Bank Runs, Deposit Insurance, and Bank Regulation - Part II - Contagious Bank Failures

Sep 21, 2026

Uploaded by Patrick

Examines how “sunspot” and contagious bank runs can emerge in a banking system and evaluates whether federal deposit insurance is necessary to prevent failures spreading from one bank to otherwise solvent institutions. Develops the concept of self-fulfilling runs based on depositor expectations, contrasts textbook hysteria with rational withdrawals tied to financial fundamentals, and reviews historical evidence from U.S. banking eras regarding the role of extraneous events.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages5

Reading Time13 min

AccessFree to download

Contagious Bank Failures

Why Bank Runs Can Be Contagious

by Charles T.Carlstrom Contagious bank falures re ofenthought to be a possible conse-quence of a banking system withoutfederal deposit insurance.This articleconsiders whether federal depositinsurance is necessary to preventthese types of bank runs. Part I,which was presented in the Feb-ruary 1 Economic Commentary,de-scribed some of the costs and benefitsof providing deposit insurance andconcluded that an analysis of conta-gious bank failures is necessary inorder to understand these benefits. Part ll continues with an examinationof contagious bank runs and a discus·sion of how the market handled bank-ing panics prior to the FederalReserve System and the Federal De-posit Insurance Corporation. ■ Contagious Bank FailuresThe last apparent difference betweenbanks and other businesses is the pos-sibility for a rumor or a failure ofanother bank to ignite bank runs andcause the failure of financially soundbanks. These types of bank failures are termed"sunspots"because,if depositors trulybelieve that a bank's solvency dependson events unrelated to market funda- Bank Runs,Deposit Insurance,and Bank Regulation,Part II mentals—such as the amount of solaractivity—a bank's solvency would,infact,depend on the amount of solaractivity.In the typical example,asunspot is the failure of one bank or agroup of banks,which ignites rumorsthat other banks might also fail It would seem irrational for depositors to run on a solvent bank.How.ever,because a bank's liquidity andsolvency depend in part on the num-ber of depositors wishing to withdrawmoney,it is rational for each deposi-tor to queue up if he expects otherdepositors also to stand on line. Sunspot bank runs are also said to be"bubble"phenomena.One of themost famous examples of a bubbleinvolved tulip bulbs in Holland dur-ing the seventeenth century.Investorsfrantically bought tulip bulbs,expect-ing their price to rise,which in turncaused their price to rise.2 Sunspot bank runs are like bubbles inthat they are self-fulfilling prophecies.To determine the correct regulatoryresponse to this apparent market fail-ure,one must first inquire empiricallyhow frequently bank failures arecaused by sunspots and then ask whatis special about banking that allowsthese types of phenomena to arise. February 15,1988 Part I of this article,presented in theFebruary 1 Economic Commentary,described some of the costs andbenefits of providing federal depositinsurance.The major benefit of pro-viding deposit insurance is the pre-vention of contagious bank runs-abank failure that spreads to solventbanks.Part II,presented here,dis-cusses why bank runs may be conta-gious and examines some of theways in which private clearinghousesprotected against widespread bankfailures.The article concludes thatfederally provided deposit insurancemay not be necessary in order toprotect against such bank runs. Determining how often bank runs arecaused by sunspots—extraneousevents—is difficult to do with any degree of statistical accuracy.However,we can examine whether bank failureswere the products of the same type ofdeposit and withdrawal behavior dur-ing both panics and nonpanics Gorton tests this hypothesis for bankfailures during the U.S.National Bank-ing Era(1863 to 1914)and showsthat the factors affecting deposits andwithdrawals were similar in periodsof widespread bank failures and inperiods when banking failures werenot widespread.His results suggestthat "banking panics during theNational Banking Era were systematicresponses by depositors to changingperceptions of risk." Corroborating evidence that extrane-ous events did not seem to cause asubstantial number ofbank failuresprior to the Great Depression is givenby Benston,et al.They show that theaverage annual rate of bank failuresfor the 1875 to 1919 period was 0.82percent,versus 1.01 percent for non-financial firms.If banks are likeother firms except for the possibilityof contagious bank runs,one wouldexpect the failure rate of banks to beat least as

FAQs

What are “sunspot” bank failures in this article?

They are bank runs triggered not by market fundamentals but by rumors or failures of other banks that cause depositors to believe solvency depends on unrelated events.

Why does the article argue that depositor runs can be rational?

Because each depositor’s decision depends on the number of other depositors withdrawing—queueing becomes rational when everyone expects others to line up.

How does the article assess whether federal deposit insurance is necessary?

It compares the expected contagion benefits against evidence suggesting contagious sunspot runs may not be widespread, concluding that federally provided insurance may not be strictly necessary.

Sep 21, 2026

Uploaded by Patrick