17. Full reserve banking

Though rarely discussed in modern mainstream economics textbooks, there are economists who have designed monetary systems without the ballooning and shrinking money supply characteristics of fractional reserve banking. After the Wall Street Crash of 1929 that preceded the Great Depression in the 1930s there was serious consideration of an alternative monetary system known as 100% reserve banking or full reserve banking. There are many subtle variations of this system that have been proposed over the years, perhaps most famously in The Chicago Plan and related book 100% Money created from the works of Irving Fisher, Frederick Soddy and Henry Simons. In more modern times, similar systems have been proposed by Nobel-Prize-winning economist Milton Friedman and more recently by a group called Positive Money in the UK. There also many groups around the world that advocate for some form of full reserve banking. For example The Alliance for Just Money (US), Monetive (Germany), Positiva Pengar (Sweden) and Ons Geld (Netherlands).

The crisis of 2007/8 has rekindled interest in full reserves. In 2012 a paper by Jaromir Benes and Michael Kumhof produced for the International Monetary Fund analysed the likely effects of adopting Fisher’s plan and concluded “Our analytical and simulation results fully validate Fisher’s (1936) claims.”

What is full reserve banking?

As discussed in chapter 1, full reserve banking is a monetary system that exclusively employs 'everlasting tokens' and does not allow banks to create 'spendable IOUs'. This means that the total money supply is automatically fixed, unless the government decides to print more.

A more formal definition of full reserve banking is remarkably hard to come by and uncertainties about its definition have led to much confusion and misinformation. So it is worth being a little pedantic for a moment about what it means to have full, as opposed to fractional, reserves.

The phrase 'fractional reserves' means that the banks only have a fraction of the amount of everlasting tokens that would be required should all the bank's depositors wish to withdraw their spendable IOUs simultaneously. Note that a bank customer "withdrawing" their spendable IOUs is essentially converting the numbers in their account balances into cash. This explanation of fractional reserves has led many to assume that the definition of full reserves must be based on the following:

An overly simple description of full reserve banking
A system where banks maintain the full amount of depositors' money on hand all the times, so that if all depositors demanded their money simultaneously, the banks would have fully enough to satisfy them.

Under this definition, it would be impossible for banks to ever lend out any of the depositors money. Instead they would just have to store the money in their vaults and act as a passive warehouse or piggybank. Obviously banks could not offer any interest on anyone's savings as they would have no means of making any money from those deposits. What's more, businesses would no longer be able to go to banks to borrow money. This would be a disastrous system. Having said that, it is still common to find textbooks that employ this flawed definition of full reserves and follow it up with the inevitable conclusion that full reserve banking is a bad idea.

Before we come on to a better definition of full reserves, it is worth taking a quick detour and considering how a loan would work in a world with money, but without banks. Say that Mary had £20,000 in cash (which she keeps in a safe or under her mattress). Now imagine that she has a job that pays well enough to pay her day-to-day expenses but she would like a cushion of £3,000 to cover any emergencies or special purchases that she can envisage over the coming year. This means that the remaining £17,000 is currently spare. Say that her friend Bob approaches her wanting to borrow £17,000 in order to buy equipment that will enable him to start up a new business venture. He estimates that it will take a year before the business is making enough money so that the loan can be repaid with interest. After carefully examining the plans, Mary goes ahead and makes the loan. One week later she walks past a car showroom and sees a shiny new car that takes her fancy. It costs £15,000. She thinks to herself, "Wow that's a really nice car. I wish I could buy it". Consider for a moment - would it be fair for Mary to demand that Bob give the money back immediately, just one week into his business start up? Hopefully you can appreciate that this would be completely unreasonable. Bob has already spent the money. Lending money for a fixed period in return for interest without the possibility of getting the money back until that period has expired is a completely natural scenario. In the real world with banks, the idea of putting money aside for investments while not being allowed access to it for some fixed period, it is known as a 'time deposit'. We can summarise the situation by stating that in the case of a fixed term loan, the lender is not entitled to demand their money back whilst the term is still ongoing. Keep this in mind when we consider a better definition of full reserves as follows...

A better description of full reserve banking
A system where banks plan on having enough money on hand, so that if all depositors demanded all the money they are currently entitled to, simultaneously, the banks would have fully enough to satisfy their demands.

