Coverage has focused on the proposed changes in bank-size categories. But the proposed Tier 1-capital benchmarks for community development may be the far more consequential story.
The reported story
Most news coverage of the 2026 Community Reinvestment Act Notice of Proposed Rulemaking has focused on the proposed changes in bank-size categories.
That focus is understandable. Under the proposal, a small bank would be an institution with assets of no more than $1 billion, an intermediate bank would have more than $1 billion but no more than $10 billion, and a large bank would have more than $10 billion in assets.
Those classifications determine which tests apply, which data must be collected and how a bank will be examined.
But that may not be the biggest story in the 407-page proposal.
What’s the big story?
Buried in the proposal is a quantitative standard based on Tier 1 capital for community development lending, investments and grants.
Under the quantitative option proposed by the OCC and FDIC:
- Annually, a large bank would need community development loans equal to at least 0.625% of allocated Tier 1 capital in each assessment area.
- Annually, the same large bank would need community development investments and grants, collectively, equal to another 0.625% of allocated Tier 1 capital.
- Annually, an intermediate bank would need community development loans, investments and grants equal to at least 1.25% of allocated Tier 1 capital.
You read that correctly.
The agencies are proposing explicit numerical standards that appear extraordinarily low when compared with actual historical community development lending.
The potential consequence is not that banks would be prohibited from doing more. Of course they could do more. The danger is that a published regulatory number can quickly become an operating target—and these numbers may reset expectations dramatically downward.
An important clarification: What do the benchmarks actually do?
The proposed percentages are part of what the agencies call a “geographic flexibility standard.”
They would determine whether a bank could receive consideration at the State, multistate MSA or bank level for community development activities conducted outside its assessment areas. The proposal presents the percentages as one option; a second option would retain a qualitative standard based on whether the bank has an adequate level of responsive community development activity. Both options stipulate that a bank must engage in sufficient community development lending and investing to be considered as “adequate” within the bank’s defined communities before any community development activity outside those boundaries will be considered.
Effectively, the proposed percentages are a prescribed minimum for a bank to attain a satisfactory CRA rating for its CD lending and CD investing (including grants) in any assessment area.
That makes them important.
The agencies say the quantitative standards generally reflect the minimum level of community development activity a bank would be expected to conduct—absent other considerations—to avoid a “needs to improve” rating. They also state that the percentages are not sufficient by themselves to determine a bank’s community development performance.
That distinction is important. But so is the regulatory signal.
Once an agency publishes a numerical minimum and says examiners cannot require a higher level for the purpose of receiving broader geographic consideration, banks will inevitably compare their programs with that number.
The question is whether the proposed number bears any reasonable relationship to historical performance.
How do the proposed standards compare with historical performance?
There has never been a clear official numerical standard for community development activity under the legacy CRA framework.
In the 2023 CRA rulemaking, the agencies acknowledged that thresholds might improve consistency but said available data were insufficient to establish them. That rule was later enjoined, and the agencies subsequently proposed returning to the prior framework.
There is, however, substantial public data about community development lending. CRA-reporting institutions disclose the number and dollar amount of their community development loans annually.
GeoDataVision reviewed the 2024 CRA data for 565 OCC- and FDIC-regulated reporting banks. Those institutions reported:
- 23,477 community development loans originated;
- approximately $111.97 billion in community development lending; and
- approximately $18.42 trillion in combined assets.
Reported community development lending therefore equaled approximately 0.61% of the banks’ combined assets. Almost half of the institutions in the analysis reported community development lending exceeding 1% of assets.
The broader official FFIEC data, which also includes Federal Reserve-supervised reporters and purchased loans, shows the same enormous scale. Across all 731 CRA reporters, 646 institutions reported 29,361 community development loan originations totaling approximately $137.2 billion during 2024.
Why does the Tier 1-capital denominator matter?
Tier 1 capital is only a fraction of a bank’s total assets.
Consider a simplified example in which Tier 1 capital equals 10% of assets. A community development loan benchmark equal to 0.625% of Tier 1 capital would equal only 0.0625% of assets.
What happens when the proposed formula is applied to the 2024 data?
