<<  <  >  >>

Page 10

individual insolvencies could be minimized, the likelihood

of contagion or systemic risk would be greatly reduced.  As

discussed below, delays that have permitted financial insti-

tutions to become deeply insolvent before closure are pri-

marily due to regulatory, not market, failure.12

Second, before the introduction of the lender of last

resort in the United States, the failure rate of banks was

actually lower than that of nonfinancial firms, and losses

to depositors and other bank creditors were lower than to

creditors of nonfinancial firms.13  In addition, U.S. banks

held higher capital-to-asset ratios prior to safety net reg-

ulations.  Recent international experiences suggest that

banks substitute government deposit insurance or public cap-

ital for private capital.14  Again, the safety net may have

made banks more, not less, fragile.

Third, Charles Calomiris and Joseph Mason have examined

in detail the bank panic that took place in Chicago during

June 1932.15  Although there did seem to be some temporary

confusion about the quality of bank assets and a short-lived

general depositor run, Calomiris and Mason do not find any

evidence of failure of banks that were solvent at the begin-

ning of the panic.  The runs were directed primarily against

the weakest banks, which were the ones that failed.  Thus,

even during the heights of bank panics of the Great Depres-

sion, depositor runs do not appear to have generated "conta-

gion" or "systemic" problems that caused otherwise solvent

institutions to fail.

Fourth, historically, bankers developed innovative con-

tracts to attenuate the likelihood of panic runs.  One exam-

ple is the "option clause" that came to be a standard provi-

sion in private bank notes circulating in Scotland during

its 18th-century "free-banking" era.16  The option clause

gave bank directors the right to suspend specie payment for

up to six months, but the bank then promised to pay a high

rate of interest on the notes during the period of suspen-

sion.  This clause allowed the banks to stop "panic" runs

and to have more time to adjust to negative liquidity shocks

that might occur, thereby avoiding fire-sale losses.  Also,

banks in Scotland had some form of extended or unlimited

liability covering their notes, rather than simple limited

liability.  Those notes were widely and voluntarily accept-

ed, and the Scottish banking system showed much greater sta-

bility than did the English system during this period.

Finally, there is little historical evidence that per-

mitting banks to expand their portfolios to include equity

reduces stability.  Eugene White shows that, during the