The assessment cut is not the change that matters

The FDIC has proposed cutting deposit insurance assessments by roughly $4.0 billion a year. For a community bank the arithmetic is easy, the reaction is predictable, and both are beside the point. The rate cut is a reversion. The change that will still be shaping bills in five years is a threshold.

The proposal was published on 30 June and the comment file closes on 31 August. It does four things: it moves the dividing line between small and large institutions from $10 billion in assets to $30 billion and indexes it to inflation; it cuts initial base assessment rate schedules by 2 basis points for small institutions and 1 basis point for large and highly complex ones; it creates a resolution readiness adjustment worth up to 1 basis point for large and highly complex institutions that elect it; and it clears out obsolete provisions.

Where the $4.0 billion comes from

Estimated annual decrease in assessments, $ billions

Only the first bar is a figure the FDIC prints on its own. The rule gives $129 million for the threshold change, about $2.7 billion for the threshold and the rate cuts together, and about $4.0 billion for the whole proposal assuming three quarters of large and highly complex institutions take the full adjustment. The other two bars are the differences between those three numbers, which makes them approximate in the same way the totals are.

The rate cut gives back what 2023 took

The 2 basis points being removed are the 2 basis points that were added. The FDIC raised initial base rate schedules effective 1 January 2023 as part of a statutorily required Restoration Plan, after the Deposit Insurance Fund reserve ratio fell below the 1.35 per cent statutory minimum. The ratio reached 1.36 per cent at 30 June 2025, the Restoration Plan ended, and it stood at 1.43 per cent at 31 March 2026, progressing toward the FDIC's long-term goal of a 2 per cent designated reserve ratio.

So the schedule for established small institutions goes back roughly to where it sat before the fund was in trouble.

Those are annual rates on the assessment base, which is average consolidated total assets less average tangible equity. Two basis points on a $500 million base is $100,000 a year. On $2 billion it is $400,000. Useful money, and worth being precise about what kind of money it is.

A rate that went up because the fund fell can go up again for the same reason. This is a variable cost that happens to be falling.

The saving is contingent on the fund, and it scales with the balance sheet, which means a bank that grows into it will see the benefit grow and a bank that budgets it as a permanent line has mislabelled it. The honest treatment is a rate that moves with the reserve ratio, forecast alongside it.

The threshold is the structural change

Moving the small and large dividing line from $10 billion to $30 billion reclassifies 76 institutions. The number of banks priced as large or highly complex falls to 74.

Those 76 do not simply pay a different rate. They are priced by a different machine. Small institutions are priced on a formula over Call Report measures and the CAMELS composite. Large institutions are priced on a scorecard. For a bank crossing that line, the question changes from what the scorecard says about it to which of its own reported ratios now drive the bill, and the second question is one a bank can actually work on.

The FDIC's stated basis for the move is a similarity finding: it compared institutions between $1 billion and $10 billion with the 76 in the $10 billion to $30 billion band and reports that on average both groups showed approximately similar shares of deposit and loan types, and similar leverage ratios, as of 31 December 2025.

Three details in the mechanics are worth having straight.

• Reclassification down is immediate. A bank priced as large immediately before the effective date that reports under $30 billion becomes small on that date. It does not have to report under the threshold for four consecutive quarters first. In both directions afterwards, the four-quarter test applies.
• The line moves. The threshold is adjusted every four years for inflation on a pre-determined indexing methodology, calculated on cumulative CPI-W data through August of the adjustment year and effective for the assessment period beginning 1 October. A bank sitting at $27 billion today is not looking at a fixed line, and neither is one at $31 billion.
• Not everyone pays less. The FDIC projects that four institutions, about 0.1 per cent, would see assessments rise, three of them by 1 per cent or more of pre-tax income. Against 74.3 per cent of profitable institutions seeing a reduction of at least that size, this is a small tail. It is still a reason to compute the number rather than assume its sign.

The adjustment that prices resolvability

The resolution readiness adjustment is a large-bank item and community banks can skip it, but it is the most revealing part of the document.

It is worth 0.5 basis points for passing a virtual data room test and 0.5 basis points for providing prescribed data access. The test gives the institution 48 hours to populate the room, and it is satisfied only if the FDIC judges the material sufficient for a potential bidder to conduct adequate due diligence, including that the financial information provided is consistent with the general ledger.

That is not a risk adjustment. Nothing in it estimates the probability that the institution fails. It prices how quickly the institution could be sold if it did, which is a different question, and the FDIC is now willing to pay a basis point for the answer.

The mechanism makes the incentive sharper than the word "voluntary" suggests. Under current rules, minimum initial base rates for large and highly complex institutions step down from 5 basis points to 2 and then 1 as the reserve ratio reaches 2 per cent and 2.5 per cent. Under the proposal they step down from 4 to 3 and then 2, because the adjustment has been folded into the schedule itself. A participating bank arrives at the same place. A bank that declines pays more than it would have under the old schedule.

What to do with it before the end of the month

Comments close on 31 August 2026, under RIN 3064-AG27. After that the useful work is arithmetic rather than advocacy.

If you sit between $10 billion and $30 billion, the delta is not the interesting output. Run your own numbers through the small bank method and find out which reported ratios would become the drivers of your assessment rate, because those are the ones that will be worth managing. If you are near the line in either direction, model the four-quarter test and the indexed threshold together, since the classification is a moving target rather than a status. If you are comfortably below it, recompute the saving on your own assessment base and enter it in the budget as a variable tied to the reserve ratio.

The limit

None of this is in force. It is a proposed rule, the comment file is open, and the schedules in the regulatory text carry their effective date as a bracketed blank: beginning [the quarter in which a final rule becomes effective].

That blank is the most honest line in the document. The rates are drafted, the fund is recovering, and the date is not yet a date. A bank can plan against the direction with some confidence and should not plan against the timing at all.

---

Sources

• Assessments Thresholds, Rate Schedules, and Adjustments (91 FR 39794) · Federal Deposit Insurance Corporation, notice of proposed rulemaking