Monday, March 18, 2013

Free Banking and Bank Runs

The European Union has demanded that Cyprus impose a tax on bank deposits as a condition of a bank bailout.   The Cyprus government has promised to compensate bank depositors with stock in banks' guaranteed by natural gas revenue.

This tax is claimed to be a violation of the government's deposit insurance guarantee.   Advocates of deposit insurance are predicting bank runs throughout the European Union.  

There were runs on Cypriot ATM machines over the weekend, which have run out of currency.    The banks in Cyprus are closed until Tuesday.  

I oppose the existence of deposit insurance.   In today's world, where deposit insurance has become commonplace, my view is the less the better.

What is the alternative?

Free Bankers have suggested two closely related solutions to bank runs.

The first is the option clause.

The option clause is a clause in a deposit contract.   It gives the issuer, the bank, the option of postponing repayment of the deposit in exchange for the payment of an interest penalty.  The penalty is paid to the depositor.

Option clauses were banned long ago for conventional bank deposits.    (However, it may be that these old regulations have been repealed by accident.)   As far as I know, such clauses are legal for deposit-like instruments such as overnight repurchase agreements.   They have not been used.

In the Scottish banking system of the 18th century, banknotes (hand-to-hand currency issued by banks,) included an option clause.   The banknotes were generally redeemable in gold or silver coin on demand, but the issuing bank could postpone payment if it chose.    If it did postpone payment, then it would would pay interest on the banknotes until it resumed redeeming them with gold and silver coins.

From the point of view of those holding and using banknotes, they promised to pay gold or silver on demand, or else, after a time along with bonus interest payments.   The interest compensated those using the banknotes for the inconvenience from having to wait to obtain gold or silver coins.

From the point of view of an issuing bank, it borrowed money at zero interest, as long as it stood ready pay off any banknotes presented on demand, or else paid interest on the funds until such a time it could redeem the funds.    The interest paid imposed a penalty on the bank if it failed to provide continual redeemability.

An option clause for checkable deposits is more or less the same thing.    The issuing bank promises to pay off checkable deposits on demand with "money," but an option clause would allow it to postpone payment while requiring it pay a higher interest rate on deposits.   As long as the bank is able to continue to pay off deposits, it pays a lower interest rate.   If it is unable to make those payments, it pays a higher penalty interest rate.  From the depositor's point of view, the checkable deposit is payable on demand with a more modest interest rate or else, after a delay, with bonus interest during the period of delay.

Similar clauses can  (and should) apply to other demand or overnight debt instruments.   These would include overnight repurchase agreements, overnight commercial paper. and even money market mutual funds.  (With mutual funds, the management fee is higher if redeemability of shares is maintained.   If not, the management fee is reduced.)

While an option clause is pretty much essential for the sorts of demand liabilities used as media of exchange, there are advantages to having such a clause for time deposits too.   For example, suppose a bank is funding thirty year mortgages with 3 month certificates of deposit.    The certificate of deposit could be due in three months at a lower interest rate, but after some longer period of time, with a higher interest rate.   

Traditionally, the "option," in the clause was the banker's.     When the banker decides he is in danger of default, he exercises the option, and begins owing penalty interest until he can return to making payments on demand.   However, it might be better if there were also a provision for those at the end of the line, including a deposit insurer (if such exists) to insist that the option be exercised.

Presumably, insolvent banks would typically exercise the option clause.   While the penalty interest payments deepen the hole, what do they have to lose?   Depositors would be stuck with funds in the insolvent institution while it undertakes various "go for broke" strategies.

This brings up the second alternative to deposit insurance, one that would come into play anytime a bank chooses to exercise the option clause.    Once the option has been exercised, if any depositor asks, the bank should be subject to an outside examination for solvency.  

If a bank is determined to be insolvent, then it should be immediately reorganized.   The principle should be that the existing stockholders get nothing and all of the bank's creditors, including all of the depositors, have a substantial "haircut" and receive shares of stock in proportion to their initial claims against the bank.   The bank is  instantly recapitalized.   The bank is now solvent.   The deposits, still subject to the exercised option, are paying bonus interest to the depositors.   The depositors (and other creditors) are now also the new owners of the bank.

It would be wise for the new owners of the bank to sell off most of the stock--it is simple diversification.   And then they can deposit the money the receive for the stock back into what now should be a solvent bank.   Since the bank was insolvent before the reorganization, the depositors would typically end up with a smaller balances than they had before.   The bank would use the new deposits to fund the purchase of sound assets.

Leaving aside some possible economy-wide shortage of base money, the newly capitalized bank would receive the funds needed to resume payments and so return to paying the lower, ordinary interest on deposits.   Fixing the solvency problem would almost certainly correct any liquidity problem.

There are two thought experiments relevant to the option clause and rapid reorganisation approach.   One is a problem with  a single bank.   A single bank is subject to a run, or perhaps some mismanagement that leaves it short on reserve balances or vault cash.     Selling liquid assets or short term borrowing would be its first recourse.   If a bank is perceived sound, it should have little difficulty in obtaining all the funds it needs.

Considering the banking system as a whole, when a single bank has a liquidity problem,  then other banks are receiving extra funds.   For example, a bank's customers are spending by writing checks.   Those checks are received by other banks and deposited, requiring a shift of funds through the clearing system.  While this causes a liquidity problem for the bank losing funds, the clearing process is providing funds to the other banks.  

Similarly, the currency that is withdrawn and spent by customers of one bank is received by customers of other banks and deposited.   One bank loses funds, but other banks gain funds.

Even a "classic" run, where people line up at a bank to obtain currency will likely involve them then depositing their funds in another bank.

These other banks have the funds needed to purchase short term assets the bank losing funds must sell.  Those other banks have extra fund to lend to the bank losing funds.