The key difference between this definition and the previous one is the phrase “currently entitled to”. This caters for the fact that in a full reserve system, people can choose to put their spare money into 'time deposit accounts'. These accounts may be for fixed terms, or may have notice periods that must be served. In either case, depositors will not be entitled to take their money out until the notice periods are served or the terms have completed. Armed with this definition, we can now see that it is indeed possible for full reserve banks to lend out money deposited with them. Banks can lend out depositor’s money for agreed periods.

Just to be clear, full reserve banking will involve people having two types of bank account:

1. 'demand deposit' accounts where the depositor can withdraw their money at any time and where the bank acts as a warehouse. This is the type of account where Mary would keep her £3000 of emergency money, ready to be withdrawn at any instant. Banks would not pay Mary any interest on this money because it would have no way of profiting from it. Indeed it may even charge Mary for looking after it for her.

2. 'time deposit' accounts where the depositor leaves their money in for a minimum duration or with a minimum notice period, where the bank can take these funds and lend them to borrowers. This is where Mary could store her £17,000 of 'spare' money. This money could be lent out to borrowers at interest and so the bank can make money from it and could pay some of that interest to Mary. The bank would make its money on the difference between the two rates.

Will a full reserve bank always have enough reserves?

The intention of a full reserve system is that, in the unlikely event of all depositors wishing to withdraw the money (they are currently entitled to) from a particular bank simultaneously, that bank could fulfill their demands. However, it is not guaranteed. If a bank made poor lending decisions, and enough borrowers failed to repay their loans then it may fall short. There is nothing built into a full reserve system to protect against this possibility. Having said that, the likelihood of this scenario actually occurring is very small. The proportion of customers asking to remove their money would have to be very high to cause a problem - compare this to the situation with fractional reserves where only a small proportion of withdrawing customers would be needed to trigger a bank failure, and that's even if the bank had made no bad loans at all. The difference between full and fractional reserves in the likelihood of a bank being found short of money to pay its depositors is enormous.

Full reserve banking: fully defined?

So now we have a better idea of what full reserve banking is about, but this does not mean that all aspects of the monetary system and banking regulations are fully defined. There are many aspects of the system that the phrase "full reserve" does not cover and it is perfectly possible for proponents of full reserves to disagree about what additional rules and regulations their proposals include. In the following sections we shall consider some of these optional ingredients.

Maturity matching (or not)

The 'time deposit' systems described so far have the property of strict maturity matching. In other words, the time period the depositor agrees to lend to the bank, dictates the maximum duration that the bank can lend that money out. This may however be overly restrictive. There are projects for which it is necessary to borrow money over a great many years before it can be repaid and it may be hard to find depositors that are willing to have their money tied up for such long periods. It may be necessary to string together a chain of lenders A, B, C etc, each lending for some component of the total time. This does indeed carry a risk inasmuch as only the first lender in the chain will be signed up from the start. Each subsequent lender will need to be acquired around the time that the previous lender wants to withdraw their loan, i.e. when lender A demands their money back, then the bank will use lender B's money to pay them. If no lender B can be found, then the bank will be unable to repay lender A, at least at that instant.

The process of making loans where the durations agreed to by depositors are not strictly matched to the duration requested by borrowers, is known as ‘maturity transformation’. Maturity transformation is a matter of degree, i.e. the number of people in the lending chain A, B, C etc could be restricted to some maximum allowable number. A high number would be riskier, a low number safer. Sadly, some economists erroneously believe that full reserve banking forbids any degree of maturity transformation at all, and condemn it on that basis.

Government guarantees (or not)

With a system of fractional reserve banking, it is considered essential that governments guarantee people's deposits to avoid the chaos of bank runs. Under the inherently safer system of full reserve banking however, arguments for or against guarantees become less obvious. Those arguing against guarantees do so on the grounds that it avoids the possibility of "moral hazard" where banks do not have the proper incentives to be careful in their choice of loans. The argument is that if governments fully guarantee depositors money, then the banks managing those deposits will take too many risks with that money, secure in the knowledge that, should their decisions prove bad, the governments will step in and rescue them. This problem will be compounded by the fact that when potential customers are choosing which bank to trust with their savings, they won't have any incentive to investigate the soundness of the bank’s lending policies.

Those arguing in favour of bank guarantees may point out that a significant fraction, perhaps even the majority, of the general public are not equipped to distinguish a safe bank from an unsafe one. After all, banks generally are at pains to convey an image of soundness and stability in their promotional material.

