In their 2014 quarterly bulletin: “Money creation in the modern economy” The Bank of England sought to clarify a common misconception about how money is created in the modern economy. The misconception is as follows: in order to create money, banks depend on savers to deposit their cash into their bank accounts first. These deposits then enable the banks to make further loans into the economy, and after these loans are made, money supply is increased. In other words, it would not be possible to increase the money supply without the initial deposits received from savers.
Let’s first Define Money
Two definitions of money will be used in this article. It is important to understand the distinction between each type of money:
Base money refers to notes and coin, and bank reserves held by central banks on behalf of banks in the economy;
Broad money refers to all forms of money actually circulating in an economy (including notes and coin, and assets which can easily be converted into cash to purchase goods and services). Broad money affects the general population of a country directly. An increase in broad money for example would lead to an increase in consumer inflation, all things remaining equal.
The distinction between the two types of money is illustrated in the diagram below:
Understanding the Basics: Fractional Reserve Lending In many jurisdictions, when banks receive deposits from their account holders, they would legally be obliged to keep a fraction of those deposits in cash (or bank reserves with the central bank), and would only then be permitted to lend-out the remainder of those deposits into the general economy. The amount that banks are obliged to hold in cash or reserves is referred to as the reserve requirement. For example, a bank with a 10% reserve requirement would legally be required to keep $10 in cash or reserves for every $100 in deposits they would receive. Hence banks are not required to hold the full value of depositors’ money in the form of cash at all times. This is referred to as the system of fractional reserve lending. The reserve requirement would ensure that banks would probabilistically have enough cash on hand to meet the demand for cash withdrawals from their deposit holders at all times, preventing loss of confidence in the banking system.
The Multiplier Effect, and the Wrong Way of Thinking: Let’s take the 10% reserve requirement mentioned earlier: If an individual deposited $1m into a bank account, the bank would be required to keep $100,000 of that deposit in cash or reserves, and would then be free to create new loans of $900,000 from that deposit, to loan out to other individuals in the economy. These new loans would most likely be deposited into other accounts held with different banks. These other banks would in-turn loan out 90% of the deposits they received, thus repeating the process over and over again. Eventually, the total amount of dollars created from the original $1m deposit would be $10m in total, circulating in the economy (using some arithmetic). Although the multiplier effect and fractional reserve lending are useful concepts to learn, they don’t reflect how money supply is expanded in reality.
The Real Way in which Money is Created: An alternative, more correct way to think about the multiplier effect, would be to consider an example of a bank with $1m in excess reserves: This bank would be in a position to create up to $10m of additional loans in the economy at outset (since doing so would mean they still have sufficient reserves to maintain a 10% reserve requirement). These additional loans would be deposited into accounts held with other banks, who could then on-lend a fraction of these additional deposits created by the original bank, or even lend out more money if possible, thereby creating a multiplier effect in money supply. The key subtlety here is that the original loans created did not necessarily depend on an initial depositor depositing new savings with the original bank. The excess reserves of $1m could have been created by the original bank raising equity or debt in the markets, or by the bank selling government treasuries to the central bank or selling assets in the repo market. In countries with no reserve requirement (e.g. USA, Canada, Sweden etc.), the situation would be even more flexible for the bank, in that they would not have even needed to raise the $1m in reserves in the first place. From the outset the banks would be able to just create new loans, and hence new deposits and new money. The only constraint to them doing so would be the risk vs return propositions of those new loans from the bank’s profitability and solvency standpoints. To illustrate money creation further, consider the example of a bank that grants a home loan to a home buyer:
Step 1 A home buyer’s bank would approve a home loan application from the home buyer. Once this occurs the home buyer’s bank would credit the home buyer’s account with the value of the home loan (it may work differently in practice, with the home buyer’s bank directly crediting the home seller’s account, but for the sake of simplicity we will assume here that the bank provides the loan to the home buyer, who then in turn purchases the home from the seller, with the proceeds). The home loan is an additional deposit in the bank account of the home buyer. Once this new deposit is created, new money (broad money) has effectively been created in the financial system. This new deposit represents a liability on the home buyer’s bank’s balance sheet. This is because the deposit holder (the home buyer) has a claim against the bank, i.e. they can withdraw the balance of their account in cash, and this represents a claim against the bank, who would then need to honour this withdrawal request. The loan itself is an asset to the bank as it represents the bank’s claim for repayment of capital and interest from the home buyer. Thus from the perspective of the home buyer’s bank, an asset has been created (the home loan), as well as a corresponding liability (the deposit in the home buyer’s account).
