
LIMITS
Non-member and brokered deposits are reported separately and are not in this split. Certificate maturities are not in the bulk file, so we cannot say when the repricing lands, only that it will.
Average share balance
018 / 083 · total shares per member
MEANS
How much the typical member keeps at the institution, and therefore what kind of membership it is. A $2,000 average is a payroll-account membership: paychecks in, paychecks out, small balances, small loans, fee income that matters. A $15,000 average is a saver base: fewer transactions, larger certificates, members who watch rates.
WHEN HIGH
High average balances go with certificate-heavy funding and rate sensitivity. A saver base is stable in a flat rate environment and runs to whoever pays more when rates move; the balance-sheet strength is real and the loyalty is conditional.
WHEN LOW
Low average balances mean the membership is the community's working population, which is the franchise the charter was written for and a thin one to run a branch on. Fee income, overdraft, and small consumer loans carry the earnings; margin does not.
VS PEERS
The system median is a little above $6,000 per member. A small credit union well above it is a sponsor group with high earners or a savings club; well below it is a payroll shop or a community charter serving lower-income members.
OUR LINE
no standalone flag · context only
LIMITS
An average hides the distribution. Ten large depositors at a small credit union can pull the average up while most members hold a few hundred dollars, and those ten are the run risk.
Participation flows · bought and sold
purchases SL0037 and sales, year-to-date annualized · participation charge-offs 550F against outstanding
MEANS
Which side of the participation market the institution is on. A buyer has more deposits than loans and is deploying them into paper someone else wrote. A seller has more loans than deposits, or a concentration it needs to reduce, and is moving paper out. The posture tells you what the balance sheet is trying to fix.
WHEN HIGH
Buying above 25% of net worth a year is a due-diligence obligation under 701.22: independent credit analysis before purchase and ongoing monitoring after, documented, regardless of what the seller provides. A six-employee credit union cannot staff that. Participation charge-offs above 1% say the paper bought is performing worse than the book around it, which is what happened system-wide in 2025.
WHEN LOW
Zero on both sides is normal for most small credit unions. A seller with nothing bought is a liquidity-tight lender exporting its excess originations, usually vehicle paper into larger buyers.
VS PEERS
The market is roughly $19 billion a year and no filing names a counterparty. The buyer's tape shows a balance; the risk is in the originator's book. Read the Participation Desk for the market view and the flow map.
OUR LINE
no standalone flag · context only · purchased balance above 25% of net worth is surfaced on the participation buyers screen
LIMITS
Annualizing year-to-date flows makes first-quarter figures noisy. The 5300 shows amounts, not pool count, collateral, or servicer; the 701.22 file is where those live, and it is not public.