The financial system is a very complicated ensemble of regulation, banking practices, and inter-dependent agents striving to achieve their own goals. Presently, the situation in money markets is notably distinct from that in other markets as the existence of a central bank on the one hand does smooth out market shocks, but on the other creates a bright-as-day monopoly on the supply of money. In this post we explore an alternative - that of free banking - the state of affairs where banks are unregulated, utterly free to enter and exit the market, and have to comply only with the capricious demands of clients, not regulators. Nowadays, free banking is largely ignored due to how conspicuously different it is compared to the central bank model. Yet, there are precious insights to be learned from it.

I recently watched a nice documentary by the Cobden centre about how banks create money ex nihilo, out of thin air. I decided to read up on this issue and quickly found myself referring to the good-old Austrian school. This post will cover the main aspects of free banking - what does it mean to be free, what mechanisms constrain banks, how does inflation arise in these settings. We will not cover central banking which is a completely different beast altogether. I'll be referring extensively to Rothbard's The Mystery of Banking and to some extent Huerta de Soto's Money, Bank Credit, and Economic Cycles.

To start, consider barter. In barter, producers directly exchange the surpluses of their goods for the surpluses of other goods made by others. The price of one good is always relative to another and is given simply by the ratio of the goods exchanged. Markets consist of the pairings of people willing to exchange objects of interest. While simple in principle, a barter economy suffers from the problems of:

  1. Double coincidence of wants - it is difficult to find one who desires exactly what you offer,
  2. Indivisiblity of goods - some goods are indivisible and hence more difficult to exchange,
  3. Lack of business calculation - it is difficult to calculate profit and loss in the context of exchanging heterogeneous real goods.

To address these issues one can adopt a particular commodity as a medium of exchange. People will then seek it in order to later exchange it for those things which are of direct interest. Thus, this commodity becomes an intermediate good facilitating the exchange of other goods. When such a commodity becomes used by virtually every market participant, it is called a money. People now exchange their own products first for money, and then they exchange that money for the other products which they desire.

Some commodities are more likely to become a money than others. Typical conditions which need to be satisfied are that the commodity should be heavy in demand, highly divisible, durable, and should have high value per unit weight. To be heavy in demand implies that it should be scarce, with limited supply. Historically, different commodities like salt, sugat, cattle, tea, tobacco have been used as money. The most dominant however have been gold and silver.

Additionally, one can define different units of weight of that money. A pound sterling is exactly one pound of silver. A dollar is one ounce of silver. Thus, since they have been defined in this way, the exchange rate between pounds and dollars is fixed. When different countries define different units of weight of a money, their exchange rates are automatically fixed.

However, using the names of these units rather than their equivalent weight provides opportunities for debasement. Consider the following example. A monarch collects all circulating coins in order to stamp his face on them, and subsequently give them back to their owners. However, during the recoinage, the monarch redefines for example the pound to be only half a pound of silver. He then mints twice the number of coins, gives half of them to their owners and keeps the other half for himself. The total grams of silver has remained the same, but the money supply in terms of nominal units has increased. The purchasing power of each unit will go down.

The price of each good is determined, naturally, by supply and demand. If the price drops, consumers are willing to buy more while producers will supply less. If the money supply is fixed and the total amount of money stays constant, an increase in the demand for one good will necessarily be counterbalanced by a decrease in the demand of another one. The only way in which the aggregate demand for all goods can increase simultaneously is if the incomes of all consumers increase, which can happen only if the money supply increases.

Thus, sustained inflation, or the rise in overall prices, can result either from a general decrease in the supply of all goods and services, or a general increase in demand. Outside of major crises, historically supply has been increasing which rules out this option. Hence, it is the increase in demand, caused by an increase in the money supply, that causes chronic inflation.

If prices are high, the same quantity of money can be exchanged for fewer goods. Based on this, one can intuitively define the purchasing power of money (PPM) as the inverse of the price level. We now consider the market for money itself. The demand for money is inversely-proportional to the PPM, because if the PPM is higher, you need fewer money to afford the same goods. The supply of money is constant. The equilibrium price level and therefore the equilibrium PPM are precisely those at which the total money demanded is equal to the total money supplied. Then, an increase in the money supply lowers the PPM and raises prices. An increase in the money demand raises PPM and lowers prices.

Since increasing the money supply increases the overall price level, it doesn't really matter what the money supply is. However, the act of increasing it does create winners and losers. The new money has to be injected into the economy somehow. Those that receive it and spend it first, before prices have risen, will benefit at the expense of those that spend it last. Those that are not reached by the ripple effects will be hurt the most. Counterfeiting, the process of increasing the money supply by creating artificial fake money, has a similar effect.

Consider what happens when government paper money is introduced. Historically, since kings have had a monopoly in coin mintage for decades and since people have gotten used to the name of the currency unit, rather than its "weight", governments could successfully, albeit with a lot of convincing, introduce paper notes and call them "dollars". Initially, the government has to convince people that the new paper "dollar" is backed up by gold or silver but once this is done, it can starting printing (effectively counterfeiting) mammoth quantities of money to finance its spending.

Keeping the paper dollar redeemable to some extent constrains the government in printing too much money. After the public has been habituated with the paper dollar, the government may make the dollar irredeemable, severing the link with the gold. Through long campaigns the public will then have to be convinced that this is actually a good thing. The end result though is an institution that can finance its spending by easily printing huge amounts of money. Likewise, the value of each dollar becomes dependent on the "human factor" of the government's reputation.