Step 2 The home buyer uses the loan proceeds to buy the home. They transfer the cash in their bank account to the bank account of the seller. If the buyer and the seller bank at the same bank, then the buyer’s bank’s balance sheet would remain unchanged after step 1. The only thing that would happen would be that the bank would just reallocate the deposit from the buyer’s account to the seller’s account. If however, the seller banked with a different bank, then the home buyer’s bank would move the deposit from the buyer’s account to the seller’s account at the new bank. They would simultaneously also need to transfer the bank reserves corresponding to the deposit to the new bank. For example, at a 10% reserve ratio, if the home price was $1m, and thus $1m was transferred from the buyer’s bank to the seller’s bank, the buyer’s bank would transfer $1m in deposits from the buyer’s bank account to the seller’s bank account at the new bank. The buyer’s bank would simultaneously transfer $100,000 in cash to the seller’s bank, or transfer reserves from their reserve account at the Fed to the seller’s bank’s account at the Fed. From the home buyer’s standpoint, initially their balance sheet showed an increase in assets (as their account was credited with a cash deposit equal to the value of the home loan), and also an increase in liabilities (equal to the value of the home loan itself, which would need to be repaid eventually). Once the home buyer paid the seller for the new house, they would move this deposit to the seller’s bank account. The buyer’s balance sheet would then have the house itself as a new asset, and a corresponding liability equal to the value of the home loan.
Step 3: The home seller would receive the value of the house deposited into their bank account. This would increase their assets in the form of a cash deposit. Subsequently they would transfer the house to the buyer. They have effectively swopped one asset (a house) for another (a cash deposit). In aggregate, the broad money supply in the economy increased by the value of the new home loan. As the home loan is repaid, the broad money supply in the economy would gradually be reduced. The base money supply would have stayed the same as no new reserves would have been created, only transferred from the buyer’s bank to the seller’s bank.
What is Quantitative Easing?
Through quantitative easing (QE) central banks aimed to buy mainly government bonds from non-bank institutions such as pension funds for example. It was hoped that doing so would achieve two main goals:
Goal 1: Reduce interest rates in the economy, thereby encouraging individuals and businesses to increase their demand for debt. This was hoped to prop-up real estate prices, and stimulate further spending for durable goods in the economy. The Fed in the USA particularly focused on buying government treasuries at the 10-year maturity mark. 10-year US treasuries are often used as a proxy for mortgage rates, and as an important benchmark for the pricing of other financial instruments in the economy.
Goal 2: Buying government treasuries from pension funds and other non-banking institutions would allow these institutions to reallocate their portfolio holdings, and buy other assets in the economy, such as corporate bonds and the shares of other companies. Theoretically, such a move would raise the corporate bond prices (lowering their yields) and also cause a general appreciation in the share prices of corporate entities in the economy (making capital raising via equity issuances easier). The combined net effect would thus be a decrease in the cost of capital in the economy (reduced debt costs and reduced equity costs). Lower cost of capital was hoped to encourage economic growth via encouraging firms to borrow and spend more on capital investment, further stimulating economic recovery.
How does QE Work?
Under the QE program, the US-Fed would buy government treasuries from banking and non-banking companies on a monthly basis. The structure of the transactions would differ depending on whether QE was done with banking or non-banking companies.
QE with Banks:
If the Fed wished to buy treasuries from banks (more specifically: primary-dealer banks), it would do so directly with them. Primary dealer banks are a collection of banks that have been selected to be the market makers for the Fed when it buys or sells government treasuries (open market operations). These banks provide a seller for the Fed when the Fed wishes to buy treasuries, and a buyer when the Fed wishes to sell the treasuries it holds. The primary dealer banks hold bank accounts at the Fed, where their bank reserves are deposited. Recall, these bank reserves together with all notes and coin in circulation are referred to as base money. Also recall that broad money refers to all forms of money circulating in an economy (notes and coin, as well as other assets which can easily be converted into cash to purchase goods and services). It is important to note that bank reserves held at the Fed cannot be spent in the economy, only transferred between organizations that also have bank accounts with the Fed, namely: the government and other primary dealer banks.
