Subject: Banking & Finance
Paper: NZZ
Reading time: 4 Min

Need for instruments against a bank run

An American law professor makes a proposal on how a withdrawal in panic of bank deposits could be curbed

With the fall of Credit Suisse, a new term emerged in the public opinion - "bank run". The term refers to what happens when many bank customers or financial institutions withdraw their deposits at the same time because they fear for a bank's solvency. The greater the outflow of money, the more likely a default becomes, which in turn causes more customers to panic and to close their accounts. A snowball effect that turns into an avalanche. In extreme cases, the bank's liquidity is not sufficient to cover the cash drain. In the week leading up to 18 March, customers werereportedly withdrawing around 10 billion Swiss francs per day.

Enacted as a reaction to the financial crisis of 2008, the current legislation tries to prevent a bank from running its business on too little capital; for example, because it has taken on higher risks than expected. As has become evident, these rules are insufficient to deal with a bank run due to a panic among depositors. The bank's liquidity remains at risk.

Contractual instruments exist that can mitigate a bank run. However, these are barely used. Time deposits are interest-bearing bank accounts with a predetermined maturity date. Time deposits generally earn slightly higher interest than a normal account. However, depositors must leave their funds deposited for the fixed term in order to receive such an interest. The longer the term, the higher the interest payment. In contrast, depositors who withdraw their funds early are penalised.

Depositors can also protect themselves by investing their assets in securities instead of cash. In this case, the securities are segregated from the bank's assets. But this still does not solve the problems of the bank, which is dependent on liquidity to run its business.

Danger of misaligned incentives

Ben Bernanke, Douglas Diamond and Philip Dybvig received the Nobel Prize last year for their research on banking and financial crises. They point out two further instruments that can be used to avoid a bank run: deposit insurance and the so-called "lender of last resort". However, both measures lead to false incentives by tempting a bank that knows about its protection to take excessive risks. The unsuccessful attempts of the Swiss National Bank (SNB) also show the limits of the lender of last resort: SNB's supply of liquidity to Credit Suisse did not help, even if the SNB was willing to waive corresponding collateral. In such a situation, measures of this kind reinforce the loss of confidence of its depositors by pointing out the difficulties of a bank.

Jeffrey Gordon, a professor at Columbia University, proposes a different solution to counteract this. According to him, a bank should offer three types of deposit accounts: The first is a retail account, it offers daily liquidity but subject to a deposit threshold. The second, a commercial transaction account, is used by an operating company to cover its regular transaction needs, such as payroll or payments to suppliers. This account should also provide daily liquidity, but without any threshold. The third account should be a so-called storage account for individuals above a certain threshold, for operating companies and their non-operating cash holdings, and for institutions holding funds for potential investment or distribution.

Withdrawal with discount

The funds in such storage accounts are held as a time deposit with a specified term and an appropriate interest rate. The depositors have the right of immediate withdrawal, but at a discount for early redemption. In addition, these depositors must hold an appropriate number of shares in the bank. The aim is for a significant proportion of the bank's shares to be owned by the depositors, a kind of mutual ownership. Crucially, those who exercise the right to early redemption lose their shares in the bank at the same time. This increases the relative holding of the remaining shareholders in the bank's capital. This prospect of appreciation should therefore increase their incentive not to panic and leave their deposits with the bank.

The downside of Gordon's proposal is that the prevention of a bank run comes again at the expense of depositors. Their liquidity is significantly reduced, and at the same time they are forced to hold shares in their bank and take a capital risk. Such a solution leads to a fundamental change in the relationship between depositors and banking institutions. If such a solution is introduced, it also remains open how depositors can be prevented from simply migrating to a system that does not have such a feature.

It is noteworthy that the proposed solutions are not aimed at remedying a loss of confidence among depositors, but only at slowing down the subsequent panic-like withdrawal of their deposits. However, given the creation of a banking colossus that some believe is "too big to fail" [edit: too big to bail], Gordon's and other proposals should also be seriously discussed.