Bowman splits Fed bank supervision into five state-line regions, each with one accountable leader

"In practice, these committees became a source for plausible deniability and a disincentive for examiners to take prompt and decisive action to address identified risks," Bowman said at the St. Louis Fed's annual community banking research conference ^[4][1]. Her complaint is about who signs a decision. For decades, she said, a web of dozens of committees produced delayed enforcement and a lack of accountability ^[5]. She said the independent review whose preliminary findings she received last month showed "a mismatch between authority for decision-making and accountability for supervisory decisions" ^[6].

Supervision will no longer be organized by Federal Reserve Bank district ^[2]. The new regions follow the structure the Conference of State Bank Supervisors already uses. Bowman, a former Kansas state bank commissioner, explained the choice in charter terms ^[14][13]. "After all, our banks are 'state member banks,' and each state conducts supervision jointly with us and, ultimately, holds the state bank charter," she said ^[14]. With regions drawn on state lines, each state supervisor deals with one Fed regional leader who answers for all supervisory activity in that territory ^[2].

The promise of speed is harder to test. The speech as reported does not include figures on how long findings took under the committees, so the case that five regional heads will act faster rests on Bowman's account ^[5]. If the design works, examiners issue early findings without the old delays ^[3]. That would meet Bowman's own standard: "A bank should not learn about and then be held accountable for changed expectations during an examination," she said ^[7]. If it does not, the delay moves up a level. Five leaders now hold decisions that dozens of committees used to share ^[5], and a cautious regional head can sit on a finding as long as any committee did. A third possibility is that banks' costs depend less on the org chart than on the threshold changes in the same speech.

The Fed will propose raising fixed-dollar asset thresholds to account for inflation and economic growth, with a mechanism to update them every five years ^[10]. Jaret Seiberg of TD Cowen said he believes the Fed has the leeway to do it, and that the new lines could sit 50% higher ^[11]. "That means the $100 billion threshold would be $150 billion, the $250 billion would be $375 billion," Seiberg wrote in a note to clients ^[12]. On that estimate, a bank with assets between $100 billion and $150 billion falls below the first line ^[12]. The note as published says the $700 billion line "would be more than $1 billion". Fifty percent above $700 billion is $1.05 trillion ^[15].

The community-bank line has been assets under $10 billion for 15 years ^[9]. On a five-year reset it would have been revisited three times in that span ^[16]. Bowman framed the coming change around business models: certain firms with traditional, non-complex models will count as community banks, so they are no longer pushed into supervisory tiers built for complex firms ^[8]. She did not say how the new test will work ^[8].

I think the thresholds and the definition move more money than the reorganization does. They decide which tier of supervision a bank pays for, and the regions decide who signs a finding. The counter-case is that earlier findings with a name attached lower the cost of exam-time surprises for every bank the Fed supervises, at any size. That case would beat mine if the Fed starts publishing the time from examination to supervisory finding by region, and the number falls.