We’ve spent two weeks looking at fractional reserve banking, and I pointed out that although this is the model presented by textbooks, it bears very little resemblance to the real world. I highlighted a couple of reasons why 2 weeks ago, and this week we’re going to improve this model with one I’m calling Bankmoneyland (it’s not catchy, but it gets the point across).
Bankmoneyland is like Multilplyland, but here the banks have invented cheques. This means that rather than having to go to the bank and withdraw cash to spend it, the citizens of Bankmoneyland can simply carry around a chequebook and write cheques. Cash is no longer the only means of payment: the mere accounting record that you have money in your bank account has become a form of money that you can spend by writing a cheque. A cheque is almost a form of banknote (hence the name of this model).
The recipient of the cheque pays it into the bank, and at the end of the day each bank has a large stock of cheques that it needs to clear. Each bank will have thousands of cheques written by customers of each other bank. Say Bank A has £1m worth of Bank B cheques that have been received by its customers, and Bank B has £1.1m of Bank A cheques. To clear the balance, Bank A simply has to pay £0.1m to Bank B. £2.1m of transactions has been cleared by a transfer of just £0.1m in cash between two banks.
In the days before electronic banking, banks had got this down to a fine art. For example, the New York clearing house in the late 19th and early 20th century would deal with cheques from the major American banks. Not only did they net out the transactions between individual banks, they netted out all the transactions between all banks, so that each bank ended up with just one single payment to receive or make that represented their balance from all the cheques presented to the Clearing House that day. It’s incredible to think – thousands upon thousands of cheques, adding up to millions or even billions of dollars, could be cleared at the end of the day by a single payment between each bank and the clearing house, for a small fraction of the total value of all the cheques.
The economist Charles Dunbar eulogised about this system in his 1922 book “The Theory and History of Money and Banking”:
“This medium of payment acquires great perfection wherever the Clearing House system is adopted… The bank deposit, circulated by means of checks, is the most convenient medium of payment yet devised. A stroke of the pen transfers it in whatever amount is needed for the largest transaction, and this transfer instantly becomes the basis for fresh operations…”
So think back to the example in Multiplyland of our saver, Bob, paying in £100. The bank kept £10 in reserve and loaned out £90, which was spent, banked, and then the next bank kept £9 and loaned out £81… It took time for the £100 of savings to multiply up to £1,000. What happens in Bankmoneyland is that Bob banks his £100, the bank puts all the cash in its reserves and loans out £1,000 by writing a cheque – the multiplication is instant. Secondly, because people use cheques far more than they use cash, there is far less demand for cash, so banks can hold far less cash without worrying about not having enough to meet the demand for cash withdrawals.
Big deal, you think, so what if the reserve requirement falls from 10% to 5%? A drop in the reserve ratio from 10% to 5% doesn’t represent a 5% increase in the money supply, it represents a doubling of the money supply. Suppose the reserves stand at £100bn. A reserve ratio of 10% then enables a money supply of £1,000bn, a 5% reserve ratio enables a money supply of £2,000bn – double the amount. In 2012 the Eurozone reduced the reserve requirement from 2% to 1%. It doesn’t sound much, but it represented the potential to double the money supply.
Bankmoneyland illustrates how the clearing process enables a huge value of transactions in the economy to be cleared by a much smaller transfer of reserves between banks, and that this in turn enables the banks to hold tiny levels of reserves. As the reserve ratio gets smaller, the capacity to increase the money supply increases exponentially. For practical purposes, the banks become unlimited in their ability to expand the money supply. (Reserve requirements are not, in reality, a particularly significant regulatory requirement on banks, and hence the UK does not even have one.) Banks control the money supply, not the government.
Bankmoneyland highlights another issue. Textbooks focus on saving deposits being leant out. But in fact, even in a cash-based economy there is no reason for the bank to distinguish between saving deposits and current account deposits. It’s not like the banks have a little safe with your name on it and all your money stored! All the cash received by the bank is held in one place. The balance between saving and current accounts will affect the demand for cash, and banks still offer “time deposit” saving accounts that pay a higher rate of interest but place limitations on your ability to withdraw (such as being required to give notice, or paying a penalty for early withdrawal). These are typically used by savers with significant currency, and do enable banks to manage their liquidity (in the modern economy this has nothing to do with cash, which is irrelevant). But nonetheless, the fundamental issue is that even in the theoretical fractional reserve system, all types of deposits represent money that can be loaned. Yes, if a bank has a higher proportion of saving to current accounts it will feel more confident to lend more and keep a smaller fraction in reserve, but even a bank with only current accounts and no saving accounts could still use its reserves to make loans.
So even if we had a system of fractional reserve banking, customers would not need to be saving for banks to be able to make loans – the connection between the two presented in textbooks is conceptually misleading. And let’s not forget that we don’t have a system of fractional reserve banking. Even the version presented as Bankmoneyland is still way off what happens in reality (we’ll get to that in 3 weeks – I’m looking forward to it).
In the real world today, electronic transactions are usually cleared between by banks in real time, but it is only at the end of the day that banks have to balance their books, and if they are in debit they have a simple way to cover this (which we will look at in a few weeks when we come to understand what LIBOR is and why the LIBOR scandal mattered so much). So the huge value of transactions in the economy every day are still cleared by much smaller net payments between the banks, as shown in the Bankmoneyland model. This will be helpful to remember when we start looking at the real-world banking system.