Now, let's shift our focus on banking. In general, banks borrow money at a low interest rate and lend out money at a higher interest rate. We can distinguish between two fundamental banking contracts.

  1. In a monetary loan contract, full availability over the loaned money is transferred from the lender to the borrower. The contract requires the establishment of a term for the return of the loan and calculation and payment of interest, since present goods are exchanged for future goods. The borrower's obligation is to return the monetary loan at the end.
  2. In a monetary irregular deposit contract there is complete, continuous availability in favour of the depositor. There is no interest, since present goods are not exchanged for future ones. The depositor's main motivation is the on-demand custody of the tantundem (deposited goods) until desired by the depositor.

Loan banking is a healthy endeavour. In the classic setting, the bank will loan money originating from savings. It may as well issue bonds and loan the received cash. None of these cases represent an increase in the money supply because both the bonds and the initial capital come from savings. When the borrower returns the loan along with interest this is also not an increase in money supply, because the payment comes from the profit of the borrower. Naturally, if the bank makes risky loans it may go bankrupt in which case the equity- and debtholders will be in trouble.

Deposit banking refers to safeguarding a deposited good and keeping it available on demand. It actually originates from e.g. wheat or oil warehouses where one can deposit a given quantity to keep and pay a small amount proportional to the storage time. In the context of a bank, it would act as a gold coin/bullion/grain warehouse. Note that the depositary has to safeguard the tantundem at all times, because the depositor might request it at any time. Using the tantundem for personal gains, such as loaning it, constitutes embezzlement.

Historically, deposit banks have maintained good reputation and this allowed the receipts for the gold they stored to become a surrogate for the gold coins themselves. Let's call these warehouse certificates. Technologically they are of two types, bank notes and demand deposits, but they serve the same purpose. With them, the total money supply does not change. Only its form changes. The stock of deposits in principle should not be considered as debt in the balance sheet of the deposit bank. However, suppose it is listed there. It then takes a small step for the deposit banker to counterfeit new warehouse certificates and lend them out without having the actual gold deposits to back them up.

This situation is very dangerous. Suppose the deposit bank safeguards real gold worth \$1000. Hence, it has issued a warehouse receipt for \$1000. However, if it counterfeits another receipt worth \$1000 and lends it out, the actual money supply will increase out of thin air, but the reserves in the bank to back up both warehouse receipts worth \$2000 will be only \$1000. This is called fractional reserve banking since the bank has less than 100% of its warehouse receipts backed up by cash. It is inflationary.

Thus, commercial banks create money out of thin air by loaning money which is not backed up by an equivalent amount in reserves. In the process of loaning, they create a new deposit account, assign it to the borrower, and the borrower can do whatever he wants with this virtual money, as long as he is to repay it at the end of the term.

Now, finally we consider the situation of free banking. In free banking, banks are treated like any other businesses. There are no government regulations and therefore banks can do whatever they want. They may engage in fractional reserve banking or any other similar mechanism. However, there are constraints imposed that will punish those banks which act in a way deemed negative by the public.

One of these constraints is reputation when setting up the bank. It is difficult for an unknown bank that has just set up office to start handling out massive amounts of loans because the public is weary and has no reason to trust the new bank. Similarly, since there are no regulations, it is likely that the barriers to entry are minimal and therefore the public have seen multiple banks which turn out to be fraudulent.

One of the biggest reality-checks on banks is simply the confidence in whether the money they safeguard is redeemable. A loss of confidence can have disastrous avalanche-like feedback effects, creating bank runs and bankrupting multiple banks in a short period of time. In the process, of course, multiple clients are hurt, but overall this makes them more cautious and the banks more careful.

A second constraint results from the competition between banks and the limited clientele of each one. Suppose bank A gives an unbacked loan to person B, who in turn spends it on something, paying person C. Now, it matters whether C is a member of bank A or not. If he is, then bank A can simply re-shuffle the demand deposit of person A in its own books to that of person C, in which case the inflationary loaning can continue, as long as confidence is maintained. However, if C is not a client of bank A, then the bank of C, suppose it is bank D, will demand payment in actual money from bank A. In that case, immediately once it is found out that A cannot redeem the loan, it will go bankrupt. It is natural to assume that bank D will demand money (as opposed to an IOU) from bank A because they are competitors. Thus, the more banks there are, the smaller the share of clients of each one, and the stronger is the incentive to keep a 100% reserve ratio.

If one bank inflates too much, the prices within the "influence" of its clients will increase. This will make the prices from competing banks more appealing, causing other banks to demand redemption as the current clients start contracting with them. In order to stay solvent, the initial bank will have to contract its inflationary policy. The mechanism on the level of countries is similar. If one country inflates, then its price level will become higher and the prices from foreign countries will become more appealing. Gold will start flowing out of the inflating country, forcing a contraction.

It is near impossible for prolonged cartels to exist as well. If a cartel accepting IOUs without redeeming the loans forms, it will become more and more profitable for each one of the participating banks to deviate from the dangers of keeping a low reserve ratio and becoming insolvent. Thus, in a free banking setup, hard money with near 100% reserve ratio will be the norm, with virtually no chronic inflation and the boom-bust cycles that it entails.