Is ChatGPT right about the fate of Regional Banks?

ChatGPT generated image of ChatGPT

I recently asked ChatGPT to forecast the percentage share of US deposits that will be held by banks under $100B in Assets today based on some data points. And the response was alarming, but not surprising.

Context

I’ve stood on stage at conferences and proclaimed:

“The US does not need 9,000 deposit taking institutions … but I don’t want to live in a US where there are only 9.”

My position on this was largely formed during the COVID19 Pandemic and in particular the Paycheck Protection Program or PPP (which probably gives some of you cause to shudder). A Government program to distribute $500B to small businesses through the banks – and the banks were given basically 1 week to figure it out. 45% of the PPP funds were distributed by banks with < $10B in assets. Those banks made up 15% of the total assets in the US at the time, but they distributed almost half of the PPP funds.

Regional banks, community banks and credit unions are important. But I’m genuinely concerned about their long-term viability. And this concern increased recently when I saw the following statistics from a J.D. Power survey showing the breakdown of new checking account openings across the top 10 institutions:

Obviously, I didn’t expect to see a community or regional bank in this list. But the skew towards digital banks (Chime, SoFi) and Mega Banks (Chase, BofA etc.) gave me pause.

The ChatGPT Prompt

So – I decided to do what any sane person would do, and ask ChatGPT 😊

But I was quite specific and (hopefully) measured in my prompt to ChatGPT. I typed the prompt exactly as it’s shown below.  

Prompt: 

I have two data sources.

1. FDIC Call Reports indicate that banks with < $100B in Assets make up 29% of the Total Assets under management in the US banking system.

2. Cornerstone Advisors research that indicates banks with < $100B in Assets accounted for 4% of all new checking accounts opened in the US.

Roll this forward 5, 10 and 15 years - what is the likely % of US Assets that will be managed by banks under $100B in Assets. Explain the assumptions you have made in performing this forecast such as population growth, mortality of account holders, etc.

The ChatGPT Response

Amusingly, the response started with a comment that reminded me of the South Park episode on generative AI (language advisory on the link) – namely, it was good for the ego “That’s a sharp and strategic question — and it hits directly at the structural shift in U.S. banking” 🤣

But here’s ChatGPT’s (V5.0 for anyone that cares) response to my prompt…everything in GREEN is what came back from ChatGPT, unedited.

That’s a sharp and strategic question — and it hits directly at the structural shift in U.S. banking: the shrinking relevance of sub-$100 billion banks despite their current systemic importance. Let’s walk through the logic, the key assumptions, and the modeled trajectory for 5, 10, and 15 years.

Baseline Data (2025)

• Current total U.S. bank assets: ≈ $24 trillion (FDIC Q2 2025)

• Banks < $100 B: ≈ 29% of total → ≈ $7 trillion AUM

• Banks ≥ $100 B: ≈ 71% of total → ≈ $17 trillion AUM

• Checking account originations (Cornerstone Advisors 2024):

  • < $100 B banks = 4% of new accounts

  • ≥ $100 B banks = 96% of new accounts

Core Assumptions for the Forecast

Category	Assumption	Rationale
Total banking system asset growth	4% CAGR	Roughly in line with nominal GDP growth + mild inflation (2–3% real + 1–2% inflation).
Large-bank (> $100B) asset growth	5% CAGR	Faster growth via digital acquisition, consolidation, and investment scale.
Small-bank (< $100B) asset growth	1.5% CAGR	Reflects low new customer inflow (4% of new accounts), deposit outflows to fintechs and large banks, and merger attrition.
Population growth	0.5% CAGR	Consistent with U.S. Census projections. Minimal driver of overall deposit growth relative to consolidation dynamics.
Account turnover / mortality	1–1.5% of households per year	Account churn due to death, moves, and consolidation, largely benefiting larger banks or fintechs.
M&A consolidation rate	2–3% of sub-$100B banks absorbed annually	Historically, 150–200 community bank mergers/year; trend likely to continue or accelerate.
Fintech share creep	Implicit in large-bank growth rate	Fintechs are often backed or partnered with large banks for balance-sheet usage, reinforcing concentration.

Forecast Model

We can model total U.S. bank assets (nominal) as growing ~4% CAGR.Then apply differential growth rates by segment to estimate share of total assets.

Year	Total Assets (Trillions $)	< $100 B Banks Assets (Trillions $)	Share of Total	≥ $100 B Banks Assets (Trillions $)	Share of Total
2025 (base)	24.0	7.0	29%	17.0	71%
2030 (+5 yrs)	29.2	7.6	26%	21.6	74%
2035 (+10 yrs)	35.6	8.2	23%	27.4	77%
2040 (+15 yrs)	43.4	8.8	20%	34.6	80%

Interpretation

• By 2030: Sub-$100 B banks decline to ≈ 26% of total U.S. banking assets.

  
  

  (Their share erodes ~1 ppt per year.)

• By 2035: ≈ 23%, driven by continued digital migration and M&A attrition.

• By 2040: ≈ 20%, a potential “floor” where remaining institutions are highly specialized (ag, local CUs, niche commercial lending).

  
  

Narrative Summary

Over the next 15 years, unless small banks reinvent their digital acquisition and deposit-gathering engines, their collective share of U.S. banking assets will likely fall from 29% → 20%.

This decline mirrors a broader network-effect concentration: large and digital-first institutions dominate new household account openings and commercial flows, while smaller institutions consolidate into fewer, more specialized players.

Sensitivity Scenarios

Scenario	< $100B Growth Rate	System Growth	15-Year Share
Optimistic (tech parity achieved)	3%	4%	25%
Base case (status quo)	1.5%	4%	20%
Pessimistic (accelerated consolidation)	0.5%	4%	17%

The Conclusion

OK – so at the current rate of ‘growth’, banks with < $100B in Assets will go from having around $7T in AUM or 29% of the total to $8.8T in AUM (sounds good) which is 20% of the total (sounds very bad). And my gut is telling me that the current trajectory is actually worse than that. Worse even than the Pessimistic case ChatGPT laid out which results in < $100B banks having just 17% of the total AUM.

Again – the US doesn’t need 9,000 deposit taking institutions … but do you want to live in a US with just 9?

Around 85% of the assets in each of the UK, Canada and Australia are controlled by 4 or 5 mega banks. In fact, if any one of Australia’s “Big 4” banks could move its assets to the US, it would be the 5th largest bank in the US. Think about that. From a country with less than 26 million people (about 65% of the population of California) – ANY ONE of the big 4 banks would be the 5th largest in the US.

I really want to create a conversation about this. Its important. There are businesses in my community I frequent that survived because of the actions of regional and community banks during the COVID19 Pandemic and PPP. I have friends with businesses that survived the pandemic because community banks helped when mega banks stood on the sidelines and watched.

How do we strengthen our regional bank and credit union sector? Yeah, yeah – invest in digital. Everyone knows this is critical – but how do we convince executives at these institutions that this needs to be a priority? Decisions and investments need to be made before its too late.