This suggests that a single bank with a liquidity problem severe enough so that it will need to exercise the option clause is likely to have a solvency problem.   It maybe be solvent in fact, but the problem exists because their depositors and other potential lenders, including other banks, have doubts.  

Such a troubled bank would end up exercising the option clause.   And then it is subject to the solvency examination.   If it is insolvent, then it is reorganized.   And once reorganized the formerly troubled bank is now well-capitalized and should have no problem borrowing from other banks to resume redeemability and reduce its interest expense.

If it turns out that the examination shows the bank to be solvent, but it still cannot borrow funds to meet it's obligations to make payments, then it is important that the bank's liabilities be negotiable.   Even if they are not usually transferable, once the bank has exercised its option, that should no longer apply.   This would allow depositors needing funds to sell their deposits, most likely to another bank, and continue to make payments.

The other scenario is where the liquidity problem is general.   There is a run on the banking system, or some other scramble for base money.   Banks that usually have no trouble selling short and safe assets or else borrowing from other banks find that no one else in a position to buy what have been highly liquid assets in the past or else to provide short term loans.    The presumption that there is some solvency problem is no longer justified.

In this situation, all of the banks end up exercising the option clause at more or less the same time.   All of the banks would be paying penalty interest rates on deposits.   All of the depositors would be receiving those funds as a bonus interest rate compensating them for an inability to obtain funds.

Aside from the penalty/bonus interest on deposits, the result would be similar to the periodic currency suspensions in the U.S. during the 19th century.   There is no need for a disruption of interbank clearings, though interbank settlements during a suspension effectively involve interbank lending.   In the 19th century these were coordinated by the clearinghouses and took the form of clearinghouse certificates.   Banks with adverse clearings borrowed them and banks with favorable clearings ended up owning them.  

Of course, with a market monetarist policy, a generalized liquidity problem would be short lived.   If the demand for base money rises, then the quantity should increase.   As those receiving new base money deposit it in these sound banks, the liquidity problem would disappear and banks would again resume payments.

Sadly, it is all to conceivable that all of the banks might be insolvent, (which might be the cause of the liquidity problems.)   Still, as with the individual banks, the result of a systemic run due to systemic insolvency would be a prompt exercise of the option clause.  

This would trigger the outside examination of the banks.  Since they are insolvent by assumption, the result would be reorganization of all the banks.    The deposit holders end up with fewer deposits and instead end up with stock in their banks.   The banks then, should expand deposits and credit to replenish deposit holdings.   (Hopefully, the banks would purchase sound assets this time around!)

As usual, I want to make a plug for private issue of hand-to-hand currency.   If there is a systemic crisis, and all banks exercise the option clause, and the only sort of currency is base money, then the result would be a shortage of currency.   (While there are worse possibilities than a currency shortage, the problem doesn't develop if the banks are issuing banknotes.)   Like all of the deposit money, the currency would also be subject to the option clause, and could continue to be used to make payments during the period of suspension.

Unfortunately, if all of the banks are insolvent, one of the more disruptive aspects of reorganization would be the need to issue new hand-to-hand currency and exchanging it for the old currency.   The old currency would be accepted at a discount.

Now, perhaps this system seems awfully disruptive and inconvenient.   Instead, why not just have the government guarantee all of the deposits, so that there are never bank runs.   And then, if there is any other liquidity problems, have a central bank to serve as lender of last resort.   With careful regulation of the banking system, there will never be much problem with insolvency.   We can depend on selfless government regulators, supervised by the statesmen elected by well-informed and public spirited voters. 

And if somehow, despite wise regulations, there is insolvency, then the government can just borrow all the funds need to cover the losses and protect the deposits.   What if the if the losses are so great that the taxpayers can't afford to pay the interest on all of the debt?   Then have foreign taxpayers fund a bailout.....

And we are back to the problem of the banks on Cyprus.    My only question is about the supposed "theft" of deposits, is what is happening to the stockholders in Cypriot banks?   If they face a 100% tax on their stock, then I can understand why Cypriot depositors are unhappy, but they are not being robbed.   They just made a bad investment.    The future is uncertain, so all investment involves risk.


  1. Basically, if we lived in a world you were describing, I would fully agree with you. If market outcome was that the bank must fail, also if there was no deposit insurance across the EU and the world, things would be pretty clear. When government intervenes by creating governmental deposit insurance it basically approved reckless depositing and reckless bank practices.

    Looking at the decision from current situation, removing deposit insurance in one case and not in other can adversely impact other institutions, in other countries that dont have problem. On the other side, i call it a robbery because, government promised there are no "bad investments", and on the other side it was not the market outcome, where depositors and banks agreed upon this type of resolution in mutual contract, but it was decided in Brussels, by politicians, who also could have decided otherwise. Its bad crisis management where future electoral outcomes played important role, as always. It would be perfect if people considered deposit as an investment that also carries certain risk which they bear.


  2. This is a great post.

    Do you think there is any potential for private deposit insurance as a third option?

  3. I think private deposit insurance should be something in addition to the option clause and rapid reorganization. If a run develops, there should still be an option clause to stop the run. And if the institution is insolvent, it should still be reorganized.

    The private deposit insurer then gets the stock that goes with the insured deposits and pays the depositors part (or all) of the difference between the value of the stock and the loss in the depositor balance.

    I would think that a deposit insurer would need some way of compelling the bank to exercise the option clause. Of course, the bank would need some kind of protection against abuse.

  4. I would think that had we had free banking all along that banks would have evolved away from backing there money with gold, silver or any other commodity but would be backed by their own credit based on the assess that they own. Their currency would be their most senior asset. The par could be provided by some particular bank's currency based on the number of their notes in circulation.