This archived content originally appeared at, the predecessor site to, and does not carry the sponsorship of the Cato Institute.

What modern free banks might look like

It has not been discussed so far in this blog what modern free banks might look like. Partly that is because several of us who blog here have the same picture in mind. It would be good for the rest of you to know what it is and decide what you think of it.

What activities would banks be allowed to engage in? Anything they wanted. In particular, they would be allowed to combine banking with commerce, so if Wal-Mart or Facebook wanted to start a bank or if a bank wanted by them, it could. If banks wanted to combine banking with brokerage or insurance, they could. In practice, some combinations of commercial banking with other lines of businesses would be common and others would not. I would not expect any bank-restaurant combinations, for example, because the businesses are so dissimilar. Even much closer combinations such as bank-plus-hedge fund might be rare because the businesses are not as alike as many people think.

How would banks be regulated? There would be no special regulations on banking, and in particular no minimum capital requirements or requirements that assets be invested in specific ways unless the bank itself promised to do so. There would continue to be some special features of law that apply to banking, just as there are some special features that that apply to other industries (fishing, trucking, warehousing, retailing, etc.) because of industry-specific features. In banking, two such features might be requirements to ensure that financial statements state certain assets and liabilities in detail to give an accurate picture of the business, and a bankruptcy regime that, one would hope, improves on the cumbersome procedures currently in place in most countries.

Advocates of free banking see no need for special regulations because they think bankruptcy would provide the ultimate check on imprudent behavior, as it does in other industries. Whether that would be the case in reality, or whether on the contrary governments would still treat some financial institutions as too big to fail, is a hugely important question that is a subject for another time.

What liability arrangements would banks have? Banks would have the choice of operating under limited liability of stockholders, unlimited liability, or combination arrangements where some shareholders would have limited liability and others would have unlimited liability. As long as the liability arrangement was clear to people who did business with the banks, there would be no special need to force banks to follow a uniform arrangement. Today, almost all banks have limited liability. Until the mid 1800s, limited liability was a privilege granted only to select companies, often in return for bribes or other favors. Then the principle became established that limited liability should be a matter of choice for stockholders. Partnerships with unlimited liability do persist, though, such as the venerable English bank Coutts and Company and many hedge funds.

What would be the monetary base? One possibility is the current fiat monetary base, frozen, as Milton Friedman proposed in an essay in a 1984 book called To Promote Prosperity. Under this proposal, banks would be free to issue notes, so over time, Federal Reserve notes would likely go out of circulation as currency and be used mainly by banks as reserves.

Another possibility is a return to gold. The gold standard has received much scorn from economists who don’t understand that the gold standard under free banking works differently from the gold standard under central banking.

Other possibilities seem far less likely. Banks could adopt a standard based on some unit not currently in use, such as the economic value of a frequent flyer mile or a British thermal unit or a basket of goods. Or they could issue currencies not tied in any set way to a good, as central banks to now and as Friedrich Hayek imagined in his pamphlet Denationalisation of Money. In a system of Hayekian banks there would be nothing corresponding to the monetary base as it now exists. No such arrangement has ever existed in any historical free banking system.

These are my educated guesses based on historical experience. Advocates of free banking would be content to allow whatever arrangements emerge through competition, and not to push the monetary system towards in one direction or other through restrictions. If people wanted to use gold, or frequent flyer miles, or adopt a Hayekian system of competing fiat currencies, it would vain for economists to protest that they were wrong to do so. It would be like saying that people are wrong to want to speak English instead of the supposedly more “rational” Esperanto (or even the truly more rational Interlingua [see page 307 of this for a speech in Interlingua by Leland Yeager, who has also written on laissez faire banking]).

Would there be fractional reserve banks or 100% reserve banks? In principle, both side by side. In practice, I know of no past banking system where the two have long coexisted. Fractional reserve banking has always outcompeted 100% reserve banking. By 100% reserve banking which I mean an arrangement where the bank holds assets for clients as a kind of warehouse and does not grant credit.

What about coins? Like the business of issuing notes, the business of issuing coins would be open to all comers. Banks might issue competing brands of coins; they might issue a common coinage through a cooperative arrangement; or they might leave coinage to other issuers. Coins might have substantial value as metal, or, as I think more likely, they might be mere tokens. Historically, where people trusted in the financial system, coins having substantial value as metal tended to be reduced to minor importance as currency over time because it is inconvenient to carry a lot of metal around in purses and cash registers. The tokens might not be redeemable in notes and deposits, or, as I think more likely, they might be. Again, historically tokens that were redeemable typically found greater acceptance than those that were not.