In Defense of the “Two Bank” Banking Reform Proposal

IOU in a piggy bank” by Images_of_Money is marked with CC BY 2.0.

In his recent essay for, titled Building a Better Banking System, Alasdair Macleod agrees with the Austrian view that sound money would help prevent a looming “combined currency, asset, and banking crisis”, yet challenges the practicality of the “Two Bank” banking reform proposal that was popularized by Murray N. Rothbard in his 1983 book The Mystery of Banking. Macleod instead proposes that the key reform would be to eliminate the limited liability protection for bank stockholders. The prospect of having a bank’s owners’ personal assets seized to satisfy a bank’s depositors would be sufficient to instil discipline and probity in the place of the current moral hazard risk fostered by limited liability and government deposit insurance. The Austrian position is that the “Two Bank” reform would eliminate one source of currency expansion; i.e., the ability of banks and banks alone to create money ex nihilo (out of thin air) via their privilege to engage in fractional reserve banking (FRB). Macleod believes that eliminating this privilege would cause credit to dry up and the economy to suffer. (The central bank’s ability to create reserves ex nihilo is the other source of money expansion and is not directly addressed by Macleod or the “Two Bank” reform proponents.)

There is much more to Macleod’s essay. It deserves to be read and reread very carefully, for there is much to learn. Before we return to his proposal to remove the limited liability for ownership for banks, let us quickly review Rothbard’s “Two Bank” proposal.

What is the “Two Bank” proposal?

Under the Rothbard proposal, banking would be divided into two separate legal entities–the Deposit Bank and the Loan Bank. Funds deposited in the Deposit Bank would be backed one-hundred percent by reserves, whether gold, silver, or even fiat currency. The Deposit Bank would not be allowed by law to lend these funds. The deposit would be a form of bailment, in which the ownership of the property does not change hands. The depositor still owns the amount of his deposit, and the Deposit Bank is merely the custodian. The depositor would use these funds just as he does now–to pay obligations–using any of the modern methods of money transfer such as check, debit card, or reserve certificates issued by the Deposit Bank or even the central bank. Of course, the Deposit Bank would earn its funds only from fees for its services, since it would not be allowed by law to lend these funds to earn interest income.

Macleod considers that these fees would be a major obstacle to public acceptance of the system, since today most checking accounts are free and may even earn some small amount of interest for the depositor. However, Deposit Banks would be offering a valuable service in a competitive environment. The cost of handling mostly electronic transfers of funds is very small, and banks today actual earn money from every debit card transaction that they process. In fact some banks require that the depositor engage in a certain number of debit card transactions each month in order to get fee discounts. In addition, for many years banks have earned substantial fee income for handling their customers’ overdraft transactions. Banks also earn substantial fee income from their ATM networks for handling non-customer transactions. As a result for many years banks have been offering free checking accounts with no minimum balance in order to capture what can be substantial fee income. Therefore, there is every reason to believe that checking accounts still would be widely accepted by the general population.

The Loan Bank

 A depositor of the Deposit Bank who wishes to put his money to work would transfer some of his funds to the Loan Bank in order to earn interest income. At this point the Loan Bank owns the funds for some agreed upon period of time. The Loan Bank invests these funds–it matters not whether to single projects, such as an oil well, a basket of loans similar to a mutual fund portfolio, or the bank’s general loan portfolio–with a promise to return them to the customer with interest.

Two facts become clear. One, lending funds to the Loan Bank, which are further lent to the Loan Bank’s customers, does not affect the money supply. And, two, the lent funds carry some risk to the lender (the Loan Bank) and the customer who lent his deposit funds to the Loan Bank. As is the case today, some loans may not be repaid, will be repaid only partially, and perhaps not at the agreed upon time. Depending upon the terms of the agreement between the customer and the Loan Bank, the Loan Bank may absorb these losses with its capital account and repay the customer in full and on time, or the customer may lose all or part of his lent funds and/or not be repaid on time.

Such is the inescapable nature of banking and lending. All loans carry some risk of non-repayment.

Two Essential Reforms

There are two essential reforms that are key to the Rothbardian “Two Bank” system. One, the Deposit Bank is subject to the same commercial code as any other business. It does not enjoy any special protection for using its customers’ funds. If the Deposit Bank lent or leased the customers’ deposits, it would engage in fraud. Two, the Loan Bank does not enjoy the privilege of engaging in fractional reserve banking, whereby it can create deposit money out of thin air. All loans would be funded out of customer savings. (The Deposit Bank’s customer lends funds to the Loan Bank, which further lends these same funds to its borrower. The Deposit Bank’s reserves remain unchanged; only the ownership of the funds changes.) The interest rate would balance the demand for loans with the supply of customer savings.