Compare the Tier 1 Capital benchmarks with the 0.61% community-development-loan-to-assets ratio in the GeoDataVision analysis of actual 2024 CD lending.
CD loans originated reported by FDIC- & OCC-regulated banks: $111.97 billion
Required CD loan volume dictated by the Tier 1 benchmarks: $11.69 billion
The actual historical lending results would be almost ten times the proposed minimum.
That is not a small difference. It is an entirely different scale of activity.
JPMorgan Chase offers a striking example. The bank reported approximately $10.526 billion in community development loans originated during 2024. Applying the proposed 0.625% formula to its Tier 1 capital produces a benchmark amount of only $1.848 billion.
Again, the proposed percentage is not a cap. It does not compel JPMorgan Chase—or any other bank—to reduce its activity. But it is a signal from regulators as to what minimum level of activity they will accept to consider a bank’s CD activity as “satisfactory”.
But what happens over time when a bank’s historical activity is many times higher than a newly published regulatory minimum?
That is the question the agencies and the banking industry need to confront.
The difference may be even greater than the reported data suggests
The 2024 CRA lending data reflects loans originated during that year.
The NPR proposes that examiners also consider prior-period community development loans held on a bank’s balance sheet, creating treatment similar to community development investments. The agencies say this could encourage longer-term lending by recognizing both current activity and qualifying loans from earlier periods.
That means historical performance under the proposed approach would include more than the $111.97 billion of current-year lending in the GeoDataVision analysis.
Outstanding prior-period loans would be added.
Consequently, the gap between actual activity receiving consideration and the quantitative formula could be even larger than the current-year comparison indicates.
Why should communities care?
Explicit standards can be highly desirable.
For decades, banks have been forced to estimate examiner expectations without a transparent numerical reference point. Greater clarity and consistency would benefit banks, regulators and communities.
The problem is not the concept of a published standard.
The problem is whether the standard is realistic.
A benchmark set far below actual historical activity could become a magnet pulling performance downward. Budgets are reviewed. Programs are compared with regulatory expectations. Management asks how much activity is necessary. A number that begins as a geographic-flexibility threshold can become, in practice, the number around which institutions build their programs.
If that happens, affordable housing projects, small-business initiatives, community facilities, economic-development programs and nonprofit organizations could ultimately receive less financing.
What should the regulators do?
Before adopting any quantitative standard, the OCC and FDIC should:
- Publish the historical data used to select 0.625% and 1.25%.
- Compare the proposed percentages with actual community development lending, investment and grant activity.
- Provide bank-level examples showing how the formula would work in practice.
- Explain why Tier 1 capital is the appropriate denominator.
- Clarify precisely when the percentage is only a geographic safe harbor and when it may influence a performance conclusion.
- Test whether the benchmark could unintentionally encourage institutions to reduce activity that already substantially exceeds it.
The agencies should also reconcile their present confidence in explicit percentages with the earlier conclusion that available community development data was insufficient to establish reliable thresholds.
Conclusion
The proposed bank-size changes are important. But they are not the only—or necessarily the biggest—story in the 2026 CRA proposal.
The proposed community development percentages could become some of the most consequential numbers in the entire rule.
GeoDataVision’s analysis indicates that the amount produced by the proposed formula is almost 90% below actual 2024 community development lending among the OCC- and FDIC-regulated reporters studied. Once prior-period loans, investments and grants are considered, the difference could be even greater.
The agencies may not intend these standards to reduce community development activity. But regulatory incentives do not depend solely on intent.
They depend on the numbers institutions are given and how those numbers will be used.
Comments on the proposal are due October 13, 2026. Banks, community organizations and other interested parties should examine the formula carefully and insist that any final quantitative standard be calibrated to actual historical experience.
Communities have too much at stake for anything less.
References
- OCC Bulletin 2026-35: Community Reinvestment Act NPR
- Federal Register: 2026 Community Reinvestment Act proposal, 91 FR 52114
- FFIEC findings from the 2024 CRA data
- FFIEC 2024 community development lending table
- FFIEC 2024 CRA flat files
- FDIC statement discussing the 2023 rule’s data limitations
- US Banks Ranked by Tier1 Capital
