What the Evidence Actually Says About Stricter Supervision

This series has examined U.S. banking supervision of troubled and failing insured banks through one question: why can stricter supervision close a failing bank faster instead of saving it? The answer is a connected mechanism, not a paradox once the pieces are separated. Supervision can expose losses, change the book-capital signal used for legal action, and impose discipline on a bank that might otherwise continue operating. The same process can lower the cost of each resolution while tightening credit for surviving banks. Readers can now judge claims about supervisory strictness by asking what information changed, which outcome was measured, whether the comparison is causal, and what the evidence leaves unresolved.

The most important qualification comes first. The New York Fed staff report does not conclude that every banking system should become stricter, that every distressed bank should be closed immediately, or that a higher failure count is automatically good. It identifies how supervision operated in a defined sample during the Global Financial Crisis and sets that mechanism against its costs and limits.

The complete mechanism: from uncertainty to a legal decision

A failing bank does not move directly from “bad loan” to “closed.” Between those states lies a chain of measurement and judgment. A loan may be worth less than its recorded amount while its eventual recovery remains uncertain. The bank estimates provisions and charge-offs; an examiner reviews those estimates; a Call Report records income and equity; a capital ratio places the institution in a regulatory category; and the supervisor decides whether monitoring, restrictions, enforcement, or closure is warranted.

That chain explains why supervision is both information production and discipline. The information function makes the bank’s condition more reliable. The discipline function changes what the bank is allowed or required to do once that condition is recognized. If the bank is allowed to keep optimistic values on its books, a legal capital gate may not be crossed. If an examination requires a conservative revision, the bank may suddenly look weaker by the measure that authorizes formal action. The examination did not create the original loan loss. It changed the point at which the loss became usable evidence.

The distinction between book equity and economic capital remains essential. Book equity is a regulatory and accounting signal, not a complete measure of what the bank would be worth if every asset were sold today. Yet a supervisory system cannot base every intervention on an unobservable counterfactual. It needs measures that can be reported, reviewed, and applied consistently. The weakness of the signal is therefore also part of its usefulness: it creates a common language for action, even though it can lag or misstate economic value.

The evidence shows that capital becomes especially decisive in distress. In the New York Fed study’s all-bank data, banks with CET1 ratios above 5% essentially never failed in the plotted sample, while the probability of closure approached 70% when equity-to-assets fell below 2%. Failing banks’ equity-to-assets ratios declined by about eight percentage points over the five years before failure, ending near 3% immediately before closure. Their average CAMELS rating rose by 2.8 points, and failing banks were usually rated 4 or 5 before they closed.

Those figures should not be treated as universal legal thresholds. They show the empirical shape of the relationship in the study: as reported capital becomes very low, the range of plausible supervisory discretion narrows. CAMELS adds information about asset quality, management, earnings, liquidity, and sensitivity to market risk. But in a troubled bank, capital, asset quality, and earnings deteriorate together as losses are recognized. The composite rating becomes closely tied to hard capitalization rather than functioning as a wholly independent judgment.

This is also why an examination can have effects that persist beyond its original meeting. A downgrade can lead to more frequent or longer examinations. Those examinations can produce additional revisions, which can reinforce the downgrade and move the bank out of a routine supervisory path. The result is feedback: distress attracts scrutiny, scrutiny produces information, and the information changes the bank’s supervisory state. A strict early assessment is not merely a harsher label attached to a fixed institution.

The legal framework gives that changing state consequences. Section 38 of the Federal Deposit Insurance Act defines five capital categories and establishes Prompt Corrective Action as a way to resolve insured-institution problems at the least possible long-term loss to the Deposit Insurance Fund. An undercapitalized institution faces close monitoring, a capital restoration plan, and limits on growth, acquisitions, branches, and new lines of business.