A key issue is that guaranteeing deposits can be done in a wide variety of ways. Guaranteeing deposits is not necessarily synonymous with rescuing the bank. It could simply be a matter of refunding the depositors with whatever they lost, but otherwise letting the bank fail. All management would lose their jobs, all shareholders and bondholders would lose their investments, just as in any other business failure. With this kind of guarantee in place, there is far less in the way of moral hazard. The bank's owners would be highly incentivised to ensure that the bank was run safely. It could also be the case that the deposit guarantees are deliberately lower than 100%, it could be some slightly lesser percentage, or subject to some excess. All of these things would further reduce problems with moral hazard, whilst still avoiding the nightmare scenario of people losing all their savings.

Would a fixed money supply be a good thing?

In the system described so far, without any further money printing, the money supply would remain constant. This means that, assuming technological progress in manufacturing, there would be continuous deflation; i.e., a constant amount of money chasing an ever increasing supply of goods leading to ever cheaper goods. Whilst this idea superficially appears attractive we showed in chapter 5 that low or negative inflation is highly problematic and positive inflation is far better for the economy. For this reason we suggest that under a full reserve system, the government print money at a rate that will lead to positive inflation. This means that the government will be free to require less tax revenue than its spending commitments in perpetuity.

At this point some people may become concerned that this is putting the fox in charge of the chickens and that we have a recipe for hyperinflation. But the current system is little different in this regard... allowing governments to borrow enormous amounts can and does also lead to crises, just a different kind. What's more, in the next section we will present an idea which will mitigate against any government's desire to print arbitrarily large amounts of new money.

Also with full reserves, inflation is dramatically more controllable anyway.

How much new money should be created?

The rate of new money creation, or monetary policy is a highly contentious issue among economists with all sorts of schemes having been tried... what we propose is a fixed percentage growth. The arguments in support of this idea revolve around interest rates, and so before we continue, we need a quick digression on interest rates in a full reserve system.

Interest rates under full reserve banking

Unlike in the case of fractional reserve banking, when someone asks to borrow money from a full reserve bank, that bank will have to ensure that the money from depositors in the time deposit accounts is available. In order for a bank to maximise its profits, it will want to set the interest rate it charges to such a level that all, or nearly all, the money it has available for loans, is indeed lent out. If the interest rate it charged was set too high, then not all the depositor's money would get lent out. If it was set too low then there would be unsatisfied potential borrowers unable to get loans. Some say that under a system of full reserve banking, banks will automatically set a 'natural rate' of interest, balancing the supply and demand for loans. If it were the case that a continuous stream of new money were created, then banks would forever have more money to lend out than would be the case without the money creation and so this interest rate would be influenced by the rate of new money creation. The higher the creation rate, the lower the interest rate. The other thing that affects the interest rate is of course the enthusiasm for taking out loans. So for any given rate of printing new money, we would expect that the greater the enthusiasm for loans, the higher the interest rate the banks will set to balance supply and demand.

So now we can say that with a fixed rate of new money creation, interest rates will be a function of the enthusiasm for borrowing. If the enthusiasm rises, interest rates will rise. If enthusiasm falls, interest rates will fall. At this point some economists get nervous about the idea of interest rates rising or falling in a way which is not under the direct control of the government and or central bank. They argue that the central bank protects the market from wildly fluctuating rates, though one only need examine the past few decades of UK interest rates (see below) to see pretty wild fluctuations even when the central bank was in charge. Indeed the ratio of the highest to lowest rate is 170-fold (highest = 17% lowest = 0.1%). This suggests that the current system can scarcely be said to provide stable rates at all. So any claim that central-bank-controlled interest is more stable than what would come about through the free market should be taken with a pinch of salt.

The Bank of England policy rate 1946-2024

Full reserve banking - Has many opponents

After centuries of fractional reserve banking, the idea of changing to full reserve feels so radical that surely it must be a crack-pot idea. It seems that many economists or politicians, when presented with the idea of full reserves feel that they should be able to brush it off with scarcely a thought, perhaps a one sentence quip. Surely it can not be worth considering seriously can it? Look at this list of seemingly damning criticisms that are frequently put forward:

• The money supply will immediately implode

• Banks won’t be able to lend money

• You can’t prevent people from using IOUs as money

• Governments will be unable to determine the right quantity of money

• We tried monetarism and it didn’t work

Before we continue it is probably worth considering for a moment some extremely eminent economists that have advocated for full reserves including Irving Fisher, Milton Friedman, Ludwig von Mises, Murray Rothbard, James Tobin, Laurence J. Kotlikoff, Mervyn King, John Kay, Jesús Huerta de Soto, Michael Kumhof. You can be certain that all of these economists will have encountered most of these potential criticisms and were still undeterred. So now let us tackle these criticisms head on:

1. The money supply will immediately implode

If the transition to full reserve banking was simply done by banning banks from creating new spendable IOUs and only allowing them to lend notes and coins then as pre-existing loans (from before the new law) were repaid, the money supply would indeed start to shrink dramatically. This is well understood by all economists that have proposed full reserves and so all the proposals have included transition mechanisms where new everlasting tokens would be created as spendable IOUs expire, thereby preventing the money supply from falling at all.