When the Fed conducts QE with primary dealer banks it purchases US-treasuries from them, and in return it credits their bank accounts held at the Fed with additional bank reserves. Since these reserves cannot be spent in the economy, no additional broad money is created in the economy when QE is performed with banks.
It could be argued that by increasing their reserves, banks would be further incentivised to grant more loans in the economy, thus increasing broad money supply, and stimulating economic activity. However, in many financial markets the constraint on additional bank lending has not been a shortage of bank reserves (particularly in markets like the USA, which have a 0% reserve requirement). So, it can be concluded that QE with primary dealer banks would have a limited effect on broad money supply.
QE with Non-Banks
If the Fed wished to perform QE with non-banking firms it would need to do so by using its primary dealers as intermediaries. The non-banking firms would sell their government treasuries to the primary dealer banks, who would simultaneously sell them to the Fed. The primary dealer banks would then receive additional bank reserves from the Fed as payment (deposited into their Fed bank accounts). Finally, the primary dealer banks would pay the non-banking firms by crediting their bank accounts with additional cash deposits.
The bank reserves received by the primary dealers cannot be spent in the economy, hence these do not lead to an increase in broad money. However, the payment made by the primary dealers to the non-banking firms would be made in cash deposits into their bank accounts. These cash deposits can be spent in the economy. Hence, when QE is performed with non-banking firms, additional broad money would be created.
After the non-banking firms sold their government treasuries to the Fed, they would need to rebalance their investment holdings/portfolios. It is hoped that they would replace the government treasuries sold with other assets in the economy (shares of other companies, real estate, or other corporate bonds etc). As mentioned earlier, by purchasing shares, the values of equities in the economy would be expected to rise. This would lower the cost of companies raising funds in the equity markets, as they would need to issue fewer shares to raise a desired amount of funding than before QE. If non-banking firms bought corporate bonds, their prices would increase, thereby driving down their yields. Hence overall, QE would reduce the cost of capital in the economy. A reduction in cost of capital could be thought to increase the desire of firms to engage in capital expansion activities, thereby stimulating economic growth.
Changing the Shape of the Yield Curve When QE is performed, the main government treasuries bought would be the 10-year tenor notes. These notes are used as important benchmarks in the pricing of other rates in the economy such as mortgage rates. Thus, by focusing on these treasuries, the Fed is able to influence many other long-term rates in the economy. This has an unintended consequence however: QE is undertaken when short-term rates are already low, but economic activity is still weak despite this. Naturally, the yield curve (term structure of interest rates) should be upward sloping, as the longer the tenor of debt, the more risky it is, and hence the higher compensation required from lenders. By reducing rates at the 10-year mark, the Fed may cause a situation where short term rates reach 0% or even become negative. This is a situation to be avoided, as it would further distort the demand and supply dynamics of capital (negative nominal interest rates would mean providers of capital would be making a guaranteed nominal losses for providing capital, and users of capital would be rewarded for borrowing capital irrespective of whether or not they had a plan to generate a return on that capital). To prevent short-term interest rates from becoming negative, the Fed would need to create an interest rate floor in the market. They would do so by offering to pay banks interest on their reserve balances held at the Fed. This interest is referred to as Interest on Reserve Balances (IORB), and would be a rate paid daily (on reserves held at the Fed overnight). The Fed also created a reverse-repo market, which works as follows: Companies with excess capital, who wanted to earn a yield on that capital in the short-term, would buy government treasuries from the Fed, and promise to sell those treasuries back to the Fed shortly thereafter (from overnight, to 1 month later). The Fed would buy these treasuries back at a premium, thereby providing the firms with a short-term yield on this repurchase (repo) transaction. Currently, the IORB rate is at 0.15%, and the reverse repo rate is set at 0.05%. What this means is that in the short-term, no provider of capital would provide their capital at a rate below 0.05%. Why would they, since they could participate in the reverse repo market and earn 0.05% minimum? Furthermore, no primary dealer bank would lend reserves to another primary dealer bank at a rate below 0.15%, when they could just leave their reserves on deposit with the