Overview

This chapter examines the credit cost of stricter U.S. banking supervision during the Global Financial Crisis. Loss recognition can bind capital and reduce lending to sound borrowers, but it can also stop evergreening and remove loans unlikely to be repaid; balance-sheet data cannot fully separate those channels. Among surviving banks, strict exposure was associated with 5.8% lower total-loan growth, 3.1% fewer branches, and 2.1% fewer employees from 2008q2 to 2013q4. The loan-growth gap mostly faded by 2018q4, while the branch effect remained detectable. Effects were larger for high-risk banks, and survivor-only measurement makes the lending estimate conservative. The evidence identifies a crisis-era trade-off, not a complete welfare verdict.

Learning outcomes

Recognized losses can reduce useful credit through capital constraints while also ending evergreening of loans unlikely to be repaid.
During 2008q2–2013q4, strict exposure was associated with 5.8% lower total-loan growth, 3.1% fewer branches, and 2.1% fewer employees.
The loan-growth gap mostly faded by 2018q4, but branch effects remained detectable and contractions were much larger among high-risk banks.

Keywords

CreditLendingCapitalEvergreeningBranchesCrisis

The Cost of Cleaning Up: Credit in a Crisis

This chapter examines the credit cost of stricter supervision in the same Global Financial Crisis setting. The specific question is whether forcing a weak bank to recognize losses merely removes bad loans or also reduces the supply of useful credit. The answer matters because the capital discipline that can make a failing bank easier to resolve can also constrain banks that remain open. Readers should be able to separate those two channels, track their different time horizons, and see why the burden is greatest for banks that entered the crisis closest to a capital constraint.

The trade-off is not between supervision and no supervision. A bank that continues lending to a borrower unlikely to repay may be postponing a loss rather than supporting productive activity. But a bank that cuts sound lending because recognized losses leave too little capital can deepen a recession. Both outcomes can appear in a falling loan balance, so the balance-sheet number alone cannot decide what happened.

The same capital signal has two credit channels

The first channel is a capital constraint. When supervisors require a bank to increase provisions or recognize charge-offs, current income falls and reported book equity becomes smaller. Unless new capital arrives quickly, the bank has less room to expand assets while maintaining its regulatory capital ratios. It may respond by shrinking the loan book, selling assets, retaining earnings, or reducing other inputs to lending. A stricter examination can therefore affect borrowers who were not responsible for the original losses, because the bank’s capacity to originate or renew loans is tied to its measured capital.

Capital serves as a loss-absorbing buffer, but it also limits the balance sheet that can be built on top of that buffer. A bank does not simply choose a loan volume independently of equity. New loans add assets, and the regulatory framework requires enough capital relative to the bank’s measured exposures. When the denominator grows or the numerator falls, the institution must either raise capital, retain more earnings, reduce risk-weighted assets, or accept greater regulatory pressure. The New York Fed study finds no detected effect of stricter supervision on newly issued equity, while it does find lower dividends. Retaining earnings helps, but it may not offset the amount of capital that a more conservative loss assessment removes from the reported balance sheet.

This mechanism also explains why the effect need not be uniform across loans. A bank can preserve lending to the safest relationships while cutting higher-risk or capital-intensive exposures. It can refuse new applications, decline renewals, sell loans, or let existing balances run off. Each response changes the observed loan-growth measure, but the distributional consequences differ. The evidence identifies the overall contraction and its concentration among weaker banks; it does not specify the exact borrower-level allocation of every reduction.

The second channel is loan cleansing. Banks sometimes continue a troubled relationship by extending or refinancing credit to a borrower that is unlikely to repay. This practice, often called evergreening, can keep a loan from being recorded as nonperforming and postpone the recognition of the bank’s loss. A strict supervisor who requires the bank to classify the loan more realistically may cause lending to that borrower to stop. The resulting contraction is not necessarily a loss of productive credit. It can be the visible end of lending that was sustaining an unviable claim.

The two channels can occur together in the same institution. Suppose a bank has a weak property loan and a healthy small-business portfolio. A stricter review can force the property loss into the accounts, reducing capital. The bank may then stop renewing the property loan, which is cleansing, and also reduce new small-business lending, which is a credit-supply cost.

Published by DeepMox

FAQ

How can stricter supervision reduce lending without every lost loan being productive credit?

It can reduce lending through two channels: a binding capital constraint that cuts the capacity to originate or renew loans, and loan cleansing that ends evergreening of borrowers unlikely to repay. The observed loan decline combines these effects, and the data do not fully allocate each reduction.

Did the loan contraction last after the Global Financial Crisis?

The loan-growth gap between high- and low-strictness surviving banks mostly disappeared by 2018q4, although the branch reduction remained statistically detectable. Convergence does not show that every borrower regained the same credit or terms.

Which banks bore the largest credit effect of stricter supervision?

The largest effects were concentrated among banks in the upper quintile of predicted failure risk based on 2008q2 information. For that group, total and residential-real-estate loan growth were each about 13 percentage points lower through the crisis, while the commercial-real-estate interaction was about 17 points lower.

Previous9 minHow Stricter U.S. Banking Supervision Moves the Failure Clock

Next15 minWhat the Evidence Shows About U.S. Banking Supervision

U.S. Banking Supervision: Credit in a Crisis

The Cost of Cleaning Up: Credit in a Crisis

The same capital signal has two credit channels

FAQ