2. Banks won’t be able to lend money

This criticism comes in two flavours.

1. The critics claim that banks can’t lend because the banks have to keep all depositors money in hand ready in case the depositors wish to make a withdrawal. The refutation of this criticism has already been discussed.

2. A more subtle variation of this critique came from Paul Krugman in an article, published in the New York Times, entitled “Is A Banking Ban The Answer?” where he relies on the financial-regulator’s definition of a bank as an institution that creates spendable IOUs. Under that definition, an institution that only lends out pre-existing everlasting tokens in the form of ‘time deposits’ is not allowed to call itself a bank. Of course under a full reserve monetary system the legal definition of a bank would change.

3. You can’t prevent people from using IOUs as money

Critics have latched on to the notion that full reserve banking is about preventing-IOUs-being-used-as-money. They have then noted that it is nearly impossible to prevent two people making an exchange of goods in return for some kind of IOU. For example, imagine person A gave a gold ring to a pawnbroker and in return received a ticket saying “the bearer of this ticket is entitled to a specific gold ring”. That ticket is an IOU from the pawnbroker. Now imagine that person A wanted to get their hands on a bicycle owned by their neighbour… if he knew his neighbour well enough, and the neighbour knew about the ring, he may be able to exchange his pawnbroker ticket for the bicycle. So here we have an IOU acting as money and the legal system would have no reasonable way of preventing it. Note though that it was critical that the receiver of the IOU had to know about and be confident of both the issuer of the IOU and the thing that the IOU is promising to deliver. These two burdens placed upon the potential receiver of the IOU mean that it is less likely that this IOU will ever become generally accepted. As discussed in chapter 1, the degree to which something can be considered money is dependent on the degree to which it is generally accepted. So pawnshop receipts are incredibly poor quality money that will be accepted almost nowhere.

At this point full-reserve critics may step in and say that the pawnbroker example involved a very quirky case where the thing that could be redeemed required very specific knowledge. What about much better known things like government bonds. Government bonds are an IOU from the government, and indeed many more people know about and have confidence in these particular IOUs. But even so, try buying some groceries with a government bond certificate. Government bonds could indeed in theory be used in exchange but again they are poor quality money.

Note that even today with our current monetary system there are multiple things that can be and are used in exchanges, take bitcoin as another example. All these things are possible to use but none of them are popular, so none of them would be a significant barrier to the introduction of full reserves.

4. Governments will be unable to determine the right quantity of money

Under a full reserve system, the amount of money in existence will be under the direct control of the government. If they want to have some small positive inflation then they will want to continuously create more of it, but exactly how much? Some proponents of full reserves have suggested government committees having regular meetings to decide the size of each new injection of money (much like central banks have regular meetings to set interest rates in our current system). Critics argue that governments are likely to get this wrong. On this point I am included to agree - but are they likely to get it any more wrong that the central banks setting interest rates? As suggested earlier in this chapter, I advocate for a fixed percentage rate of growth which does not entail any regular meetings anyway.

4. We tried monetarism and it didn’t work

Some people equate full reserve banking with monetarism in as much they both reflect the idea of paying attention to and controlling the total quantity of money. These people will point to that period in the early 80s where Milton Friedman’s brand of monetary policy was beloved by Margaret Thatcher and Ronald Reagan. This period was marked by high unemployment and recessions. During this period there was no move to full reserve banking and the only mechanism that banks had to influence the money supply was by raising or lowering interest rates. There was no effort made to distinguish between bank lending for productive vs non-productive purposes (as advocated in this book). This meant that any time there was any kind of bubble in an asset class purchased with borrowed money (as discussed many times in earlier chapters), that bubble could only be extinguished by discouraging both productive and non-productive lending at the same time. A guaranteed disaster. There is more than one way to control the money supply. Just because one bad mechanism failed, it does not mean that all mechanisms will fail.

Did you like contents of this chapter? If not click here.