In an FEE conference lecture, Lawrence White argues that fractional reserve banks can exist legally (under libertarian law) and are viable businesses because people primarily use banks to exchange deposit balances with one another rather than for safe storage of money. I think this argument appears to have some merit on first glance but ultimately breaks down on closer inspection.
It is true that bank customers primarily use their bank account as a money-exchange device - just calculate the average volume of check/debit or other ledger payments versus cash withdrawals. The percentage of cash flow which is ledger payments (rather than cash deposits or withdrawals) is almost 100%, and there is no reason to believe that it would be significantly less in a free banking economy. However, the root problem of paying interest on demand deposits is that, as Rothbard puts it, the bank is technically insolvent, since its liabilities have zero maturity and its assets have maturities out to 30, 60, 90 days (typically) and loans which have an even longer time horizon. A fractional-reserve bank, no matter how useful and attractive it might at first seem to customers, even under libertarian law, would always be liable to bank runs. During times of uncertainty, bank runs would occur and such banks would, in fact, collapse.
It is often pointed out by defenders of fractional reserves that all businesses are liable to go bankrupt due to mistakes or failure to foresee bad economic conditions. Collapse of fractional reserve banks, on this view, is no different - no one could have foreseen that economic uncertainty would occur and, as a result, bank panics and collapses. However, this is a false claim - successful businesses and individuals do in fact predict and prepare for “rainy days” and hold liquid assets for the purpose of self-insuring against unforeseeable (therefore, uninsurable) calamities. Over time, a “market level” of such holdings will emerge, such that, successful businesses are those which typically hold X% of their assets liquid in the event of unforeseeable economic conditions.
The same would hold true of time-deposit structuring in a free banking economy - banks would need to be able to calculate their exposure to unforeseeable risks and, over time, those banks which fail to make the time structure of their assets and liabilities solvent will collapse and serve as “object lessons” to the industry. In essence, the fractional-reserve banker is making guesses about the time horizons of his customer’s deposits… “I guess they won’t be demanding this money for X days, so I can loan it out to ABC Corp. for X days and provide some interest-sharing to my depositors to incentivize deposits to my bank.” Banks which use time deposits do not have to take on the risk of mistaken guesses about the time horizons of its customers… the customers themselves estimate their time horizons and bear the risks of miscalculation. This distributed knowledge is certainly more accurate than the centralized knowledge of the banker and, in any case, ensures that the risks of economic uncertainty are being borne by those who actually want it. If you want to take on the risk of locking your money away in a time deposit, you can do so and assume the risk of bankruptcy if you are unable to meet your own liabilities in time… but other customers of the bank do not have to bear any of the risk you are taking on to yourself.
White suggests that the fact that people primarily use their deposits for purposes of exchanging with other account-holders implies that fractional reserves banking is a viable business model - but this fails to take into account the time horizons problem and treats monies of different maturities as if they are homogeneous. Essentially, White is claiming that people would want to perform ledger transfers between accounts whose deposits have been loaned out at various maturities. Imagine I have a time deposit with 25 days remaining to maturity. I cannot exchange dollars from this time deposit at a 1:1 ratio with deposits of zero maturity for the same reason you cannot sell a bond at face value - no one will ever pay the full price of the bond in the present because time preference is never zero. If I wanted to buy a $2500 flat screen with money from my 25-day maturity time deposit, I would have to adjust for the interest rate over 25 days. Let’s say the interest comes to 0.5% (roughly 6% per annum). I would have to add $12.50 to the “zero maturity” purchase price In order to pay the TV seller out of my 25-day to maturity time deposit. The $2500 television would cost $2512.50 to purchase out of my 25-days to maturity time deposit.
Only deposits with identical maturities could be exchanged at a 1:1 ratio. This is the result of time preference and arbitrage. Any other arrangement would cause someone to bear losses and provide risk-free profits to someone else. Banking institutions could separate deposits into classes on the basis of maturity and neither banking institutions nor risk-averse depositors would want to bear the risks which other depositors choose to take on without compensation that is in direct proportion to the risk being borne, the fractional reserve system would fall into disuse after a few banking panics which leave mostly non-fractional reserve banks standing. Fractional reserve banks induce each customer to share in the common risk pool of all depositors. The profit-sharing paid from the bank’s interest proceeds is pro rata to the size of each interest-bearing deposit but all depositors bear equally the risks which the bank’s management takes. If the bank’s assets become illiquid and it is unable to meet its obligations, all depositors lose some or all of their money, even if they wanted the bank to hold their money at zero maturity, on demand. Full-reserve banks could pay as much or more interest on time deposits without pooling risks between unlike depositors.
Fractional reserves should not be prohibited in libertarian law, IMO (I reject the argument that they necessarily constitute “fraud”) but they also would not be very prevalent (again IMO). I think White’s argument is neither here nor there in this regard because he fails to take into account that money at different maturities would not exchange dollar for dollar in a free banking economy.
Any thoughts?
Clayton -