Macleod quotes Ludwig von Mises from “Senior’s Lecture on Monetary Problems (1931)” that the danger arises when the central bank intervenes (presumably via deposit insurance) to protect the banks from their own irresponsible lending and that fluctuations in bank credit are behind the boom/bust cycle. The former danger is cured by making the Deposit Bank purely a custodian of the depositors’ funds. The latter is cured by prohibiting the extension of bank credit that is not funded by savings; i.e., eliminating fractional reserve banking.

The “Two Bank” reform and Roman Law

Macleod goes to great length to explain the origin of credit in the West as emanating from Roman Law. It is a very interesting discussion and well worth studying. In Roman Law there was a distinction between ownership of property and a duty to repay it, which I understand as the difference in the status of money in the Deposit Bank (still owned by the customer) and the Loan Bank (owned by the Loan Bank but with a duty to repay it according to contract). According to Macleod, these are at the heart of the “difficulties of overturning the entrenched legal position of Roman, or natural law.” But it appears that Rothbard’s “Two Bank” reform is compatible with Roman Law. Ownership would not be transferred when a customer places funds in the Deposit Bank, but he would transfer ownership of the funds to the Loan Bank and the Loan Bank would have a duty to repay the loan. The risk of non-repayment clearly is with the latter and not the former. It is acceptance of this risk that allows the Loan Bank customer to earn interest. There is no such thing as a risk-free loan or risk-free investment.

Macleod sees one further difficulty of the “Two Bank” reform. Only a fraction of the population own gold or silver; therefore, Macleod fears that most of the citizenry would not be able to participate in the system. First of all, there is no requirement that tying the currency to gold is a necessary step in creating the “Two Bank” reform. In fact, it is more than likely that initially the central bank would still exist and money would still be fiat, with fiat currency used as reserves just as they are used as reserves now. Secondly, eventually Congress could abolish the Fed and distribute its gold reserves among the banks according to their level of demand deposits, thereby establishing a fixed relationship between the dollar and gold. At this point the dollar would simply be a shorthand for a certain amount of gold. Eventually, the term “dollar” could be dropped and the unit of currency would become a certain amount of gold. It’s been done before. This was the situation in the US from after the Civil War (1865) to the founding of the Federal Reserve System (1913). The US enjoyed one of its most expansionist periods in its history and no central bank existed at the time.

Regarding Removing Limited Liability Protection

Macleod’s proposal to eliminate limited liability protection for banks has much merit. But other businesses enjoy this protection, so why remove it only for the banking business? An alternative approach would allow banks to eschew limited liability in order to attract customers who are not willing to accept as much risk. If the Loan Banks’ owners accepted full/unlimited liability, there would be much less (but not zero) risk of non-repayment of funds. Therefore, the Loan Banks’ customers must be willing to accept a lower return. Those who want higher returns could open accounts at Loan Banks that have limited liability. The risk of non-repayment of the customers’ loans to the Loan Banks would be somewhat greater, because the banks’ owners’ personal assets could not be accessed in order to repay their customers. In either case I am certain that both Macleod and the Rothbardians would recommend that governments liquidate their so-called deposit insurance protection programs, which are not compatible with any free market approach.


In conclusion, the Rothbardian “Two Bank” reform is compatible with the free market. It would eliminate the source of the boom/bust credit cycle by eliminating the ability of banks to create money ex nihilo via lending. Bank lending (by the Loan Bank) would not affect the money supply, because only savings would be lent. Customer deposits (into the Deposit Bank) would be backed one-hundred percent by reserves, as is compatible with ordinary commercial law. The only bank examination required of the Deposit Bank would be periodic audits to ensure that the bank had sufficient reserves for its deposits. Expensive and intrusive examinations and regulations by banking authorities would be eliminated, because their reason for existence would be gone; i.e., the liquidation of deposit insurance. Much great bank expense would be reduced, and competition eventually would pass this reduced expense along to the public via more services and more interest return. In short, bankers could return to their primary task of being good bankers and not have the additional task of being good followers of ever increasing regulations.

© Patrick Barron 2022 Website