Bank Runs

The danger of bank runs has been overstated. For one thing, a bank run is unlikely to cause insolvency. Suppose that depositors, worried about their bank's solvency, start a run and switch their deposits to other banks. If their concerns about the bank's solvency are unjustified, other banks in the same market area would generally gain from recycling funds they receive back to the bank experiencing the run. They would do this by making loans to the bank or by purchasing the bank's assets at non-fire-sale prices. Thus, a run is highly unlikely to make a solvent bank insolvent.

Of course, if the depositors' fears are justified and the bank is economically insolvent, other banks would be unlikely to throw good money after bad by recycling their funds to the insolvent bank. As a result the bank could not replenish its liquidity and would be forced into default. But the run would not have caused the insolvency; the insolvency had already been incurred, but not fully recognized. The recognition of the existing insolvency caused the run.

Runs are feared even more because of their potential spillover to other banks. The likelihood of this happening depends on what the "running" depositors do with their funds. They have three choices:

If the run is either type 1 or 2, no great harm is done. The deposits and reserves are reshuffled among the banks, possibly including overseas banks, but they do not leave the banking system. Temporary loan disruptions may occur because borrowers have to transfer from deposit-losing to deposit-gaining banks, and interest rates and exchange rates may change. But these costs are not the calamities that people often associate with bank runs.

Higher costs could occur in a type 3 run. Currency (an important component of bank reserves) would be removed from the banking system. This would cause a multiple contraction in aggregate money and credit, which would dampen economic activity in other sectors. In addition, almost all banks would sell assets to replenish their liquidity, and few banks would buy. Fire-sale losses would be large, and the number of bank failures would increase.

In practice, bank failures have been relatively infrequent. From the end of the Civil War through 1920 (after the Federal Reserve was established in 1913 but before the Federal Deposit Insurance Corporation in 1933), the bank failure rate was lower than that of nonbanks. The failure rate increased sharply in the twenties and again between 1929 and 1933, when nearly 40 percent of the nation's banks failed. Yet, from 1875 through 1933, losses from failures averaged only 0.2 percent of total deposits in the banking system annually. Losses to depositors at failed banks averaged only a fraction of the annual losses suffered by bondholders of failed nonbanking firms.

A survey of all failures of national banks from 1865 through 1936 by J. F. T. O'Connor, who was comptroller of the currency from 1933 through 1938, concluded that runs were a contributing cause in less than 15 percent of the three thousand failures. The fact that the number of runs on individual banks was far greater than this means that most runs did not lead to failures.

The evidence suggests that most bank runs were type 1 or 2, and few were of the contagious type 3. Because a type 3 run—a run on the banking system—causes an outflow of currency, such a run can be identified by an increase in the ratio of currency to the money supply (most of the various measures of the money supply consist of currency in the hands of the public plus different types of bank deposits). Increases in this ratio have occurred in only four periods since the Civil War, and in only two—1893 and 1929 to 1933—did an unusually large number of banks fail. Thus, market forces and the banking system on its own successfully insulated runs on individual banks in most periods. Moreover, even in the 1893 and 1929-33 incidents, the evidence is unclear whether the increase in bank failures caused the economic downturn or the economic downturn caused the bank failures.