Fed and earn 0.15%. Recall, the reserves held at the Fed can only be transferred between primary dealer banks, and cannot be leant-out into the economy. The Fed would therefore have created a price floor for short-term interest rates, preventing them from reaching 0% or below as a result of QE. As an aside, banks assess the position of their reserve balances on a daily basis. On certain days some banks in the banking system may have excess reserves, whereas others may have a deficit in reserves. The banks with excess reserves would typically lend these reserves to other banks in what is called the interbank market. If a bank experiencing a deficit in reserves could not raise sufficient reserves in the interbank market, it could turn to the Fed for assistance. The Fed would then act as a lender of last resort to the bank in deficit, and lend them the reserves that they required in order to remain sound from a risk perspective. The Fed would charge a rate of interest on this reserve loan to the bank in deficit, and this rate is referred to as the “discount” rate. Not to be confused with discount rates in general (which are used to price treasury bills). The Fed would want to incentivise banks to meet their reserve requirements by first borrowing from their fellow banks. Also, a bank would only struggle to raise sufficient reserves in the inter-bank market if its fellow banks would be unwilling to loan it reserves. This would typically indicate a credit quality problem with the bank experiencing a deficit in reserves. For these reasons, the Fed’s discount rate would be set at a rate higher than the interbank lending rate. The interbank lending rate is referred to as the “Federal Funds Rate”, and is set at a target rate by the Fed periodically. The Federal Funds Rate affects the rates at which banks are able to raise short-term capital from one another, and is therefore an important benchmark used in the pricing of all other rates of all financial instruments in the economy. By controlling the Federal Funds Rate, the Fed can effectively control all other rates in the economy.
The Federal Funds Rate acts as a price ceiling for short term interest rates in the economy, since banks would only use the Fed as a lender of last resort. In other words, banks would try to raise short-term capital from every other means possible, before approaching the most expensive source of short-term funding: the Fed.
In this article we have seen that money is not created from savers depositing their excess cash into their bank accounts. Instead money is created when banks issue new loans or new deposits to businesses or individuals;
We have seen that many developed financial markets have zero reserve requirements, and for this reason, the banks in such markets are able to create additional loans without having reserve requirements as a lending constraint;
Broad money was defined as the money in circulation in an economy, whereas base money was defined as the money held in the form of notes and coin, and bank reserves held at the central bank;
An example of money creation was shown, whereby a home buyer purchased a home using a home loan. From this transaction the balance sheet of the home buyer increased (they acquired a new house as an asset, and simultaneously incurred an additional liability in the form of a home loan). The balance sheet of the buyer’s bank increased if both the buyer and the seller of the house banked at the same bank. If not, the buyer’s and seller’s bank’s balance sheets were affected in the following ways: the buyer’s bank acquired a new asset (the home loan), and the seller’s bank acquired a new liability (the cash deposit paid to the seller’s bank account arising from the sales proceeds of the house). The new money (broad money) was created when the home loan was granted, and this new money was reduced or destroyed gradually as the home loan was paid off;
The concept of QE was introduced, whereby central banks aimed to buy government treasuries from banking and non-banking firms in the hope of reducing interest rates for longer-term maturities, and indirectly attempting to stimulate economic activity;
When QE is performed with banks (primary dealer banks) then no additional broad money is created, only base money is created. However, when QE is performed with non-banking firms, both broad money and base money are created;
One of the unintended consequences of QE would be for short-term rates to drop to zero, or even become negative. In order to prevent this, the Fed would set a price floor on short-term rates: IORB for overnight reserve balances, and the reverse repo rate for non-primary dealer banks and other firms in the economy;
Banks lend and borrow from each other in the interbank market. When a bank is unable to raise short-term funding in the interbank market it may turn to the Fed as a lender of last resort. The Fed would provide a loan to a bank at the discount rate. The discount rate would be set at a level higher than interbank lending rates to incentivise banks to raise funds from other banks in the interbank market first. In this manner the discount rate acts as a price ceiling for short-term rates.