Look at how a large bank builds a commercial credit analyst and the timeline is right there in the job posting. KeyBank's Commercial Credit Analyst Program runs six to eight weeks of classroom work covering financial accounting, financial analysis, risk rating rationale, and credit policy, followed by a twelve month rotation on the credit surveillance team. Only after that rotation is the analyst considered for a seat on a portfolio management team. Call it fourteen months from first day to first real assignment, and that is at an institution with a formal program, a training budget, and a bench.
Most community banks do not have any of that. They have a chief credit officer who learned the work from someone who has since retired, and a new analyst sitting next to her.
I want to be careful about how this gets framed, because the easy version of this story is wrong. The easy version says banks cannot hire. The data does not support it. The 2025 CSBS Annual Survey of Community Banks ranks staff retention fifth among internal risks, with 68 percent of bankers calling it extremely or very important. That is a real number, but it is down from 85 percent in the 2022 survey, and CSBS attributes the decline to labor market shortages easing. Hiring got easier. The problem did not go away. It moved.
What the Fourteen Months Actually Buys
Here is the part worth sitting with. Ask a credit officer what a first-year analyst spends her time on and you will hear a fairly consistent answer. She is spreading rent rolls. She is keying T-12 operating statements into a template. She is reconciling a personal financial statement against a K-1 that does not tie. She is chasing the third version of an appraisal because the second one was missing an exhibit. A commercial real estate loan file arrives as seventy or more distinct documents, and converting that pile into something a committee can act on is where the first year goes.
That work has to happen. It is also almost entirely mechanical, and it is not what makes someone a credit person. Credit judgment comes from seeing enough deals to know which sponsor story holds up and which one is a spreadsheet with a good haircut. It comes from watching a deal get taken apart in committee. It comes from being wrong once, in public, on a loan that later went sideways.
So the sequence most banks run is this. Spend the scarcest fourteen months you have teaching someone to do the part of the job that does not compound. Then, right as she reaches the part that does, lose her.
The Turnover Is Real, and the Reason Is Not What You Think
Crowe's bank compensation survey, drawn from 388 financial services organizations, found officer level turnover doubled in two years, from 3 percent in 2021 to more than 6 percent in 2023. In a business where a seasoned credit officer represents a decade of accumulated pattern recognition, doubling is not a rounding error.
The reason bankers gave is the sentence I keep coming back to. The top driver of departures was lack of career development, at 45 percent. Compensation came second, at 42 percent. People are not primarily leaving over money. They are leaving because year one looked like data entry and year two looked like more of it.
And the pipeline behind them is thinning. The Bureau of Labor Statistics counts roughly 67,800 credit analysts in the United States as of 2024 and projects the occupation to decline through 2034, while still generating about 3,700 openings a year. Read those two numbers together. Nearly every opening in this field is a replacement, not a new seat. The profession is not growing. It is churning.
The Clearest Tell
David O'Connell at Datos Insights has made the point that the formal credit training programs the large banks ran have mostly been shutting down since the early 2000s, and the lenders and underwriters who came up through them are now reaching retirement. That is qualitative, but you can see it confirmed structurally.
The Massachusetts Bankers Association now runs a Credit Analyst Apprenticeship Program. It is an eight month program, and it is on its sixth cohort.
Think about what it means that a state banking association has to run the apprenticeship. It means individual member banks concluded they could not carry that training load alone. When an industry starts pooling its apprenticeships, the institutional knowledge transfer inside the individual firms has already broken.
The Version of This That Actually Works
None of this is an argument for fewer underwriters. It is an argument about what you spend their first year on.
If the mechanical layer of the loan file is handled before the analyst opens it, if the documents arrive already extracted, normalized, and tied out with every figure traceable back to the page it came from, then her first year looks different. She spends it on the exceptions rather than the extraction. She spends it reading the deal, testing the sponsor's assumptions, and learning to defend a recommendation in front of people who will push back. She reaches the compounding part of the job in months instead of years.
That is a better outcome for the bank on two separate counts, and the second one is the one that gets missed. The obvious benefit is throughput. The less obvious benefit is retention, and it is arguably the larger one, because 45 percent of the people who left cited career development. A first year spent on judgment rather than keying is not a perk. It is the retention strategy.
It also changes what an examiner sees. An analyst who spent her first year reasoning through exceptions rather than transcribing documents produces a cleaner, more defensible file, which is the same discipline we wrote about in the community bank's playbook for AI-powered CRE underwriting, and it sits inside the supervisory expectations we mapped in what the OCC, FDIC, and Federal Reserve actually say about AI in commercial loan underwriting.
There is a version of this argument that overreaches, and I want to name it rather than let a reader find it. Automation does not create credit judgment. It creates the time and the reps in which judgment can form. An analyst who never has to touch a rent roll will not automatically become a better credit person than one who spent two years in the weeds. What she will have is more deals seen, more committee cycles observed, and more exceptions reasoned through in the same span of calendar time. Judgment is a function of reps. The fastest way to more reps is to stop spending them on transcription.
The banks that work this out over the next few years will not be the ones with the biggest training budgets. They will be the ones who figured out that the fourteen months is the scarcest asset on the balance sheet, and stopped spending it on data entry.
If you are carrying meaningful CRE concentration and running it through two or three analysts, I would rather show you what that first year could look like than describe it.
Send me two or three times that work for your team over the next two weeks and I will walk you through it. See how community and regional banks are using LenderBox, or request a demo.
